ABSTRACT 
 
Title of Dissertation: REVISION OF THE GENERA OF THE 
RHAGIONIDAE OF THE WORLD (DIPTERA: 
BRACHYCERA) 
 
 Peter H. Kerr, Doctor of Philosophy, 2004 
 
Dissertation Directed by: Charles Mitter 
 Norman E. Woodley 
 
As a group, flies represent one of the most prolific and important elements of our 
natural world. The order Diptera comprises approximately 150,000 species in 
approximately 142 families. My research focuses on the Rhagionidae (formerly 
known as the Leptidae), a family of flies considered to contain some of the most 
primitive living members of the dipteran suborder Brachycera and believed to have 
diversified as early as 170 million years ago. The taxonomic classification of the 
Rhagionidae has been unstable for decades because there are few morphological 
characters that can be used to support hypotheses of relationship among its members. 
Much of the morphology in this group, however, has not been examined 
systematically.  
 
An independent estimate of phylogeny for the group is carried out and presented here, 
based on 208 morphological characters for 43 ingroup species and molecular 
characters consisting of 3200+ bp sequences of 28S rDNA of 34 ingroup species. The 
goal of this work is to better understand how the genera of the Rhagionidae relate to 
one another and to their kin within the infraorder Tabanomorpha. Ultimately, this 
knowledge is fundamental for developing a stable classification system for the group.  
 
The Rhagionidae are recognized as a monophyletic group containing four subfamilies 
containing a total of 17 extant genera. The subfamily Spaniinae is defined by a 
special modification of tergite 9 of the female genitalia, which is shared by members 
of Omphalophora, Ptiolina, Spania, Spaniopsis, and Symphoromyia. Omphalophora 
Becker is resurrected from synonymy with Ptiolina. Spaniinae is defined by having 
scale-like thoracic hairs, as in Chrysopilus, Schizella and Stylospania. Arthroceratinae 
contains a single enigmatic genus, Arthroceras. Most females belonging to these 
three subfamilies have spermathecal duct accessory glands. These structures are 
reported here for the first time and are unique in Tabanomorpha. The Rhagioninae is 
the most primitive subfamily of the Rhagionidae. The saw sclerite in the larval 
mandible may be synapomorphic for this subfamily. Members of Rhagioninae 
include Atherimorpha, Desmomyia, Rhagio, and Sierramyia gen nov. Rhagina 
Malloch is recognized as a junior synonym of Rhagio. The Bolbomyiidae are 
recognized at the family level for the first time. Alloleptis tersus is incertae sedis 
within Tabanomorpha. 
 
Two new species are described: Schizella woodleyi (from Luz?n, Philippines) and 
Sierramyia chiapasensis (from Chiapas, Mexico). A key is given to the genera of the 
Rhagionidae with dichotomies leading to all families of Tabanomorpha. Genera of 
Austroleptidae, Bolbomyiidae, and Rhagionidae are diagnosed and described, with a 
list of included species for each genus.  
 
REVISION OF THE GENERA OF THE RHAGIONIDAE OF THE WORLD  
(DIPTERA: BRACHYCERA) 
 
by 
 
Peter H. Kerr 
 
Dissertation submitted to the Faculty of the Graduate School of the  
University of Maryland, College Park in partial fulfillment  
of the requirements for the degree of  
Doctor of Philosophy 
2004 
 
 
 
Advisory Committee: 
Associate Professor Charles Mitter 
Adjunct Professor Norman E. Woodley 
Adjunct Professor Wayne N. Mathis 
Associate Professor Jeffrey W. Shultz 
Associate Professor Charles F. Delwiche 
 
 
 
 
?Copyright by 
Peter H. Kerr 
2004 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
ii 
 
This dissertation should not be considered a scientific publication for the 
purposes of zoological nomenclature. 
 
iii 
 
 
 
In honor of Cynthia Virginia Kerr Gould, 
my loving grandmother, 
who taught me to appreciate a fascinating world. 
 
  
 
 
iv 
Acknowledgements 
Financial support for this research was provided by the Williston Fund (Smithsonian 
Institution), the American Museum of Natural History, the Washington Biologists 
Field Club, the CanaColl Foundation, the Graduate School and Department of 
Entomology at the University of Maryland, and the National Science Foundation 
Partnerships for Enhancing Expertise in Taxonomy (PEET 952-1773).  
 
I am greatly indebted to the following people and institutions that generously made 
specimens available for this study. These include David A. Grimaldi, American 
Museum of Natural History, New York (AMNH); Dan J. Bickel, Max Moulds and 
David K. McAlpine, Australia Museum, Sydney (AMSA); David K. Yeates, 
Australian National Insect Collection, Canberra (ANIC); Bernhard Merz, Mus?um 
d'Histoire Naturelle, Geneva; John E. Chainey, Natural History Museum, London 
(BMNH); Neal L. Evenhuis, B. P. Bishop Museum, Honolulu (BPBM); Keve 
Ribardo, California Academy of Sciences, San Francisco (CAS); Jeffrey Cumming, 
Canadian National Collection of Insects, Ottawa (CNC); James K. Liebherr, Cornell 
University, Ithaca (CU); Stephen A. Marshall and Matthias Buck, University of 
Guelph, Ontario (DEBU); Manuel Zumbado, Instituto Nacional de Biodiversidad, 
Heredia, Costa Rica (INBIO); Donald W. Webb, Illinois Natural History Survey, 
Champaign (INHS); Gregory W. Courtney, Iowa State Insect Collection, University 
of Iowa State, Ames (ISUC); Brian V. Brown, Natural History Museum of Los 
Angeles County, Los Angeles (LACM); Philip D. Perkins, Museum of Comparative 
Zoology, Harvard University, Cambridge (MCZ); Pekka Vilkamaa, Zoological 
 
v 
Museum, University of Helsinki, Helsinki (MZHF); Thomas Pape, Naturhistoriska 
Riksmuseet, Stockholm (NHRS); David A. Barraclough and Brian R. Stuckenberg, 
Natal Museum, Pietermaritzburg, South Africa (NMSA); and  Hans-Peter Tschorsnig, 
Staatliches Museum f?r Naturkunde, Stuttgart (SMNS). 
 
Molecular work on the Rhagionidae would not have been possible without the help of 
the following people: Brian Wiegmann and Sonja Scheffer generously provided the 
28S rDNA primers used in this work; in addition to supplying me with unpublished 
xylophagid sequences, Brian Wiegmann sent me Dasyomma specimens, and a 
Chrysopilus species from Queensland in alcohol; Donald W. Webb provided the 
critical taxon Ptiolina fasciata from Saskatchewan; David Yeates provided all 
Spaniopsis specimens; Richard Bashford of Forestry Tasmania provided a species of 
Atherimorpha and Chrysopilus from the WARRA Long Term Research Site in 
Tasmania, Australia; Malin Kerr assisted in the collection of South American 
Atherimorpha species; Wayne N. Mathis provided Symphoromyia sp. from Kenai 
Fjords National Park, Alaska; and Amnon Frieberg kindly sent me the larvae of a new 
species of Vermileo from Israel.  
 
I appreciate the patience, encouragement and guidance of each person on my thesis 
committee: Charles Mitter, Norman E. Woodley, Wayne N. Mathis, Jeffrey W. 
Shultz, and Charles F. Delwiche. I feel very fortunate to have had the opportunity to 
do this research in the company of many fine scientists at the Smithsonian Institution 
and University of Maryland. It was my special honor to work alongside Norman 
 
vi 
Woodley, Wayne Mathis, F. Christian Thompson, Allen Norrbom, Raymond Gagne, 
and Oliver Flynt on the Diptera floor at the United States National Museum (USNM). 
Holly Williams, Lucrecia Rodriguez, and Jonathan Eibl were also a critical part of my 
experience at the USNM and I thank them for their assistance. The Entomology 
Department at the University of Maryland always provided a rich, welcoming, 
inspiring, and supportive atmosphere of which I am grateful and proud. I especially 
wish to thank Charles Mitter, who served as my primary advisor and offered 
optimism and encouragement in the face of apparent failures all along the way. His 
stubborn faith in me kept me from giving up on this project. I also want to thank 
Jeffrey Shultz, Charles Delwiche, Ray St. Legier and David Hawthorne for sharing 
their wisdom as I completed coursework and confronted the challenges that came 
with my research.  I want to thank fellow entomology students for the million favors I 
asked of them and for our many interesting discussions on systematic theory and 
methodology: Alessandra Baptista, Chris Desjardins, Isaac Winkler, Ethan Kane, 
Andrew Mitchell, Roger Hutchings, and Nick Baer. I also want to thank the students 
in the Charles Delwiche lab: Malin Kerr, Tsetso Bachvaroff, Mathew Cimino, John 
Hall, Ken Karol, Jeff Lewandowski, and Maria Virginia Sanchez Puerta for sharing 
their expertise and enthusiasm in molecular techniques and analysis. Gregory Hess? 
enthusiastic work to keep the entomology department computer lab up-to-date, with 
terrifically fast and stable equipment was invaluable. 
 
Finally I wish to thank my wife Malin Kerr, who kept me sane, healthy and happy 
(and in love) through the whole process. 
 
vii 
ABSTRACT.................................................................................................................. i 
List of Figures............................................................................................................. xi 
Chapter 1. General Introduction to the Genera of the Rhagionidae and Related 
Taxa.............................................................................................................................. 1 
Overview................................................................................................................... 1 
Recent Classification ................................................................................................ 5 
History of the Concept of Rhagionidae .................................................................. 11 
Biology.................................................................................................................... 37 
Diversity.................................................................................................................. 42 
Fossil Record .......................................................................................................... 43 
Biogeography.......................................................................................................... 46 
Phylogenetic Hypotheses........................................................................................ 47 
Chapter 2. Phylogenetic Analysis of the Genera of the Rhagionidae and Related 
Taxa............................................................................................................................51 
Introduction............................................................................................................. 51 
Morphological Treatment ....................................................................................... 55 
Morphological Treatment ....................................................................................... 55 
Introduction......................................................................................................... 55 
Materials and Methods........................................................................................ 57 
Laboratory Methods........................................................................................ 57 
Taxon Sampling.............................................................................................. 59 
Phylogenetic Analysis..................................................................................... 63 
Results................................................................................................................. 63 
Morphological Characters and Character State Coding.................................. 63 
Maximum Parsimony Analysis of Morphological Data ............................... 187 
Discussion......................................................................................................... 196 
Monophyly of the Rhagionidae .................................................................... 196 
Pelecorhynchidae .......................................................................................... 198 
Definition of Chrysopilinae, Arthrocerinae, and Spaniinae.......................... 199 
Ptiolina and Omphalophora ......................................................................... 201 
Rhagioninae .................................................................................................. 213 
Bolbomyia and Austroleptis .......................................................................... 214 
Molecular Treatment............................................................................................. 215 
Introduction....................................................................................................... 215 
Materials and Methods...................................................................................... 217 
Taxon Sampling............................................................................................ 217 
Extraction...................................................................................................... 220 
Amplification ................................................................................................ 220 
Sequencing.................................................................................................... 222 
Sequence Alignment ..................................................................................... 222 
Phylogenetic Analysis................................................................................... 230 
Tests of alternative hypotheses using Maximum Parsimony........................ 231 
Results............................................................................................................... 232 
Alignment region classification .................................................................... 232 
 
viii 
Effects of alignment...................................................................................... 233 
Maximum Parsimony Analysis of Molecular Dataset.................................. 240 
Maximum Likelihood Analysis of Molecular Dataset.................................. 245 
Discussion......................................................................................................... 248 
Exploring the effects of alignment................................................................ 248 
Evaluating the Alignment Sets...................................................................... 249 
Assessment of the Most Parsimonious Trees................................................ 250 
Assessment of the Most Likely Trees........................................................... 251 
Monophyly of Rhagionidae .......................................................................... 251 
Monophyly and composition of Pelecorhynchidae....................................... 256 
Monophyly and composition of Chrysopilinae and Spaniinae..................... 256 
Bolbomyia ..................................................................................................... 257 
Austroleptis ................................................................................................... 259 
Vermileonidae............................................................................................... 259 
Combined Analysis............................................................................................... 261 
Introduction....................................................................................................... 261 
Materials and Methods...................................................................................... 262 
Taxon Sampling............................................................................................ 262 
Combined Phylogenetic Analysis................................................................. 264 
Supertree analysis ......................................................................................... 265 
Results............................................................................................................... 266 
Maximum Parsimony Analysis of Combined Data ...................................... 266 
Subset Taxon Sampling of Morphological Data........................................... 266 
Subset Taxon Sampling of Molecular Data.................................................. 268 
Partition Homogeneity Test .......................................................................... 270 
Combined Sample of Morphology and Molecular Data............................... 270 
Bayes Analysis of Combined Dataset........................................................... 273 
Subset Taxon Sampling of Morphological Data........................................... 273 
Subset Taxon Sampling of Molecular Data.................................................. 275 
Combined Analysis of Morphology and Molecular Data............................. 277 
Supertrees Analysis....................................................................................... 278 
Matrix Representation using Parsimony (MRP)........................................... 278 
Discussion......................................................................................................... 281 
Combined analysis........................................................................................ 281 
More taxa or more characters? Both!............................................................ 281 
Bayes analysis............................................................................................... 282 
Supertree analysis ......................................................................................... 282 
Conclusions of the Phylogenetic Analysis............................................................ 284 
Chapter 2 Notes..................................................................................................... 287 
Chapter 3. Definition of Rhagionid Genera and Related Taxa, and their Family 
Groups...................................................................................................................... 290 
Introduction........................................................................................................... 290 
Key to Adult Genera of Rhagionidae.................................................................... 290 
Austroleptidae....................................................................................................... 298 
Genus AUSTROLEPTIS Hardy ....................................................................... 298 
List of included species................................................................................. 302 
 
ix 
Bolbomyiidae status revised................................................................................. 304 
Genus BOLBOMYIA Loew............................................................................. 304 
List of included species................................................................................. 308 
Rhagionidae .......................................................................................................... 309 
Genus ALLOLEPTIS Nagatomi and Saigusa................................................... 309 
List of included species................................................................................. 311 
Genus ARTHROCERAS Williston .................................................................. 311 
List of included species................................................................................. 315 
Genus ARTHROTELES Bezzi......................................................................... 316 
List of included species................................................................................. 319 
Genus ATHERIMORPHA White..................................................................... 320 
List of included species................................................................................. 325 
Genus CHRYSOPILUS Macquart.................................................................... 326 
List of included species................................................................................. 332 
Genus DESMOMYIA Brunetti......................................................................... 342 
List of included species................................................................................. 345 
Genus LITOLEPTIS Chillcott .......................................................................... 345 
List of included species................................................................................. 348 
Genus OMPHALOPHORA Becker.................................................................. 349 
List of included species................................................................................. 353 
Genus PTIOLINA Zetterstedt........................................................................... 354 
List of included species................................................................................. 359 
Genus RHAGIO Fabricius................................................................................ 360 
List of included species................................................................................. 365 
Genus SCHIZELLA Bezzi ............................................................................... 371 
Key to Schizella species................................................................................ 375 
Schizella woodleyi new species.................................................................... 376 
List of included species................................................................................. 379 
Genus SIERRAMYIA, NEW GENUS............................................................. 379 
Sierramyia chiapasensis Kerr, new species.................................................. 383 
Genus SPANIA Meigen.................................................................................... 386 
List of included species................................................................................. 390 
Genus SPANIOPSIS White .............................................................................. 391 
List of included species................................................................................. 395 
Genus STYLOSPANIA Frey............................................................................ 395 
List of included species................................................................................. 398 
Genus SYMPHOROMYIA Frauenfeld ............................................................ 398 
List of included species................................................................................. 404 
Notes. .................................................................................................................... 405 
Appendix A. Figures. .............................................................................................. 414 
References................................................................................................................ 587 
 
 
x 
List of Tables. 
 
Table 1. Presence of Rhagionidae in the fossil record................................................ 45 
Table 2. Taxon sampling, morphological analysis. .................................................... 61 
Table 3. Species with incomplete sampling................................................................ 62 
Table 4. A list of the morphological characters and their state coding. Characters not 
used in the phylogenetic analyses are in italics................................................. 167 
Table 5. Determination of Ptiolina species into Ptiolina and Omphalophora 
subgroups. ......................................................................................................... 211 
Table 6. Taxon sampling for molecular analyses. Classification consistent with 
Woodley (1989) / Stuckenberg (2001). ............................................................ 219 
Table 7. 28S rDNA primers (from Wiegmann et al., 2000). .................................... 221 
Table 8. Relationships among the major taxa, evident in the Maximum Parsimony 
majority rule consensus trees of different alignment sets (AS). ....................... 234 
Table 9. Relationships among the major taxa, evident in the Maximum Likelihood 
trees of different alignment sets (AS). .............................................................. 235 
Table 10. Average pairwise genetic distance between each species and the rest of the 
sampled set. All characters included except the ?unalign? character set (AS #1).  
The mean of all pairwise distances is 0. 06139372........................................... 237 
Table 11. Average pairwise genetic distance between each clade and the rest of the 
sampled set.  All characters included except the ?unalign? character set (AS #1).
........................................................................................................................... 238 
Table 12. Overlap between most parsimonious consensus trees and bootstrap 
consensus trees. The highest overlap between all compatible majority rules 
consensus trees and bootstrap consensus trees occurs in alignment sets 3, 4, and 
6. The highest overlap between strict consensus and bootstrap consensus trees 
occurs in alignment set #4................................................................................. 239 
Table 13. Clades of interest. ..................................................................................... 241 
Table 14. Taxonomic abbreviations.......................................................................... 243 
Table 15. Hypotheses tested in light of molecular evidence. ................................... 253 
Table 16. Combined analysis dataset; morphological and molecular taxon overlap.263 
 
 
 
 
xi 
List of Figures. 
Figure 1. Brachycerous Diptera tree of life (adapted from Yeates and Wiegmann, 
1999). Rhagionidae are located within Tabanomorpha, at the base of the tree. ... 3 
Figure 2. Accumulation of Genus Names in Rhagionidae. The period between 1878 to 
1934 marked an expansive period for the concept of Rhagionidae, as the 
boundaries of this family were exceptionally porous.. ....................................... 18 
Figure 3. Strict consensus of 24 most parsimonious trees generated from the 
morphological matrix........................................................................................ 188 
Figure 4. Primers used for sequencing of 28S rDNA gene....................................... 222 
Figure 5. Maximum Parsimony analysis of 28S rDNA data (AS #4).  Strict consensus 
of 6 most parsimonious trees.. .......................................................................... 244 
Figure 6. Maximum Likelihood analysis of 28S rDNA data (AS #4)...................... 247 
Figure 7. Maximum Parsimony analysis of pruned morphological matrix. ............. 267 
Figure 8. Maximum Parsimony analysis of pruned molecular matrix...................... 269 
Figure 9. Maximum Parsimony analysis of combined matrix.................................. 272 
Figure 10. Bayesian tree for pruned morphology matrix.......................................... 274 
Figure 11. Bayesian tree generated from molecular characters of data matrix  
composed of taxa appropriate for a combined molecular and morphological 
analysis.............................................................................................................. 276 
Figure 12. Bayesian tree for combined morphology and molecular data. Posterior 
probabilities are noted above branches. ............................................................ 278 
Figure 13. Strict consensus of 958299 trees from MRP analysis. ............................ 280 
Figure 14. Schematic of the male genitalia, showing primary structures................. 414 
Figure 15. Dorsal outline of hypandrium for illustration of Character 118, origin of 
gonocoxal apodeme.. ........................................................................................ 414 
Figure 16.  Aedeagus, lateral view. (A) Chrysopilus sp., with lateral ejaculatory 
apodemes [USNMENT00025242] (B) Suragina concinna, with aedeagal tines 
[USNMENT00025980] (C) Bolbomyia nana, with aedeagal tines. 
[USNMENT00024051] (D) Arthroceras pollinosum, with lateral ejaculatory 
apodemes and aedeagal tines [USNMENT00022737]. .................................... 415 
Figure 17. Lateral view of Austroleptis multimaculata [USNMENT00025739] ..... 416 
Figure 18. SEM image of antenna of Austroleptis multimaculata............................ 416 
Figure 19. Mouthparts of Austroleptis multimaculata [USNMENT00025905]. (A) 
Labellum, ventral view (B) Hypopharynx and associated structures, lateral view 
(C) Palps and associated structures................................................................... 417 
Figure 20. Wing in species of Austroleptis. (A) A. multimaculata 
[USNMENT00025745] (B) Austroleptis sp. (South America) 
[USNMENT00022609]..................................................................................... 418 
Figure 21. Male genitalia of Austroleptis multimaculata [USNMENT00025739]. (A) 
epandrium, dorsal view (B) hypandrium, dorsal view...................................... 419 
Figure 22. Internal structures of female genitalia of Austroleptis sp. 
[USNMENT00025761] (A) Sternite 9 and associated structures, dorsal view (B) 
Spermatheca...................................................................................................... 420 
 
xii 
Figure 23. Internal structures of female genitalia of Austroleptis multimaculata 
[USNMENT00024145] (A) Sternite 9 and associated structures, dorsal view (B) 
Detail of genital chamber and ejaculatory process of spermathecal ducts. ...... 421 
Figure 24. Female tergite 10 and cercus of Austroleptis sp. [USNMENT00025761]
........................................................................................................................... 422 
Figure 25. Habitus of Bolbomyia nana. (A) male, [USNMENT00025987] (B) female, 
[USNMENT00022909]..................................................................................... 423 
Figure 26. Head and antennal form of Bolbomyia nana, male 
[USNMENT00022909]..................................................................................... 424 
Figure 27. Mouthparts of Bolbomyia nana [USNMENT00025904]. (A) Labellum and 
palps (B) Hypopharynx and associated structures. ........................................... 425 
Figure 28. Male genitalia of Bolbomyia nana [USNMENT00024051]. (A) epandrium, 
dorsal view (B) hypandrium, dorsal view......................................................... 426 
Figure 29. Internal structures of the female genitalia of Bolbomyia nana. (A) Sternite 
9 [USNMENT00022946] (B) Spermathecae [USNMENT00022920]............. 427 
Figure 30. Female terminalia of Bolbomyia nana. (A) Dorsal view 
[USNMENT00022920] (B) Dorsal view of tergite 10, lateral view of left cercus 
[USNMENT00022946]..................................................................................... 428 
Figure 31. Female sternite 8 of Bolbomyia nana [USNMENT00022920]............... 429 
Figure 32. Antenna in species of Arthroceras. (A) A. fulvicorne, female, lateral view 
[USNMENT00022731] (B) A. gadi, male, oblique lateral view 
[USNMENT00022629]..................................................................................... 430 
Figure 33. Antenna in species of Arthroceras. (A) A. pollinosum, female, dorsal view 
[USNMENT00025222] (B) A. leptis, male, oblique lateral view 
[USNMENT00022614]..................................................................................... 431 
Figure 34. Mouthparts of Arthroceras pollinosum [USNMENT00022625]. (A) 
Labellum, ventral view (B) Hypopharynx, cibarial pump, and associated 
structures (C) Palps and lacinia, ventral view................................................... 432 
Figure 35. Wing in species of Arthroceras. (A) A. fulvicorne, female, dorsal view 
[USNMENT00022731] (B) A. gadi, female, ventral view 
[USNMENT00022628]..................................................................................... 433 
Figure 36. Wings in species of Arthroceras, dorsal view. (A) A. pollinosum, female 
[USNMENT00025222]. (B) A. leptis, male [USNMENT00022614] .............. 434 
Figure 37. Epandria (male genitalia) in species of Arthroceras, dorsal view. (A) A. 
fulvicorne [USNMENT00022737] (B) A. leptis [USNMENT00022613]........ 435 
Figure 38. Dorsal view of hypandrium (male genitalia) in species of Arthroceras. (A) 
A. fulvicorne [USNMENT00022737] (B) A. leptis [USNMENT00022613] ... 436 
Figure 39. Aedeagus (male genitalia) in species of Arthroceras. Aedeagal tines (= at) 
are indicated. (A) A. fulvicorne, lateral view [USNMENT00022601] (B) A. 
pollinosum, dorsolateral view [USNMENT00022737] .................................... 437 
Figure 40. Female genitalia in Arthroceras fulvicorne USNM ENT 00025219. (A) 
Sternite 9, with associated spermathecal structures from the dorsal perspective. 
(B) Detail of spermathecal duct, with accessory gland..................................... 438 
Figure 41. Female terminalia structures in Arthroceras fulvicorne 
[USNMENT00022736]. (A) Cerci and associated structures, partially dissected 
showing sternite 9 from dorsal perpective (B) Sternite 8, ventral view ........... 439 
 
xiii 
Figure 42. Arthroteles cinerea, female [USNMENT00023229] .............................. 440 
Figure 43.  Antenna in species of Arthroteles, lateral view. (A) A. cinerea, female 
[USNMENT00023229] (B) A. bombyliiformis, male [USNMENT00025009] 441 
Figure 44. SEM image of head of Arthroteles bombyliiformis................................. 442 
Figure 45. Wing in species of Arthroteles. (A) A. cinerea, male, ventral view 
[USNMENT00023231] (B) A. bombyliiformis, male, dorsal view 
[USNMENT000233230]................................................................................... 443 
Figure 46. Male genitalia of Arthroteles bombyliiformis [USNMENT00024986]. (A) 
Epandrium, ventral view (B) Hypandrium, dorsal view................................... 444 
Figure 47. Sternite 9 (female genitalia) and associated structures of Arthroteles 
bombyliiformis. ................................................................................................. 445 
Figure 48. External structures of the female terminalia of Arthroteles bombyliiformis 
[USNMENT00025017]. (A) Tergite 10, dorsal view; and cerci, splayed flat, 
lateral view (B) Sternite 8, dorsal view............................................................. 446 
Figure 49. Habitus of Atherimorpha sp. (South America), female 
[USNMENT00023370]..................................................................................... 447 
Figure 50. Antenna of Atherimorpha atrifemur........................................................ 447 
Figure 51. Macerated heads showing mouthparts of Atherimorpha nemoralis. (A) 
Posterior view [USNMENT28402] (B) Lateral view [USNMENT00025903] 448 
Figure 52. Mouthparts of Atherimorpha nemoralis [USNMENT00025903]. (A) 
Labellum (B) Hypopharynx and associated structures (C) Palps and associated 
structures ........................................................................................................... 449 
Figure 53. Epandrium in species of Atherimorpha. Dorsal view. (A) A. triangularis 
[USNMENT00028418] (B) Atherimorpha sp. (South America) 
[USNMENT00028520]..................................................................................... 450 
Figure 54. Hypandrium in species of Atherimorpha. Dorsal view. (A) A. triangularis 
[USNMENT00028418] (B) Atherimorpha sp. (South America) 
[USNMENT00028520]..................................................................................... 451 
Figure 55. Internal structure of the female genitalia in Atherimorpha nemoralis. (A) 
Dorsal view [USNMENT00025317] (B) Oblique lateral view 
[USNMENT00025109]..................................................................................... 452 
Figure 56. External structures of the female terminalia of Atherimorpha nemoralis. 
(A) Dorsal view [USNMENT00025317] (B) Oblique lateral view 
[USNMENT00025109]..................................................................................... 453 
Figure 57. Female sternite 8 of Atherimorpha nemoralis [USNMENT00025109]. 
Dorsal view ....................................................................................................... 454 
Figure 58. Antennal form in species of Chrysopilus, lateral view. (A) C. ornatus, 
female [USNMENT00025947] (B) C. quadratus, female [USNMENT00025948]
........................................................................................................................... 455 
Figure 59. Wing of Chrysopilus species, dorsal view. (A) C. ornatus 
[USNMENT00025947] (B) C. quadratus [USNMENT00025948] ................. 456 
Figure 60. Epandrium in species of Chrysopilus, dorsal view. (A) C. quadratus 
[USNMENT00025951] (B) Chrysopilus sp. (New Caledonia) 
[USNMENT00025952] (C) Chrysopilus thoracicus [USNMENT00025242] . 457 
Figure 61. Hypandrium in species of Chrysopilus, dorsal view. (A) C. quadratus 
[USNMENT00025951] (B) Chrysopilus rotundipennis [USNMENT00025980] 
 
xiv 
(C) Chrysopilus sp. (New Caledonia) [USNMENT00025952] (D) Chrysopilus 
thoracicus [USNMENT00025242]................................................................... 458 
Figure 62. Internal structures of the female genitalia in Chrysopilus thoracicus 
[USNMENT00025875], dorsal view ................................................................ 459 
Figure 63. Internal structures of the female genitalia in Chrysopilus sp. 
[USNMENT00025877], ventral view. (A) Detail of furrowed and sclerotized 
ejaculatory process of the spermathecal ducts (B) Spermatheca ...................... 460 
Figure 64. Spermathecae and spermathecal duct accessory gland in Chrysopilus 
thoracicus [USNMENT00025875]................................................................... 461 
Figure 65. Partially dissected female terminalia of Chrysopilus testaceipes 
[USNMENT00025876], dorsal view ................................................................ 462 
Figure 66. Male genitalia of Desmomyia thereviformis [USNMENT00025267]. (A) 
Epandrium, ventral view (B) Hypandrium, dorsal view................................... 463 
Figure 67. Sternite 9 (female genitalia) and associated structures of Desmomyia 
thereviformis [USNMENT00025628] .............................................................. 464 
Figure 68. Spermathecae of Desmomyia thereviformis [USNMENT00025628] ..... 465 
Figure 69. Litoleptis alaskensis, male. (A) Habitus, lateral view 
[USNMENT00024416, holotype] (B) head, anterior view 
[USNMENT00024417, paratype]..................................................................... 466 
Figure 70. Antennal form in species of Litoleptis. (A) L. alaskensis, male 
[USNMENT00024416, holotype] (B) L. alaskensis, male 
[USNMENT00024417, paratype] (C) L. chilensis [no USNM barcode, slide 
mounted holotype] ............................................................................................ 467 
Figure 71. Wing in species of Litoleptis. (A) L. alaskensis, male 
[USNMENT00024417, paratype] (B) L. chilensis, male [no USNM barcode, 
slide mounted holotype].................................................................................... 468 
Figure 72. Male genitalia of Litoleptis alaskensis, holotype [USNMENT00024416]. 
(A) Epandrium, oblique dorsal view (B) Hypandrium, dorsal view................. 469 
Figure 73. Antennal form in species of Omphalophora. (A) O. fasciata, male 
[USNMENT00025460] (B) O. majuscula, female [USNMENT00025471] .... 470 
Figure 74. Antennal form in species of Omphalophora and Ptiolina. (A) O. 
nigripilosa, male [USNMENT00025204 HOLOTYPE] (B) P. nitida, male 
[USNMENT00022957]..................................................................................... 471 
Figure 75. Wing in species of Omphalophora, ventral view. (A) O. fasciata 
[USNMENT00025460] (B) O. majuscula [USNMENT00025471]................. 472 
Figure 76. Wing in species of Omphalophora and Ptiolina, ventral view. (A) O. 
nigripilosa [USNMENT00025205, allotype] (B) P. nitida 
[USNMENT00022957]..................................................................................... 473 
Figure 77. Epandrium (male genitalia) in species of Omphalophora, dorsal view. (A) 
O. fasciata [USNMENT00025461] (B) O. lapponica [USNMENT00025921]474 
Figure 78. Epandrium (male genitalia) of Omphalophora majuscula 
[USNMENT00025472]..................................................................................... 475 
Figure 79. Hypandrium (male genitalia) in species of Omphalophora, dorsal view. 
(A) O. fasciata [USNMENT00025461] (B) O. lapponica 
[USNMENT00025921]..................................................................................... 476 
 
xv 
Figure 80. Hypandrium (male genitalia) of Omphalophora majuscula 
[USNMENT00025472]..................................................................................... 477 
Figure 81. Dissected female terminalia of Omphalophora majuscula, dorsal view 
[USNMENT00025474]..................................................................................... 478 
Figure 82. Spermatheca and spermathecal duct accessory glands of Omphalophora 
majuscule [USNMENT00025474] ................................................................... 479 
Figure 83. Antennal form in species of Ptiolina. (A) P. edeta, male 
[USNMENT00022657] (B) P. edeta, female [USNMENT00023016] ............ 480 
Figure 84. Antennal form in species of Ptiolina. (A) P. obscura, male 
[USNMENT00023018] (B) P. zonata, female [USNMENT00022842] .......... 481 
Figure 85. Labellum and theca in species of Ptiolina. (A) Ptiolina sp., male, posterior 
view [USNMENT00022845] (B) P. zonata, male, lateral view 
[USNMENT00023001]..................................................................................... 482 
Figure 86. Hypopharynx, cibarial pump, and associated structures in species of 
Ptiolina, posterior view. (A) Ptiolina sp., male [USNMENT00022845] (B) P. 
zonata, male [USNMENT00023001] ............................................................... 483 
Figure 87. Palp and associated structures in Ptiolina. (A) Ptiolina sp., male 
[USNMENT00022845] (B) P. zonata, male [USNMENT00023001] ............. 484 
Figure 88. Wing in species of Ptiolina, ventral view. (A) P. obscura 
[USNMENT00023020] (B) P. zonata [USNMENT00022842] ....................... 485 
Figure 89. Epandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
edeta [USNMENT00023011] (B) P. nigra [USNMENT00025932] ............... 486 
Figure 90. Epandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
nitida [USNMENT00025933] (B) P. obscura [USNMENT00025937]........... 487 
Figure 91. Epandrium (male genitalia) of Ptiolina zonata [USNMENT00022846].  
Dorsal view. Note: subepandrial sclerite flipped upside-down, facing wrong 
direction. ........................................................................................................... 488 
Figure 92. Hypandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
edeta [USNMENT00023011] (B) P. nigra [USNMENT00025932] ............... 489 
Figure 93. Hypandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
nitida [USNMENT00025933] (B) P. obscura [USNMENT00025937]........... 490 
Figure 94. Hypandrium (male genitalia) of Ptiolina zonata [USNMENT00022846]
........................................................................................................................... 491 
Figure 95. Dissected female terminalia in species of Ptiolina, dorsal view. (A) P. 
zonata [USNMENT00022841] (B) P. mallochi [USNMENT00022953] ........ 492 
Figure 96. Spermatheca and spermathecal duct accessory glands in species of 
Ptiolina. (A) P. zonata [USNMENT00022841] (B) P. mallochi 
[USNMENT00022953]..................................................................................... 493 
Figure 97. Male genitalia of Rhagina incurvatus [USNMENT00022728]. (A) 
Epandrium (B) Hypandrium ............................................................................. 494 
Figure 98. Female terminalia of Rhagina incurvatus [USNM ENT 00025853]. (A) 
Dissected, showing spermathecae (B) Detail of cerci....................................... 495 
Figure 99. Internal structures of the female genitalia of Rhagina incurvatus [USNM 
ENT00025853]. (A) Dorsal view of sternite 9 (B) Spermatheca ..................... 496 
Figure 100. Habitus of Rhagio mystaceus, a common species in the Maryland/ DC 
area.................................................................................................................... 497 
 
xvi 
Figure 101. Antennal form in species of Rhagio, lateral view. (A) R. punctipennis, 
male [USNMENT00025946] (B) R. scolopaceus, male [USNMENT00023193]
........................................................................................................................... 498 
Figure 102. SEM image of antenna of Rhagio mystaceus........................................ 499 
Figure 103. Mouthparts in Rhagio mystaceus. (A) Hypopharynx and associated 
structures, showing clypeus and scape and pedicel of antenna 
[USNMENT00025900] (B) Labellum [USNMENT00025908]. Scale = 0.1 mm.
........................................................................................................................... 500 
Figure 104. Wing of species of Rhagio, dorsal view. (A) R. punctipennis, male 
[USNMENT00025946] (B) R. scolopaceus, male [USNMENT00023193] .... 501 
Figure 105. Epandrium in species of Rhagio, dorsal view. (A) R. mystaceus 
[USNMENT00025230] (B) R. punctipennis [USNMENT00025950] ............. 502 
Figure 106. Hypandrium in species of Rhagio, dorsal view. (A) R. plumbeus 
[USNMENT00025949] (B) R. punctipennis [USNMENT00025950] ............. 503 
Figure 107. Aedeagus (male genitalia) of Rhagio mystaceus [USNMENT00025230]
........................................................................................................................... 504 
Figure 108. Female genitalia of Rhagio palpalis [USNMENT00025879]. Dorsal 
view. (A) Partially dissected terminalia (B) Sternite 9 ..................................... 505 
Figure 109. Anterior end of sternite 9 and base of spermathecal ducts at genital 
chamber in species of Rhagio. (A) R. incisus [USNMENT00025873] (B) R. 
mystaceus [USNMENT00025244] ................................................................... 506 
Figure 110. Sternite 8 of Rhagio incisus [USNMENT00025873]............................ 507 
Figure 111. Cercus in species of Rhagio. (A) R. hirtus [USNMENT00023150] (B) R. 
incisus [USNMENT00025873] (C) R. palpalis [USNMENT00025879], missing 
left second cercus segment................................................................................ 508 
Figure 112. Lateral view of Schizella furcicornis [USNMENT00025863].............. 509 
Figure 113. Schizella woodleyi Kerr, new species. Head, anterior view.................. 510 
Figure 114. Schizella woodleyi Kerr, new species. Wing......................................... 510 
Figure 115. Schizella woodleyi Kerr, new species. Epandrium, dorsal view ........... 511 
Figure 116. Schizella woodleyi Kerr, new species. Epandrium, posterior view....... 511 
Figure 117. Schizella woodleyi Kerr, new species. Hypandrium, dorsal view......... 512 
Figure 118. Lateral view of head, Sierramyia chiapasensis..................................... 513 
Figure 119. Antenna of Sierramyia chiapasensis..................................................... 513 
Figure 120. Female genitalia of Sierramyia chiapensis [USNMENT00022656]. (A) 
Terminal segments, lateral view (B) Sternite 9................................................. 514 
Figure 121. Female sternite 8 of Sierramyia chiapensis [USNMENT00022656], 
dorsal view ........................................................................................................ 515 
Figure 122. Antenna of Spania nigra, male [USNMENT00025865]. (A) lateral view 
(B) anterodorsal view........................................................................................ 516 
Figure 123. Wing in Spania nigra, male, dorsal view [USNMENT00025865]....... 517 
Figure 124. Male genitalia of Spania nigra [USNMENT00025868]. Dorsal view. (A) 
Epandrium (B) Hypandrium ............................................................................. 518 
Figure 125. Female terminalia of Spania nigra [USNMENT00024389]................. 519 
Figure 126. Spermatheca (imploded) and spermathecal duct accessory gland of 
Spania nigra [USNMENT00024389]............................................................... 520 
 
xvii 
Figure 127. Antennal form in species of Spaniopsis. Lateral view. (A) S. clelandi, 
female [USNMENT00025409] (B) S. longicornis, female 
[USNMENT00022643]..................................................................................... 521 
Figure 128. Antennal form in species of Spaniopsis. Lateral view. (A) S. mackerrasi, 
male [USNMENT00025556] (B) S. marginipennis, female 
[USNMENT00025412 ? top specimen on pin] ................................................ 522 
Figure 129. Antennal form in species of Spaniopsis. Lateral view. (A) S. rieki, female 
[USNMENT00022653] (B) S. tabaniformis, female [USNMENT00025568]. 523 
Figure 130. Antenna of Spaniopsis vexans. Female, lateral view 
[USNMENT00025408]..................................................................................... 524 
Figure 131. Mouthparts of Spaniopsis clelandi [USNMENT25907]. (A) Labellum, 
ventral view (B) Hypopharynx, cibarial pump, and associated structures, lateral 
view (C) Palp, mandible, and lacinia; lateral view ........................................... 525 
Figure 132. Wing in species of Spaniopsis, ventral view. (A) S. clelandi 
[USNMENT00022649] (B) S. mackerrasi [USNMENT00025589] ................ 526 
Figure 133. Wing in species of Spaniopsis. (A) S. longicornis, ventral view 
[USNMENT00025293] (B) S. marginipennis, dorsal view 
[USNMENT00025252]..................................................................................... 527 
Figure 134. Wing in species of Spaniopsis, ventral view. (A) S. rieki 
[USNMENT00022655] (B) S. tabaniformis [USNMENT00025568].............. 528 
Figure 135. Wing in Spaniopsis vexans, dorsal view [USNMENT00025407] ........ 529 
Figure 136. Epandrium in species of Spaniopsis, ventral view. (A) S. clelandi 
[USNMENT00025396] (B) S. marginipennis [USNMENT00025249] ........... 530 
Figure 137. Hypandrium in species of Spaniopsis, dorsal view. (A) S. clelandi 
[USNMENT00025396] (B) S. marginipennis [USNMENT00025249] ........... 531 
Figure 138. Spermathecae in species of Spaniopsis. (A) S. clelandi 
[USNMENT00025398] (B) S. marginipennis [USNMENT00025251] ........... 532 
Figure 139. Female terminalia in species of Spaniopsis, dorsal view. (A) S. clelandi 
[USNMENT00025398] (B) S. marginipennis [USNMENT00025251] ........... 533 
Figure 140. Stylospania lancifera [USNMENT00025234, holotype]...................... 534 
Figure 141.  Antennal form in Symphoromyia crassicornis. (A) S. crassicornis, 
female [USNMENT00023210] (B) S. crassicornis, male [USNMENT00023208]
........................................................................................................................... 535 
Figure 142. Antennal form in Symphoromyia cruenta (A) female 
[USNMENT00025941] (B) male [USNMENT00025942] .............................. 536 
Figure 143. Antennal form in Symphoromyia hirta, lateral view (A) male 
[USNMENT00028585] (B) female [USNMENT00028622] ........................... 537 
Figure 144. Mouthparts of Symphoromyia sp. [USNMENT00025909]. (A) Labellum, 
oblique ventral view (B) Hypopharynx, cibarial pump, and associated structures, 
lateral view (C) Palp and lacinia, lateral view .................................................. 538 
Figure 145. Wing in species of Symphoromyia. (A) S. cruenta, female 
[USNMENT00025941] (B) S. crassicornis, female [USNMENT00023211].. 539 
Figure 146. Wing in species of Symphoromyia. (A) S. hirta, female 
[USNMENT00028605] (B) S. flavipalpis, female [USNMENT00025944] .... 540 
 
xviii 
Figure 147. Epandrium (male genitalia) in species of Symphoromyia. (A) S. 
crassicornis, dorsal view [USNMENT00028629] (B) S. hirta, ventral view 
[USNMENT00025791]..................................................................................... 541 
Figure 148. Hypandrium (male genitalia) in species of Symphoromyia, dorsal view. 
(A) S. crassicornis [USNMENT00028629] (B) S. hirta [USNMENT00025791]
........................................................................................................................... 542 
Figure 149. Female genitalia in Symphoromyia hirta [USNMENT00028587]........ 543 
Figure 150. Female genitalia in Symphoromyia plagens [USNMENT00025792]... 544 
Figure 151. Dissected terminalia in Symphoromyia species. (A) S. hirta 
[USNMENT00028587] (B) S. plagens [USNMENT00025792]...................... 545 
Figure 152. Spermathecae and spermathecal duct accessory glands of Symphoromyia 
plagens [USNMENT00025792] ....................................................................... 546 
Figure 153. Spermathecae and spermathecal duct accessory glands of Symphoromyia 
hirta [USNMENT00028587]............................................................................ 547 
Figure 154. Sternite 8 in species of Symphoromyia, dorsal view. (A) S. hirta 
[USNMENT00028587] (B) S. plagens [USNMENT00025792]...................... 548 
Figure 155. Mouthparts of Glutops rossi [USNMENT00025333]. (A) Labellum, 
lateral view (B) Hypopharynx, posterior view (C) Palps ................................. 549 
Figure 156. Epandrium in species of Glutops. Dorsal view. (A) G. punctatus 
[USNMENT00025327] (B)  G. rossi [USNMENT00025231] (missing tergite 
10) ..................................................................................................................... 550 
Figure 157. Hypandrium in species of Glutops. Dorsal view. (A) G. punctatus 
[USNMENT00025327] (B) G. rossi [USNMENT00025231].......................... 551 
Figure 158. Female genitalia of Glutops singularis [USNMENT00025338]. (A) 
Dissected terminalia (B) Spermathecae ............................................................ 552 
Figure 159. (A) Female sternites 7 and 8, showing long intersegmental region of 
Glutops singularis [USNMENT00025338] (B) Detail of sternite 8, lateral view 
(same specimen)................................................................................................ 553 
Figure 160. SEM image of antenna of Pelecorhynchus fusconiger.......................... 554 
Figure 161. Labellum in species of Pelecorhynchus. (A) P. fusconiger 
[USNMENT00025360] (B) P. personatus [USNMENT00025385] ................ 555 
Figure 162. Internal mouthparts in species of Pelecorhynchus, lateral view. (A) P. 
fusconiger [USNMENT00025360] (B) P. personatus [USNMENT00025385]
........................................................................................................................... 556 
Figure 163. Palps and associated structures in species of Pelecorhynchus. (A) P. 
fusconiger [USNMENT00025360] (B) P. personatus [USNMENT00025385]
........................................................................................................................... 557 
Figure 164. Hypandrium in species of Pelecorhynchus. (A) P. fusconiger, ventral 
view [USNMENT00025897] (B) P. personatus, dorsal view 
[USNMENT00025896]..................................................................................... 558 
Figure 165. Epandrium in Pelecorhynchus personatus [USNMENT00025896]. 
Dorsal view ....................................................................................................... 559 
Figure 166. Female terminalia in species of Pelecorhynchus elegans 
[USNMENT00025881], lateral view. (A) Partially dissected (B) Detail of cercus
........................................................................................................................... 560 
 
xix 
Figure 167. (A) Spermatheca of Pelecorhynchus fusconiger [USNMENT00025883] 
(B) Sternite 8, showing intersegmental membrane between sternites 7 and 8 in 
Pelecorhynchus personatus [USNMENT00025882] ....................................... 561 
Figure 168. Tergite 8 of female Pelecorhynchus, showing ducts. (A) P. elegans 
[USNMENT00025880], anterior view (B) P. personatus 
[USNMENT00025882], ventral view............................................................... 562 
Figure 169. Left lateral view of Pseudoerinna jonesi [USNMENT00025319]........ 563 
Figure 170. Internal structures of female genitalia of Pseudoerinna jonesi 
[USNMENT00025319]. (A) Sternite 9 and associated structures (B) 
Spermatheca...................................................................................................... 564 
Figure 171. External structures of female genitalia of Pseudoerinna jonesi 
[USNMENT00025319]. (A) Cerci (B) Sternite 8, with proceeding 
intersegmental membrane ................................................................................. 565 
Figure 172. Lampromyia sp. [USNMENT00025955]. (A) Epandrium, ventral view 
(B) Hypandrium, dorsal view ........................................................................... 566 
Figure 173. Male genitalia of Vermileo vermileo [USNMENT00025956]. (A) 
Epandrium, ventral view (B) Hypandrium, dorsal view................................... 567 
Figure 174. Structures of the female terminalia of Vermileo vermileo 
[USNMENT00025793]. (A) Partially dissected terminalia, showing highly 
autapomorphic common spermathecal duct (B) Spermatheca.......................... 568 
Figure 175. Thoracic structure of Chrysopilus quadratus, lateral view. (A) Entire 
thorax (B) Detail showing reduced proepimeron (D) Detail of postspiracle and 
halter. ................................................................................................................ 569 
Figure 176. Thoracic structure of Rhagio mystaceus, lateral view. (A) Entire thorax 
(B) Detail showing anepisternum and setose proepimeron (D) Detail of 
postspiracle and halter....................................................................................... 570 
Figure 177. Postspiracular scale of Pelecorhynchus fusconiger............................... 571 
Figure 178. SEM of haltere of Austroleptis multimaculata...................................... 572 
Figure 179. SEM of halter of Bolbomyia nana......................................................... 573 
Figure 180. SEM of haltere in Arthroceras leptis..................................................... 574 
Figure 181. SEM of halter of Arthroteles bombyliiformis........................................ 575 
Figure 182. SEM images of haltere in Desmomyia thereviformis............................ 576 
Figure 183. SEM images of halter in Ptiolina species.............................................. 577 
Figure 184. SEM images of halter of  Ptiolina edeta. .............................................. 578 
Figure 185. Anterior view of larval head Symphoromyia sp. Abbreviations: ant = 
antenna; lb = labrum; lbt = labral teeth; lc = lacinia; mb = mandibular brush; mh 
= mandibular hook; mx = maxilla; mp = maxillary palpus. ............................. 579 
Figure 186. SEM images of the larval head. (A) Chrysopilus sp. (B) Rhagio sp. (C, D) 
Symphoromyia sp. (E) Pelecorhynchus sp. (F) Glutops rossi........................... 580 
Figure 187. SEM images of the larval hind segment, posterior view (except in B, 
where oblique ventral view). (A) Symphoromyia sp. (B) Rhagio sp. (C, D) 
Pelecorhynchus sp. ........................................................................................... 581 
Figure 188. Habitus of Ptiolina sp. larva, lateral view. ............................................ 581 
Figure 189. Larva of Ptiolina sp. Anterior view, bright field illumination. ............. 582 
Figure 190. Lateral view of undissected Ptiolina sp. larva. Notice scalloping of first 
thoracic segment. .............................................................................................. 582 
 
xx 
Figure 191. Lateral view of Ptiolina sp. larval head. Notice surface texture in absence 
of mandibular brush. Bright field illumination. ................................................ 583 
Figure 192. Oblique dorsal view of dissected Ptiolina sp. larval head. Bright field 
illumination. ...................................................................................................... 583 
Figure 193. Adoral view of basal mandibular sclerite and mandibular hook of Rhagio 
sp., showing saw sclerite. Nomarski illumination. ........................................... 584 
Figure 194. Mandibular hook of Symphoromyia sp., showing adoral groove. 
Nomarski illumination. ..................................................................................... 584 
Figure 195. Larva of Dasyomma sp., lateral view, bright field illumination. Narrow, 
sclerotized palp evident..................................................................................... 585 
Figure 196. Larva of Dasyomma sp., lateral view, bright field illumination. Rod 
associated with mandibular brush and inner groove of mandibular hook evident.
........................................................................................................................... 585 
Figure 197. Larva of Tabanus sp., with head capsule extended from (pulled out of) 
first thoracic segment. Lateral view.................................................................. 586 
 
 
 1 
 
Chapter 1. General Introduction to the Genera of the Rhagionidae and Related 
Taxa. 
Overview 
The order Diptera comprises approximately 128,000 species in approximately 142 
families (Brown, 2001). As a group, flies represent one of the most prolific and 
important elements of our natural world (Skevington and Dange, 2002). Through 
their history, they have evolved a fantastic array of divergent forms that have allowed 
them to fill diverse ecological roles. As pollinators, scavengers, parasites, disease 
vectors, insect predators and a source of food to many other organisms, flies are an 
integral, essential part of virtually every ecosystem in the world (Vockeroth, 2002). 
Secondary to the origin of the halter and the structural flight mechanisms that define 
the Diptera as a monophyletic order, the innovations of larval mandible and adult 
antennal morphologies arose in the Late Triassic and spurred the biological radiation 
of flies that are now classed in the suborder Brachycera and represent most of the 
dipteran diversity in the world today (Rohdendorf, 1991; Krzeminski, 1992; 
Krzeminski, 1998; Yeates and Lambkin, 1998; Yeates and Wiegmann, 1999; 
Mostovski and Jarzembowski, 2000; Wiegmann et al., 2003).   
 
Very little of the origin of the Brachycera is known. The sister taxon is uncertain and 
the arrangement of the basal branches of its phylogenetic tree is mostly speculative 
(Hennig, 1968; Hennig, 1973; Rohdendorf, 1974; Wood and Borkent, 1989; 
Rohdendorf, 1991; Oosterbroek and Courtney, 1995; Amorim and Silva, 2002). My 
research focuses on the Rhagionidae (formerly known as the Leptidae), an extant 
 
 2 
 
family of flies considered to contain some of the most primitive living members of 
the Brachycera and for this reason, receives much attention by those interested in 
higher level Dipteran systematics (e.g., Hennig, 1967; Hennig, 1973; Nagatomi, 
1977; Kovalev, 1982; Woodley, 1989; Griffiths, 1994; Friedrich and Tautz, 1997; 
Grimaldi and Cumming, 1999; Wiegmann et al., 2000; Yeates, 2002). The family 
lacks any synapomorphies to join its members, however, and the phylogenetic 
relationships of the genera have not been examined in a systematic phylogenetic 
framework (Felsenstein, 2004). Understanding the phylogeny of the genera of 
Rhagionidae will provide stability to the classification of this family and may provide 
insights into the early evolution of Brachycera.   
 
 
 
 3 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
Figure 1. Brachycerous Diptera tree of life (adapted from Yeates and Wiegmann, 
1999). Rhagionidae are located within Tabanomorpha, at the base of the tree. 
 
 4 
 
 
 
 5 
 
Recent Classification 
Rhagionidae belong to the infraorder Tabanomorpha that currently includes from 
three to seven other families. The limits of Tabanomorpha are not entirely resolved, 
but all agree that in addition to Rhagionidae, the Tabanomorpha contain the following 
taxa (as family or subfamily lineages): Athericidae, Pelecorhynchidae, Spaniidae, and 
Tabanidae. Bolbomyia Loew may represent an additional family (Sinclair et al., 
1993), although this has not been formally proposed. These taxa are what I will call 
the ?core tabanomorph taxa.? Austroleptis may also be included here, but for now will 
not, since there is speculation that it may belong in the Xylophagomorpha (Sinclair et 
al., 1993). Most of the genera that make up the core tabanomorphs, aside from 
Tabanidae, were originally placed in the Rhagionidae. Characters that have been used 
to unite the Tabanomorpha are the apomorphic presence of a brush on the larval 
mandible, larval head retractile, adult with convex bulbous clypeus, and 
ventrolaterally expanded first segment of the cercus in adult females (Woodley, 1989; 
Sinclair, 1992; Wiegmann et al., 2000). It has been suggested that the Acroceridae 
and Nemestrinidae (Nemestrinoidea) may also belong within the Tabanomorpha 
(Nagatomi, 1992; Griffiths, 1994; Stuckenberg, 2001), however no systematic study 
has supported this notion. Molecular evidence supports Tabanomorpha exclusive of 
the Nemestrinoidea (Wiegmann et al., 2000; Wiegmann et al., 2003), as does a recent 
synthetic morphological study by Yeates (2002). The Xylophagomorpha 
(Xylophagidae) have been shown to be sister to the Tabanomorpha, with weak 
support (Wiegmann et al., 2000), and some authors prefer to combine these taxa at a 
higher level, maintaining the name Tabanomorpha in a more inclusive sense 
 
 6 
 
(Griffiths, 1994; Stuckenberg, 2001). Since so many of the non-tabanid tabanomorph 
taxa were once placed in the Rhagionidae, and in many cases, their taxonomy 
continues to affect the placement of rhagionid genera, a brief note on the taxonomy of 
each of these groups is presented.  
 
In 1973, Stuckenberg created Athericidae, showing that Atherix and its allies were 
more closely related to Tabanidae than to the remaining members of Rhagionidae, 
where these genera were originally placed.  The sister group relationship between the 
genera of Athericidae and Tabanidae is now clear, based on strong morphological and 
molecular evidence (Stuckenberg, 1973; Woodley, 1989; Sinclair, 1992; Sinclair et 
al., 1993; Wiegmann et al., 2000; Stuckenberg, 2001; Wiegmann et al., 2003). 
 
Hardy originally described Austroleptis as a leptid (=Rhagionidae) (Hardy, 1920a). 
However, in 1953, it was placed in the Xylophagidae (Steyskal, 1953) and more 
recently, others have speculated that it belongs within the Xylophagomorpha on 
account of its larvae having been reared from wood, as is the case for most 
Xylophagidae (Colless and McAlpine, 1991; Sinclair et al., 1993). Nagatomi (1982a; 
1991) considered Austroleptis a basal taxon of the Rhagionidae and erected the 
subfamily Austroleptinae to account for it, although he did not rule out the idea that it 
may represent a lineage outside of Rhagionidae. In 2001, Stuckenberg elevated 
Austroleptinae to family level, however could not provide evidence to support, nor 
did he provide speculation, as to the sister group of this lineage or its relationship to 
other lower Brachyceran families. 
 
 7 
 
 
Bolbomyia Loew was originally described as a xylophagid, from specimens preserved 
in amber (Loew, 1850). Chillcott (1961) located this genus within the Rhagionidae, 
but James (1965) preferred its original placement in the Xylophagidae. In 1982, 
Bolbomyia was treated by Nagatomi (1982a) as a rhagionid however its placement 
has since been questioned. Sinclair et al. (1993) concluded that Bolbomyia may be 
more closely related to Athericidae and Tabanidae than to the rest of the rhagionid 
genera, based on the shared presence of the aedeagal tine in the male genitalia. 
Stuckenberg (2001) rejected this argument, however, and retained Bolbomyia within 
the Rhagionidae, placing it within its own subfamily, the Bolbomyiinae. 
 
Macquart established the genus Pelecorhynchus in 1850, for an Australian species, P. 
maculipennis Macquart. This species is a synonym of P. personatus (Walker) that 
was originally described as a member of the genus Silvius Meigen (Tabanidae). 
Pelecorhynchus species were placed in Tabanidae, within their own subfamily 
created by Enderlein in 1922, the Pelecorhynchinae. Pelecorhynchus remained in the 
Tabanidae until 1942 when Mackerras and Fuller created the family 
Pelecorhynchidae. Steyskal (1953) considered Pelecorhynchus closely related to 
Coenomyia and placed these genera together in the Coenomyiidae (now 
Xylophagidae), along with Arthroteles (Rhagionidae) and Stratioleptis 
(=Odontosabula, Xylophagidae). Although this may appear as a major departure from 
previous classification, the coenomyiid lineage at that time was still considered a 
close relative of the Tabanidae (Steyskal, 1953; Hardy, 1955). In 1970, Teskey 
 
 8 
 
removed Glutops from the Xylophagidae (Coenomyiidae sensu Steyskal) to the 
Pelecorhynchidae (sensu Mackerras & Fuller, 1942) on account of similarities of 
larval morphology between Pelecorhynchus and Glutops. Krivosheina (1971), 
however, erected the family Glutopidae to accommodate Glutops. Kovalev (1981) 
synonymized Glutopidae with Rhagionidae, however, and Nagatomi (1982) 
recognized the Glutopinae, placing Pseudoerrina with Glutops, as a subfamily of the 
Rhagionidae. Pelecorhynchidae remained a family represented by the single genus, 
Pelecorhynchus, according to Nagatomi (1982). Stuckenberg, however, recognized 
all three genera as members of the Pelecorhynchinae, a subfamily of the Rhagionidae 
(2001). Molecular evidence supports Teskey?s assertion that Glutops and 
Pelecorhynchus form a monophyletic group (Wiegmann et al., 2000; Wiegmann et 
al., 2003). Furthermore, the molecular evidence suggests that the Pelecorhynchidae 
are sister to the Athericidae and Tabanidae lineage (Wiegmann et al., 2000; 
Wiegmann et al., 2003). Larval morphology also putatively supports this placement 
(Woodley, 1989; Sinclair, 1992; although see Stuckenberg, 2001). There have not 
been any adult morphological synapomorphies proposed to support the monophyly of 
Glutops, Pelecorhynchus, and Pseudoerrina. The larva for Pseudoerrina is not 
known. 
 
The origins of the Spaniidae begin with Frey, who in 1954 established the subfamily 
Spaniinae for Bolbomyia Loew (as Cechenia Frey), Spania Meigen, Ptiolina 
Zetterstedt and Omphalophora Becker on the basis of having a bare laterotergite and 
short stylate antennae (Frey, 1954). At the time, these were distinguished as the 
 
 9 
 
smallest rhagionids known. Nagatomi (1982) removed Bolbomyia to the Rhagioninae 
and considered the structure of the female terminalia as the most important source of 
characters to define the subfamily Spaniinae. Nagatomi asserted that a wide amount 
of separation between female first cerci, the lack of a ventral process on the basal 
cercus of the female, and tergite 10 short or absent were distinguishing characters for 
the Spaniinae. On the basis of these features and an antenna with a tapering, stylate 
first flagellomere, Nagatomi added Spaniopsis to the subfamily. He also added 
Litoleptis, but since he did not have a Litoleptis female available for examination, this 
placement was presumably based on a single antennal character. There is some 
degree of homoplasy among all of these characters, however, and Stuckenberg (2001) 
specifically questioned the usefulness of the female postabdomen for phylogenetic 
inquiry. Nonetheless, Stuckenberg (2001) raised Spaniinae to family rank without any 
changes from the arrangement proposed by Nagatomi (1982). No unambiguous 
synapomorphies for the group have been recognized.  
 
The taxa that comprise the family Vermileonidae were also originally placed in the 
Rhagionidae, and later recognized as a separate subfamily (Williston, 1886; Hennig, 
1967; Hennig, 1973). In 1977, Nagatomi (1977) raised this group to family rank and 
speculated that the family was either basal to the tabanomorph families 
(Pelecorhynchidae, Rhagionidae, Athericidae, and Tabanidae) or basal to the entire 
lower Brachycera. Nagatomi preferred the latter arrangement, judging the 
vermileonid lineage to be very old, however he did not use explicit methods to make 
this determination and conceded that the phylogenetic placement of this family 
 
 10 
 
grouping remained unclear (Nagatomi, 1977). Teskey (1981c) regarded the family as 
being related to the Asiloidea, based on characters of the male terminalia and larval 
mouthparts. This placement was rejected by Woodley (1989), however, who 
tentatively placed the family within Tabanomorpha, incertae sedis based on the 
retractile head of the vermileonid larva. Griffiths (1994) proposed a new infraorder, 
Vermileonomorpha, to account for the family. Molecular evidence presented by 
Wiegmann et al. (2000) locates the Vermileonidae as sister to the genera of the 
Rhagionidae, within the Tabanomorpha clade, although this result did not definitively 
preclude Vermileonidae being located outside of the Tabanomorpha (Wiegmann et 
al., 2000; Stuckenberg, 2001). 
  
For simplicity, I recognize the Tabanomorpha classification, including the genera of 
families contained therein, of Woodley (1989) as current. Woodley?s conservative 
assessment of the Rhagionidae includes Austroleptis, Bolbomyia, and the Spaniidae, 
exclusive of the Pelecorhynchidae. Tabanomorpha is defined as containing the 
Athericidae, Pelecorhynchidae, Rhagionidae, Tabanidae, and Vermileonidae. 
Xylophagidae is removed to the Xylophagomorpha. I add here, however, a few 
exceptions to Woodley?s classification scheme (1989). Sierramyia gen. nov. is used 
as a replacement name for Neorhagio Lindner. Also, Archicera Szil?dy is treated as a 
junior synonym of Spania Meigen. A more detailed history of the development of the 
family concept of Rhagionidae follows. 
 
 
 11 
 
History of the Concept of Rhagionidae 
The taxonomic history of Rhagionidae involves nearly all of the great workers in 
Diptera, including Linnaeus, Fabricius, Latreille, Macquart, Westwood, Becker, and 
Hennig, among others (Linnaeus, 1758; Fabricius, 1775; Latreille, 1802; Latreille, 
1804; Fabricius, 1805; Latreille, 1809; Macquart, 1834; Macquart, 1840; Westwood, 
1840a; Westwood, 1840b; Macquart, 1850; Macquart, 1855; Westwood, 1876; 
Becker, 1900b; Becker, 1900c; Becker, 1900a; Becker, 1921; Hennig, 1952; Hennig, 
1955; Hennig, 1967; Hennig, 1968; Hennig, 1972; Hennig, 1973). In the following 
account, I report the development of the Rhagionidae family concept, beginning with 
the original species placed in the group. In parentheses, I indicate the currently valid 
binomial combination for each species mentioned, if different from its original form. 
Once the basis for the family is established, I track family boundaries at the genus 
level. Junior synonyms are followed by their currently valid senior synonyms, in 
parentheses.  
 
The genus Rhagio was established by Fabricius in 1775 for four species originally 
placed in Musca by Linnaeus in 1758: M. scolopaceus Linnaeus (=Rhagio 
scolopaceus (Linnaeus)), R. vermileo Fabricius (=R. tringarius (Linnaeus)), R. 
tringarius Fabricius (=nomina dubia), and Rhagio diadema Fabricius (=nomina 
dubia). Although Linnaeus originally used these same species epithets in combination 
with Musca, Rhagio tringarius and R. diadema of Fabricius? work were based on 
misidentifications of the original material and did not pertain to the original Linnaean 
 
 12 
 
concepts. Rhagio tringarius and R. diadema, therefore, were retained as original 
Fabricius names.  
 
At this time, antenna and mouthparts were the primary characters relied upon for the 
classification of flies. Fabricius? diagnosis of Rhagio was an example of this. It read 
simply, ?the mouthparts of the group are without a sheath [= mandibles]. Two palpi, 
setose, extend near the base of the proboscis. The antennae are cylindrical, setose.? 
Although these features are common to species of Rhagio, they scarcely preclude 
other, non-related groups and the vague premise on which this genus was established 
became immediately problematic for classification above the genus level. In 1802, 
Latreille erected the family ?Rhagionides?, containing the original Rhagio species and 
R. atratus Fabricius (= nomina dubia), plus Dolichopus ungulatus Linnaeus 
(Dolichopodidae), and Dolichopus virens Scopoli (=Liancalus virens (Scopoli), 
Dolichopodidae) (Latreille, 1802). [R. diadema Fabricius was not mentioned in this 
work however it is presumed that the species remained in Rhagio at this time.] In 
1803, Meigen created Atherix for what he believed was Rhagio diadema Fabricius 
(but was actually originally described as Sylvicola melancholia Harris, according to 
Coquillett (1910)). Atherix remained within the family. In subsequent work, Latreille 
placed ?Rhagionides? within a higher context, in the Proboscidea, the largest of four 
Dipteran higher-level groups, which contain 13 other families including muscids, 
syrphids, and tipulids. Latreille?s definition of the ?Rhagionides? at this time included 
Rhagio Fabricius, Atherix Meigen (Athericidae), Pachystomus Latreille (= 
Xylophagus Meigen), and Ortochile Latreille (Dolichopodidae) (Latreille, 1809). 
 
 13 
 
Latreille placed Dolichopus in its own family, yet retained another dolichopodid, 
Ortochile Latreille, within the ?Rhagionides?. 
 
Samouelle was the first to use the family group name correctly, as Rhagionidae, in 
1819. However, since Fabricius emended the genus name of Rhagio to Leptis in 1805, 
most authors used Leptis for the genus name, and consequently, Leptidae as the 
family group name. Fabricius changed the name Rhagio because he argued that it was 
too similar to Rhagium, a cerambycid beetle. Although Latreille was aware of 
Fabricius? change of Rhagio to Leptis, he was one of the few authors to continue to 
use the name Rhagio (Latreille, 1809: 287). 
 
In 1820, Meigen documented the expansion of the family ?Leptides?, which had 
grown to 35 species (Meigen, 1820) in three genera: Leptis, Atherix, and Clinocera 
(Empididae). The following year, Wiedemann maintained Clinocera in the family, 
however not without some reservation (Wiedemann, 1821). As a footnote next to 
Clinocera, Wiedemann wrote, ?in the meantime I maintain this genus at this position, 
perhaps it will prevail after productive investigations.?  
 
In 1826, Macquart monographed the family and moved Clinocera Meigen to 
Empididae (Macquart, 1826). This improved the cohesiveness of the rhagionid 
concept. He also added a new genus, Chrysopilus, to distinguish several species 
formerly in Rhagio. The family concept now included Rhagio, Atherix, and 
Chrysopilus. Macquart?s follow-up work in the group came in 1834, where he gave a 
 
 14 
 
brief introduction and description of each genus, for which he attempted to establish a 
classificatory framework. Leptidae were placed as a tribe in the family Brachystomes 
(subdivision Tetrachoeta), which contained 10 other tribes: ?Mydasiens?, ?Asiliques?, 
?Hybotides? (=Empididae) Empides?, ?Vesiculeux? (=Acroceridae), ?Nemestrinides?, 
?Xylotomes? (=Therevidae), ?Bombyliers?, ?Syrphies?, and ?Dolichopodes?. He placed 
leptids between the therevids and acrocerids. In addition to Leptis, Chrysopilus, and 
Atherix, he added the genus Spania Meigen (originally described in 1830 without a 
family designation), Clinocera, and Vermileo Macquart (Vermileonidae). He created 
the genus Vermileo for Musca vermileo Linnaeus (not to be confused with Rhagio 
vermileo Fabricius). In 1840, Macquart expanded the concept of the group further by 
adding genera from outside of Europe. Lampromyia Meigen (Vermileonidae), 
originally placed in Bombyliidae, but on account of its similarity to Vermileo, was 
moved to Leptidae. Dasyomma (Athericidae), a new genus from Chile, was also 
added (Macquart, 1840). 
 
That same year, Westwood placed Leptidae in Division ?Brachocera? (defined by 
having antennae short, not having more than three distinct joints and palpi  1- or 2-
jointed) in the stirps Tanystoma (defined by having antenna ends with arista and pupa 
incomplete) which contains families Tabanidae, Bombyliidae, Anthracidae 
(=Bombyliidae), Acroceridae, Empididae, Tachydromiidae (=Empididae), Hybotidae 
(=Empididae), Asilidae, Mydasidae, Therevidae, Leptidae, Dolichopidae, and 
Scenopinidae (Westwood, 1840a).  The Stratiomyidae, Beridae (=Stratiomyidae), and 
Coenomyiidae, were placed in another stirps, Notocantha, on account of having 
 
 15 
 
articulated flagellum. Clinocera was retained in Leptidae as a ?bridge? to 
Dolichopodidae, due to its wing venation (Westwood, 1840a). The genus Ptiolina 
was named for a new species and was added to the family in 1842 by Zetterstedt, who 
worked on the Diptera of Scandinavia (Zetterstedt, 1842).  
 
Throughout the 1850?s Francis Walker consistently changed the generic limits of the 
Leptidae. In 1851, working on the Diptera of Great Britain, he placed the Leptids in 
Brachycera with all other families except Hypocera (=Phoridae), and Eproboscidea 
(=Hippoboscidae, Nycteribiidae), and the nematocerans (Walker, 1851). The regional 
leptid fauna included five genera: Leptis, Chrysopila (=Chrysopilus Macquart), 
Atherix, Ptiolina, and Spania. In 1856, Walker added Syneches Walker (currently 
Empididae) (Walker, 1856). However, in this publication, he confused Leptis with 
Chrysopilus so that new Chrysopilus species were described as members of Leptis. 
Leptis and Rhagio, consequently, were treated as separate genera and Chrysopilus 
was not mentioned. Three years later, he added another new genus, Suragina Walker 
(Athericidae) (Walker, 1859).  
 
In 1856, J.M.F. Bigot introduced a vastly broader concept of the Leptidae than what 
previous workers had determined. According to Bigot, the family comprised 12 
genera, nine of which were newly located. In addition to Rhagio, Ptiolina, and 
Psammorycter (=Vermileo), he added Anthalia Zetterstedt (Empididae), Baryphora 
Loew (Therevidae), Chauna Loew (Stratiomyidae), Exeretonevra Macquart 
(Xylophagidae), Lampromyia Meigen (Vermileonidae), Leptipalpus Rondani 
 
 16 
 
(=Chrysopilus), Microcera Zetterstedt (=Heleodromia, Empididae), Pelechoidecera 
Bigot (=Atherix), and  Wiedemannia Zetterstedt (Empididae) (Bigot, 1856). Spania 
was actually also included, as a junior synonym of Ptiolina. 
 
The first attempt at cataloging the Diptera of North America was published during 
this era (Osten Sacken, 1858). In his treatment, Osten Sacken did not recognize the 
work of Bigot. According to his interpretation, the Leptidae remained a relatively 
small family in the Nearctic Region, confined to 26 species in three genera: Leptis, 
Syneches, and Atherix. The difference between the number of genera placed in the 
family by Osten Sacken and the number of genera placed in the family by Bigot 
(1856) indicated more than simply a difference of faunal diversity in the two regions. 
The family concept was interpreted in different ways in different parts of the world. A 
transatlantic rift in the recognition of family delimitations marked this era, 
exacerbated by limited-scope regional treatments.  
 
In 1862, Loew (Loew, 1862) published monographs of Diptera of North America 
where he outlined a different concept of the Rhagionidae. His interpretation was 
similar to that of Schiner (1862), including Leptis, Atherix, Vermileo, Chrysopilus, 
and Spania (sensu Walker, including Ptiolina). In addition to these genera, he added 
Dasyomma, Nodutis Meigen (=Atherix Meigen, Athericidae), and Triptotricha Loew 
(= Dialysis Walker, Xylophagidae). As Bigot (1856), Loew made a point to exclude 
the genus Syneches from Leptidae, citing this placement as ?one of the many errors 
which we meet with in the writings of Mr. Walker.?  
 
 17 
 
  
In 1867, Frauenfeld recognized that the interpretation of the genus name Ptiolina 
Zetterstedt had been inconsistent among researchers (Frauenfeld, 1867). The 
interpretation of the genus by Schiner (1862), Walker (1851), and Haliday (1833), 
was significantly different from that originally set forth by Zetterstedt in his Insecta 
Lapponica (Frauenfeld, 1867). Frauenfeld corrected the confusion by giving the name 
Symphoromyia to the natural group of species that mistakenly were placed together 
under the name Ptiolina by Schiner and others. 
 
In a list of North American species of Diptera, Osten Sacken added Pheneus Walker 
(=Vermileo Macquart, Vermileonidae) to Leptidae, which was reduced from eight 
(sensu Loew, 1862) to six genera in North America (Osten Sacken, 1858; Osten 
Sacken, 1874a; Osten Sacken, 1874b). The misinterpretation of the genus Ptiolina 
Zetterstedt continued, however, as the name was listed as a senior synonym of Spania 
Meigen (Osten Sacken, 1874b). 
 
In 1878, Burgess described a new genus, Glutops. In discussing the systematic 
position of this genus, he argued that if Arthropeas was admitted into the Leptidae, 
then Glutops would naturally follow (Burgess, 1878).  Arthropeas was placed 
provisionally among the Xylophagidae based on its antennae. Burgess noted however, 
that Osten Sacken was inclined to subordinate the structure of the antennae in 
Arthropeas to its general habitus, which was undoubtedly that of a leptid. Indeed, 
many genera would follow under this interpretation because it left the family without 
 
 18 
 
any defining character. This marked the beginning of treating Rhagionidae as the 
explicit, universal catchall for incertae sedis genera of the lower Brachycera (Fig. 2). 
 
Accumulation of Genus Names in Rhagionidae
0
20
40
60
80
100
120
1775 1800 1825 1850 1875 1900 1925 1950 1975 2000
 
Figure 2. Accumulation of Genus Names in Rhagionidae. The period between 1878 to 
1934 marked an expansive period for the concept of Rhagionidae, as the 
boundaries of this family were exceptionally porous.  The x-axis indicates year 
and the y-axis indicates the accumulation of genus names placed within 
Rhagionidae. 
 
Twenty years after publishing his catalogue of North American Diptera, Osten 
Sacken came out with a second edition, revised and updated (Osten Sacken, 1878). 
Although Osten Sacken attempted to check specimens and recognize synonymies, his 
taxonomic judgment was rather skewed by personal biases. For instance, he rejected 
priority of names given by Walker of the British Museum because of a lack of trust in 
 
 19 
 
this researcher?s work. He preferred to give priority of such names to colleague and 
co-author Loew. He took the opportunity to explain his aversion to Walker in a 
lengthy criticism: ?Mr. Walker?s identifications of the species of former authors are 
often, I may say in most cases, incorrect... These facts are given as a warning for 
entomologists not to trouble themselves too much about the interpretation of Mr. 
Walker?s descriptions, because in most cases, they will find themselves misled by the 
very data furnished by him.?  Furthermore, Osten Sacken did not utilize the emerging 
implications of evolutionary theory on taxonomy and classification. Echoing 
perceptions of pre-Darwinian authorities, he expressed the desire to place families in 
?a satisfactory linear arrangement.?  
 
Osten Sacken divided the family into sections Psammorycterina Loew and Leptina 
Osten Sacken (Osten Sacken, 1878).  Section Psammorycterina included  two genera, 
Triptotricha (=Dialysis, Xylophagidae) and Pheneus (=Vermileo, Vermileonidae). 
Section Leptina included six genera, Chrysopilus, Leptis, Ptiolina, Spania, Atherix, 
and Glutops. He did not follow Burgess? suggestion to place Arthropeas in Leptidae, 
instead locating this genus within Coenomyiidae. In subsequent work, Osten Sacken 
made Leptynoma Westwood a junior synonym of Lampromyia Macquart, and in 
agreement with Macquart and Schiner, placed the genus in Leptidae (Osten Sacken, 
1883).  
 
During this period, Williston noted that the addition of Triptotricha (=Dialysis, 
Xylophagidae) and Glutops to Leptidae provided ?exceptions to the distinctive 
 
 20 
 
characters of the family, and will make the limits very hard to define.? (Williston, 
1886). The work of Bigot in 1887 immediately shows this to be true (Bigot, 1887). 
Similar to his 1856 treatise, Bigot chose to redefine the family in the broadest terms. 
The limits of Rhagionidae were unclear enough to allow the inclusion of an ever 
greater diversity of taxa from other families in Lower Brachycera. In Bigot?s 
interpretation, the family comprised 23 genera, at least 12 of which were newly 
placed into the family (Bigot, 1887). Most of these taxa were placed in the 
Empididae. These new additions included Apolysis Loew (Bombyliidae), 
Bergenstammyia Mik (=Clinocera Meigen, Empididae), Chamadaepsia Mik 
(=Wiedemannia Zetterstedt, Empididae), Dialysis Walker, Eucelidia Mik 
(=Wiedemannia Zetterstedt, Empididae), Eurytion Jaennicke (=Ptiolina Zetterstedt), 
Heleodromyia Haliday (Empididae), Kowarzia Mik (=Clinocera Meigen, 
Empididae), Macellopalpus Bigot (=Chrysopilus Macquart), Roederia Mik 
(=Wiedemannia Zetterstedt, Empididae), Ruppiella Wiedemann (Therevidae), and 
Philolutra Mik (=Wiedemannia Zetterstedt, Empididae). He does not include generic 
names indiscriminately, however. He excluded Spania Meigen and Ptiolina 
Zetterstedt, considering them incertae sedis because of the poor description of these 
genera. He also excluded Pheneus Walker, Nodutis Meigen, Exeretonevra Macquart, 
and Pelechoidocera Bigot. 
 
At this point discussion arose as to whether include Hilarimorpha Schiner in the 
Leptidae, a genus that very much resembles a genus now placed in the family, 
Litoleptis Chillcott. Schiner originally described Hilarimorpha as an empidid, but 
 
 21 
 
later changed his opinion and claimed it was a leptid (Schiner, 1860).  Primarily 
basing their opinions on larval characters, Mik and Brauer came out against this 
revised opinion of Schiner (Brauer, 1880; Mik, 1881). Williston also argued that 
Hilarimorpha did not belong in Leptidae (Williston, 1885; Williston, 1888). This led 
to a telling insight into the family concept, provided by Osten Sacken (1890) in 
arguing the Hilarimorpha was a leptid, ?Prof. Mik says that Hilarimorpha has only 
four posterior cells, and therefore cannot be a Leptid, which ought to have five. 
[Agnotomyia is] a Leptid with the exceptional number of four posterior cells. Prof. 
Mik further says that Hilarimorpha has no pulvilliform empodium. The genus 
Lampromyia affords an instance that a pulvilliform empodium may not be developed 
in a Leptid. Again he says that the posterior branch of the fork of the third vein in the 
Leptidae is always behind the apex of the wing and not before it, as in Hilarimorpha. 
But Spania is a Leptid, and yet that vein ends before the apex of its wing.?  The 
family had grown to the point where there were no longer any defining characters 
which held the group together. 
 
This led Williston, in 1896, to subordinate Xylophagidae and Coenomyiidae to 
subfamilies of the Leptidae on account that the ?sole character which can be used to 
distinguish the families-the structure of the third antennal joint- divides the group 
unnaturally, throwing with the Xylophagidae forms whose affinities are greatest with 
the Leptidae, notwithstanding the antennal character? (Williston, 1896). Thus, the 
leptid fauna was listed in his Manual of North American Diptera as having 15 genera 
(increased from six genera in the previous treatment) divided into three subfamilies: 
 
 22 
 
subfamily Xylophaginae, containing Arthropeas Loew, Coenomyia Latreille, 
Rachicerus Walker, and Xylophagus Meigen; subfamily Arthroceratinae, containing 
Subulomyia Williston (for Subula Meigen, occupied name (=Xylomyia Bezzi)), 
Glutops Burgess, and Arthroceras Williston; and subfamily Leptinae, containing 
Dialysis Walker, Triptotricha Loew (= Dialysis Walker), Leptis, Chrysopila, Spania, 
Ptiolina., Atherix, and Symphoromyia. He placed Hilarimorpha Schiner in the 
Empididae (Williston, 1896). 
 
Becker, along with other European dipterists, rejected the notion that Xylophagidae 
and Coenomyiidae were subfamilies in the Leptidae (Becker, 1900a). In his 1900 
treatment, however, he included Hilarimorpha Schiner and also Arthropeas Loew, 
with some reservation (Becker, 1900a).  
 
Bezzi treated the Palearctic Brachycera in 1903, making significant modifications of 
the ?Leptididae? (Bezzi, 1903).  In a complicated new interpretation of the group, he 
retained the synonymous names Rhagio Fabricius and Leptis Fabricius by making 
Vermileo Macquart a junior synonym of Rhagio. The newly defined Rhagio concept 
was then placed with Lampromyia Macquart in subfamily Rhagioninae. Leptis 
Fabricius, meanwhile was placed with Atherix Meigen in another subfamily, 
Leptidinae. Another subfamily was erected, the Chrysopilinae, to contain 
Omphalophora Becker (=Ptiolina), Chrysopilus, Symphoromyia, Ptiolina, Spania, 
and Hilarimorpha. Subsequent authors largely ignored these changes. 
 
 
 23 
 
By 1905, the leptid fauna in the Catalog of North American Diptera contained 19 
genera (Aldrich, 1905). Bolbomyia Loew, Chiromyza Wiedemann (Stratiomyidae), 
Coenomyia Latreille (Coenomyiidae), Mythicomyia Coquillett (Bombyliidae), 
Rachicerus Walker (Xylophagidae), and Xylomyia Rondani (Xylomyiidae) were new 
additions to the North American list. Pheneus Walker (=Vermileo) was reinstated and 
Ptiolina Zetterstedt was treated as a junior synonym of Spania Meigen. Above the 
family level, Aldrich classified leptids, inclusive of Xylophagidae and Coenomyiidae, 
with Stratiomyidae, Tabanidae, and Acanthomeridae (= Pantophthalmidae) in the 
Orthorrhaphous Brachycera. Having no defining features, the Leptidae became the 
well established home to basal brachycerans that were not obviously stratiomyids, 
tabanids, or pantophthalmids. 
 
A year later, Lameere (1906) recommended that Nemestrinidae be subordinated to 
subfamily rank in Leptidae. This added the following nemestrinids to the Leptidae: 
Hirmoneura Meigen, Rhynchocephalus Fischer (= Nemestrinus Latreille), Fallenia 
Meigen, and Nemestrinus Latreille. Verrall, in reference to this work, would later say 
?throughout the paper [Lameere] exhibits such an absolute want of personal study of 
the Diptera and of their literature as to render his writings worse than useless for 
students? (Verrall, 1909). No subsequent publications followed the proposal of 
subordinating Nemestrinidae. 
 
In an updated version of Manual of North American Diptera, Williston reviewed the 
higher-level classification schemes for Diptera up to that point  ?in order that the tyro 
 
 24 
 
in their study may not reach the erroneous conclusion that any system is 
authoritative? (Williston, 1908). He argued that a modification of the classification 
system must reflect a modern evolutionary framework. This was not easy, however, 
and he conceded that  ?nearly all these lines of evolution are polyphyletic, resulting in 
numerous cases of parallel resemblances which must be taken into account in any 
attempt at true classification. In other words, the evolution of characters in the 
different lines of descent does not proceed pari passu, and opinions will always differ 
as to the different values to be assigned to the specialized characters.? At this time, 
Williston continued to support a broad definition of the family. His definition of the 
family remained virtually unaltered from the 1896 treatment (Williston, 1908).  
 
The catalog of world Diptera was also published in 1908, authored by Calman 
Kert?sz (Kert?sz, 1908). He treated the family as Rhagionidae, rather Leptidae, 
diligently noting that Leptis was a junior synonym of Rhagio because the emendation 
of Fabricius was unjustified. His treatment contained 20 genera in 274 species, 
exclusive of the xylophagids and coenomyiids. The following genera were included; 
Arthroceras, Atherix, Chrysopilus, Dasyomma (Athericidae), Dialysis, Glutops 
(Pelecorhynchidae), Hilarimorpha (Hilarimorphidae), Lampromyia (Vermileonidae), 
Mythicomyia (Bombyliidae), Pheneus (=Vermileo, Vermileonidae), Ptiolina, Rhagio, 
Spania, Suragina (Athericidae), Symphoromyia, Triptotricha (=Dialysis, 
Xylophagidae), and Vermileo.  
 
 
 25 
 
Verrall adopted an intermediate view in regards to the status of Leptidae, accepting 
the opinions of Kert?sz, yet broadening the concept to include the xylophagids and 
coenomyiids and a new genus, Atrichops, for Atherix crassicornis Meigen 
(Athericidae) (Verrall, 1909). Above the family level, Verrall proposed a 
classification scheme that placed the Stratiomyidae, Acanthomeridae, Leptidae, 
Tabanidae, Nemestrinidae, and Cyrtidae together in the Eremochaeta.  ?My reasons 
for this are the evident gradations through Beris and Xylomyia to Xylophagus and 
Coenomyia... The Acanthomeridae also are obviously intermediate between the 
Stratiomyidae and Leptidae and may lead on to the Tabanidae. There may be 
stepping-stones, though I write with little confidence, from the Tabanidae through the 
long proboscis of the Pangoninae to the Nemestrinidae, and then the superfamily 
would end with the Cyrtidae, from which ?Natura non facit saltum? to such 
tromopterous Bombyliidae as the bare humpbacked Glabellula, Platypygus, etc.?  His 
classification suffered by adhering to the notions that evolution proceeds in a linear 
fashion, discounting the tree-like division of relatedness between natural groups.  
 
During the first part of the twentieth century, most of the contributions to the 
development of Rhagionidae (for the most part, still called Leptidae) came from 
collections made outside of Europe and North America. Brunetti, working in India, 
created the new genus Desmomyia and placed it in the subfamily Arthroceratinae on 
account of its similarity to Arthroceras (Brunetti, 1912), although this was probably 
based on a misconception of Arthroceras, since Desmomyia is quite different from 
Arthroceras, most notably in antennal form. Desmomyia may easily be confused with 
 
 26 
 
Rhagio, which it highly resembles. While working on the Philippine Diptera, Bezzi 
described the new genus Schizella (Bezzi, 1917). Based on his collections in 
Tasmania, White created three new genera, Atherimorpha, Spaniopsis, and Clesthenia 
(Therevidae) (White, 1914) and Hardy described another genus, Austroleptis, from 
Australia (Hardy, 1920a). In his revision of Neotropical Rhagionidae, Lindner 
described the genus, Neorhagio (=Atherimorpha), from Chile (Lindner, 1924) and the 
following year, described the genus Therevirhagio (=Atherimorpha), from Australia 
(Lindner, 1925). During this time, Lindner also described the genus, Bicalcar 
(=Chrysopilus), from a specimen collected in South America by Loew for which he 
established the subfamily Bicalcarinae (Lindner, 1923).  
 
In 1926, Bezzi treated the South African Rhagionidae, excluding what were 
considered the subfamilies Hilarimorphinae, Xylophaginae, and the Coenomyinae 
(Bezzi, 1926). According to his interpretation, the family consisted of seven genera, 
divided into four subfamilies (Vermileoninae, Rhagioninae, Chrysopilinae, and 
Arthrotelinae). Along with Lampromyia, Atrichops, Atherix, Atherimorpha, and 
Chrysopilus, two newly described genera were added to the family; Pachybates 
(Athericidae) and Arthroteles, for which he created the new subfamily Arthrotelinae. 
 
The debate of whether or not to include the Hilarimorphidae, Xylophagidae, and the 
Coenomyidae as subfamilies of the Rhagionidae or as distinct families continued 
during this period. Unable to resolve the debate by examining adults, researchers 
turned to the larvae and pupal stages of these organisms. In 1926, based on 
 
 27 
 
examination of rhagionid larvae and pupae, Greene concluded that the groups were 
indivisible (Greene, 1926). His conclusion was based mainly on the form of the 
spiracle in the pupa, which is constant for each family. ?Should this group be 
divisible into three families one would expect to find three distinct forms of spiracles, 
but in the material available only one type occurs, which very strongly confirms my 
impression that these insects do not belong to more than one family.? This result 
conflicted with Malloch who decided that the Xylophagids and Leptids were not only 
separate but also belonged in different superfamilies, based on larval and pupal 
characters (Malloch, 1917). Greene?s work marked one of the last times 
Xylophagidae and Coenomyidae were subordinated to subfamily rank within 
Rhagionidae. 
 
In his revision of the North American Rhagionidae, Leonard deferred judgement of 
whether or not to include Xylophagidae and Coenomyidae in the Rhagionidae 
(Leonard, 1930). He stated: ?With the present evidence at hand and without making a 
more critical study of the family as a whole and its relation to the more closely allied 
families I do not feel qualified? in delimiting the family concept. Therefore, he relied 
on the Rhagionidae concept, sensu latu, of Williston (1908) and included the groups 
as subfamilies. In a follow up paper he conceded, ?I have not felt myself sufficiently 
well acquainted with the delimiting characters of several families nearly related to the 
Rhagionidae (sensu latu) to attempt to say very definately to which families several 
genera should really belong? (Leonard, 1931).  
 
 
 28 
 
In 1931, Malloch gave a history of the Rhagio name and testified that ?Rhagionidae? 
was the correct term for the family concept (Malloch, 1931). However, his paper was 
not recognized by English-speaking authors outside of the United States, many of 
whom continued to use the name Leptidae for the remainder of their careers (that last 
into the 1960s). In Malloch?s 1932 treatment, Rhagioninae include Atherimorpha 
White, Dasyomma Macquart, Dasyommina (=Dasyomma), Rhagina, Chrysopilus. 
Arthroceratinae includes Austroleptis Hardy, Heterostomus Bigot, Arthroceras 
Williston, and Arthropeas Loew (Malloch, 1932b).  
 
Szil?dy published a catalogue of the Palearctic rhagionids in 1934 (Szil?dy, 1934). He 
divided the family into three subfamilies, Bicalcarinae (including Bicalcar 
(=Chrysopilus)), Vermileoninae (including Vermileo and Lampromyia), and 
Rhagioninae (including Rhagio (and new subgenera Rhagionella and Rhagiella), 
Atrichops (Athericidae), Atherix (Athericidae), Hilarimorpha (Hilarimorphidae), 
Symphoromyia, Chrysopilus, Omphalophora (=Ptiolina), and Ptiolina (including 
Spania and new subgenera Cekendia and Archicera). He also added two new genera 
to this subfamily, Sapporomyia (=Chrysopilus) and Atherigia (Athericidae). 
 
Hennig reviewed dipteran systematics, placing a special emphasis on the 
phylogenetic value of larval characters (Hennig, 1952). He transferred Xylophagus to 
a separate family, the Erinnidae (=Xylophagidae). Soon after, Steyskal performed a 
similar review of larval characters  and reached conclusions consistent with those of 
Hennig (Steyskal, 1953). Groups such as the Xylophaginae were distinct enough to 
 
 29 
 
merit recognition as separate families. He placed 17 genera in the Rhagionidae; 
Atherix (Athericidae), Atrichops (Athericidae), Bicalcar (=Chrysopilus), Chrysopilus, 
Dasyomma (Athericidae), Dialysis, Lampromyia, Omphalophora (=Ptiolina), 
Paraphoromyia (=Symphoromyia), Ptiolina, Rhagina (=Rhagio), Rhagio, Spania, 
Spaniopsis, Suragina (Athericidae), Symphoromyia, and Vermileo (Vermileonidae) 
(Steyskal, 1953). He transferred Atherimorpha and Austroleptis to the Xylophagidae 
and put Arthroceras, Arthropeas, Arthroteles, Bolbomyia, and Glutops in the 
Coenomyiidae. 
  
In 1961, Chillcott revised the genus Bolbomyia and placed the genus in the 
Rhagionidae, although he noted: ?The correct assignment of this rather primitive 
genus is difficult. The combination of the two-jointed sytle, tibial spurs on all three 
legs, primitive wing venation and genitalia, and pulvilliform tarsal empodium does 
not fit any of the present subfamilies of Rhagionidae? (Chillcott, 1961).  
  
Several years later, Chilcott described a new rhagionid genus from Alaska, Litoleptis 
(Chillcott, 1963). In this work, he also redescribed the genus Cekendia and 
synonymyzed the genus with Bolbomyia. His study of these genera led him to believe 
that a natural group within the Rhagionidae comprised the following genera: 
Litoleptis, Bolbomyia, Hilarimopha, Archicera and Austroleptis.  
  
In the Catalogue of North American Diptera (James, 1965), Hilarimorpha was placed 
in its own family, Hilarimorphidae while Arthroceras, Arthropeas, Bolbomyia, 
 
 30 
 
Coenomyia, Glutops, Rachicerus, and Xylophagus were placed in the Xylophagidae. 
Shortly afterwards, Nagatomi argued that Arthroceras was more closely related to 
Rhagio and its allies than to Coenomyia or Xylophagus  and transferred the genus to 
the Rhagionidae (Nagatomi, 1966). 
  
In 1966, Stuckenberg reported that the family Rhagionidae included five genera of 
the Neotropical region (Atherimorpha White, Austroleptis Hardy, Chrysopilus 
Macquart, Dasyomma Macquart, and Neorhagio Lindner)  (Stuckenberg, 1966). In 
this work he added another genus from Brazil, Xeritha, a close relative of Atherix. 
James later increased this number to seven, although his interpretation of the family 
was quite different from that of Stuckenberg (James, 1968). James excluded 
Atherimorpha, Austroleptis, and Xeritha from the Rhagionidae, and included 
Vermileo, Atherix, Chrysopilus, Dasyomma, Neorhagio, Rhagio, and Suragina. 
  
In 1970, Teskey proposed that Glutops Burgess be placed in the Pelecorhynchidae 
(Teskey, 1970a). His observations were based on newly discovered larval and pupal 
characters of the genus, which were remarkably similar to the same stages of 
Pelecorhynchus. The following year, however, Krivosheina proposed that Glutops be 
placed in its own family, based on a comparative-morphological study of all 
developmental stages (Krivosheina, 1971).  
  
In 1973, Hennig established a new Diptera classification, which included a treatment 
of the Rhagionidae (Hennig, 1973). Hennig placed the genera of the Rhagionidae into 
 
 31 
 
three subfamilies as follows: Arthroceratinae: Arthroceras, Arthroteles, 
Atherimorpha, Austroleptis, Bolbomyia, Glutops, Heterostomus; Rhagioninae, 
Atherix group: Atherix; Rhagio group: Chrysopilus, Dialysis, Rhagio, Schizella, 
Symphoromyia; Spania group: Litoleptis, Omphalophora, Ptiolina, and Spania; 
Vermileoninae:  Lampromyia, Vermileo and Vermitigris. Perhaps inspired by his 
work, both the Atherix group and its allies and the Vermileoninae were later 
recognized as families. Stuckenberg (1973) established Athericidae to include the 
genera Atherix Meigen, Atrichops Verrall, Suragina Walker, Pachybates Bezzi, 
Trichacantha Stuckenberg, Dasyomma Macquart, and Xeritha Stuckenberg. In 1977, 
Nagatomi  erected Vermileonidae for Lampromyia Macquart, Vermileo Macquart, 
and Vermitigris Wheeler on account of their highly autapomorphic morphology and 
life history (Nagatomi, 1977). 
 
Although the limits of the rhagionid concept decreased significantly with the 
establishment of Athericidae and Vermileonidae, the make-up of what was left of 
Rhagionidae remained taxonomically unstable and the family lacked a coherent 
definition. In 1978, for example, Webb transferred Dialysis back to Rhagionidae in 
his revision of the genus (Webb, 1978). Three years later, in the Manual of Nearctic 
Diptera, James (1981) returned Dialysis to the Xylophagidae. Similarly, the 
placement of Austroleptis, Bolbomyia, Glutops, Hilarimorpha, and Pseudoerinna 
continued to be debated among authors (Krivosheina, 1971; Nagatomi, 1977; 
Kovalev, 1981; Teskey, 1981b; Griffiths, 1994; Grimaldi and Cumming, 1999; 
Stuckenberg, 2001).  
 
 32 
 
  
Nagatomi (1982) presented an updated concept of the family, including an intuitive 
tree to represent his ideas of the rhagionid phylogeny as part of a landmark 
monograph of the Rhagionidae (Fig. 1). Pelecorhynchus was shown as the sister 
group to the rest of the Rhagionidae, which under his concept, included genera that 
had typically been placed in the Pelecorhynchidae; Glutops and Pseudoerinna. These 
genera were considered the most primitive in Nagatomi?s scheme. Austroleptis was 
also recognized as an early offshoot of the rhagionid lineage, which includes 
Alloleptis, Arthroceras, Arthroteles, Atherimorpha, Bolbomyia, Chrysopilus, 
Desmomyia, Litoleptis, Neorhagio, Ptiolina, Rhagina, Rhagio, Schizella, 
Solomomyia, Spania, Spaniopsis, Stylospania, and Symphoromyia. Although this 
arrangement was not arrived at using a cladistic approach, and therefore the 
monophyly of Rhagionidae remained untested, Nagatomi?s work was a significant 
advance in understanding the extant diversity of the group.   
 
At this time, cladistic phylogenetic methods were being developed and applied for the 
first time, creating a new era of inquiry into the relationships among genera of the 
Tabanomorpha, and more specifically, of the Rhagionidae. Naturally, as researchers 
applied more rigorous approaches to the systematic study of the Rhagionidae, new 
phylogenetic hypotheses emerged. Woodley (1989) was the first to approach the 
group on modern cladistic grounds, providing a summary of the family, its generic 
composition, and the placement of the family into a larger context, within 
Tabanomorpha and Brachycera. His support for phylogenetic relationships relied 
 
 33 
 
principally on larval characters developed by other workers (Malloch, 1917; Steyskal, 
1953; Tsacas, 1962; Krivosheina, 1967; Roberts, 1969; Teskey, 1970a; Hennig, 1973; 
Thomas, 1974; Webb, 1983). Larvae for many rhagionid genera are not known or 
poorly described, but Woodley?s work provided a stabilizing force for rhagionid 
classification. For the most part, Woodley?s concept of Rhagionidae overlapped with 
Nagatomi?s 1982 concept, although the genera of the Pelecorhynchidae sensu lato 
(Glutops, Pelecorhynchus, and Pseudoerinna) were recognized as a separate family, 
sister to the Athericidae and Tabanidae, and not part of the Rhagionidae in any sense. 
 
Sinclair (1992) presented a treatise of the larval mandible and associated mouthpart 
structures in Diptera in order to evaluate, among other ideas, Woodley?s hypothesis of 
the orthorrhaphous brachyceran phylogeny (Woodley, 1989). In this work, Sinclair 
confirmed the larval characters that Schremmer (1951), Teskey (1969; 1970a; 1981a), 
Hennig (1973) had originally studied and that Woodley had used to construct his 
cladogram. Sinclair considered the larval mandible of Rhagionidae as a ?groundplan 
condition? of Brachycera and did not find any evidence to support rhagionid 
monophyly (Sinclair, 1992).  However, based on the association of an articulated rod 
with the larval mandibular brush, Sinclair supported the sister group relationship 
between Pelecorhynchidae and Athericidae + Tabanidae that Woodley had 
recognized. This contradicted Nagatomi (1982a), who continued to publish an 
expanded definition of Rhagionidae that included Glutops and Pseudoerinna 
(Nagatomi, 1982c; Nagatomi, 1982b; Nagatomi, 1982e; Nagatomi, 1982d; Nagatomi, 
1984; Nagatomi and Soroida, 1985; Nagatomi, 1991; Nagatomi, 1992).  
 
 34 
 
 
Consistent with the trend of decreasing the breadth of the rhagionid concept, which 
started with Stuckenberg (1973; removal of the Athericidae) and Nagatomi (1977; 
removal of the Vermileonidae), and continued through the work of Woodley (1989; 
removal of Pelecorhynchidae sensu lato), Sinclair, Cumming, and Wood (1993) 
examined the male genitalia in lower Brachycera and determined that Bolbomyia was 
more closely related to Athericidae and Tabanidae than to the rest of the rhagionid 
genera. Bolbomyia is a very small fly that has a flattened clypeus, a tibial spur on its 
fore leg, and aberrant wing venation and its placement has always been somewhat 
controversial. At first glance, it certainly does not recall the typical rhagioniform 
habitus. At various times, it has been considered a xylophagid (Loew, 1850; James, 
1968), but evidence for this placement has been considered weak and its eventual 
placement within Rhagionidae apparently occurred for no better reason than its lack 
of defining (synapomorphic) characters. Sinclair et al. (1993) argued that the presence 
of aedeagal tines was an important but overlooked character that in fact provided 
strong evidence for the proper placement of Bolbomyia. Rather than sharing common 
ancestry with the rhagionid or xylophagid genera, they asserted that the genus has a 
close relationship with Athericidae and Tabanidae (Sinclair et al., 1993). 
 
In a study principally devoted to new fossil brachyceran forms, Grimaldi and 
Cumming (1999) noted that wing venation may be an important source of 
phylogenetically informative characters. They proposed that the relative position of 
fork R
4
-R
5
 with respect to cell dm and the curvatures of R
5
 (straight) and R
4
 (with a 
 
 35 
 
sharp bend at its base) may serve as synapomorphies for most Rhagionidae. They 
noted that some of the most controversial members of the family, namely 
Austroleptis, Bolbomyia, and Litoleptis, do not exhibit this wing morphology and 
there was reason to believe, based on the shared loss of wing vein CuA
1
, that these 
genera form a distinct monophyletic entity that includes some extinct forms (Grimaldi 
and Cumming, 1999). 
 
The development of molecular techniques allowed for a new approach to solving 
these issues in Diptera (e.g., Pelandakis and Solignac, 1993; Pawlowski et al., 1996; 
Friedrich and Tautz, 1997; Carreno and Barta, 1998). Wiegmann et al. (2000) were 
the first to use this type of data in an effort to answer specific questions regarding the 
higher-level relationships in Tabanomorpha. In their analysis using 28S rDNA, the 
Pelecorhynchidae were recovered as sister to Athericidae and Tabanidae, a result 
consistent with Sinclair (1992) and Woodley (1989). The Rhagionidae were 
recovered as a monophyletic group, sister to the Vermileonidae. The most contentious 
genera of the Rhagionidae (e.g., Bolbomyia and Austroleptis) were not sampled, 
however. Within the Rhagionidae, Rhagio was recovered basal to Symphoromyia, 
which was sister to Ptiolina and Chrysopilus.  
 
Stuckenberg flatly rejected most of these notions in a bold reassessment of the family 
(2001). He argued that Bolbomyia and the entire Pelecorhynchidae form a part of 
Rhagionidae, but in a new sense that excluded taxa that he believed belong in the 
separate, newly recognized families Austroleptidae and Spaniidae. Although it was a 
 
 36 
 
notable departure from contemporary treatments of the group, Stuckenberg?s 
hypothesis closely parallels the ideas of Nagatomi (1982). Austroleptinae Nagatomi 
and Spaniinae Nagatomi were simply elevated to the family level. Although 
Stuckenberg asserted that the Pelecorhynchidae were ?always ranked by Nagatomi as 
a subfamily of Rhagionidae, a conclusion he defended categorically,? 
Pelecorhynchinae sensu Stuckenberg was, in fact, a novel arrangement. Nagatomi 
considered Pelecorhynchidae to contain a single genus, Pelecorhynchus, sister to the 
Rhagionidae. The other genera often considered to be pelecorhynchids, Glutops and 
Pseudoerinna (Teskey, 1970a; Teskey, 1970b; Woodley, 1989; Sinclair et al., 1993; 
Wiegmann et al., 2000; Wiegmann et al., 2003), were placed by Nagatomi within the 
rhagionid subfamily Glutopinae (Nagatomi, 1982a). Stuckenberg was therefore the 
first to locate Pelecorhynchus within Rhagionidae.  Also, Stuckenberg was the first to 
recognize the subfamily Bolbomyiinae, to account for the unusual genus Bolbomyia 
(Stuckenberg, 2001). 
 
A common refrain from these most recent papers is that a comprehensive 
monographic treatment of the rhagioniform genera is sorely needed. Current, 
conflicting classifications are either based on intuition developed from years of 
taxonomic experience in the group (e.g., Nagatomi, 1982a; Stuckenberg, 2001), 
primarily or exclusively single character systems (e.g., Krivosheina, 1991; Sinclair, 
1992; Sinclair et al., 1993; Grimaldi and Cumming, 1999; Wiegmann et al., 2000; 
Wiegmann et al., 2003), systematic studies involving incomplete taxon sampling 
(e.g., Wiegmann et al., 2000; Stuckenberg, 2001), analysis derived from the published 
 
 37 
 
literature (e.g., Woodley, 1989; Amorim and Silva, 2002; Yeates, 2002), or 
incomplete combinations of these components. Therefore, although great strides have 
been made in the development of the concept of the family Rhagionidae and related 
genera, a definitive, well-supported classification has not been realized.   
Biology 
Adults. 
Rhagionids are medium sized flies, ranging from approximately 4- 20mm in length, 
that are typically collected by sweeping vegetation in sheltered, rather moist, forest 
clearings or woodlands which are often of high elevation and/or mountainous. 
 
Perhaps the most well known rhagionids are Symphoromyia and Spaniopsis whose 
adult females take blood meals from vertebrate hosts. In various areas throughout its 
range, Symphoromyia has been regarded as a bothersome pest, particularly in 
mountainous or high latitude regions (Knab and Cooley, 1912; Cockerell, 1923; 
Frohne, 1953b; Frohne, 1953a; Frohne, 1959; Shemanchuk and Wintraub, 1961; 
among many others, see Turner, 1979), and in some cases, the bite of Symphoromyia 
species has caused inflammation, swelling, and even severe allergic reaction (Knab 
and Cooley, 1912; Turner, 1979; Chvala, 1983). It is interesting to note, however, that 
biting is largely restricted to North America, even though the distribution of 
Symphoromyia extends through Asia and Europe (Chvala, 1983). Even within the 
Nearctic region, some species of Symphoromyia seem not to attack people 
(Sommerman, 1962; Turner and Chillcott, 1973; Turner, 1979). Symphoromyia adults 
tend to be most active during the summer months, however, occasionally may be seen 
 
 38 
 
in flight during early spring. Species of Spaniopsis may also be pests and all are 
known to bite (Ferguson, 1915; Colless and McAlpine, 1991). Spaniopsis reportedly 
prefers shady, humid habitats, often at high elevation sites (Paramonov, 1962). 
Spaniopsis adults may be collected in Australia between November and May. Despite 
the fact that species of Symphoromyia and Spaniopsis may be pestiferous, they are not 
considered medically or economically important. None serve as vectors of disease. 
 
Arthroteles Bezzi is noteworthy for its specialized flower-feeding behavior, restricted 
to the mountain ranges of the Western Cape Province and the escarpment in eastern 
South Africa (Stuckenberg, 1956). Stuckenberg (1956) reports that Arthroteles 
cinerea resembles bombyliids in flight and is most often collected on the flower 
heads of Helichrysum spp. (Asteraceae). Interestingly, Arthroteles apparently has 
specific-specific, or nearly species-specific periods of activity. Historically, A. 
cinerea adults are collected in March, whereas A. bombyliiformis are collected in 
August or September. A. orophila is active in November. The flight of A. longipalpis 
occurs in July, and may overlap to some degree with A. bombyliiformis. 
 
In South America, Atherimorpha is principally associated with Nothofagus-
dominated woodlands, although a few species of Atherimorpha are found in arid and 
scrubby habitats to the north of Santiago and in Brazil where Nothofagus is not found 
(Malloch, 1932a). In Australia, Atherimorpha species may inhabit dense scrub and or 
wet montane forests (sometimes in association with an Australian species of 
Nothofagus). Few notes are available on the biology of South African Atherimorpha, 
 
 39 
 
but records show that they are also mostly collected in mountainous regions. In all 
habitats, it appears that Atherimorpha species gravitate towards small, slow moving 
streams, which presumably, provide moist soil substrate for their larval development. 
 
Austroleptis Hardy is another genus usually confined to mountainous regions and is 
reportedly a visitor of flowers (Colless and McAlpine, 1991). At Cradle Mountain 
National Park, in central Tasmania, I collected Austroleptis multimaculata males and 
females on the leaves of flowering Richea scoparia, however I did not see the insects 
feed. Less is known about the South American members of this genus (Nagatomi and 
Nagatomi, 1987), which are rarely collected.  
 
Chillcott (1963) reports that females of Bolbomyia are frequently collected from 
flowers, whereas the males are usually at rest on nearby vegetation. D. Webb (pers. 
comm.) collected Bolbomyia nana in a small forest clearing with a fern understory in 
North Carolina. Jeff Cumming, Richard Vockeroth, and others including myself, have 
had success sweeping Bolbomyia nana from low-lying vegetation in small forest 
clearings at King Mountain, Gatineau National Park, in Quebec. Details of its life 
history are not known. The flight period appears to be exceptionally short, lasting 
only a couple weeks or it appears, perhaps as little as a few days per year (depending 
on weather conditions).  
 
Adult Rhagio Fabricius and Chrysopilus Macquart have been reported as predaceous 
on other insects (Kellogg, 1908; Leonard, 1930; Paramonov, 1962; Narchuk, 1969; 
 
 40 
 
Narchuk, 1988) but this has never been confirmed and is unlikely given their gawky 
movements and the generalized morphology of their mouthparts. Rhagio scolopacea 
has even been reported as a bloodfeeder (Heim and Leprevost, 1892; Ferguson, 1915; 
Lindner, 1925) but these accounts are certainly false. It is remarkable, actually, how 
little is known regarding the adult stage of these common, widespread genera. Rhagio 
adults are generally active between April and September. Chrysopilus adults may be 
found throughout the year in tropical habitats, and become more seasonal, relative to 
their latitudinal displacement from the equator. In temperate climates, they are most 
common throughout the summer in both the northern and southern hemispheres. 
 
There is nothing written specifically regarding the adult biology of Arthroceras, 
Litoleptis, Ptiolina, or Spania as far as I know. The flight period for these genera may 
be exceptionally short, perhaps on the scale of a few weeks. Most often the flight 
period of these flies coincides with the spring or early summer, although Arthroceras 
appear to have a longer window of activity, in some years between April and August.  
 
Larvae. 
Among the rhagionid genera, only Chrysopilus, Rhagio, Symphoromyia, and Ptiolina 
have described larvae.  
 
Rhagionid larvae are most commonly cited as predators on a variety of insects (James 
and Turner, 1981; Foote, 1991). Foote, for instance, specifically identifies 
Symphoromyia as an insect predator (Foote, 1991). However, in a very careful study 
 
 41 
 
of the immatures of Symphoromyia species in Alaska, Sommerman (1962) noted that 
Symphoromyia larvae are slow moving and apparently are not predaceous. Similarly, 
Ptiolina are slow moving, ?shining green in life,? and feed on mosses, according to 
Brindle (1959). Others have also noted an association of Ptiolina and moss (Brauer, 
1883; Lane and Anderson, 1982). Ptiolina nigrina Wahlberg, however, apparently 
feeds on the liverwort species Marchantia polymorpha (Nartshuk, 1995). Ptiolina 
species have weakened and reduced mouthparts, a condition that is obviously 
unsuitable for predation (pers. obs.). Although the larvae of Spania nigra Meigen 
have not been characterized, Mik (1896) reportedly found a Spania nigra larva in the 
thallus of Pellia neesiana, another liverwort species (Mik, 1896). It is interesting to 
note that although the food source for Symphoromyia is not known, species of this 
genus are associated with moss and I suspect that it may similarly provide at least 
some, if not all of their food intake (Sommerman, 1962). 
 
Larval Chrysopilus Macquart may be aquatic, associated with streamside vegetation, 
or, like Rhagio Fabricius, may be found in moist soils that are rich in organic matter. 
Both are predators of oligochaetes and soft-bodied insect larvae (Tsacas, 1962; 
Roberts, 1969; Thomas, 1978a; Thomas, 1978b; Thomas, 1997). In addition to this, 
Paramonov (1962) notes that Chrysopilus larvae eat the eggs of Schistocerca and 
Dociostaurus (Orthoptera). The lifespan of Chrysopilus auratus larvae is long, lasting 
well over a year (Alexandre, 1970 in Thomas, 1997), although the pupal period may 
last less than two weeks (Tsacas, 1962). This is type of cycle is presumed to be the 
 
 42 
 
case for most rhagionid genera, as it is for most Athericidae and Tabanidae (Thomas, 
1997). 
 
Austroleptis larvae are uncharacterized, but have been reared inadvertently from 
rotting wood (Colless & McAlpine, 1991). 
 
Paramonov (1962) asserts that the larvae of Rhagionidae ?penetrate into and devour 
beetles? some larvae are saprophagous or, very rarely, coprophagous.? No 
references, either to authors or specific genera, are made however. His claims 
regarding the Rhagionidae, as here composed, are not confirmed and in my opinion, 
should be disregarded. 
 
Diversity 
The Rhagionidae, according to the definition of Woodley (1989), currently contain 
approximately 604 extant species in 19 genera. The largest genera are Atherimorpha, 
Chrysopilus, Ptiolina, Rhagio, and Symphoromyia, which comprise 88% of the total 
species. Chrysopilus is by far the largest genus with 302 species. In order of 
decreasing species diversity (species number in parenthesis), the currently-recognized 
rhagionid genera follow: Rhagio (161), Atherimorpha (50), Symphoromyia (35), 
Ptiolina (23), Austroleptis (8), Arthroceras (7), Arthroteles (4), Bolbomyia (4), 
Rhagina (4), Litoleptis (3), Schizella (3), Desmomyia (2), Alloleptis (1), Sierramyia 
(1), Solomomyia (1), Spania (1), Spatulina (1), and Stylospania (1). There are dozens 
of undescribed species of Atherimorpha from South America and Chrysopilus and 
 
 43 
 
Symphoromyia also are likely to have many more species than is currently 
recognized. There are at least several undescribed Austroleptis species from South 
America, as well. A list of all extant rhagionid species is given at the end of the 
generic treatments in chapter three. 
 
Since Rhagionidae lack diagnostic features to define its members, it is difficult to 
assess the diversity of fossils that may pertain to the group. Even for extant species, 
the family is largely known as a grouping of convenience for unusual lower 
brachycerans that may or may not belong to the same lineage. This provides fertile 
dumping grounds for fossil genera in particular, many of which are established from 
poorly preserved specimen remnants. Currently, there are 44 extinct species in 23 
fossil genera that are classified as rhagionids (Evenhuis, 1994). Even if some of these 
genera are misplaced, it appears likely that living representatives display only a 
fraction of the diversity of form once present in the Rhagionidae (and certainly in the 
Tabanomorpha). 
 
Fossil Record 
The order of appearance of fossils that are now classified as rhagionids (Evenhuis, 
1994) is summarized in Table 1.  Evidence suggests that the Rhagionidae (or 
rhagionid-like forms) were very abundant in ancient times. In Baltic amber, for 
instance, the Rhagionidae are reported to be ?by far? the most common of the lower 
brachycerous families (Larsson, 1978).  Compression fossil assemblages in Siberia 
 
 44 
 
also suggest faunal dominance for much of the history of Brachycerous dipteran life 
on earth (M. Mostovski, pers. comm.).  
 
The rhagionid lineage is an old assemblage of genera, which appears to have reached 
a significant level of diversity by at least the Middle Jurassic (Kovalev, 1981; 
Evenhuis, 1994). The oldest fossil currently placed in the Rhagionidae is 
Palaeobolbomyia sibirica, which is compressed in sediments dated at 187 MYA 
(Kovalev, 1982; Stuckenberg, 2001). Molecular dating using 28S ribosomal DNA 
estimates that basal divergences of the Tabanomorpha lineage began approximately 
170 MYA (Wiegmann et al., 2003).  
 
A study of rhagionid fossil taxonomy in light of new phylogenetic information is a 
topic for future work. Can characters derived from fossil and extant flies be 
compared? If so, are they consistent with patterns found in extant groups? Fossil wing 
vein patterns are usually preserved, and therefore, are often the most highly relied 
upon feature of a compression fossil. Do wing characters serve as a faithful predictor 
of phylogenetic relationship in this group?  
 
 45 
 
 
Table 1. Presence of Rhagionidae in the fossil record. 
Era Period Epoch Extant Genera Extinct Genera 
Pliocene   
Miocene  Dipterites 
Oligocene Rhagio  
Eocene Bolbomyia, 
Chrysopilus, 
Symphoromyia 
 
Cenozoic Tertiary 
Paleocene   
Upper  Zarzia 
Middle   
Cretaceous 
Lower Atherimorpha, 
Ptiolina 
Jersambromyia,  
Mesobolbomyia,  
Mongolomyia, 
Palaeochrysopilus 
Upper  Mesorhagiophryne,  
Mesostratiomyia, 
Palaeoarthroteles, 
Palaeostratiomyia, 
Protorhagio,  
Rhagiophryne, 
Scelerhagio, 
Stratiomyiopsis 
Mesozoic 
Jurassic 
Middle  Ija, 
Jurabrachyceron, 
Kubekovia, 
Paleobolbomyia, 
Palaeobrachyceron 
 
 
 46 
 
Biogeography 
The largest genus of the Rhagionidae, Chrysopilus Macquart, is found throughout the 
globe, including one species that extends into the tundra (Nartshuk, 1995), and 
another as high as 4000masl in the Neotropics (pers. obs.).  The Palaearctic and 
Oriental regions are the most species-rich areas for Chrysopilus, although there are 
certainly many more species in the Neotropical region than are currently recognized. 
  
Rhagio Fabricius is distributed throughout the Holarctic reaching its southernmost 
extension in the Oriental region where it is found in Java and Sumatra.  Putative close 
relatives of Rhagio such as Rhagina Malloch and Desmomyia Brunetti have localized 
distributions within the range of Rhagio. Rhagina is restricted to China and Java, and 
Desmomyia is known only from India (Brunetti, 1912; Yang et al., 1997). Sierramyia 
may also be related to these genera. It has been collected only from mountainous 
regions of Mexico.  
  
Atherimorpha White is distributed in a typical Gondwanan fashion, found in 
Australia, Patagonia, and South Africa. Its putative sister taxon, Arthroteles Bezzi 
(Stuckenberg in Nagatomi and Nagatomi, 1990) is endemic to South Africa. Fossil 
evidence suggests that the biogeographical history of this lineage may actually be 
more complicated than simple Gondwanan vicariance. A fossil specimen reported to 
have strong affinities to Arthroteles, Palaeoarthroteles mesozoicus, was found and 
described from Transbaikalia (Russia), in sediments aged between the Upper Jurassic 
and Lower Cretaceous (Kovalev and Mostovski, 1997). 
 
 47 
 
  
Bolbomyia Loew was originally described from the Baltic region as a fossil in amber, 
and living members of the genus inhabit the Russian Far East, Canada, and the USA.  
Arthroceras Williston and Spania Meigen exhibit similar distributions, and into Asia 
(Nagatomi, 1966). 
 
Litoleptis is unusual in having a pan-Pacific distribution, with species endemic to 
Alaska, the Philippines, China, and Chile (Chillcott, 1963; Hennig, 1972; Yang et al., 
1997). 
  
Studies of current distributional patterns of living genera in light of new phylogenetic 
evidence may provide insights into the biogeographic history of lineages within the 
Tabanomorpha. Fossils are an important element, however, in providing evidence to 
refute hypotheses of vicariance and illuminate other aspects of biogeographic patterns 
that are germane to the study of Rhagionidae as a whole (Rosen, 1990). The group 
appears old and anciently abundant enough to have been distributed throughout 
Laurasia and Gondwanaland before the breakup of these continents.  Therefore, at 
least some current biogeographic patterns may be explained as the survival of relict 
populations of formerly cosmopolitan species. 
 
Phylogenetic Hypotheses 
Although most authorities have incorporated their ideas of phylogeny into a 
classification scheme, others have proposed groupings without formalizing their 
 
 48 
 
recommendations for changes in taxonomic rank or placement. In this section, I will 
briefly review the most recent and pertinent phylogenetic ideas, sometimes 
represented by classifications alone, and how they differ from one another regarding 
Rhagionidae and other genera within the Tabanomorpha. These are the phylogenetic 
hypotheses I intend to test in this study. 
 
1. Monophyly and Position of Pelecorhynchidae.   
a. Glutops, Pelecorhynchus, and Pseudoerrina a paraphyletic unit 
(Krivosheina, 1971; Nagatomi, 1982a). 
b. Glutops, Pelecorhynchus, and Pseudoerrina a monophyletic unit 
(Woodley, 1989; Sinclair, 1992; Stuckenberg, 2001).  
c. All three genera a clade, within Rhagionidae (Stuckenberg, 2001). 
d. All three genera a clade, sister to Athericidae + Tabanidae (Woodley, 
1989; Sinclair, 1992; Wiegmann et al., 2000; Wiegmann et al., 2003). 
e. Glutops and Pseudoerrina sister clade to the rest of Rhagionidae 
(Nagatomi, 1982a). 
f. Glutops a clade outside of Rhagionidae (Krivosheina, 1971). 
 
2. Position of Vermileonidae. 
a. Vermileonidae outside of Tabanomorpha, suitable for separate infraordinal 
rank recognition (Nagatomi, 1977; Nagatomi, 1991; Griffiths, 1994; 
Stuckenberg, 2001). 
b. At the base of Tabanomorpha (Nagatomi, 1977; Sinclair et al., 1993). 
 
 49 
 
c. Sister of Rhagionidae within Tabanomorpha (Wiegmann et al., 2000; 
Yeates, 2002; Wiegmann et al., 2003). 
 
3. Position of Bolbomyia. 
a. A xylophagid (Loew, 1850; James, 1965). 
b. A member of Rhagioninae (Nagatomi, 1982a). 
c. A member of a separate subfamily, Bolbomyiinae, within Rhagionidae 
(Stuckenberg, 2001). 
d. Together with Austroleptis and Litoleptis, within or outside of 
Rhagionidae (Grimaldi and Cumming, 1999). 
e. Sister taxon to Athericidae + Tabanidae (Sinclair et al., 1993). 
 
4. Position of Austroleptis 
a. A separate subfamily, Austroleptinae, within Rhagionidae (Nagatomi, 
1982a). 
b. A monotypic family, within Tabanomorpha (Stuckenberg, 2001). 
c. A xylophagomorph (James, 1965; Colless and McAlpine, 1991; Sinclair et 
al., 1993). 
d. Together with Bolbomyia and Litoleptis, within or outside of Rhagionidae 
(Grimaldi and Cumming, 1999). 
 
5. Monophyly and Position of Spaniidae. 
 
 50 
 
a. Spaniinae (-idae) form a monophyletic unit within Rhagionidae 
(Nagatomi, 1982a). 
b. Spaniidae a clade outside of Rhagionidae (Stuckenberg, 2001). 
 
6. Monophyly of Chrysopilus. 
a. Chrysopilus monophyletic with respect to Schizella and Solomomyia 
(Nagatomi, 1982a). 
 
7. Monophyly of Rhagio.  
a. Rhagio monophyletic with respect to Rhagina and Desmomyia (Nagatomi, 
1982a). 
 
8. Monophyly of Atherimorpha. 
a. Atherimorpha monophyletic with respect to Arthroteles (Stuckenberg in 
Nagatomi and Nagatomi, 1990).  
 
 
 51 
 
Chapter 2. Phylogenetic Analysis of the Genera of the Rhagionidae and Related 
Taxa.  
Introduction 
Shared inheritance of both genotypic and phenotypic characters through common 
ancestry is our evidence of evolution and the relatedness of living organisms. The 
following research attempts to gather these characters together, systematically, to 
understand the relationships among rhagionid genera and their place in lower 
Brachycera.   
 
Traditional morphological techniques have generated many robust hypotheses on the 
phylogeny of many organisms, attesting to the rigor of morphological study. However 
previous work on the rhagioniform Diptera has been limited to particularly accessible 
character systems, such as antennal, wing, and thoracic morphologies, and may have 
produced a somewhat distorted view of underlying phylogenetic patterns. A 
comprehensive, systematic approach is taken here to develop and evaluate characters 
of the head including the internal mouthparts, thorax, abdomen, male genitalia, and 
female genitalia. One of the major aims of this effort is to develop new, 
phylogenetically informative characters, as well as adopt and evaluate character 
systems used in the past.  
 
The value of morphology for phylogenetic analysis has recently been in some dispute, 
with the popularity and acknowledged utility of molecular systematic methods for 
addressing phylogenetic problems (Baker and Gatesy, 2002; Scotland et al., 2003; 
 
 52 
 
Jenner, 2004). Phylogenetic analyses based on morphology alone may lack statistical 
power to assess branch support adequately simply due to the few number characters 
that available for morphological analysis. The decision of whether to study 
morphology or molecular characters is, for the most part however, a false dichotomy. 
It is readily apparent that the two data types are complimentary, as each serves as an 
independent source of evidence regarding the phylogeny of the group. The 
hypotheses generated from each source may be compared against one another and the 
data from each may also be combined as part of a total evidence approach, to yield a 
combined hypothesis. Wiegmann et al. (2000) showed that 28S nuclear ribosomal 
DNA that may be used to track divergences among rhagionid and related genera. His 
work provides the foundation for further work that I develop here, to complement and 
compare against independent estimates of phylogeny for the group generated from 
morphological data.   
 
The three sections of this chapter are thus laid out as followed. The opening section 
treats the morphology of the group, in which over two hundred characters were 
generated and scored for 34 species of Rhagionidae sensu Woodley (1989) in 18 
genera (Alloleptis, Arthroceras, Arthroteles,  Atherimorpha, Austroleptis, Bolbomyia, 
Chrysopilus, Desmomyia, Litoleptis, Ptiolina, Rhagio, Schizella, Sierramyia 
(=?Neorhagio? in Nagatomi, 1982a), Spania, Spaniopsis, Stylospania, 
Symphoromyia), 3 species of Pelecorhynchidae sensu Woodley (1989) in 3 genera 
(Glutops rossi, Pseudoerrina jonesi, Pelecorhynchus personatus), 2 species of 
Vermileonidae in 2 genera (Lampromyia canariensis, Vermileo vermileo), 2 species 
 
 53 
 
of Athericidae in 2 genera (Atherix pachypus and Dasyomma atratulum), and 2 
species of Tabanidae in 2 genera (Dichelacera marginata and Tabanus atratus). Four 
species of xylophagid taxa (Arthropeas americana, Coenomyia ferruginea, Dialysis 
rufithorax, Xylophagus lugens) were used as outgroup taxa. This data is analyzed 
using maximum parsimony. 
 
The second section of this chapter is a molecular treatment of the group. An 
approximately 3000bp stretch of 28S rDNA is amplified for molecular phylogenetic 
analysis for 22 species of Rhagionidae sensu Woodley (1989) in 8 genera 
(Atherimorpha, Austroleptis, Bolbomyia, Chrysopilus, Ptiolina, Rhagio, Spaniopsis, 
Symphoromyia), 2 species of Pelecorhynchidae sensu Woodley (1989) in 2 genera 
(Glutops rossi and Pelecorhynchus personatus), 3 species of Vermileonidae in 2 
genera (Leptynoma and Vermileo), 3 species of Athericidae in 2 genera (Atherix and 
Dasyomma), and 4 species of Tabanidae in 2 genera (Chrysops and Tabanus). 
Sequences of Pachygaster leachii (Stratiomyiidae), Pantophthalmus sp. 
(Pantophthalmidae), and xylophagid taxa (Arthropeas magnum, Coenomyia 
ferruginea, Dialysis elongate, Exeretonevra angustifrons, Heterostomus sp., and 
Xylophagus abdominalis) were taken from GenBank and used as outgroup taxa. This 
data is analyzed using maximum parsimony and maximum likelihood optimality 
criteria. 
 
The third section of this chapter combines the morphological and molecular data of 
the previous sections using two techniques to arrive at a best-estimate phylogeny of 
 
 54 
 
the group. The first technique uses a concatenated morphological and molecular data 
set for simultaneous analysis using maximum parsimony and Bayesian inference. The 
second approach is a type of supertree analysis, in which a parsimony analysis is 
conducted on a matrix representation of the trees generated from the morphological 
and molecular data. This analysis gives an overall picture of the hypotheses generated 
from both morphological and molecular analyses in a single tree diagram, in which all 
taxa are included. 
 
A re-classification of the group based on the phylogenetic analyses of this chapter is 
addressed in Chapter 3.  
 
 
 
 55 
 
Morphological Treatment 
Introduction 
A thorough understanding of the morphological diversity present in the Rhagionidae 
and related groups is a necessary step in satisfying three primary aims of this study. 
The first aim is to find new, phylogenetically informative characters particularly in 
character systems that have been overlooked in the past, such as the internal 
structures of the female terminalia and characters revealed by using SEM technology. 
The second aim is to evaluate characters and/or character systems that have been used 
in the past, and develop these character systems further by using a greater taxon 
sampling. A large matrix consisting of new and well-developed characters and a full 
complement of taxa is essential in accomplishing the third aim, which is the 
application of modern cladistic methods to develop a rigorous hypothesis of 
phylogenetic relationship for the group based on morphology (Hillis et al., 1994; 
Rannala et al., 1998). 
 
Thus far, groupings based on morphological characters have led to discordant 
classification schemes. Pelecorhynchus Macquart, for example, has been placed as 
sister to the Rhagionidae on the basis of male and female genitalic characters 
(Nagatomi, 1977), within the Rhagionidae based on female abdominal characters 
(Stuckenberg, 2001), and, along with Glutops and Pseudoerrina, sister to  the 
Athericidae and Tabanidae based on larval characters (Teskey, 1970; Sinclair, 1992; 
Woodley, 1989). Similarly, Bolbomyia Loew may be located among the 
Xylophagidae because of its flattened clypeus (James, 1965), among the Rhagionidae 
 
 56 
 
on account of an elongated intersegmental region in the female abdomen (Nagatomi, 
1982a; Stuckenberg, 2001), together with Austroleptis and Litoleptis (inside or 
outside of Rhagionidae) because it lacks wing vein M
3
 (Grimaldi & Cumming, 1999), 
or as sister to Athericidae and Tabanidae on account of having aedeagal tines in the 
male genitalia (Sinclair et al., 1993).  
 
In taxa such as Austroleptis Hardy, highly autapomorphic morphology has caused 
higher level taxonomic instability (Nagatomi, 1982a; Nagatomi, 1984). Austroleptis 
has evolved in such a way that it shares few character states with possible relatives, 
obscuring its relation to the rest of Lower Brachycera. Thus, ecological information, 
such as larval feeding habits, has been used as a surrogate to direct morphological 
evidence as a basis for proposed classification (Colless & McAlpine, 1991; Sinclair et 
al., 1993). Stuckenberg preferred to use the derived state of Austroleptis as evidence 
for supporting its own, family-level recognition (Stuckenberg in Nagatomi, 1982a; 
Stuckenberg, 2001). Similarly, the autapomorphic morphology of Litoleptis Chillcott 
has led to speculation regarding its classification (Chillcott, 1963; Hennig, 1972; 
Grimaldi & Cumming, 1999). 
 
Morphology may be an insufficient source of information for tracking tabanomorph 
lineage divergences that may be as much as 170 million years old (Wiegmann et al., 
2003). Advances in molecular systematics and statistics-based methods of 
phylogenetic inference (e.g., Hillis et al., 1994; Huelsenbeck, 1995; Huelsenbeck and 
Crandall, 1997; Huelsenbeck and Rannala, 1997; Rannala et al., 1998; Huelsenbeck 
 
 57 
 
and Bollback, 2001; Huelsenbeck et al., 2001; Huelsenbeck et al., 2002; Felsenstein, 
2004) have already shown promise in helping to resolve ancient divergences and will 
certainly gain more favor as morphological inquiries are exhausted (Hillis and Wiens, 
2000; Baker and Gatesy, 2002). Study of tabanomorph morphology however is far 
from exhausted. Most of the disagreement in the classification of taxa such as 
Pelecorhynchus, Bolbomyia, and Austroleptis is the result of studying a limited set of 
characters and/or taxa. A better understanding of tabanomorph morphology clearly is 
needed to refine classifications based on morphology and to provide a larger context 
for molecular studies. 
Materials and Methods 
Laboratory Methods 
The abdomens of pinned, dried specimens were dabbed with a small amount of 
approximately 5% KOH to soften the tissue and allow for the entire abdomen 
(females) or terminal segments (males) to be gently torn off. For mouthpart 
dissections, the entire head is taken. The abdomen or head was then placed in 10% 
KOH solution in a vial warmed in a hot water bath (~95?C) for approximately 10 
minutes. The material was removed and rinsed thoroughly with water. For males, the 
terminalia were dissected by separating the epandrium from the gonocoxites. This 
was done either in water or glycerin. For females, the lateral membrane separating the 
dorsal (tergites) and ventral (sternites) sclerites of the abdomen was split to expose 
the internal tissues. The specimens were then placed in a saturated solution of 
chlorozol black in water for approximately 5-7 minutes for staining. Excess dye was 
removed in water and further dissections of the female terminalia were carried out in 
 
 58 
 
water, 80% ethanol, or glycerin. Preparations were placed on a microscope slide, in a 
small pool of glycerin, covered with a cover slip, and photographed digitally. Digital 
images were captured using a JVC KY-F70 top mounted digital camera, and 
enhanced using AutoMontage photo imaging software. For long-term preservation, 
terminalia are stored in glycerin in a genitalia vial and are mounted on the pin 
underneath the specimen and label(s). 
 
Scanning electron microscopy (SEM) was carried out for larval specimens to 
illustrate surface structure in Pelecorhynchus sp., Glutops rossi Pechuman, Vermileo 
sp., Rhagio sp., Symphoromyia sp., and Chrysopilus sp. A cross section cut was made 
with micro scissors to excise both the anterior segments (containing the head) and the 
posterior segments containing the terminal segment. These sections were rinsed in 
hydrogen peroxide solution (H
2
O
2
), transferred to 100% EtOH, then soaked twice for 
several minutes in pure hexamethyldisilazane (HMDS) under a fume hood and 
allowed to dry. The specimens were then affixed to SEM flat-disc mounts with water-
soluble Elmer?s glue. Adults were mounted directly onto SEM mounts. SEM imaging 
was done at the Smithsonian Institution Scanning Electron Microscopy Core, in the 
National Museum of Natural History, with the assistance of Scott Whittaker, 
Smithsonian SEM Lab Manager. All images were taken using the Philips XL-30 
ESEM with LaB6 filament and the SIS AnalySIS Image analysis package, in the low 
vacuum mode using water vapor. 
 
 
 59 
 
The larval characters are used to help resolve taxa at the genus level. Species-level 
differences of tabanomorph larvae are very poorly documented and as a result, larvae 
included in this study were identifiable to genus level only. Larval characters scored 
for Atherix pachypus were taken from several undetermined Atherix species and 
ambiguities represent apparently interspecific differences between specimens. 
Similarly, all Chrysopilus, all Rhagio, and all Symphoromyia species were scored 
from the same set of congeneric larval specimens. Character coding, therefore, is 
identical for species within these genera and do not help to resolve infrageneric 
relationships. A single larval specimen identified as Ptiolina sp. was used to score P. 
mallochi and P. zonata.  
Taxon Sampling 
The ingroup taxa sampled here include representatives from all of the genera in 
Rhagionidae recognized by Nagatomi (1982), with the exception of the monotypic 
genus Solomomyia, which is characterized as a close relative of Chrysopilus 
(Nagatomi, 1982a) and Alloleptis, a genus only known from a male specimen 
(Nagatomi, 1982a). Outgroup taxa included representatives from all families within 
Tabanomorpha, as well as several genera within the Xylophagomorpha. Table 11 
shows the species used in the morphological phylogenetic analysis, along with their 
recent family designations, and the geographic distribution of the specimens 
examined for each species. The scoring of several species was restricted to limited 
availability of specimens. These species are listed in Table 12, with explanations.  
 
 
 60 
 
The breadth of taxon sampling was determined on the basis of availability of 
specimens for study and the importance of taxa for testing specific hypotheses of 
relationship. These hypotheses include the monophyly and position of 
Pelecorhynchidae (Krivosheina, 1971; Nagatomi, 1982a; Stuckenberg, 2001; 
Woodley, 1989; Sinclair, 1992; Wiegmann et al., 2000); the monophyly and position 
of Spaniinae (Nagatomi, 1982a; Stuckenberg, 2001); the monophyly of Chrysopilus 
(Bezzi, 1917; Nagatomi, 1982a), Rhagio (Malloch, 1932a; Nagatomi, 1982a; Yang 
and Nagatomi, 1992; Yang et al., 1997), and Atherimorpha (Nagatomi, 1982a, 1990); 
the position of Vermileonidae (Nagatomi, 1977, 1991; Griffiths, 1994; Stuckenberg, 
2001; Sinclair et al., 1993; Wiegmann et al., 2000; Yeates, 2002); the position of 
Bolbomyia (James, 1965; Nagatomi, 1982a; Sinclair et al., 1993; Grimaldi & 
Cumming, 1999; Stuckenberg, 2001); and the position of Austroleptis (Nagatomi, 
1982a; Colless & McAlpine, 1991; Stuckenberg, 2001; Sinclair et al., 1993; 
McAlpine). Species were used as terminals, an approach that is best for 
reconstructing phylogenetic relationships (Yeates, 1995; Wiens, 1998). In most cases, 
multiple exemplars were used to sample species diversity, especially for large genera 
within Rhagionidae, and to increase the accuracy of phylogenetic inference (Hillis, 
1996; Rennala et al., 1998). Where possible, highly autapomorphic exemplars were 
avoided.  
 
A USNM ENT barcode label was attached to the pin of all specimens examined, 
except most non-dissected specimens belonging to the genera Rhagio and 
Chrysopilus. The barcodes provide a unique identifier to each specimen, and are 
 
 61 
 
tracked using a FilemakerPro database. The database of rhagionid specimens is 
published on the National Museum of Natural History, Entomology Department web 
server and may be viewed at 
http://entomology.si.edu/Entomology/Rhagionidae/Search.lasso.  
 
Table 2. Taxon sampling, morphological analysis. 
Taxa Recent Family 
Placement 
Region Geographic Distribution 
of specimens examined 
Alloleptis tersus  Rhagionidae OR Indonesia: Sulawesi 
Arthroceras pollinosum  Rhagionidae NA USA: CA, CO, OR, NM, WA, 
WN 
Arthroteles bombyliiformis Rhagionidae AT South Africa: Cape Province 
Atherimorpha atrifemur Rhagionidae NT Chile: Chilo?, Llanquihue, 
Malleco, Osorno Provinces 
Atherimorpha nemoralis  Rhagionidae NT Chile: Arauco, Caut?n, Chilo?, 
Llanquihue, Malleco, Osorno, 
Valdivia Provinces 
Atherimorpha vernalis  Rhagionidae AU Australia: Tasmania 
Atherix pachypus  Athericidae NA USA: CO, MT 
Austroleptis multimaculata  Rhagionidae/ 
Austroleptidae 
AU Australia: Tasmania 
Bolbomyia nana  Rhagionidae NA Canada: Ontario, Quebec 
USA: MD, MI, NY, PA, VA 
Chrysopilus ferruginosus  Rhagionidae OR Philippines: Luzon 
Chrysopilus panamensis  Rhagionidae NT Costa Rica: Lim?n 
Chrysopilus quadratus  Rhagionidae NA USA: MD, NH, PA 
Chrysopilus thoracicus  Rhagionidae NA USA: MD, TN 
Dasyomma atratulum  Athericidae NT Chile: Chilo? 
Desmomyia thereviformis Rhagionidae OR India: E. Punjab. 
Dichelacera marginata  Tabanidae NT Colombia: Antioqua 
Glutops rossi  Rhagionidae/ 
Pelecorhynchidae/ 
Glutopidae 
NA Canada: Alberta, British 
Columbia 
Lampromyia canariensis  Vermileonidae AT Spain: Canary Islands 
Litoleptis alaskensis Rhagionidae/ 
Spaniidae 
NA USA: AK 
Pelecorhynchus personatus  Rhagionidae/ 
Pelecorhynchidae 
AU Australia: NSW, Queensland 
Pseudoerrina jonesi  Rhagionidae/ 
Pelecorhynchidae 
NA USA: WN 
Ptiolina lapponica  Rhagionidae/ 
Spaniidae 
PA Finland: Petsamo, Ponoj 
Ptiolina majuscula  Rhagionidae/ 
Spaniidae 
NA USA: AK 
Ptiolina mallochi  Rhagionidae/ 
Spaniidae 
NA USA: AK 
 
 62 
 
Taxa Recent Family 
Placement 
Region Geographic Distribution 
of specimens examined 
Ptiolina zonata  Rhagionidae/ 
Spaniidae 
NA Canada: Manitoba, NWT, 
Yukon 
USA: AK, WA 
Rhagio costatus  Rhagionidae NA USA: CA, WA
Rhagio hirtus  Rhagionidae NA USA: IN, MD, NH, PA, VA 
Rhagio mystaceus  Rhagionidae NA USA: MI, MD, NH, NJ, VA, 
VT 
Rhagio plumbeus  Rhagionidae NA USA: PA, MN 
Rhagio scolopaceus  Rhagionidae PA Switzerland: Grisons, Zurich 
Rhagio sinensis  Rhagionidae OR China: Fukien 
Rhagio vertebratus  Rhagionidae NA USA: PA, ME 
Rhagina incurvatus  Rhagionidae OR Indonesia: Java
Schizella furcicornis  Rhagionidae OR Philippines: Luzon, Mindanao 
Sierramyia chiapasensis  Rhagionidae NT Mexico: Chiapas 
Spania nigra  Rhagionidae/ 
Spaniidae 
PA Austria: Tirol 
Switzerland: Om 
USA: WN 
Spaniopsis longicornis  Rhagionidae/ 
Spaniidae 
AU Australia: NSW 
Spaniopsis clelandi  Rhagionidae/ 
Spaniidae 
AU Australia: ACT, NSW 
Stylospania lancifera  Rhagionidae OR Philippines: Samar 
Suragina concinna  Athericidae NA USA: TX 
Symphoromyia hirta  Rhagionidae NA USA: CT, MD, PA, VA 
Symphoromyia cruenta  Rhagionidae NA USA: CA 
Tabanus atratus  Tabanidae NA USA: GA 
Vermileo vermileo  Vermileonidae PA Isreal: Nahal Tut 
Spain: Balearic Islands 
Arthropeas americana  Xylophagidae NA USA: MI 
Coenomyia ferruginea  Xylophagidae/ 
Coenomyiidae 
NA USA: MD, MN 
Dialysis rufithorax  Xylophagidae NA USA: MD, VA 
Xylophagus lugens  Xylophagidae NA USA: MD, VA 
 
Table 3. Species with incomplete sampling. 
Species Gender unavailable 
for direct examination 
Comment 
Alloleptis tersus Male and Female The male of this species was scored using the literature 
(Nagatomi, 1984). The female of Alloleptis tersus is 
unknown. 
Litoleptis alaskensis Female A female of Litoleptis alaskensis was deposited in the 
Canadian National Collection, but subsequently loaned 
to the B. P. Bishop Museum. The Bishop Museum does 
not have a record of this and the specimen is not 
currently in Hawaii. The female of this species has 
been examined by Nagatomi (1982), however it is 
undescribed in the literature. 
Pseudoerrina jonesi Male The male of this species was scored using Nagatomi 
 
 63 
 
Species Gender unavailable 
for direct examination 
Comment 
(1984). 
Stylospania lancifera Female This species is known from a single individual, which 
is male. The holotype was used for scoring.  
 
Phylogenetic Analysis 
Phylogenetic analyses were performed using two computer programs, on both PC and 
Macintosh operating systems. PAUP* 4.0b10 (Swofford, 2001) was used for 
maximum parsimony (MP) phylogenetic analyses. MrBayes 3.0b4 (Ronquist and 
Huelsenbeck, 2003) was used for Bayesian inference (BI).  
 
For parsimony analyses, 1000 heuristic search replicates were performed with 
random-taxon-addition, tree bisection reconnection (TBR) branch swapping, steepest 
decent and ?MulTrees? options in effect. All characters were treated unordered and 
assigned equal weights. Scores of all MP trees were verified to avoid artifactually 
inflated sets of MP trees. In order to gauge the internal consistency of the data, 
bootstrap analyses (Felsenstein, 1985) were carried out. MP bootstrap analyses were 
performed with 1000 repetitions, each with 30 random-taxon-addition TBR heuristic 
searches. MacClade 4.03 (Maddison and Maddison, 2000) was used to analyze 
character change and support inherent in the phylogenetic tree.  
Results 
Morphological Characters and Character State Coding 
The morphology, its scoring, and the particular character states present among the 
sampled taxa are detailed below. I begin with the head of the adult and work 
 
 64 
 
posteriorly, covering aspects of the thorax, wing, legs, abdomen, male genitalia, and 
external and internal female genitalia. I follow this with a discussion of the larval 
characters. General notes and a summary of each major character system precede a 
more detailed discussion of each of the characters that were inspected over the course 
of the study. Not all characters were used for the phylogenetic analysis.  
 
Alternative scorings of the same character are possible, given differing perspectives 
on perceived character evolution, especially for complex characters. In order to 
explore this realm of possible character state codings, some of the same characters 
were scored in several different ways. This was done as an explicit demonstration of 
how characters may be scored (or how they?ve been scored in the past), based on 
differing interpretations of character evolution.  
 
Novel character codings may work for a limited taxon set, but can break down once a 
greater sampling is included. Even though these character codings ultimately fail, the 
effort in developing them is not necessarily for naught. The way in which they fail 
may be informative and for this reason, are preserved and discussed below.   
 
In total, over two hundred characters were generated and scored for the included 
taxon set (Table 13). Ultimately, each character was evaluated on its consistency, 
clarity, and independence from other characters. Inferior, non-independent character 
state codings were eliminated for the phylogenetic analysis, which used 
approximately 150 characters.  
 
 65 
 
 
Adult head.  
[#2; note 1]. Clypeus. The bulbous clypeus is a putative synapomorphy for 
Tabanomorpha. All Rhagionidae sensu lato have this character, except Austroleptis. 
In Austroleptis, the clypeus is recessed, as in Xylophagidae. The clypeus in Litoleptis 
is nearly flat, as it is in Bolbomyia. For both of these genera, the clypeus was scored 
as bulbous, although it is only slightly so. I looked at many Chrysopilus species and 
there are differences within this genus in the form of the clypeus; some are produced 
anteriorly much more than in other Chrysopilus species. An attempt was made to 
score relative to the sharpness of the break between the eye and clypeus, but this is 
variable within large genera such as Chrysopilus and Symphoromyia and the lack of 
distinct differences between the variation of states among genera discouraged me 
from developing this aspect of the clypeus further. 
 
[#3]. Scape. In some species of Tabanomorpha, the scape is much larger than pedicel. 
This occurs across a broad range of species, including members of Atherimorpha, 
Desmomyia, Dichelacera, Glutops, Lampromyia, Pelecorhynchus, Symphoromyia, 
and Vermileo. The scape is larger than the pedicel in two South American 
Atherimorpha species (A. albohirta and A. praefica (subgenus Philippoleptis 
Malloch)), however these species are apparently autapomorphic for this condition and 
are not included in this taxon sample. 
 
 
 66 
 
[#4]. First flagellomere (lateral compression). In most taxa, the amount of lateral 
compression of the first flagellomere is easily scored. Species of Symphoromyia, 
Chrysopilus, and Rhagio have first flagellomere clearly laterally compressed, whereas 
in species of Arthroteles, Arthroceras, Glutops, Pseudoerrina, Pelecorhynchus, and 
Xylophagidae, the first flagellomere is rounded in cross section. There are cases, 
however, where congenerics may differ in having laterally compressed or rounded 
first flagellomere. For example Ptiolina (Omphalophora) majuscula has the first 
flagellomere rounded, whereas most Ptiolina species have a clearly flattened first 
flagellomere. Variation of this sort is also present in Atherimorpha. Some difficulties 
may exist in scoring this character, such as in Stylospania, where the cross section of 
the first flagellomere is laterally compressed somewhat, and oval in cross section. In 
these instances, the scoring may be subtle. Stylospania lancifera is scored as having a 
laterally compressed first flagellomere. 
 
[#5]. Antennal composition (not used in phylogenetic analysis). Stuckenberg 
(1999) has hypothesized that the brachycerous antenna has evolved in a progressive 
fusion of segments, resulting in a transformation series that steadily concentrates 8 
flagellomeres into a single, stylate flagellomere. As an example, he used character 
transformations apparent within the Vermileonidae. The fundamental flaw in his 
argument is that character homology is defined by the common ancestry of character 
states, arising from a single character. The reduction of the flagellomere in 
Brachycera may be the result of similar evolutionary pressures, but the 
transformations, for example, in Therevidae, Xylophagidae, and Empididae are 
 
 67 
 
independent events and therefore, are analogous, and not homologous, to the 
transformation series apparent in Vermileonidae. Due to the complex nature of the 
evolution of the antenna, it is difficult to assess homology confidently, but I prefer to 
score the antenna as part of two characters rather than one. The character as scored 
here, I believe, may be susceptible for mistaking analogy for homology and is not 
used in the phylogenetic analysis. The modified characters are below. 
 
[#6]. Antennal flagellum (presence of break). Hennig (1972) proposed that a break 
between the first flagellomere and those distal to it could be a synapomorphy for the 
Rhagionidae. There may be some confusion in this character, however. For example, 
Arthroteles is scored as having a distinct break whereas Arthroceras is scored as 
gradually tapering, even though the antennae are similar to one another. In 
Arthroteles, however, the first flagellomere is distinct in that it is not only enlarged, 
but the flagellomeres distal to it are smaller, cylindrical, and not tapering. In 
Arthroceras, each flagellomere tapers distally, starting with the first flagellomere. 
This is also the case in Glutops.  
 
[#7]. Segment(s) distal of first flagellomere. Where there is an abrupt change in 
shape of the antenna after the first flagellomere, distal segments may be of three 
types: 1) as segmented flagellomeres (as in Alloleptis, Arthroteles, Atherimorpha, 
Austroleptis, and Bolbomyia) 2) stylate (as in Litoleptis, Ptiolina, Spania, and 
Spaniopsis) or 3) aristate (as in Atherix, Chrysopilus, Dialysis, Lampromyia, Rhagio, 
Suragina, Symphoromyia, and Vermileo). Nagatomi (1982) has also asserted that 
 
 68 
 
having an antenna with a tapering, stylate first flagellomere distinguishes subfamily 
Spaniinae. 
 
[#8]. Arista. The arista is scored as bare or microsetose is applicable for a subset of 
taxa (those with aristate antenna). There may be a continuum of character states, 
although most Chrysopilus species have an arista that is clearly more microsetose 
than most Rhagio species. Bare as used here indicates that the microsetae of the 
antenna are not prominent; that is, when they are shorter than the width of the arista. 
Under high magnification, such "bare" arista will reveal microsetae. [R. gracilis 
(Johnson) is as microsetose as S. limata Coquillett, but bare for most of the basal half 
and certainly less microsetose than in many Chrysopilus species.] 
 
[#9]. Eyes (microsetae). At first, it appeared that many lower brachyceran flies had 
bare eyes, but upon very close inspection, short, sparsely distributed microsetae are 
visible in most species. Alloleptis tersus, Coenomyia ferruginea and Pseudoerrina 
jonesi have eyes that are conspicuously setose. It is worth mentioning however, that 
within the genus Pseudoerrina, this character state is not invariable. In Pseudoerrina 
fuscata (Nagatomi, 1982a: 39), the eyes are practically bare, as in most lower 
brachycerans. 
 
[#10]. Eyes in male (separation). Males in these genera have eyes that either touch 
centrally (holoptic) or are separated (dichoptic). It is worth mention that although 
males of most Arthroteles species are holoptic, there is one species, Arthroteles 
 
 69 
 
longipalpus Nagatomi & Nagatomi, where the male is dichoptic. Similarly, male 
dichoptism is known to occur in Chrysopilus, Pelecorhynchus, and Rhagio, although 
all species sampled for this study have holoptic males. Male holoptism is the most 
common condition for these genera and is most likely the plesiomorphic state. 
 
[#11]. Eyes in male (dorsally flattened). Dorsally flattened head in males may be a 
morphological adaptation associated with swarming behavior. More dorsally-oriented 
eye surface area allows for greater vision in the vertical plane. The condition is found 
in Arthroteles, Austroleptis, Glutops, Symphoromyia, and many tabanids including 
Dichelacera and Tabanus. 
 
[#12]. Eyes in male (facets). Differential sizing of eye facets may also be associated 
with swarming behavior. All species with dorsally flattened heads have facets either 
gradually tapering in size ventrally or divided into upper and lower areas. However, 
species with dorsally rounded heads may exhibit any of the eye facet arrangements 
(scored here). Turner (pers. com.) has seen Chrysopilus and Rhagio species swarm in 
Washington state; Wood (pers. com.) has witnessed Glutops hilltopping. 
Symphoromyia is also a well known swarmer (refs). 
 
[#13]. Parafacials in male. In Dichelacera and Tabanus, parafacials are lightly 
protruding. It is also a distinctive feature of Desmomyia, Pelecorhynchus and 
Glutops. Some, but not all Symphoromyia species also have swollen parafacials. 
 
 
 70 
 
[#14]. Occiput. The concave occiput present in members of Tabanidae is distinctive. 
Pelecorhynchus has a similar head shape. 
 
[#15]. Head width. Teskey used the width of the head in relation to the thorax to 
characterize Glutops (1970: 1171). Most taxa surveyed here have a head that is 
approximately the same width as the thorax, although there is variation among 
species within genera (Chrysopilus, Rhagio, Spaniopsis) and even within species 
(Chrysopilus quadratus, Spaniopsis clelandi, Symphoromyia cruenta). The 
differences between approximately the same width and wider may be subtle, as in the 
case of Pseudoerrina jonesi (where the state is scored as head wider than thorax). 
Adult mouthparts. 
Nagatomi & Soroida (1985) carried out an exhaustive survey of the mouthparts of 
orthorrhaphous Brachycera and concluded that the adult mouthparts are of little value 
to help solve phylogenetic problems (1985, p. 304). The mouthparts show a high 
degree of plasticity; there are some trends, but no unambiguous changes were found 
to confirm known phylogenetic placements. It appears that mouthpart morphologies 
are generally more indicative of feeding behavior than common ancestry. Blood 
feeders Symphoromyia and Spaniopsis, for instance, share more characters with the 
Tabanidae and Athericidae than with the rest of the putative close relatives within the 
Rhagionidae or Spaniidae. 
 
[#16]. Labellum (not used in phylogenetic analysis). Some rhagionids, such as 
Sierramyia chiapasensis, have a distinctively large labellum in relation to the size of 
 
 71 
 
their palps. Arthroteles species exhibits the other extreme case, in having a much 
reduced labellum (despite its long proboscis). This character was scored for both 
sexes, when both were available, unless otherwise noted. A high amount of variation 
of this character within several species and genera suggests that labellum size is a 
fairly labile character, subject to shifting evolutionary pressures associated with 
feeding habit and/or habitat. The measure of labellum size itself is problematic since 
palp size, and other reference measures of the head, may vary independently. This 
may result in a large labellum being scored as short, for instance, in cases where the 
palps are especially long. For this reason, and for the apparent high rate of change of 
this character, it was removed from the phylogenetic analysis.  
 
[#17]. Pseudotrachae. Bolbomyia, Litoleptis, and Lampromyia are the only taxa 
surveyed that did not have pseudotracheae. Austroleptis has distinctive 
pseudotracheae that have taenidia-like ribbing. Other taxa, such as Arthroceras show 
some transverse ribbing of the pseudotracheal channels, but none to such a strong 
degree as in Austroleptis. Variation of this kind in the pseudotracheae, however are 
difficult to score discretely. Therefore, only the presence or absence of 
pseudotracheae was scored for phylogenetic analysis. 
 
[#18]. Theca (elongation). The theca is the sclerite at the ventrobasal area of the 
labellum. The length of the theca varies considerably among taxa, however two states 
were most readily apparent and easily scored. Where the theca is clearly longer than 
wide (such as in Arthroteles, Austroleptis, Dichelacera, and Lampromyia, among 
 
 72 
 
other taxa), the character is scored as elongate. The theca is scored as short, where as 
wide as long or just slightly longer than wide. Most Dasyomma species are as D. 
atratulum (scored here) in having a short theca, however in at least one Dasyomma 
species (D. coeruleum), the theca is elongate. Similarly, Symphoromyia species may 
either have short or long thecae. It is divided medially into two sclerites in most basal 
brachycerans, however it is entirely fused as a single sclerite in Atherix, Dasyomma, 
Dichelacera, and presumably in other Tabanidae. Atrichops has a divided theca, and 
therefore, having a fused theca is not an unambiguous synapomorphy for the 
Athericidae + Tabanidae. 
 
[#19]. Theca (lateral sclerite composition; alternate scoring not used for 
phylogenetic analysis). Variation is present in the way in which the theca is 
composed. The theca is often divided medially by a central suture. In this case, the 
lateral thecal surfaces are touching along the entire length medially, or are curved ")(" 
and showing some separation. At first, these states were scored independently (as 
currently scored). However both states are present within Atherimorpha and 
Spaniopsis, suggesting that the character was divided too finely and consequently, the 
two states were then lumped together (and scored in following character). This 
character is most easily viewed after the head is dissected. Therefore, there are a 
number of taxa, particularly the rare ones on loan, that are not scored for this 
character, since the dissection of the head was not carried out. This character is not 
included in the phylogenetic analysis. 
 
 
 73 
 
[#20]. Theca (lateral sclerite composition). Characters of the theca were rescored so 
that the character states were more discrete, and therefore, more easily assessed. After 
looking at the full complement of specimens available, states of the theca seemed 
more naturally divided into three states; lateral sclerites separate, lateral sclerites 
adjacent and/or separated by medial suture, or fused into single sclerite, without 
medial suture. 
 
[#21]. Palp segment number. The reduction in palp number from three to two is a 
synapomorphy for Brachycera. The number of palp segments varies across the 
Tabanomorpha, however, and it does not appear to be an especially conserved 
feature.  
 
[#22]. Palps (relative length of each segment). For genera with two palps, the length 
of palp segments relative to one another is fairly consistent among congenerics. Most 
genera have distal segment longer than proximal segment, however, the proximal 
segment is longer than the distal segment in Tabanus, Bolbomyia, Vermileo and 
Austroleptis. There is also some variation within species. Bolbomyia nana may either 
have distal segment longer than proximal segment, or each segment approximately 
the same length. In Glutops species, the segments are approximately the same length 
in the females, whereas in the males, the proximal segment is longer than the distal 
segment. There is also sexual dimorphism of this character in Xylophagus lugens. In 
this species, the proximal segment is longer than the distal segment in males whereas 
in the female, the reverse is true. 
 
 74 
 
 
[#23]. Lateral ridge of clypeus. The presence of a ridge along the lateral margin of 
the clypeus is evident in Coenomyia ferruginea, Arthropeas americana, and Dialysis 
rufithorax.  
 
[#24]. Location of oral margin. The location of the mouthparts, relative to the rest of 
the head, may be variously placed. For example, the oral margin clearly confines the 
mouthparts to the anterior part of the head in Coenomyia ferruginea. Nagatomi 
(1990) mentioned this as a defining character for the Coenomyiidae. Ultimately, 
however, discrete differences of states between species across the entire data matrix 
could not be discerned and therefore, this character was not used in the phylogenetic 
analysis. 
 
[#25]. Position of stipes. Where present, the position of the stipe was scored as 
converging toward one another, directed medially; attached to or in contact with 
tentorium at the back side of the head; or directed posteriorly. The difference between 
medially and posteriorly directed was sometimes subtle and several genera exhibit 
both of these states (Rhagio and Atherimorpha). Dichelacera marginata, Glutops 
rossi, Lampromyia canariensis, Tabanus atratus, and Vermileo vermileo were scored 
as having the stipe attached to or in contact with the tentorium. Stipe shape was also 
scored, but after scoring five genera (Atherimorpha, Dichelacera Ptiolina, 
Spaniopsis, and Symphoromyia), five different states were generated. It was clear the 
 
 75 
 
shape was not only variable, it was difficult to describe accurately. For this reason, 
scoring of stipe shape was halted and not used in the phylogenetic analysis. 
 
[#26]. Cardo. The cardo in Pelecorhynchus personatus is conspicuously inflated. The 
cardo is also enlarged in Pseudoerrina jonesi and Symphoromyia cruenta. This 
structure is apparently absent in Dasyomma atratulum, Dialysis rufithorax, 
Chrysopilus species and Suragina concinna. 
 
[#27]. Lacinia length. Lacinia were scored relative to the length of the palps. This 
character was much easier to score when the head was dissected. In undissected 
specimens, the lacinia were sometimes visible, however, most often the lacinia were 
concealed and no scoring was possible. Since the heads of rare species were not 
dissected, there are a number of taxa unscored for this character. Pseudoerrina, 
Tabanus, Vermileo, Lampromyia, Suragina, and Dialysis were scored from 
illustrations (Nagatomi and Soroida, 1985). 
 
[#28]. Lacinia serration. This character was scored for undissected specimens where 
the lacinia were visible, however, most often the lacinia were concealed. A number of 
species representing rare genera are not scored. Serrated lacinia may be associated 
with blood feeding. It is present in known blood feeders Symphoromyia, Spaniopsis, 
and Dasyomma, however it is also present in Sylvicola and the limited taxon sampling 
of this character makes it difficult to assert this correlation with confidence. It is not a 
 
 76 
 
required feature for blood feeders, in any case, as Tabanus and Dichelacera 
apparently do not have serrated lacinia. 
 
[#29]. Mandibles. Mandibles are required for blood feeding. Not surprisingly, 
mandibles are present in all the blood feeding flies (Atherix, Dasyomma, Dichelacera, 
Spaniopsis, Suragina, Symphoromyia, and Tabanus) and are absent in all non-blood 
feeding flies. 
 
[not scored] Epipharynx and hypopharynx. Epipharynx is a straw-like appendage 
that varies in length somewhat, but not within discrete units of measurement. 
Therefore, no attempt was made to score this feature for phylogenetic analysis. 
Generally, the epipharynx form is canal-shaped, with a blunt apical opening. The 
hypopharynx subtends the epipharynx and is complementary in form. The epipharynx 
of Bolbomyia nana is unusual in that it comes to a sharp point.  
 
[not scored] Cibarial pump (general form). The cibarial pump is an internal 
structure, located ventrad of the frons, connected to the base of the epipharynx 
("basipharynx" of Peterson, 1916). Variation in form varies from nearly spherical as 
in Bolbomyia nana and Austroleptis multimaculata, to nearly parallel sided and 
elongate as in species of Chrysopilus and Atherimorpha. Scoring of cibarium shape is 
difficult because the variations differ subtly, along a spectrum of possibilities. 
Generally, outgroups have more compact, rounded cibarial pumps. In both Spaniopsis 
and Symphoromyia, near the posterior margin of the cibarial pump on its ventral 
 
 77 
 
surface, there is a ?c? shaped carina, opened anteriorly. Illustrations of Austroleptis 
collessi and tabanids Nagatomyia, Silvius, and Chrysops in Nagatomi and Soroida 
(1985: p335; fig.202) show a carina in the same location, but upon inspection, they 
are of a different form. Although the presence of this carina is noted here, it was not 
scored for phylogenetic analysis because it is a subtle character likely subject to 
scoring error.  
 
[#30]. Cibarial pump length. The cibarial pump is located internal to the clypeus 
and may either be relatively short or long. Since the width of the cibarial pump is held 
relatively constant, the length is defined by its proportion to the width. Cibarial pump 
length is scored as short when the width is the same as or greater than the length. 
When clearly longer than wide, the cibarial pump length is scored as long. 
Interestingly, the smallest taxa scored for this feature (Austroleptis, Bolbomyia, 
Litoleptis, Ptiolina, and Spania) all have short cibarial pumps. It is possible that the 
physical forces associated with scaling have an effect on this character. However, 
sampling for this character was limited. Spaniopsis clelandi, a small to moderately 
sized fly (appx. 4?5.5 mm) also has a short cibarium. 
 
[#31]. Cornu length. The cornu are the dorsolateral extensions of the cibarial pump. 
Cornu are generally wispy and elongate, about as long as the cibarial pump. The 
length of the cornu was scored relative to the length of the cibarial pump. Since the 
length of the cibarial pump is free to vary independently, there is the possibility that a 
long cornu may be scored as short due to the presence of an especially long cibarial 
 
 78 
 
pump (and vice-versa). However, since there are probably some allometric constraints 
of the cornu with respect to the cibarial pump, a particularly long cornu should be 
revealed by such scoring. Indeed, taxa with short cibarial pumps did not, as a 
consequence, automatically have a long cornu. Bolbomyia nana is the only taxon with 
a short cibarial pump and long cornu. Taxon sampling, however, was limited. 
 
[#32]. Cornu attachment. Typically, the cornu extends freely, aside the pharyngeal 
pump. In Spaniopsis species, however, the cornu is fused to the pharyngeal pump. 
Nagatomi and Soroida (1985) illustrate Atherix ibis and Suragina caerulescens as 
having a fused cornu also, however I did not see this in any of the athericids over the 
course of my study. All other taxa examined have cornu that extend beyond the 
pharyngeal pump, in line with the cibarium. Austroleptis is autapomorphic in having 
apically microsetose cornu, however this was not scored for the phylogenetic 
analysis. 
 
[#33]. Pharyngeal pump width. The pharyngeal pump is located internally, its 
anterior arm bound by the dorsal cornu of the cibarial pump. The pharyngeal pump 
takes two basic forms. One is displayed in Chrysopilus, Rhagio, Ptiolina, Bolbomyia, 
and other genera, including outgroups, where the pharyngeal pump is stretches 
between the dorsal cornu at its anterior-most extent, but for most of its length, is 
narrow. Alternatively, the pharyngeal pump may be anteriorly broad then narrowed 
posteriorly (as a bicycle saddle, turned backwards). The pharyngeal pump of 
 
 79 
 
Arthroceras pollinosum is nearly the intermediate of these two forms. It is scored as 
anteriorly broad. 
 
[#34]. Pharyngeal cup (presence of concavity). Where anteriorly broad, the 
pharyngeal pump is either mostly flat or with margins upturned to form a cupped 
structure. All blood feeders and a few other taxa cupped have a pharyngeal pump that 
for most or all of their length, is upturned at its margins. Interestingly, the 
pelecorhynchids show a broad, cupped pharyngeal pump form. Is this due to shared 
ancestry with the Tabanidae and Athericidae? There are some inconsistencies in the 
character distribution, however, as Spania, a putatively close relative to 
Symphoromyia and Spaniopsis, does not have a similarly shaped broad, cupped 
pharyngeal pump. Arthroceras, a non-blood feeder, also has a broad, cupped 
pharyngeal pump. 
 
[#35]. Pharyngeal pump length. The length of the pharyngeal pump was scored 
relative to the length of the cibarial pump. The cibarial pump is free to vary and thus 
the relative metric may not be constant across taxa. There may also be problems with 
non-independence between the length of the pharyngeal pump and the length of the 
cibarial pump. A long cibarial pump (for sucking) may require a long pharyngeal (for 
swallowing). States are 1) longer than length of cibarial pump 2) approximately same 
length as cibarial pump 3) approximately half length of cibarial pump 4) very short, 
reduced. 
 
 
 80 
 
Adult thorax. 
[#36]. Dorsum vittae. Vittae are broad longitudinal stripes on the dorsal surface of 
the thorax. There is considerable variation of this character within Tabanidae, as well 
as within many genera (Arthroceras, Coenomyia, Glutops, Ptiolina, Rhagio, 
Spaniopsis and Suragina). There is also variation within species. In Arthroteles 
bombyliiformis, for example, vittae are more visible in females than in males. In 
Austroleptis multimaculata (and other Austroleptis species), vittae are only present in 
females. In Coenomyia ferruginea, specimens may be either vittate or not, regardless 
of gender. 
 
[#37]. Dorsum setation. Some species of Atherimorpha are distinguished by having 
rows of dorsal setae that are stronger and longer than other setae of the dorsum. In 
Austroleptis collessi, dorsocentral bristles are lightly present. All other taxa surveyed, 
however, have dorsal setae of approximately the same length. 
 
[#38]. Presence of scale-like colored setae. Flattened, scale-like setae reflecting 
blue, red, or golden colors and are present in Chrysopilus, Schizella, and apparently in 
Stylospania (although it is difficult to determine). Pelecorhynchus may have similarly 
colored setae, but they are not flattened and are of a different nature. 
 
[#39]. Proepimeron presence. The proepimeron is variously developed among 
genera within Tabanomorpha, yet consistent within genera. The proepimeron is 
located posterior of the proepisternum and anterior of the anepisternum, ventral to the 
 
 81 
 
area containing the anterior spiracle. Sometimes the proepimeron is difficult to locate, 
on account that it may be separated from the proepisternum and the anepisternum by 
a discrete, superficial suture. In Xylophagus, the proepimeron is apparently absent. In 
most taxa (Arthroceras, Atherimorpha, Atherix, Austroleptis, Bolbomyia, and others), 
the proepimeron is roughly rectangular, approximately twice as long (dorsoventrally) 
as wide (anterior-posteriorly). In Chrysopilus, Schizella, and Stylospania, however, 
the proepimeron is reduced and slender (as in Fig. 175). 
 
[#40]. Proepimeron setation. The proepimeron may be with or without setae. Rhagio 
and Desmomyia are the only Rhagionidae sensu lato with setose proepimera. Other 
genera with setae-bearing proepimera are Arthropeas, Atherix, Coenomyia, 
Dichelacera, Suragina, and Tabanus. 
 
[#41]. Anepisternum setation (alternative coding not included in phylogenetic 
analysis). As indicated by the large number of character states for this feature, there 
is a considerable amount of variation as to the degree in which the anepisternum is 
setose. This character may be better scored in two characters, presence of setae and 
the nature of these setae. This is a preliminary scoring that is modified below (not 
included in phylogenetic analysis). 
 
[#42]. Anepisternum setation (alternative coding not included in phylogenetic 
analysis). Originally, I coded several more character states, detailing various setal 
patterns on the anepisternum (setose on upper margin only; setose on dorsal and 
 
 82 
 
posterior margins; setose throughout posterior half; setose throughout upper half 
including entire anterior and posterior margins). However, there is a great amount of 
variability within genera and I think lumping these characters together is more 
informative than a fine-tuned approach, especially considering that the objective is to 
understand very old divergence patterns among these taxa. Not to mention, a more 
generalized approach is easier to score. This character may be ordered. If so, since 
there are four character states, some downweighting is appropriate. This is a 
preliminary scoring that is modified below (not included in phylogenetic analysis). 
 
[#43]. Anepisternum setation. There is a considerable amount of variation as to the 
patterns and degree in which the anepisternum is setose among the taxa surveyed. 
This diversity was categorized by three basic states: 1) anepisternum bare or with one 
or two setae 2) less than one-half of sclerite area with setae 3) one half or more of 
sclerite area with setae. Although all Rhagio species scored here have at least a few 
anepisternal setae, the anepisternum is bare in Rhagio maculatus. This may be 
important, as a bare anepisternum is one of the putative autapomorphies of Rhagina, 
setting it apart from Rhagio. Ptiolina may either have a bare anepisternum (P. zonata, 
P. mallochi) or an anepisternum with approximately half its surface covered with 
setae (P. lappomica, P. majuscula). Specimens of Spaniopsis longicornis may have 
the anepisternum either bare or setose. 
 
[#44]. Laterotergite form. The laterotergite is the thoracic area ventrolateral to the 
subscutellum. The laterotergite may be subdivided into dorsal (anatergite) and ventral 
 
 83 
 
(katatergite) areas. The laterotergite may either be smooth and evenly surfaced or the 
katatergite may be swollen, and thereby differentiated from the anatergite. The 
anatergite and katatergite may also be defined by a faint, superficial medial suture (as 
in some Rhagio species and Sierramyia chiapasensis). The katatergite was scored as 
swollen and differentiated from the anatergite only in cases where the katatergite was 
clearly swollen (when viewed from the lateral perspective), otherwise the laterotergite 
subdivisions were scored as indistinguishable. In many taxa (Chrysopilus, Glutops, 
Spaniopsis, Symphoromyia), a slight depression of the laterotergite is present 
medially (apparently to allow freedom of movement for the halter). In these cases, the 
condition is scored as ana- and katatergites indistinguishable. In Symphoromyia 
cruenta, the katatergite is swollen in the female, but indistinguishable from the 
anatergite in the male. 
 
[#45]. Laterotergite setation. The presence or absence of laterotergite setae is a 
major character for taxonomic keys used to distinguish genera of the Tabanomorpha. 
 
[#46]. Laterotergite setal arrangement. Laterotergite setae are generally distributed 
in three different ways. Setae may be confined exclusively to the katatergite, found 
mostly on the katatergite and extending partially onto the anatergite, or found 
throughout all parts of the laterotergite. Arthroceras species, Chrysopilus species, 
Pseudoerrina jonesi, Schizella species, Arthropeas americana, Coenomyia 
ferruginea, and Dialysis rufithorax have setae distributed through the laterotergite. 
 
 84 
 
Species of Athericidae, Tabanidae, as well as those belonging to Desmomyia, Rhagio, 
Sierramyia, and Symphoromyia have laterotergite setae restricted to the katatergite. 
 
[#47]. Microsetation between base of halter and postspiracular scale. The thoracic 
surface between the base of the halter and the postspiracular sclerite is very small, 
and the microsetae that may be present in this area are often inconspicuous. 
Especially in very small specimens, it is easy to miscode this character as microsetae 
absent. For this reason, the character is not used in phylogenetic analysis. 
 
[#48]. Flaps of posterior thoracic spiracle. The upper margin of the posterior 
thoracic spiracle may be produced in the form of a flap that apparently may be used to 
close the spiracle airway. The presence of flaps on the thoracic spiracle may be an 
adaptation to prevent desiccation, particularly for large flies. It is worth noting, 
however, that the large fly Coenomyia ferruginea does not have spiracular flaps. The 
margins of the thoracic spiracle are sometimes raised (as in Chrysopilus ferruginosus 
and Chrysopilus thoracicus) in these cases, the spiracle is scored as not having flaps. 
The presence of flaps was scored as such only when it was clear the length of the flap 
spanned the spiracle length (and could thus close the airway). 
 
[#49]. Posterior thoracic spiracle lining. The presence or absence of microsetae 
lining the margin of the thoracic spiracles is likely a morphological response to 
particular environmental conditions. Genera with large ranges such as Chrysopilus 
and Rhagio exhibited infrageneric variation in this character. Atherimorpha is also 
 
 85 
 
polymorphic for this character. However, all examined species of Arthroceras, 
Arthroteles, Ptiolina, and Spaniopsis (which comprise a greater sampling than 
included here) have bare thoracic spiracle margins. 
 
[#50]. Postspiracular scale. The presence of a postspiracular scale is a putative 
synapomorphy for Tabanidae and Athericidae (Stuckenberg, 1973). Species of 
Pelecorhynchus also have this character state. In Pseudoerrina jonesi, the 
postspiracular scale is present, but reduced to a linear ridge. It may be present or 
absent species of Glutops (it is usually present). When present in Glutops, it is a 
linear ridge, similar in form to the scale in Pseudoerrina jonesi. The postspiracular 
sclerite is also broadly raised in species of Rhagio, although with a different 
orientation than what is found in the Tabanids, Athericids, and Pelecorhynchids (Figs. 
175-177). 
 
[#51]. Postspiracular sclerite setation. The postspiracular sclerite is setose only 
among xylophagid taxa. A possible exception is Exeretonevra and Heterostomus (not 
included in this matrix), where although the postspiracular sclerite is setose, setae are 
found only on the dorsal margin of the postspiracular sclerite. 
 
[#52]. Setation posterior to postspiracular sclerite. In some taxa, there is a small, 
isolated tuft of setae conspicuously present posterior to the postspiracular sclerite, on 
what may be called the metanepisternum, between the halter and hind coxa. This 
feature is conspicuously present in species of Pelecorhynchus. It is also present in 
 
 86 
 
species of Arthroceras, Arthropeas americana, some (but not all) Chrysopilus, 
Coenomyia ferruginea, Glutops species, Pseudoerrina jonesi, Symphoromyia species, 
and Xylophagus lugens. 
 
[#53]. Proscutellum. Stuckenberg (2001) has asserted that the presence or absence of 
the proscutellum has phylogenetic significance at the generic level. As he describes it, 
the proscutellum is an arcing suture that creates a lenticular-shaped swelling at the 
posterior edge of the mesoscutum. In Tabanus atratus, for example, it is easily seen. 
It appears to be polymorphic in Ptiolina mallochi, Spania nigra, and Vermileo 
vermileo. It is clearly present in Rhagio costatus, however, absent in other Rhagio 
species. 
 
[#54]. Subscutellum form. In species of Atherix, Dasyomma, Dichelacera, 
Pelecorhynchus, Pseudoerrina, and Tabanus have a raised, swollen area of the 
subscutellum is visible. In all other taxa surveyed, the subscutellum is smooth. 
 
[#55]. Subscutellum setation. In the xylophagid taxa, the subscutellum is setose, at 
least within the lateral margins. 
 
[#56]. Postmetacoxal bridge (not used in phylogenetic analysis). In order to see 
this character, the entire specimen must be cleared and the hind coxa removed. 
Stuckenberg (2001: 26) points out that ?in Tabanidae, the epimera are extended 
ventromedially to form either a strengthened area in the median ventral membrane or 
 
 87 
 
are joined to form a postmetacoxal bridge. In Athericidae, the modification is present 
but less visible, as a thin extension of the epimeron.? This character is difficult to 
score and requires destructive dissection. Rare taxa were not scored. This character is 
not used for the phylogenetic analysis.  
 
#57. Postmetacoxal bridge type (not used in phylogenetic analysis).  There are 
four forms of the postmetacoxal bridge; a complete broad extension, an incomplete 
broad extension, a. complete thin extension, or an incomplete thin extension. How 
these postmetacoxal bridge forms are related to one another is unclear. Since this 
character is difficult to score reliably, it is not used for the phylogenetic analysis.  
 
Wing. 
[#58]. Wing color (not used in phylogenetic analysis). Although most species are 
constant with respect to the color of the wing surface, this is a highly variable 
character among species within the same genus. It is also problematic to score in that 
degrees of lightness/darkness are vague and subjective. Furthermore, such a character 
is unlikely to retain phylogenetic information for resolving deep lineage divergences 
among the lineages. For these reasons, this character was not included in the 
phylogenetic analysis.  
 
[#59]. Wing markings (not used in phylogenetic analysis). The drawbacks of the 
previous character are, for the most part, similar to the previous character. Most 
species are constant with respect to the color of the wing markings, however the 
 
 88 
 
character is variable among species within the same genus. And just as in wing 
surface color, color patterns are unlikely to retain phylogenetic information for 
resolving deep lineage divergences among the lineages. For these reasons, this 
character was not included in the phylogenetic analysis.  
 
[#60]. Pterostigma (not used in phylogenetic analysis). Stuckenberg (2001:16) 
notes that the curvature of R
2+3
 is associated with the presence of a pterostigma within 
the Rhagionidae sensu Stuckenberg (2001). According to Stuckenberg, the 
prominence of the curvature was related to the size of the pterostigma. In some 
species of Chrysopilus, for instance, the curvature was strong when the pterostigma 
was large. The curvature of R
2+3
 that Stuckenberg points out, however, is very subtle 
(2001:17, fig. 11). I scored the presence or absence of the pterostigma as one part of 
Stuckenberg's assessment. The presence or absence of a pterostigma was judged on 
the amount of coloration between the apical area of R
1
 and R
2+3
. If there is any 
coloration at all, the pterostigma is scored as present. As with all characters scored by 
color, it seems unlikely that information regarding ancient evolutionary divergences 
may be retained. This is character is not used for the phylogenetic analysis.  
 
[#61]. Lower calypter. An invagination indicates the separation between upper and 
lower calypter. In the Tabanidae, the lower calypter is well developed. In 
Pelecorhynchus personatus, the lower calypter is present, but not as large as in 
Tabanus, Dichelacera and other tabanids. In all other taxa, the lower calypter is fully 
reduced. 
 
 89 
 
 
[#62]. Upper calypter development. The upper calypter is the membrane subtending 
the base of the wing. In Tabanidae, the upper calypter is smaller than the lower 
calypter and in Tabanus atratus, it appears absent. In Coenomyia ferruginea, Dialysis 
rufithorax, Lampromyia canariensis, Sierramyia chiapasensis, Vermileo vermileo, 
and Xylophagus lugens, the upper calypter is underdeveloped to some degree, in 
comparison to the other taxa which a have a fully rounded upper calypter. In these 
cases, where the upper calypter has a mostly straight margin, the character is scored 
as underdeveloped. 
 
[#63]. Upper calypter form (not used in phylogenetic analysis). This is a 
reformulated version of the previous character. It is not discrete, however, and not 
used for the phylogenetic analysis.  
 
[#64]. Alula development. The lack of alula is very evident in species of Sierramyia 
and among vermileonid taxa. However there is a gradation between having a fully 
reduced alula and having one that is fully developed. For the most part, however, 
most congenerics are consistently one type or another. The exception to this is 
Rhagio, where species may either have a narrow or broad alula area. Specimens with 
absolutely no convex curvature were obviously scored as reduced, with no curvature. 
Narrow curvature indicates curvature in the alula that is three times as wide as deep, 
or greater. Broad curvature indicates curvature of the alula which is less than three 
times as wide as deep, or less. These differences were scored by eye. 
 
 90 
 
 
[#65]. Alula shape. The alula assumed one of two forms. The tabanids, for example, 
show a marked posterior shift of alula area distally, so that the alula area is nearly 
triangular. If the alular area is viewed as a distribution curve, the mean is located near 
the distal margin. This state is also seen in species of Glutops, Pelecorhynchus, 
Pseudoerrina, Spania, and Spaniopsis. The alula is scored as rounded evenly where 
the alular area is rounded, so that the mean of its distribution is at or near the center. 
 
[#66]. Anal lobe (not used in phylogenetic analysis). The development of the anal 
lobe may be correlated with the reduction of the alula. In all cases where the anal lobe 
is scored as reduced (Lampromyia canariensis, Sierramyia chiapasensis, and 
Vermileo vermileo), the alula is also reduced. However, while the anal lobe is well 
developed in species of Xylophagus, these species also have a reduced alula. 
Although the scoring of this character doesn't overlap perfectly with the scoring of the 
alula, I suspect the two characters are constrained by non-independence with respect 
to one another. For this reason, the scoring of the development of the anal lobe was 
not included in the phylogenetic analysis.  
 
[#67]. Humeral crossvein strength (not used in phylogenetic analysis). The 
humeral crossvein (h) was scored for relative strength. The humeral crossvein may be 
very faint, or present as a thick vein. However, over the course of scoring this 
character, it became apparent that it is variable even at the finest taxonomic scale 
(among specimens of the same species) and therefore it became obvious that such a 
 
 91 
 
character is incapable of elucidating the branching pattern of ancient divergences. For 
this reason, it was removed for the phylogenetic analysis. 
 
[#68]. Humeral crossvein shape (not used in phylogenetic analysis). The humeral 
crossvein (h) may be variously oriented or shaped across taxa. It may be arced (or 
'bow-shaped') or straight and oriented acutely, obtusely, or perpendicularly with 
respect to the leading margin of the wing. However, as was the case for scoring the 
humeral crossvein for relative strength, it became apparent that the high variability of 
this character among conspecifics and congenerics suggests that it is not appropriate 
for use in reconstructing the phylogeny of the ingroup. There is also a large amount of 
subtlety to this character that may induce scoring error. For this reason, it was 
removed for the phylogenetic analysis.  
 
[#69]. Sc-r crossvein presence (not used in phylogenetic analysis). The Sc-r 
crossvein was sometimes present as only a concentration of setae. In this case, it was 
scored as weakly developed, even though its position was clear. There is a lot of 
variation in this character. Schizella is the only taxon where the Sc-r crossvein is 
always absent, however this may be an artifact of a small sampling of this genus (6 
specimens, 3 species). Given the variation of this character, it appears unlikely to be 
informative at the level necessary for revealing ancient evolutionary events. For this 
reason, it was removed for phylogenetic analysis. 
 
 
 92 
 
[#70]. Sc-r crossvein location. The Sc-r crossvein was identified as a potentially 
useful character for phylogenetic analysis by Stuckenberg (2001). There is slight 
variation of this character within genera that results in a difference of scoring among 
congenerics (as in Atherimorpha and Chrysopilus); some of which is related to sexual 
dimorphism (Coenomyia ferruginea). The most notable divergence is present in 
species of Lampromyia and Vermileo, where the crossvein is placed approximately 
midway between h and the origin of the radial-sector. Litoleptis and Sylvicola are also 
autapomorphic in having sc-r located on the proximal side of the humeral crossvein. 
 
[#71]. Setation of R
1
, dorsal surface. The presence of setae on dorsal surface of 
wing vein R
1
 is the common condition for most lower flies. Litoleptis, Lampromyia, 
and Vermileo, however, are exceptions to this. 
 
[#72]. Setation of R
1
, ventral surface (not used in phylogenetic analysis). The lack 
of setae on ventral surface of wing vein R
1
 is the common condition for most lower 
flies. Ventral R
1
 setae are found in some, but not all species of Atherimorpha, 
Austroleptis, and Xylophagus however. Due to the limited taxon sampling of 
Austroleptis and Xylophagus, the morphological diversity of these genera is not fully 
represented in the matrix. It is a logical to assume that the plesiomorphic condition is 
for the ventral R
1
 wing vein to lack setae, in each of these genera (Atherimorpha, 
Austroleptis, and Xylophagus). In order to avoid misinterpretation of the data due to 
limited taxon sampling, this character was not used in phylogenetic analysis. 
 
 
 93 
 
[#73]. Orientation of wing veins R
1
 and R
2+3
 at wing margin. The position of R
2+3
 
at the wing margin is scored, relative to the position of wing veins R
1
 and R
4
. A 
putative synapomorphy for members of the Athericidae is for wing veins R
1
 and R
2+3
 
to meet together at the wing margin (Stuckenberg, 1973). Yet the members belonging 
to the putative sister taxon of Athericidae, the Tabanidae, have R
2+3
 closer to R
4
 than 
to R
1
. Species of Glutops, Lampromyia, Pelecorhynchus, Pseudoerrina, and Vermileo 
also have a distally positioned R
2+3
 wing vein. Between these two extremes, the R
2+3
 
wing vein may be positioned near R
1
 (as in Arthroceras, Chrysopilus, Desmomyia, 
some Ptiolina, Schizella, Sierramyia, Spania, Spaniopsis, Symphoromyia, and the 
Xylophagidae) or displaced away from R
1
, nearly between R
1
 and R
4 
(but closer to R
1
 
than to R
4
) as in species of Arthroteles, Atherimorpha, Bolbomyia, and some species 
of Ptiolina. 
 
[#74]. Curvature of R
2+3
 (not used in phylogenetic analysis). Stuckenberg 
(2001:16) notes that the curvature of R
2+3
 is associated with the presence of a 
pterostigma within the Rhagionidae sensu Stuckenberg 2001. According to 
Stuckenberg, the prominence of the curvature was related to the size of the 
pterostigma. In some species of Chrysopilus, for instance, the curvature was strong 
when the pterostigma was large. The curvature of R
2+3
 that Stuckenberg points out, 
however, is exceptionally subtle (2001:17, fig. 11). I scored this feature to the best of 
my abilities. However in many cases, such as in Symphoromyia species, many 
Chrysopilus species and even within some Rhagio species such as R. plumbeus, it 
could have been scored faithfully either as present or absent. How much of a curve 
 
 94 
 
needs to be present in order for it to be scored as such? As a landmark, I used figure 
11 showing Rhagio scolopaceus (Stuckenberg, 2001) as the exemplary condition 
where the curve is present. If the curve was not adjacent to the position of the 
pterostigma, as in Rhagio scolopaceus and as explained by Stuckenberg, the character 
was scored as absent. I also used Spania nigra (type species of Spaniidae sensu 
Stuckenberg) as the exemplar condition where the curve is not present. Stuckenberg 
notes that the portion of R
2+3
 apical to the pterostigma does not include the sinuous 
curve that he refers to as the character. However, it appears to me that the most 
substantial difference between the R
2+3
 in Spania nigra and that in Rhagio 
scolopaceus is the portion of the wing vein distal to the pterostigma. In Spania nigra 
(as in Suragina concinna, Austroleptis species, Chrysopilus quadratus (and other 
species including Chrysopilus thoracicus and Chrysopilus panamensis) and 
Dasyomma atratulum), R
2+3
 is virtually identical to Rhagio scolopaceus in its basal 
curvature. Only when the apical portion of the wing vein veers slightly back towards 
the wingtip (and not directly towards the leading margin of the wing) does the 
curvature mentioned by Stuckenberg become readily apparent. In one specimen of 
Desmomyia thereviformis [USNMENT00025269], the wings differ in such a way that 
the left wing should be scored as R
2+3
 curvature present, while in the right wing it is 
clearly absent. 
 
[#75]. R
2+3
 at wing margin (not used in phylogenetic analysis). The curvature of 
the R
2+3
 wing vein apical of the pterostigma is highly correlated with the presence or 
absence of the special curvature of R
2+3
 pointed out by Stuckenberg (2001). In 
 
 95 
 
Ptiolina lapponica and Ptiolina majuscula, the curve is scored as absent even though 
the R
2+3
 wing vein meets the wing margin at a close angle because of the placement 
of the pterostigma (it is not positioned inside the point of curvature, a necessary 
component of Stuckenberg's scoring). This character is not used for phylogenetic 
analysis for reasons given in the discussion of the previous character.  
 
[#76]. Setation of dorsal surface of R
2+3
 (not used in phylogenetic analysis). The 
presence or absence of microsetae on the dorsal surface of wing vein R
2+3
 was scored. 
The apomorphic condition is the presence of microsetae, although this is present in 
only some species of Atherimorpha, and in Sierramyia chiapasensis. Sierramyia is 
the only genus-level taxon that is unambiguously scored as having microsetae on 
R
2+3
. The genus is represented in this study by a single specimen, however. 
Bolbomyia nana has a bare dorsal R
2+3
, however in its congener, B. wuorentausi, it is 
microsetose. The variable nature of this character suggests that it is not useful in 
deep-level phylogenetic analysis, and for this reason, it was removed. 
 
[#77]. Setation of ventral surface of R
2+3
 (not used in phylogenetic analysis). The 
presence or absence of microsetae on the ventral surface of wing vein R
2+3
 was 
scored. The apomorphic condition is the presence of microsetae. The apomorphic 
condition is found in only some species of Atherimorpha, Bolbomyia nana and in 
Sierramyia chiapasensis. Sierramyia is the only genus-level taxon that is 
unambiguously scored as having microsetae on the ventral R
2+3
 wing vein. The genus 
is represented in this study by a single specimen, however. The variable nature of this 
 
 96 
 
character suggests that it is unlikely to be informative for deep-level phylogenetic 
analysis, and for this reason, it was removed. 
 
[#78]. R
4
-R
5
 fork. Sylvicola is distinct in lacking the R
4
-R
5
 fork. Since Sylvicola is 
not used in the phylogenetic analysis, this character is non-informative. All taxa share 
the same state. 
 
[#79]. Position of R
4
-R
5
 fork. Grimaldi & Cumming (1999: 16) note that the base of 
fork R
4
-R
5
 is at the same level as the distal end of cell dm, and not distal to it in the 
Rhagionidae. This, they propose, is a potential synapomorphy for the family. The 
scoring is formulated so that it is consistent with their reasoning. Grimaldi & 
Cumming were not explicit in how this character should be scored, however, and the 
line separating proximal and distal positions can be positioned a number of ways 
(relative to wing leading margin, wing base/tip, etc.). I score this character by 
orienting the wing so that the base (the part that attaches to the thorax) is held 
horizontal to the wing tip. A vertical line could be drawn visually so that the position 
of the R
4
-R
5
 fork could be determined, relative to the distal end of cell dm. As scored, 
species of Rhagio may have the base of the R
4
-R
5
 fork proximal of (as in R. 
mystaceus), directly above (as in R. plumbeus, R. hirtus), or distal of the distal end of 
cell dm (as in R. costatus). Differences in scoring between Atherimorpha species are 
also present. 
 
 
 97 
 
[#80]. R
4
 at base. The curvature of wing vein R
4
 at its base has been proposed as a 
possible a synapomorphy for the Rhagionidae (Grimaldi & Cumming, 1999; 
Stuckenberg, 2001: 16, fig. 11). The scoring of this character is very subtle. 
Undoubtedly, many species of Chrysopilus and Rhagio exhibit an obvious and abrupt 
change of direction at the base of wing vein R
4
 (as illustrated in Stuckenberg, 2001: 
fig 11). However, the flexure at the base of wing vein R
4
 is continuously variable 
across a range of curvatures and the point at which it ceases to become 'strong' is 
unclear. For this reason, I took scored wings as strongly curved or angled, when as 
illustrated (Stuckenberg, 2001: fig. 11). An abrupt change of direction was not a 
necessary condition for this scoring, although the bend had to be severe. In Rhagio 
vertebratus and Rhagio hirtus males, the flexure was less severe than in the females. 
The character was scored as ambiguous for these species. The basal curvature of R
4
 in 
Rhagina incurvatus is much less abrupt than in species of Rhagio, however it was 
scored as strongly angled because the direction of the wing vein changes 90?. All 
other wings, with the basal curvature of R
4
 smoothly changing direction (as in species 
of Arthroteles, Atherimorpha, Atherix, Austroleptis, Bolbomyia, Dasyomma, and 
Ptiolina, among others) were scored as relaxed, not strongly curved. 
 
[#81]. R
4
 at apex. Taxa also vary in the amount of curvature of the apical portion of 
wing vein R
4
, as it meets the margin of the wing. In species of Pelecorhynchus for 
example, the R
4
 vein veers anteriorly to create a distinctive curvature. In most 
Tabanidae, the R
4
 vein also meets the margin of the wing anteriorly, however the 
curvature is less pronounced. The differences among most of the taxa, however, were 
 
 98 
 
subtle for this character. The most obvious difference, to allowing for discrete coding, 
was the orientation of R
4
 at the margin. If R
4
 is apically curved so that it is directed 
anteriorly, it is scored as curving towards the leading margin of the wing. If the apical 
portion of R
4
 was straight and directed toward the wingtip, it was scored as straight or 
nearly straight apically. In some cases, such as in Ptiolina lapponica and Rhagio 
plumbeus, R
4
 was slightly curved anteriorly at the wing margin, but not enough to be 
directed toward the leading edge of the wing. Therefore, R
4
 wing vein in these taxa 
was scored as nearly straight. In addition to Dichelacera, Pelecorhynchus, and 
Tabanus having an anteriorly-directed R
4
 wing vein, Arthropeas americana, Atherix 
pachypus, Coenomyia ferruginea, Pseudoerrina jonesi and Suragina concinna also 
exhibit this condition. Glutops rossi and Dasyomma atratulum were scored as having 
R
4
 straight (although their putative sister taxa were scored otherwise). 
 
[#82]. Position of R
4
 and R
5
 at margin. This character scores the position of R
4
 and 
R
5
 relative to the wingtip. The wing veins may be anterior to the wingtip (i.e., R
5
 is at 
or anterior to the wingtip), straddle the wingtip (i.e., R
4
 and R
5
 contain the wingtip 
between them), or be posterior to the wingtip (i.e., R
4
 is at or posterior to the 
wingtip). Intrageneric variation of this character is present in at least Austroleptis, 
Chrysopilus, Ptiolina, Rhagio and Symphoromyia. In the case of Austroleptis and 
Symphoromyia, the ambiguity at the genus level is not recorded in the matrix, due to 
limited sampling. The apparent lability of this character below the genus level 
suggests that information regarding ancient evolutionary events is likely to be lost. 
The character is retained for phylogenetic analysis, however. 
 
 99 
 
 
[#83]. Alignment of R
5
. Grimaldi & Cumming (1999: 16) note that in Rhagionidae 
"vein R
5
 is almost always straight and R
4
 arises from it with a sharp bend at its base, 
often at 90?." The first part of this statement is scored here. The latter half of this 
statement, referring to the sharp bend at its base, is scored by character number #80. 
There are inherent problems in scoring a character such as this, as in other cases 
where the amount of curvature is characterized. Variation in the flexure of R
4+5
 at the 
point where R
4
 and R
5
 originate is continuous. It is unclear at what point, exactly, 
R
4+5
 ceases to be straight, in line with R
5
. R
5
 often arises from R
4+5
 with a clear, but 
small change of direction. If the change of direction was 10? or less, the character was 
scored as straight. Sometimes, however, R
4+5
 is arced apically towards the anterior 
wing margin, as in Arthropeas americana and Coenomyia ferruginea. In this case, the 
direction of R
4+5
 may be calculated as the average vector across its length or as the 
ultimate orientation of R
4+5
 where R
4
 and R
5
 arise. Differences in how this is 
calculated will affect the scoring. Arbitrarily, I chose to measure the direction of R
4+5
 
by its approximate average vector. 
 
[#84]. M
3
 wing vein. The presence or absence of the third medial vein was scored. 
This has been proposed as a synapomorphy for a natural group composed of 
Austroleptis, Bolbomyia, and Litoleptis. This group may also include fossil taxa 
Mesobolbomyia, Pauromyia, Probolbomyia, and Zarzia (Grimaldi & Cumming, 
1999). Since it is logical to assume that an incompletely present M
3
 wing vein 
represents an intermediate step between the complete presence or absence of M
3
, the 
 
 100 
 
character may be ordered (it is not, however, in the phylogenetic analysis). The state 
of being incompletely present is defined by an M
3
 which is present, but not reaching 
the wing margin. Spania nigra specimens may have a completely or incompletely 
present M
3
 wing vein. Spaniopsis clelandi specimens also show variation in either 
having M
3
 incompletely present or absent. 
 
[#85]. Halter knob size (not used in phylogenetic analysis). The length of the halter 
knob was measured in comparison to the halter stem. Differences were arbitrarily 
limited to five separate states. States varied quite readily within genera and the 
variation of the knob/stem ratio was continuous. Since the difference between some 
states was minor and rather ambiguous (approximately 1/2 length of stem/ between 
1/3?1/2 length of stem/ between 1/2?2/3 length of stem), scoring may have been 
particularly subject to experimental error. For this reason, the character was removed 
for phylogenetic analysis. 
 
[#86]. Halter stem (not used in phylogenetic analysis). There are different 
microstructural patterns evident upon very high magnification of the base of the 
halter. SEM photos of this region in Arthroceras, Arthroteles, Bolbomyia, 
Desmomyia, and Ptiolina show these differences in detail. However there is a 
continuous gradation of states which make scoring particularly difficult. Although 
this character system was explored, it is not included in the phylogenetic analysis. 
 
 
 101 
 
[#87]. Anterodorsal setation of halter (not used in phylogenetic analysis). This is 
the first of a two-part character, examining microsetae of the halter. There are 
different microsetal patterns evident upon very high magnification of the anterodorsal 
side of the halter stem and knob. However there is a multitude of patterns and 
devising a consistent and cohesive scoring for this character is particularly 
challenging. After examination of a wide range of taxa, it appears possible such 
character data may be informative at the species level. However, it is certainly too 
variable to provide information for deeper level divergences. It is not used for the 
phylogenetic analysis. 
 
[#88]. Ventroposterior setation of halter (not used in phylogenetic analysis). This 
is the second of a two-part character, examining microsetae of the halter. There are 
different microsetal patterns evident upon very high magnification of the 
ventroposterior side of the halter stem and knob. However, just as previously noted, 
there is a multitude of patterns and devising a consistent and cohesive scoring for this 
character is particularly challenging. It is not used for the phylogenetic analysis. 
Legs. 
[#89]. Fore tibial spur. The presence or absence of the fore tibial spur was scored. 
The tibial spurs did not show any intrageneric variation; all species belonging to the 
same genus had the same tibial spur formula. Members of Arthropeas, Bolbomyia, 
Coenomyia, Dialysis, Lampromyia, Pseudoerrina, Vermileo, and Xylophagus have a 
fore tibial spur. All other taxa lack fore tibial spurs. 
 
 
 102 
 
[#90]. Mid tibial spur. The presence or absence of the mid tibial spur was scored. 
The tibial spurs do not show any intrageneric variation, all species belonging to the 
same genus have the same tibial spur formula. Litoleptis alaskensis (and its 
congeners) lack(s) a mid tibial spur. Alloleptis tersus is the only taxon that bears a 
single mid tibial spur. All other species have two mid tibial spurs. In species of 
Austroleptis, these spurs are generally shortened more than in other genera. 
 
[#91]. Hind tibial spur. The presence or absence of the hind tibial spur was scored. 
The tibial spurs do not show any intrageneric variation, all species belonging to the 
same genus have the same tibial spur formula. Species of Dichelacera, Litoleptis, 
Spania, Spaniopsis and Tabanus lack hind tibial spurs. Alloleptis tersus, and species 
of Arthroceras, Chrysopilus, Ptiolina, Schizella, Stylospania, and Symphoromyia 
have a single hind tibial spur. Species of Rhagio, Sierramyia, Desmomyia, 
Arthroteles, and Atherimorpha have two tibial spurs, as do species of Austroleptis, 
Vermileo, and members of Athericidae and Xylophagidae. 
 
[#92]. Hind coxal tubercle. The hind coxal tubercle is a small, anterior-facing 
protuberance of the hind coxa, usually visible from the antero-lateral view. It is 
present in most taxa. It is absent in species of Austroleptis, Coenomyia, Dichelacera, 
Litoleptis, Pelecorhynchus, Tabanus, Vermileo, Xylophagus, as well as in 
Pseudoerrina jonesi and Sierramyia chiapasensis. Additional tabanid species, 
belonging to Chrysops, Esenbeckia, and Haematopota were examined; these also lack 
 
 103 
 
the hind coxal tubercle. However, while absent in all tabanids, it is present in all 
athericids, including Atrichops (not in the matrix). 
 
[#93]. Hind tibial macrochaetae presence. The presence or absence of hind tibial 
macrochaetae was scored. Macrochaetae are setae that are distinguishable by their 
greater girth and length. Some macrochaetae, however, may be easily overlooked. For 
this character, macrochaetae were scored as present for all species that clearly 
exhibited hind tibial setae of two sizes on the anterior and/or posterior faces of the leg 
(the dorsal ridge of the tibia usually has thicker setae; this area was not considered for 
scoring). The macrochaetae may be short, inconspicuous, and nearly flush with the 
tibial surface, as in species of Chrysopilus, Rhagio and Arthroceras. Alternatively, 
macrochaetae may be lengthened, conspicuous, and projecting out from the tibial 
surface. Rather than divide this feature into two characters (present/absent and small, 
inconspicuous/large, conspicuous), macrochaetae were scored as a single character 
with three states. I did not wish to overemphasize the importance of having or not 
having macrochaetae since in some genera, such as Arthroteles, these setae are very 
few in number and are nearly absent altogether. The fact that some but not all Rhagio 
species have small macrochaetae also suggests that the evolutionary distance between 
the presence and absence of these setae is not particularly large. 
 
[#94]. Hind tibial macrochaetae length (not used in phylogenetic analysis). After 
thorough review of the macrochaetae, I decided to abandon this character. 
Information collected regarding this character was integrated into the previous 
 
 104 
 
character. Atherimorpha may have particularly distinctive macrochaetae; they are 
usually longer than what is found in other taxa where macrochaetae are present 
although even within Atherimorpha, this character can be quite variable. It is not used 
for the phylogenetic analysis. 
 
[#95]. Hind tibial swelling. It is reported that species of Rhagina often have a ventro-
apical swelling on the hind tibia (Yang et al., 1997: 184). The presence or absence of 
this hind tibial condition was scored. Surprisingly, neither Rhagina incurvatus nor 
another Rhagina species (indet. from Borneo) had a noticeable ventro-apical swelling 
on the hind tibia. Ultimately, no taxa were scored as having such a swelling. This 
condition appears to be a species level distinction for Rhagina sinensis Yang & 
Nagatomi. 
 
[#96]. Hind metatarsus swelling. Desmomyia is unique in having the first hind 
metatarsus of the male enlarged. 
Male Genitalia. 
[#97]. Epandrial sclerite aspect ration. The epandrial sclerite is longer than wide in 
Pelecorhynchus, Pseudoerrina, Suragina, Atherix, and Dasyomma. In all other taxa, 
the epandrial sclerite was scored as wider than long, however, this character is 
variable within the Tabanidae. In Tabanus atratus, for instance, it is wider than long, 
however in other species, such as Tabanus sulcifrons, the epandrial sclerite is longer 
than wide. The epandrial sclerite is also longer than wide in Esenbeckia incisuralis 
and Heptatoma pullecens, however, in Haematopota pechumani, Chrysops lateralis, 
 
 105 
 
and Scaptia australis, it is wider than long. In Dichelacera marginata and Hybomitra 
atrobasis the sclerite is approximately as wide as long. Dichelacera marginata was 
scored as wider than long. Spaniopsis longicornis was also scored as wider than long 
(as Spaniopsis clelandi and Spaniopsis marginipennis), but nearly as wide as long. 
 
[#98]. Anterior margin of epandrial sclerite. If the anterior emargination reaches 
near the midline of the epandrial sclerite, it is scored as strongly notched. Otherwise, 
the anterior margin of the epandrial sclerite is scored as modestly curved as long as 
there is some emargination. In cases where the anterior margin is flat, or convex 
anteriorly, it is scored as not emarginate. Within the genus Atherimorpha, species 
may be either modestly or strongly emarginate, however it is most common to have 
modest anterior emargination and all Atherimorpha species scored for the matrix 
share this state. Variation within Chrysopilus and Ptiolina is reflected in the matrix. 
The midline cut off point to define the separation between states was an arbitrary 
designation. The variation of this character is clearly continuous within and among 
genera. Pelecorhynchus personatus and Glutops rossi were the only taxa scored as 
having the epandrial sclerite not emarginate anteriorly, however in Glutops punctatus, 
there is a modestly curved emargination. 
 
[#99]. Curvature of epandrium. The epandria in most of the Tabanomorpha and all 
of the Rhagionidae sensu Woodley (1989) are nearly flat and are positioned directly 
above the hypandrium, so that they are separated by a horizontal gap. In species of 
Pelecorhynchus (but not Glutops), Vermileo, Lampromyia, Suragina, and Atherix (as 
 
 106 
 
in Atrichops), however, the epandrium is enlarged and convex dorsally, so that its 
lateral margins surrounds the cerci and hypoproct and contains the hypandrium 
ventrally. This is most easily viewed from the posterior perspective. 
 
[#100]. Cercus. The cerci may be directly attached to the epandrial sclerite or 
attached to a detached structure complex. The detached structure may be simply the 
hypoproct (without any dorsal element) or it may be a subepandrial element in the 
form of a membrane or sclerite. A difference that separates Austroleptis, Glutops, 
Pelecorhynchus, Pseudoerrina, Tabanidae, and Athericidae from all the other taxa is 
the origin of the cerci. The cerci may be separated from the epandrial sclerite by a 
membrane (as in species of Glutops) but their point of attachment is directly to the 
epandrial sclerite. In other taxa, cerci are attached to the hypoproct or subepandrial 
sclerite or membrane. In Coenomyia ferruginea, the cerci are adjacent to the epandrial 
sclerite, but are firmly attached to the hypoproct. Some connective membrane exists 
between the epandrial sclerite and the cerci in Coenomyia ferruginea, but this appears 
to be a secondary attachment. 
 
[#101]. Subepandrial sclerite. A structure- called the subepandrial sclerite (Sinclair, 
1992) or tergite 10 (Nagatomi, 1984)- may be present between the epandrial sclerite 
and the cerci. The subepandrial sclerite is scored as present only when the 
subepandrial structure is sclerotized. Sclerotization and the presence of the 
subepandrial sclerite are most easily viewed from the posterior perspective. The 
subepandrial sclerite is present in Chrysopilus, Ptiolina and Symphoromyia. For taxa 
 
 107 
 
where the cercus attaches directly to the epandrial sclerite, this character was not 
scored because in these cases, there is character non-independence. The lack of the 
subepandrial sclerite is a necessary condition to allow for the direct attachment of 
cerci to the epandrial sclerite. 
 
[#102]. Subepandrial sclerite form. The subepandrial sclerite was scored as entire 
or divided medially. In Chrysopilus and Symphoromyia, the sclerite is separated into 
two parts, whereas in Ptiolina, the structure is undivided. Taxa that lack a 
subepandrial sclerite were not scored for this character. 
 
[#103]. Setation of subepandrial sclerite. The presence or absence of subepandrial 
sclerite setae was scored. In Ptiolina and Symphoromyia, the sclerite bears setae, 
whereas in Chrysopilus, the structure is bare. Taxa that lack a subepandrial sclerite 
were not scored for this character. 
 
[#104]. Hypoproct form (not used in phylogenetic analysis). An attempt was made 
to score the form of the hypoproct. The hypoproct takes on a number of shapes, 
including sagittate, pentagonal, oval, and rectangular. The variation of this character 
was continuous, however, and apparently varied freely among species of the same 
genus (as in Rhagio). Identifying distinct states from a continuous variation of form 
was, in the end, a very challenging and perhaps a foolhardy endeavor. I am not 
confident enough in the scoring of this character to include it for phylogenetic 
 
 108 
 
analysis. The same can be said for the following character, which scores the presence 
or absence of hypoproct lobes. Neither character is used for the phylogenetic analysis. 
 
[#105]. Hypoproct, form in outline (not used in phylogenetic analysis).  The 
hypoproct may appear developed into faint lobes, distinguished by an increase of 
lateral sclerotization. Most genera with multiple species sampling exhibited both the 
presence and absence of lateral sclerotization and I suspect that this character varies 
at the species level, or perhaps even among individuals of the same species. For this 
reason, it is not included in the phylogenetic analysis. 
 
[#106]. Hypoproct setae. The presence or absence of hypoproct setae was scored. 
This character varies at the species level within at least Atherimorpha, Ptiolina, and 
Rhagio and it may be unlikely that it is an important character for defining clades 
above the subgenus or species level. However, it is retained for phylogenetic analysis 
because it is an easily scored, discrete character. 
 
[#107]. Cercus shape (not used in phylogenetic analysis). In some Rhagio species 
(R. plumbeus, for example) the cercus is nearly square or rectangular, whereas in 
most species of Chrysopilus, the form of the cercus is more eye-shaped, or leaf-
shaped. I surveyed the rest of the taxa and ceased to score this character. Although 
clear differences in the form of the cercus may exist when comparing a limited 
number of taxa, it is not feasible to assess such irregular shapes, designate them into 
discrete categories, and score them over a large matrix of relatively distantly related 
 
 109 
 
taxa. Thus, although differences in cercus shape were noted, they are not included in 
the phylogenetic analysis. 
 
[#108]. Cercus separation. The lateral distance between cerci were measured and 
scored in three states. Measurements were made by eye. Cerci were scored as 1) 
directly adjacent to one another, separation distance one quarter width of cercus or 
less, 2) partially displaced from one another, separation distance approximately half 
the width of single cercus, or 3) widely displaced from one another, separation 
distance greater than three quarters width of cercus. 
 
[#109]. Cercus orientation (not used in phylogenetic analysis). The angle at which 
the cerci are held was scored, relative to the epandrial sclerite. In some taxa, such as 
in species of Pelecorhynchus, the cerci are clearly oriented in a vertical fashion. 
Arbitrary cut off points were established for the designation of three separate states, 
although in reality, the character varied continuously. Cerci held at 0?20? relative to 
the epandrial sclerite were scored as 'horizontal,' cerci held at 21?70? relative the 
epandrial sclerite were scored as 'at angle,' and cerci held 71?90? relative the 
epandrial sclerite were scored as 'vertical.' The degree angles were approximated by 
eye. One fairly common problem was that in some species, such as Atherix pachypus, 
the cercus was flush with the epandrial sclerite dorsally yet rounded so that the 
ventral portion of the cercus was vertically oriented. How this condition related to the 
other states was unclear. Orientation of cerci was also affected by their treatment in 
the dissection process. Some may have been expanded or flattened inadvertently. For 
 
 110 
 
these reasons, ultimately, there was little confidence in the scoring for this character 
and it was removed for phylogenetic analysis. 
 
[#110]. Cercus form, from posterior view (not used in phylogenetic analysis). The 
cerci were scored as either curved or flattened, when viewed from the posterior 
perspective. The variation of this character was continuous and states were designated 
by approximate form. No distinct cut off point was devised, aside from 'mostly flat' 
versus 'mostly cupped, or curved,' which were approximated by eye. Curvature of the 
cerci may have been affected by the dissection process. Some may have been 
flattened inadvertently. For these reasons, ultimately, there was little confidence in 
the scoring for this character and it was removed for phylogenetic analysis.  
 
[#111]. Hypandrial sclerite. The presence of a free hypandrial sclerite or one that is 
partially or completely fused with the gonocoxites was scored. Hypandrium is a term 
that refers to the entire ventral portion of the male genitalia (containing the 
gonocoxites, gonostylus, aedeagus and associated structures). The term 'hypandrium' 
is also used in specific reference to the ventral sclerite attached to the gonocoxites. 
When referring to this ventral sclerite of the hypandrium here, I prefer to use the term 
hypandrial sclerite in order to avoid confusion. The hypandrial sclerite is scored as 
'free' when it is separated from the gonocoxites by a complete suture (i.e., it is not 
fused to the gonocoxites). There is an intermediate condition where the hypandrial 
sclerite is fused anteriorly and partially free from the gonocoxites posteriorly. This 
condition is found in species of Arthroceras and Symphoromyia and is scored as 
 
 111 
 
'incompletely fused.' The third state is found in Athericidae, Tabanidae, 
Xylophagidae, and species of Alloleptis, Austroleptis, Chrysopilus, Lampromyia, 
Litoleptis, Ptiolina, Schizella, Spania, Spaniopsis, and Vermileo. In these taxa, the 
hypandrial sclerite and the gonocoxites are fused so that the ventral portion of the 
hypandrium is a single unit. 
 
Since the hypandrial sclerite is separated from the gonocoxites in Nematocera and in 
many basal brachycerans including Bombyliidae, Therevidae, and Asilidae (Sinclair 
et al., 1993), as well as in taxa scored here, the state is considered generally 
considered plesiomorphic. The fusion of the hypandrial sclerite with the gonocoxites 
is a putative synapomorphy for Tabanidae + Athericidae (Woodley, 1989; Sinclair et 
al., 1993). The fact members of the outgroup Xylophagomorpha have a fused 
hypandrial sclerite, however, somewhat complicates the story of the evolution of this 
character within Tabanomorpha. Within the Pelecorhynchidae, and even within 
Pelecorhynchus, there is variation of this character. Glutops and Pseudoerrina have a 
free hypandrial sclerite, while most Pelecorhynchus have gonocoxites fused to 
hypandrium. In P. personatus, the hypandrial sclerite is free and is scored as such in 
the matrix. Yet the hypandrial sclerite in this species is very narrow and different in 
form than the typically broad, smoothly triangular sclerite (as in Glutops and other 
taxa including Pseudoerrina). The free hypandrial sclerite in P. personatus may 
represent a secondary transition. 
 
 
 112 
 
[#112]. Gonocoxal sinuous ridge. The dorsal side of the gonocoxite may have a 
sinuous ridge that leads to the gonocoxal apodeme. This is lacking in Athericidae, 
Tabanidae, Vermileonidae, and many Xylophagidae, as well as Alloleptis tersus, 
Austroleptis multimaculata, Chrysopilus quadratus (unlike most Chrysopilus 
species), Litoleptis alaskensis, and Stylospania lancifera. 
 
[#113]. Gonocoxal apodeme presence. The presence or absence of gonocoxal 
apodemes was scored. Austroleptis and Litoleptis are the only taxa where the 
gonocoxal apodemes are absent. I do not consider this a compelling synapomorphy 
for supporting common ancestry of these genera, although it is a rarely occurring 
feature. The reduction of such a structure may occur on account of a variety of 
independent evolutionary constraints. 
 
[#114]. Gonocoxal apodeme length. The length of the gonocoxal apodeme was 
scored by noting the position at its apex, relative to the anterior margin of the 
hypandrium. Although the gonocoxal apodeme is continuously varied to some degree, 
two mutually exclusive states emerged. The gonocoxal apodemes were scored as 
either 1) short or long enough to reach the anterior margin of the hypandrium or 2) 
extending well beyond the anterior margin of the hypandrium. The long length of the 
gonocoxal apodemes has been given as evidence to support Athericidae + Tabanidae 
(Stuckenberg, 1973; Sinclair et al., 1993), although the character is lacking in Xeritha 
Stuckenberg (as noted in Sinclair et al., 1993). Austroleptis multimaculata and 
 
 113 
 
Litoleptis alaskensis are not scored for this character since they lack gonocoxal 
apodemes. 
 
[#115]. Gonocoxal apodeme origin. The relative lengths of the medial and lateral 
margins of the gonocoxal apodeme were scored. Relative length of the gonocoxal 
margins may be strikingly different among taxa. Athericidae and Tabanidae, for 
example, have very short medial gonocoxal margins in comparison to the lateral 
margins, so that what may be called the parameral bridge (the transverse structure 
spanning the parameral sheath, joining the gonocoxites dorsomedially), is shifted 
anteriorly. Instead of referring to the relative placement of the parameral bridge, I 
believe that scoring the relative lengths of the inner and outer margins of the 
gonocoxal apodeme is more discrete, objective, and consistent. In addition to the 
Athericidae and Tabanidae, other taxa that have gonocoxal apodemes with shorter 
inner margins are Lampromyia canariensis, Vermileo vermileo, Pelecorhynchus 
personatus, and Xylophagus lugens. Austroleptis multimaculata and Litoleptis 
alaskensis are not scored for this character since they lack gonocoxal apodemes. 
 
[#116]. Parameral sheath. The parameral sheath surrounds the aedeagus posteriorly. 
Ventrally, the sheath is sometimes bulbous and/or produced into paired swellings 
ventrally. The character was scored as developed into bulbous sac ventrally when the 
parameral sheath was expanded so that it was flush with the gonocoxites ventrally. 
The sac was scored as having lobes only when membranous lobes were distinctly 
present. Such lobes were present in species of Arthroceras, Ptiolina, Spaniopsis, and 
 
 114 
 
Symphoromyia. Interestingly, in species of Austroleptis, there are two centrally 
located, ventral hypandrial lobes. However in Austroleptis, the lobes arise from the 
gonocoxites (instead of from the parameral sheath) and are sclerotized (instead of 
membranous). I don't believe at the sclerotized gonocoxal lobes of Austroleptis are 
homologous with the membranous lobes of the parameral sheath found in other taxa. 
The parameral sheath itself is unmodified in Austroleptis. Therefore, Austroleptis was 
scored as not having the parameral sheath developed into a sac ventrally. 
 
[#117]. Lateral ejaculatory process. Lateral aedeagal processes (sensu Sinclair, et 
al. 1993) are present in two forms. In Desmomyia thereviformis, all Rhagio species, 
Rhagina incurvatus, and Sierramyia chiapasensis, the lateral aedeagal processes are 
thin, lightly sclerotized structures that are integrated into the sperm sac. In all species 
of Arthroceras, Arthroteles, Atherimorpha, Chrysopilus, Glutops rossi, Pseudoerrina 
jonesi, all Ptiolina, Schizella furcicornis, Spania nigra, all Spaniopsis, and all 
Symphoromyia, the lateral aedeagal processes are thickened, well sclerotized 
structures that are integrated into the sperm sac basally, but extend freely apically. In 
most taxa that have aedeagal tines, such as Athericidae, Tabanidae, and Bolbomyia 
nana, no lateral aedeagal processes are present. This has led to misunderstandings of 
the lateral aedeagal structures as possibly being homologous to the aedeagal tines. It 
is instructive to note here, however, that all Arthroceras species have both aedeagal 
tines and lateral aedeagal processes. 
 
 
 115 
 
[#118]. Ejaculatory apodeme length. The length of the ejaculatory apodeme was 
scored by noting the position at its apex, relative to the anterior margin of the 
hypandrium. Although the gonocoxal apodeme is continuously varied to some degree, 
ejaculatory apodeme length was divided into four discrete categories. The gonocoxal 
apodemes were scored as 1) reduced, nearly absent, 2) short, not reaching anterior 
margin of the hypandrium, 3) moderately long, reaching the anterior margin of the 
hypandrium, or 4) long, reaching beyond the anterior margin of the hypandrium. 
Perhaps there should have been an additional state for especially long ejaculatory 
apodemes, such what is found in species of Atherix and Dasyomma. However since 
there is continuous variation in this character, it is difficult to score 'extra long' when 
there are few objective reference points. Perhaps on account of this type of variation, 
and the lack of intuitive and discrete length categories, states are generally not 
conserved at the genus level. The ejaculatory apodeme, for example, in species of 
Chrysopilus, may be short, moderately long, or long. Similarly, the length of the 
ejaculatory apodeme is not fixed in Ptiolina, Rhagio, and Symphoromyia. It is not 
used in the phylogenetic analysis. 
 
[#119]. Ejaculatory apodeme form. The form of the ejaculatory apodeme was 
scored as either tubular, laterally compressed, compressed dorso-ventrally, tripartite, 
or umbruculate (umbrella-shaped) anteriorly. Tripartite is a term used describe the 
structure when it is dorsally compressed laterally and ventrally compressed dorso-
ventrally, as in the case of Bolbomyia nana. Austroleptis species are distinctive by 
having the ejaculatory apodeme umbrella-shaped at its apex. In these and other cases, 
 
 116 
 
such as in Ptiolina fasciata where the ejaculatory apodeme is very clearly laterally 
compressed, the character is easily scored. However, for other taxa, the form of the 
ejaculatory apodeme does not fall so neatly in a single category. The ejaculatory 
apodemes of Chrysopilus ferruginosus and Ptiolina zonata, for example, are mostly 
tubular, but also laterally compressed somewhat. For these species, the character is 
scored as an ambiguity, present as both a tubular and laterally compressed apodeme. 
Furthermore, it is important that the apodeme be examined from more than a single 
perspective since a laterally-compressed apodeme will look tubular when only viewed 
from above. Generally, ejaculatory apodeme forms were consistent within genera, 
however, I don't have a great deal of confidence in this character on account of the 
difficulties inherent in its scoring. 
 
[#120]. Aedeagal tines. The presence or absence of aedeagal tines was scored. The 
tines are present in Athericidae, Tabanidae, Bolbomyia, and Arthroceras. The 
presence of aedeagal tines has been proposed as a synapomorphy to unite 
Athericidae+Tabanidae+Bolbomyia (Sinclair, et al., 1993). 
 
[#121]. Endoaedeagal process presence. The endoaedeagal process (termed by 
Sinclair et al, 1993) is a slender, sharply pointed projection that extends posteriorly 
into the sperm sac. This structure has been proposed as one of three synapomorphies 
of the male genitalia which support the monophyly of the Brachycera (Sinclair et al, 
1993). Sinclair et al. (1993) defined the origin of the endoaedeagal process at "the 
base of the aedeagal tines or the 'precursor' sclerites." Sometimes, as in some Ptiolina 
 
 117 
 
species, however, the tines or 'precursor' sclerites are missing and the point at which 
the endoaedeagal process begins is not immediately obvious. Sometimes a break (in 
the form of a partial or complete suture, or as a small gap) is present between the 
aedeagal apodeme and endoaedeagal process, and this can be used to indicate where 
the endoaedeagal process begins. However, this suture or break is not always present. 
And across the lower Brachycera, the length of a posterior extension of the 
ejaculatory apodeme varies, so that scoring can be difficult in "almost completely 
reduced" situations.  
 
In particular, within the genus Ptiolina the character is free to vary, in ways that may 
make scoring problematic. In Ptiolina majuscula and P. lapponica, for example, an 
endoaedeagal component is clearly present (anterior of the ?precursor? sclerites), but 
it is broad, blunt, and short. In P. nitida, the posterior end of the aedeagal apodeme is 
turned downward sharply, at the anterior boundary of the sperm sac and the landmark 
'precursor' sclerites are missing in this species, so the precise point at which the 
endoaedeagal component begins is unclear. In P. fasciata, the landmark ?precursor? 
sclerites are present and clearly mark the posterior apex of the aedeagal apodeme; the 
endoaedeagal component is absent in this species. P. edeta, P. mallochi, P. nigra, P. 
obscura, and P. zonata are similar to P. nitida in form.  
 
In many Chrysopilus species, the posterior end of the aedeagal apodeme is turned 
downward sharply, at the anterior boundary of the sperm sac just as it is in many 
Ptiolina species. In Chrysopilus ferruginosa and C. quadratus, there is an additional 
 
 118 
 
structure posterior to the aedeagal apodeme, where the ?precursor? sclerites usually 
are located. This may be interpreted as a fusion of the two 'precursor' sclerites or an 
unrelated structure. In any case, it is not fused with the aedeagal apodeme and does 
not penetrate the sperm sac. Therefore it is not scored as the endoaedeagal process. 
 
[#122]. Endoaedeagal process form. For taxa where the endoaedeagal process is 
present, the form of the endoaedeagal process is scored for three states. Species of 
Arthroteles, Atherimorpha, and Rhagio, and also in Desmomyia thereviformis and 
Rhagina incurvatus, the endoaedeagal process is very distinctly laterally compressed, 
taking the form of a butterknife. In other taxa, the endoaedeagal process is either 
narrowly conical, smoothly cylindrical, or dorsolaterally flattened. These states were 
rather continuous and blended with one another. Dialysis rufithorax was the only 
taxon with an unmistakable dorsoventrally flattened endoaedeagal process, and a 
separate state was designated on account of this. 
 
[#123]. Gonostylus setation (not used in phylogenetic analysis). The relative 
amount of setation on the gonostylus is scored as either heavily or lightly present. 
This is a continuous character that lacks discrete boundaries between states, although 
the relative amount of gonostylar setation is generally conserved within each genus. 
Chrysopilus species, for instance, generally have densely arranged, robust gonostylar 
setae, as do putative allies Schizella and Stylospania lancifera. The continuous nature 
of this character, however, precluded it from being scored confidently and it was not 
used for phylogenetic analysis. 
 
 119 
 
 
The gonostylus may take a number of shapes, some of which may provide useful 
information for use in phylogenetic analyses. In Glutops rossi and G. punctatus, for 
instance, gonostylus is full and rounded, with a distinctive inner margin that is s-
shaped and bearing an internally directed hook apically. Pelecorhynchus fusconiger 
has a gonostylus with very similar characteristics. The inner margin is roughly s-
shaped and at its terminus, bears an internally directed hook. However, the diversity 
of gonostylus form among Pelecorhynchus species suggests that scoring this is not 
necessarily so simple. In P. elegans, for instance, the gonostylus has an inner margin 
whose curvature is exaggerated into a thick medial swelling. In P. personatus, this 
swelling is narrow so that it nearly takes the form of a ridge. The distal area elongated 
and conical in P. personatus, however the inward facing hook is still present. In P. 
(Archeomyia) mirabilis, the inward facing ridge is produced into a long slender 
extension, making the gonostylus appear two-pronged and distally, the inward facing 
hook is missing. The transformation series of this feature, and the homology of the 
inward-facing swelling and distal hook may be intuitive within the Pelecorhynchidae, 
however, assessing the homology of shape across a broad array of taxa is far more 
daunting. I prefer to simply make note of similarities of gonostylar form that may be 
appreciated in light of a phylogenetic hypothesis rather than code these rather 
complicated forms a priori, for phylogenetic analysis. Therefore, it is not used for the 
phylogenetic analysis. 
 
 
 120 
 
Female abdomen.  
[#124]. Terminal abdominal segments (not used in phylogenetic analysis). 
Rhagionidae are known as Snipe flies, a name that refers to their evenly tapered 
abdomen. The character may vary strikingly among Tabanomorpha, where the female 
abdomen either may be evenly tapered or with a clear difference between wider basal 
segments and narrowed distal segments. If a straight line could be visualized along 
the lateral margins of the abdominal sclerites, the abdomen was scored as evenly 
tapered. In the cases where there is an abrupt size difference between basal and distal 
segments, the line is jagged in keeping with the margins of the sclerites. Species of 
Pelecorhynchus provide an example for this case. The amount of 'jag' required to 
invoke its recognition for being scored is not always self-evident. For example, in 
Glutops punctatus, the female abdomen there is an obviously abrupt size difference, 
yet in Glutops rossi, the break is much less clear. Without reference to G. singularis, 
G. rossi may appear evenly tapered. Upon closer inspection, however, there is a 
subtle but clear break between the fourth and fifth abdominal sclerites, just as in G. 
singularis (but less so). As the breaks revealing abrupt size differences between basal 
and distal segments become more subtle, the character begins to break down, as 
nearly all genera have species where abdominal sclerite width differences may be 
interpreted as either evenly or abruptly tapered. Chrysopilus, Rhagio, and Spaniopsis 
have species with smoothly tapering abdomen, although there is usually a difference 
between the basal segments of the abdomen and those distally, which have the ability 
to telescope within these segments. It is interesting to note that while in most species, 
as in Pelecorhynchus, Glutops, and Schizella furcicornis, the change in width occurs 
 
 121 
 
between the fourth and fifth segments. In Symphoromyia cruenta, the constriction 
occurs between the fifth and sixth segments. Although the current character is not 
used for phylogenetic analysis, the nature of this character may be integrated into the 
analysis by taking into account related measures, such as the shape of tergite 7 and 
the intersegmental membrane distance between segments 7 and 8. These measures are 
associated with the form and ability for telescoping segments. This character was 
removed for phylogenetic analysis. 
 
[#125]. Tergite 7. The shape of tergite 7 was scored as either clearly longer than 
wide, about as long as wide, or clearly wider than long. This determination was made 
after the sclerite was laid flat. Glutops rossi was the only taxon that is scored as about 
as long as wide. Glutops singularis also has a square tergite 7. This character is 
consistent within genera, except in Ptiolina, where tergite 7 may be wider than long 
or longer than wide. In Arthroceras pollinosum, tergite 7 is clearly longer than wide. 
This is also the case for Arthroceras fulvicorne, however Nagatomi & Iwata (1976) 
illustrate tergite 7 in Arthroceras japonicum as apparently wider than long. In 
Arthroceras leptis, tergite 7 is about as long as wide. Tergite 7 in Arthroteles cinerea 
is also illustrated by Nagatomi & Iwata (1976) as wider than long, however, I have 
examined this species and it is longer than wide, as it is in Arthroteles bombyliiformis. 
Stuckenberg (2001) noted that this may be an important character. 
 
[#126]. Intersegmental membrane between segments 7 and 8. In some taxa, there 
is a distinctly long intersegmental length between the distal segments of the female 
 
 122 
 
abdomen. Stuckenberg (2001) has pointed this out and has used it as a justification to 
support family relationships. Intersegmental length between the 7th and 8th segments 
is scored as 1) short, as throughout abdomen or 2) especially long. Where scored as 
especially long, the intersegmental region between tergites 7 and 8 is clearly longer 
than the intersegmental region between tergites 4 and 5. It is tempting to compare the 
7/8 intersegmental region to the length of sternite 8. That was not done here, as 
sternite 8 may vary independently. Thus, the 7/8 intersegmental may be longer or 
shorter than the length of sternite 8. The 7/8 intersegmental of Austroleptis 
multimaculata is clearly longer than the intersegmental membranes anterior to this 
and is in apparent conflict with how Stuckenberg (2001) has scored the present 
character for the genus. 
 
A long intersegmental length of the distal abdominal sclerites provides the ability to 
retract the terminal segments in a telescoping manner. Therefore, taxa with long distal 
intersegmentals may also be referred to as having an 'extensible abdomen.' This form 
seems to be a basic adaptation for oviposition in a terrestrial habitat, adapted for soft 
earth in saturated conditions and in leaf litter and mold, and this is where flies with 
extensible abdomen are most commonly found. The extensible abdomen occurs 
across a fairly broad spectrum of taxa within lower Brachycera, suggesting that it may 
be a very old character. In any case, I think these factors suggest that it may not be a 
reliable character for phylogenetic inference. If it is not plesiomorphic, then it at least 
appears vulnerable to change due to convergence (restraints of common habitat) 
 
 123 
 
rather than (or at least as much as) change due to common ancestry. It is, however, 
included for phylogenetic analysis. 
 
Female terminalia (external structures) 
[#127]. Tergite 9 length. In Spania nigra and species of Spaniopsis, tergite 9 is 
reduced to a very narrow sclerite. This character state is not found in any of the other 
taxa. 
 
[#128]. Tergite 9 anterolateral arms. There is a special modification of tergite 9, in 
which narrow, anteriorly-directed ventrolateral extensions (arms) are produced, 
enveloping sternite 9. The ventrolateral arms of tergite 9 are actually firmly attached 
to sternite 9 laterally, via thick membranous tissue. Surprisingly, this has not been 
noted by previous authors and appears phylogenetically informative. Where present, 
it is obvious only after careful dissection. The feature is found exclusively, in all 
species of Ptiolina, Spania, Spaniopsis, and Symphoromyia. Since this character not 
only scores the presence of these arms, but also the fact that they are firmly attached 
to sternite 9 (an additional unique aspect of this modification), extra weighting of this 
character may be appropriate. 
 
[#129]. Tergite 10 length (not used in phylogenetic analysis). Relative length of 
tergite 10 was first pointed out in Nagatomi & Iwata (1976: 36). The length of tergite 
10 was compared against the width and scored in three ways: 1) absent, 2) present, 
reduced, 3) present, not reduced (length approximately equal to measured half width) 
 
 124 
 
or 4) present, not reduced. Where the length of tergite 10 is measured to be less than 
half the width of the sclerite, it is scored as reduced. Tergite 10 is scored as not 
reduced where the length is approximately equal to, or greater than, half the width. 
Differences between the states of this character are subtle, however, and vulnerable to 
inconsistent scoring. This became clear as scoring progressed and scoring was 
stopped once this determination was made. For the taxa scored, variation within 
genera and families suggest that this character ought to be removed for phylogenetic 
analyses.  
 
[#130]. Tergite 10 form. Female tergite 10 is scored either as entire or split into two 
separate lateral sclerites. This character is not applicable for Pseudoerrina jonesi, 
where tergite 10 is absent. Nagatomi & Iwata (1976) show tergite 10 of Atherix 
basilica as entire, however in Atherix pachypus, it is split (as the illustration of 
Atherix ibis in Nagatomi & Iwata (1976:38, fig 26)). Tergite 10 is entire in 
Arthroceras pollinosum, A. leptis, A. fulvicorne, and A. subaquilum, however it is 
illustrated as partially spit in A. japonicum by Nagatomi & Iwata (1976:22, fig 12). 
 
[#131]. Sternite 8 form. In Austroleptis multimaculata (and congeners), it appears 
that the sternite 8 is divided into anterior and posterior segments. Sternite 8 in species 
of Austroleptis is shifted posteriorly so that the division of this sternite is positioned 
to allow for flexibility of the cerci and associated structures with respect to the rest of 
the abdomen. Nagatomi & Iwata (1976) have a different interpretation of this sternite 
in Austroleptis. Instead of considering it two parts of the same sclerite, they interpret 
 
 125 
 
the anterior part as sternite 8 and the posterior part as sternite 9. Since the genital fork 
is homologous to sternite 9, however, and this present in Austroleptis, the 
interpretation of Nagatomi & Iwata invokes the duplication of sternite nine and 
subsequent reduction of the anterior copy, to a simple sclerite. I reject this 
interpretation as highly unlikely. Assuming that sternite 8 is simply divided into two 
parts, all segments are present and accounted for. In some species, such as in an 
undescribed species I've examined from South America, the division between the 
anterior and posterior areas of sternite 8 is inconspicuous. One could very easily miss 
seeing this division, and mistake the compound structure as a single, elongate sclerite. 
Austroleptis is autapomorphic for this feature. 
 
[#132]. Sternite 8 cleavage. Sternite 8 is the sclerite underneath the genital furca, and 
the notch (or 'cleavage') along its posterior margin apparently allows for the male 
reproductive organ to reach the genital chamber. Originally, I tried to account for the 
diversity of form represented by the notch itself, which may be deep/shallow, 
narrow/broad, v-shaped/u-shaped, etc. But, ultimately, these differences proved to be 
exceedingly subtle and were subject to varying interpretation. There may be some 
pattern as to the form of the cleavage of the posterior margin of sternite 8, but I could 
not devise an unambiguous scoring metric that teased out this information for 
phylogenetic analysis. However, there is one major, easy-to-score feature of the 
posterior margin of sternite 8. This is simply the presence or absence of the notch. In 
Dasyomma atratulum, there is a medial suture that is very inconspicuous, yet the 
notch is scored as absent because the posterior margin itself is entire. All members of 
 
 126 
 
Athericidae, Tabanidae, and Vermileonidae of this taxon set, are scored as lacking the 
posterior margin notch. The notch is also absent in Austroleptis multimaculata, 
Spania nigra, and species of Spaniopsis.  
 
[#133]. Sternite 8 length. The length of sternite 8 was scored relative to its width, 
and placed into one of four separate categories; 1) wider than long, 2) as long as wide, 
3) longer than wide, or 4) elongated. Elongated, in this sense, means that the sclerite 
is at least twice as long as wide or longer (e.g., sternite 8 in members of 
Xylophagidae). This character was generally consistent within genera, however, the 
sternite 8 of Ptiolina species may either be as long as wide or longer than wide. This 
difference within Ptiolina is actually a consistent character that, in corroboration with 
several other characters, helps to distinguish between Omphalophora and Ptiolina. 
Spania nana and species of Spaniopsis are the only taxa that have sternite 8 wider 
than long. The two divisions of sternite 8 in Austroleptis multimaculata, when added 
together, are elongate, but I prefer to leave this character unscored as the homology of 
the state may be disputed. Some of these differences in sternite 8 morphology may be 
viewed in the illustrations by Nagatomi & Iwata (1976). 
 
[#134]. Sternite 10 form (not used in phylogenetic analysis). Sternite 10 may be 
sclerotized evenly or sclerotization may be weakened medially. Where the 
sclerotization is weakened, it appears as if the sclerite is divided into two lateral 
components, as in Bolbomyia macgillisi. The latter state is a relatively common 
feature, however, it is not always easy to detect. The lack of medial sclerotization 
 
 127 
 
may be incomplete, or the sclerotization of sternite 10 may be weak across its entire 
length. This led to poor confidence in the accuracy of the scoring. My impression, 
after examining a broad sampling of taxa, is that this character is of minor importance 
(at best) for confirming or refuting hypotheses of relationship. For these reasons, the 
character was not included in phylogenetic analysis. 
 
[#135]. Sternite 10 shape (not used in phylogenetic analysis).  The form of sternite 
10 took on a variety of shapes, which were scored as one of five states; 1) pentagonal, 
pointed posteriorly, 2) rectangular, 3) sagittate, 4) semi-circular, or 5) ovoid. Where 
sternite 10 is clearly five-sided, it is scored as roughly pentagonal. Typically, the 
pentagon shape is nearly regular. However, in Ptiolina, it is broad and clearly wider 
than long. A four-sided sternite 10, with minimal side curvature is scored as 
rectangular. Where sternite 10 is generally pentagonal in shape, pointing posteriorly, 
but bears lateral anterior lobes, it is scored as sagittate. Semi-circular sclerites are 
defined by having a straight anterior margin and a fully rounded posterior margin. If 
sternite 10 is rounded throughout, it is scored as ovoid. Where the anterior margin is 
flat and the posterior and lateral sides rounded, sternite 10 is scored as semi-circular. 
Sternite 10 may be rounded throughout and in this case, it is scored as ovoid. 
Although differences in shape among taxa are evident, these differences are 
exceedingly subtle and difficult to score consistently. Boundaries between 
semicircular, ovoid and rectangular, or between pentagonal, semi-circular, and 
sagittate, are often determined by slight differences of curvature that are rarely 
unambiguous. For this reason, the character is not used for phylogenetic analysis.  
 
 128 
 
 
[#136]. Sternite 10 position. The position of sternite 10 was scored relative to the 
first cercus segment. Significant differences exist between species, and these 
differences are not altered by handling because connective tissue is short and 
structures are virtually immobile, even after manipulation in glycerol. The characters 
scored are 1) nearly completely anterior to first cercus segment (~10% below basal 
cercus), 2) posterior half below first cercus segment (approximately 20?50% below 
first cercal segment, 80?50% below tergite 10), or 3) entirely or almost entirely 
underneath cercus segments. Since the cercus attaches to tergite 10, this is another 
way to score the relative development of tergite 10 versus sternite 10. Where tergite 
10 is reduced, the cercus is shifted anteriorly and the sternite 10 subtends a greater 
portion of the cercus. The relative length of tergite 10 was difficult to score 
confidently and for this reason, abandoned for use in phylogenetic analysis. While the 
present character is essentially a derivation of character #133 (tergite 10 absent, 
present/reduced, or present/not reduced), it is more discrete, easier to score, and thus 
in my opinion, more trustworthy. The overlap between the present character and 
character #133 isn't complete, however. In Lampromyia canariensis, for example, 
tergite 10 is missing, yet sclerite 10 is nearly completely anterior to the first cercus 
segment and is therefore coded differently than in Atherix pachypus, for example, 
where tergite 10 is also absent, but sternite 10 is produced directly below the first 
cercus segment as other athericids, tabanids, and in Vermileo vermileo. 
 
 
 129 
 
[#137]. Cercus segmentation. The cercus may be one- or two-segmented. The loss 
of the basal cercus segment is a putative synapomorphy for Athericidae + Tabanidae. 
Species of Austroleptis also have a single cercus segment. 
 
[#138]. Basal cercus elongation. The basal cercus is scored as either elongated or 
not. Basal cerci that are at least three times as long as wide are considered elongate. 
This is the condition in Arthroceras americana, Coenomyia ferruginea, Dialysis 
rufithorax, and Xylophagus lugens (the xylophagids included in this study). 
Otherwise, the basal cercus is scored as not elongate. 
 
[#139]. Basal cercus lobe (alternate character coding based on gestalt; not used 
in phylogenetic analysis). The basal cercus was scored for the presence of a ventral 
process. The degree to which the ventral process must be present in order to be scored 
as such has not been established by previous authors who have examined this feature 
(Nagatomi and Iwata, 1976), and variation in the degree to which the lobe is present 
makes consistent scoring of this character difficult. The lobe was scored as present or 
absent according to 'gestalt.' For consistency and repeatability, a more precise scoring 
method is also devised, scored, and discussed in a subsequent scoring system. The 
present system uses 'expert' judgement, however, I place little confidence that the 
boundary between presence and absence of the first cercus lobe remains fixed across 
the entire sampling taxa here, and even less confidence in the prospect that the 
scoring here would be generated independently by another worker looking at the 
same material. The presence or absence of the first cercus segment, due to the 
 
 130 
 
continuous nature of its variation, is largely subjective. For this reason, the scoring of 
this character is inappropriate for phylogenetic analysis. 
 
[#140]. Basal cercus lobe (alternate character coding based on extreme cases; not 
used in phylogenetic analysis). Among most species of Tabanomorpha (with two 
cercus segments), there is a range of continuous states with respect to the degree of 
lobing of the first cercus segment. There appears to be an obvious, discrete gap 
between this range of morphological variation, however, and an extreme morphology 
evident in a few species. The basal cercus lobes in species of Glutops, Pseudoerrina, 
and Pelecorhynchus, and in Vermileo vermileo, appear well out of the range of 
variation present among other Tabanomorpha. They are distinct in extending 
posteriorly past the point of attachment of the second cercus, are nearly parallel-sided, 
and nearly as long or longer than the second cercus.  
 
[#141]. Basal cercus lobe (alternate character coding based on area ventral of 
first cercus, relative to width of second cercus; not used in phylogenetic 
analysis). In species of Pelecorhynchus, Pseudoerrina, and Glutops, the ventral 
process is very conspicuous. However, aside from these extremes, there is continuous 
variation of this character when examined across a broad sampling of Tabanomorpha. 
Even among species in the genus Chrysopilus, many degrees of a ventral process are 
present, including the lack of such a structure. This variation clouds the point at 
which a ventral extension must be considered present. As an arbitrary but discrete 
metric, the width of the second cercus segment at point of attachment was compared 
 
 131 
 
against the length of the first cercus segment ventral to this. The ventral process was 
scored as present where the length of the first cercus ventral to the ventral margin of 
the second cercus segment was equal to or greater than the width of the entire second 
cercus segment. Others may argue the appropriateness of this demarcation, especially 
as the width of the second cercus segment is free to vary and thus, may complicate 
the scoring of the ventral process of the first cercus segment. While this criticism 
identifies an obvious flaw inherent in the measurement, I have found that the width of 
the second cercus segment is mostly constant across tabanomorph taxa. The choice of 
determining a threshold for scoring the presence of a ventral process of the first 
cercus segment (equal to the width of the second cercus segment, at point of 
attachment) is much more problematic because 'borderline' cases abound. If a lobe is, 
by definition, at least as wide as the width of the second cercus segment, some cases, 
such as in Arthroteles bombyliiformis, an apparent lobe must be scored as absent. 
Conversely, if the threshold is loosened so that the lobe must be at least half the width 
of the second segment, obvious cases where the lobe is not present, such as in 
Ptiolina, a lobe will be scored as present. Ultimately, the many ways in which the 
basal cercus may vary confounds objective scoring schemes devised for scoring the 
lobe. The basal cercus may vary in the position of attachment of the second cercus, 
the posterior extension of the basal lobe (relative to the second cercus), and the 
ventral extension of the lobe (relative to the base of the first cercus segment). The 
latter two components, in particular, appear to vary continuously. The presence of a 
ventral process in the basal cercus segment is probably associated with egg-laying 
 
 132 
 
behavior. This is a preliminary scoring of this character, modified below, and is not 
used for the phylogenetic analysis. 
 
[#142]. Second cercus position (not used in phylogenetic analysis). In the process 
of understanding the morphological diversity of the female cercus, and taking into 
account the extent of its variation, I noticed that the placement of the second cercus 
segment on the basal segment was consistent within genera and discretely different 
among genera. Often, the placement of the second cercus segment was associated 
with the presence or absence of a ventral lobe (i.e., second segment dorsally placed in 
taxa with ventral lobes present), however, not always. Since this scoring of this 
character is much more easily and objectively determined, it may provide surrogate 
phylogenetic information not accessible when focusing on the presence or absence of 
the ventral lobe itself. Not scored for taxa with elongate first cercus segment, because 
apical margin of this segment is reduced (only one point of attachment possible). This 
is a preliminary scoring of this character, modified below, and is not used for the 
phylogenetic analysis. 
 
[#143]. Basal cercus lobe. The first cercus segment postero-ventral lobe is a 
sometimes subtle feature that may distinguish itself in a number of ways. I've tried to 
break these confounding morphologies down into what I consider the major elements 
of what makes a lobe; gestalt, length of first cercus below second cercus attachment, 
and position of the attachment point of the second cercus. I've put the discrete, 
objectively scoreable states of the latter two components together into a single 
 
 133 
 
character as a way to score the presence or absence of the lobe. An additional state is 
also added to account for the distinct lobe in species of Glutops, Pelecorhynchus, 
Pseudoerrina, and Vermileo.  
 
Where the ventral length of the first cercus segment is equal to or longer than the 
width of the attachment point of the second cercus and the second cercus is attached 
dorsally (measured by the comparison of widths of the first cercus above and below 
the second cercus), then the lobe is considered present. In all of these cases, the taxon 
was also scored as having a lobe when measured by gestalt. Where the ventral length 
of the first cercus segment is shorter than the width of the attachment point of the 
second cercus and the second cercus is attached centrally or ventrally (measured by 
the comparison of widths of the first cercus above and below the second cercus), then 
the lobe is considered absent. In all of these cases, the taxon was also scored as 
lacking a lobe when measured by gestalt. In Arthroceras pollinosum, the basal cercus 
is expanded so that although the ventral length is greater than the width of the second 
cercus, but the second cercus segment is attached centrally. In A. fulvicorne, A. leptis, 
and A. subaquilum, this is also the case. The cases may be made for Arthroceras 
species in favor or against having a ventral lobe, depending on your predisposition. 
This state I've scored separately, as unclear. It appears more like a third condition in 
Arthroceras rather than ambiguous between two states. In Atherimorpha atrifemur, 
the second cercus is positioned dorsally, but the ventral portion of the basal cercus is 
not expanded. This taxon was scored as ambiguous; the lobe may be present (as in 
Atherimorpha vernalis) or absent (as in Atherimorpha nemoralis). 
 
 134 
 
 
[#144]. Ventral lobes of basal cercus. When looked at the female terminalia from 
the posterior view, the ventral apices of the basal cerci may curve inwardly to meet 
one another medially, forming a circle or ring. Stuckenberg has pointed out this 
character and has advocated its use in phylogenetic analysis (2001). Although this 
character appears to depend in some degree on the presence or absence of ventral 
lobes of the basal cercus segments, I found that the 'basal cercus inwardly curved' 
state may be present in taxa without prominent first cercus lobing (Chrysopilus spp. 
and Schizella furcicornis). Conversely, some taxa with ventrally-directed lobes 
(Symphoromyia hirta and Arthroteles bombyliiformis) do not have inwardly curving 
basal cerci. In a great majority of cases, however, if the lobe is present, it curves 
inwardly and where the lobe is absent or posteriorly produced, there is no 
ventromedial curvature. There is logically some non-independence with the present 
character and the presence/absence of lobes. Therefore, this character may be 
downweighted for phylogenetic analysis. 
 
[#145]. Basal cercus separation. The dorsal position of the basal cerci, in relation to 
one another, was scored for two states. Where the cerci are separated from one 
another dorsally by approximately the width of the second cercal segment at point of 
attachment, the state is scored as 'separated.' In other cases, the cerci are closer to one 
another dorsally and are scored as 'adjacent.' 
 
 
 135 
 
[#146]. Second cercus segment. The second cercus is scored as either elongated or 
not. Basal cerci that are at least three times as long as wide are considered elongate. 
This is the condition in species of Ptiolina, Spania nigra, Arthroceras americana, 
Coenomyia ferruginea, and Dialysis rufithorax. Otherwise, the basal cercus is scored 
as not elongate. 
 
[#147]. Cercus apical sensory pits. The terminal cercus segment may or may not 
have what is called an apical sensory pit. The sensory pit is a circular depression, 
located apically, on the lateral or posterior surface of the cercus. In many genera 
(Chrysopilus, Pelecorhynchus, Rhagio) the sensory pit may be present or absent. 
Among the species of Rhagio sampled for this dataset, all lack apical sensory pits 
however there are species with such pits, such as Rhagio incisus. In Arthroceras 
pollinosum, the apical sensory pit is very shallow and, arguably, may be scored as 
absent. However, the apical sensory pit is clearly present in all other Arthroceras 
species examined for this character (A. fulvicorne, A. leptis, A. subaquilum), in 
precisely the same location as where the shallow depression lies in A. pollinosum. 
Therefore, the apical sensory pit was scored as present in this taxon. 
Female terminalia (internal structures) 
[#148]. Tergite 8 ducts. There are a pair of thin ducts that arise from the posterior 
margin of tergite 8 in species of Glutops, Pelecorhynchus, and in Pseudoerrina 
jonesi. The ducts are very inconspicuous and are most easily seen after staining in 
chlorozol black. The ducts have not been recognized by previous authors and their 
 
 136 
 
function is unknown. The ducts terminate anteriorly in a membranous sac or a clump 
of lightly sclerotized tissue. 
 
[#149]. Number of spermathecae. The most common condition for Diptera is to 
have three spermathecae. However, Dialysis (Xylophagidae) has four and Bolbomyia 
(Rhagionidae) has two. These are the only genera I observed to differ from the 
standard condition. The nematoceran, Sylvicola, has a single spermatheca. 
 
[#150]. Shape of spermathecae (not used in phylogenetic analysis). The shape of 
the spermatheca were scored into seven states. Clubbed spermathecae were defined 
by having increased girth leading up to a knob-like end. Spherical spermathecae are 
the most easily recognizable shape; a spherical form. Swollen spermathecae are 
widest at the apex and exhibit a smooth tapering down to its base. Oval spermathecae 
have a clear point of expansion (unlike the swollen form) and are oval in shape, with 
rounded ends. Elongate oval spermathecae also have a clear point of expansion but 
have an elongate form, with parallel sides, similar in shape to a fruiting cattail (Typha 
sp.). Eye-shaped spermathecae are similar in girth and shape to oval spermathecae, 
but are sharply pointed at both tips. Pear-shaped in form is enlarged and rounded at 
the base and narrowed distally. Sometimes taxa have misshapen spermathecae that 
are unscoreable. This is probably an artifact of age or other factors that occurs at the 
level of individual specimens. For a large part, spermathecal form varies in 
confounding ways that makes it difficult to assert homology between character states 
and within many genera, the diversity of form indicates that information regarding 
 
 137 
 
deeper level phylogenetic events is probably not retained. For this reason, while the 
form of the spermathecae are noted in this character, it is not suitable for use in 
phylogenetic analyses. 
 
[#151]. Sclerotization of spermathecae. After staining, the sclerotized portions 
maintained a light brown color to varying degrees. The amount of sclerotization 
varied continuously across the taxa of this dataset, however, degree of sclerotization 
was divided into three states. Spermathecae without any trace of brown coloration 
were scored as unsclerotized. Spermathecae with faint, light brown coloration were 
scored as lightly sclerotized. Spermathecae with brown or dark brown coloration were 
scored as sclerotized or well sclerotized. The difference between sclerotized and well 
sclerotized was most often ambiguous and for this reason, the two conditions were 
lumped into a single state. Although there may be weak phylogenetic signal in this 
character, it appears susceptible to relatively short term change. In Rhagio, for 
instance, there are differences across species and even within species. Some of these 
differences among specimens of the same species may be associated with differences 
in age. 
 
[#152]. Glandular hairs of spermathecae (not used in phylogenetic analysis). At 
high magnification, small glandular hairs may be visible at the distal tip of the 
spermathecae. This character is most obvious in Symphoromyia species. However, it 
is a difficult character to score for many taxa and the scoring may be positively 
correlated with spermathecal size and ease of observation rather than actual 
 
 138 
 
presence/absence. Although such hairs are obvious in extreme cases, this character 
may be inconsistent and unreliable. For this reason, it is determined unsuitable for 
phylogenetic analysis.  
 
[#153]. Spermathecal duct length. Spermathecal ducts varied in length and this 
aspect was scored. Duct length is measured from the point of insertion of the common 
spermathecal duct into the genital chamber to the distal tip of the spermathecae. Short 
spermathecal ducts are no more than three times the length of sternite 9. Moderate 
spermathecal ducts are more than three times but less than five times the length of 
sternite 9. Long spermathecal ducts are longer than five times the length of sternite 9, 
but not so long as to be difficult to measure. Very long spermathecal ducts fold upon 
themselves many times, as a pile of spaghetti. Pseudoerrina jonesi is the only taxon 
that could not be scored confidently as having one of these four states. The ducts in 
this species are approximately three times the length of sternite 9, the boundary at 
which two states are divided. Because of this, Pseudoerrina jonesi was scored as an 
ambiguity, having both of these states. 
 
[#154]. Spermathecal duct accessory glands (presence). There is what appears to 
be an accessory gland that arises from the spermathecal duct in species of 
Arthroceras, Ptiolina, Spania, Spaniopsis, Symphoromyia, and in some species of 
Chrysopilus. The spermathecal duct accessory glands may be very inconspicuous and 
are most easily seen after staining in chlorozol black. Perhaps on account of their 
inconspicuous nature, the spermathecal duct accessory glands have not been 
 
 139 
 
recognized by previous authors. Since the female genitalia of most of these taxa have 
been examined and illustrated by previous authors, this character was probably the 
most surprising of those developed for this dataset. Furthermore, such a character is 
likely to have phylogenetic utility, since functional structures such as accessory 
glands are unlikely to arise de novo very frequently. The shape of the accessory 
glands was approximately the same throughout. In Chrysopilus, species may or may 
not have spermathecal duct accessory glands. In addition to C. panamensis (Costa 
Rica) and C. quadratus (USA), C. alaskaensis Hardy also lacks spermathecal duct 
accessory glands. C. alaskaensis is unusual in that it has robust legs, unlike any other 
Chrysopilus species I've seen. It also shows a reduction of the flattened, metallic 
thoracic setae that are typical for Chrysopilus species and their allies, such as 
Schizella. It appears to me that C. alaskaensis is an atypical species, which survives 
the ecological pressures associated with the harsh Alaskan climate and has lost the 
spermathecal duct accessory glands secondarily. In addition to C. thoracicus and C. 
ferruginosa, spermathecal duct accessory glands are also present in C. calopterus 
(Brasil), C. rhagoides (Costa Rica) and C. testaceipes (USA). It is important that this 
structure is referred to as the 'spermathecal duct accessory gland' and not simply the 
'accessory gland' so as not to confuse it with the accessory gland posterior to the 
genital chamber, which is common to all flies. 
 
[#155]. Spermathecal duct accessory glands (origin). The spermathecal duct 
accessory gland, where present, varied in its placement along the spermathecal duct. 
The spermathecal duct accessory glands of Spaniopsis and Symphoromyia arise at 
 
 140 
 
approximately along the distal third of the spermathecal ducts. For many taxa, 
however, this character was variable within genera. However this character may be 
useful for supporting intrageneric phylogenetic relationships. Species of Ptiolina have 
either one of two states. Ptiolina edeta, P. mallochi, and P. zonata have spermathecal 
duct accessory glands that arise directly from the base of the spermathecae. In 
Ptiolina fasciata, P. lapponica, and P. majuscula, however, the spermathecal duct 
accessory glands arise at approximately the distal third of the spermathecal ducts. In 
Arthroceras pollinosum, the spermathecal duct accessory glands arise approximately 
halfway along the length of the spermathecal ducts however in Arthroceras 
fulvicorne, the glands arise at approximately the proximal third of the ducts. The 
position of the spermathecal duct accessory gland was also variable in Chrysopilus, 
where it may arise at the base of the spermathecae or at approximately the distal third 
of the spermathecal duct. This character is ordered. Due to the number of characters, 
it is also slightly downweighted so that the ordering does not exert undue influence on 
the structure of the tree. 
 
[#156]. Circular ridge of ejection apparatus. In many Tabanidae, there is a circular 
ridge at the distal end of the sclerotized ejection apparatus. This circular ridge is also 
present in most Athericidae. Bolbomyia nana was also scored as having this character 
state, although the ridge is more distinctly rounded than in tabanid and athericid taxa 
and may not be homologous. Similarly, the rounded ridge in Coenomyia ferruginea 
compelled me to score this state as present in the species. The ridge is not present in 
Dasyomma atratulum, however it is plainly visible in congener, D. coeruleum. 
 
 141 
 
 
[#157]. Sclerotized ring at base of spermathecal ducts. In Spania nana and species 
of Spaniopsis, a narrow ring of sclerotized tissue is present. Sometimes this is 
difficult to see, but presence/absence can be scored confidently after close 
examination. 
 
[#158]. Spermathecal duct swelling. In Bolbomyia, there is a distinct swelling 
approximately halfway along each spermathecal duct. Although this feature was not 
found in any other taxon, it was unusual enough to merit mention. 
 
[#159]. Ejection apparatus furrows (presence; not used in phylogenetic analysis). 
The ejection apparatus of the female (Evenhuis, 1989) may be recognized as a 
thickened area of the spermathecal ducts, near their base.  Furrows of the ejection 
apparatus may be arranged perpendicular to the length of the duct so that they appear 
as rings, or the furrows may be arranged at an angle, as indicated in the next 
character. The furrows are exceptionally minute and require very high magnification 
for viewing. On account of this, I was unable to confidently determine in all cases. 
For this reason, it was removed for phylogenetic analysis. 
 
[#160]. Ejection apparatus furrows (form; not used in phylogenetic analysis). 
Where the furrows of the ejection apparatus are present, the angle of the furrows were 
scored. The furrows may either run transversely, so that the spermathecal ducts 
appear ringed, or they may run at an angle from the center. On account of the high 
 
 142 
 
magnification required to detect the ultrastructural patterns of the spermathecal duct, 
scoring was not always made with the highest confidence. For this reason, it was 
removed for phylogenetic analysis. 
 
[#161]. Insertion of spermathecal ducts. In Arthroceras pollinosum, the 
spermathecal duct is partially folded internally, into the duct junction. In all other 
species in other genera, the ducts join smoothly. 
 
[#162]. Sclerotization of ejection apparatus. The ejection apparatus of the 
spermathecal ducts may or may not be sclerotized. Upon staining with chlorozol 
black, the spermathecal ducts dye blue, except in areas of sclerotization, which light 
brown or brown. Spermathecal ducts were scored as sclerotized if any brown 
coloration was detected near their base. 
 
[#163]. Base of spermathecal ducts. The spermathecal ducts may or may not be 
thickened near the junction with the common spermathecal duct. Where thickened, 
the ducts stain to a rich blue and have noticeably thicker walls. When not thickened, 
the ducts are a pale light blue. The spermathecal duct junction may or may not be 
thickened, in the same way. These two features are scored together because there 
seems to be some dependence between the states. In cases where the spermathecal 
ducts are enlarged, the thickness of the junction appears free to vary. However, in all 
cases where the spermathecal duct is not enlarged, the spermathecal duct junction is 
also not enlarged. Pelecorhynchus personatus is scored as not having a thickened 
 
 143 
 
spermathecal duct junction, however, in P. elegans, the junction is thickened. Judging 
by the amount of variation of this character within genera, it appears rather 
homoplasious. 
 
[#164]. Common spermathecal duct (presence). The spermathecal ducts most often 
originate from a single duct that leads from the genital chamber, which I call the 
common spermathecal duct. However, in some taxa, the common spermathecal duct 
is absent. Spermathecal ducts of Bolbomyia nana, Pseudoerrina jonesi, and Suragina 
concinna, for instance, lead directly to the genital chamber. In P. jonesi and B. nana, 
the ducts lead directly to the genital chamber. In S. concinna, the ducts converge at 
their base and meet the genital chamber very close to one another. Since B. nana has 
two spermathecal ducts whereas P. jonesi and S. concinna have three, the processes 
involved in the gain/loss of the common spermathecal duct may be different between 
them. In Pelecorhynchus personatus and Glutops rossi, the common spermathecal 
duct is so short as to be nearly absent. However it was scored as present in these taxa. 
 
[#165]. Common spermathecal duct (vermileonid modification). The common 
spermathecal duct sometimes retains dye very strongly, yet the blue color of the dye 
is still mottled and uneven, as if the structure is finely matted, micropilose, or with an 
uneven surface. This condition is present in species of Vermileo and Lampromyia. In 
other genera, where the common spermathecal duct is thickened, the dye stains 
smooth throughout and the common spermathecal ducts are tubular. It may appear 
that several things are being scored here; thin or thickened and ability to retain dye or 
 
 144 
 
not. However, I strongly doubt that the thickness of common spermathecal duct 
present in the vermileonids is homologous to the thickness seen in other taxa. In the 
vermileonids, the common spermathecal duct is much enlarged and is another 
condition, apart from being simply thickened. Retaining dye easily is an artifact of 
this condition that makes it easier to identify. 
 
[#166]. Common spermathecal duct enlargement. The thickness of the common 
spermathecal duct was scored, relative to the thickness of the spermathecal ducts, and 
scored for four states: 1) thinner than individual ducts, 2) not thickened, 
approximately as thick as an individual duct, 3) thickened to approximate girth of 
individual duct volumes combined, and 4) thickened to larger girth than individual 
duct volumes combined. The continuous nature of this character made it difficult to 
define boundaries between states. The character states were devised on account of the 
facility of scoring common spermathecal duct girth compared to the girth of 
individual spermathecal ducts, rather than in recognition of patterns of variation 
among taxa. Difference in character state scoring for summation of duct volume 
("equal to duct sum", or "slightly larger than this") is very similar. This difference, 
particularly in Pelecorhynchus personatus and Tabanus atratus is nearly negligible 
and may indicate that the character is divided too finely. This is also suggested by the 
high intrageneric variation, for example, as in Chrysopilus and Rhagio. This character 
was scored only for taxa with a duct-like common spermathecal duct. Lampromyia 
canariensis and Vermileo vermileo have a modified common spermathecal duct and 
therefore, are not scored here. 
 
 145 
 
 
[#167]. Common spermathecal duct, anterior end. Arthroteles bombyliiformis has 
a ridge at the apical end of the common spermathecal duct, where the individual ducts 
arise. This is present in no other taxa. 
 
[#168]. Common spermathecal duct length. The length of the spermathecal duct is 
also scored, relative to the largest diameter of the genital chamber. There are inherent 
problems of measurements of this sort, where the frame of reference (as the genital 
chamber) are free to vary. However, I have found this metric sufficient for scoring as 
they appear, from a general, 'gestalt' perspective. On account of the specially 
modified common spermathecal duct present in Lampromyia canariensis and 
Vermileo vermileo, these taxa are not scored for this character. This character was 
ordered for phylogenetic analysis. 
 
[#169]. Genital chamber, surrounding area. The genital chamber is a membranous 
pouch from which the spermathecal ducts arise. Generally, the area of this pouch is 
surrounded by membrane that is attached to the surrounding sternite 9 sclerite. 
However, in species of Ptiolina, Spania, Spaniopsis, and Symphoromyia, the area of 
the genital chamber is tightly defined by medial sclerotization of sternite 9. In 
Arthropeas americana and Xylophagus lugens, the genital chamber is hemmed in 
laterally by sternite 9 sclerotization, but a narrow membrane between the chamber 
and the sclerite precludes it from being scored as Ptiolina and its allies. 
 
 
 146 
 
[#170]. Genital chamber shape. When surrounded by membrane, the exact 
boundaries of the genital chamber, itself composed of membranous tissue, are often 
poorly defined. I found that scoring membranous boundaries of the genital chamber 
was particularly susceptible to subjective interpretation and experimental error or 
inconsistency. Therefore, this character was scored only for taxa where the genital 
chamber area was defined by sternite 9 sclerotization. In these cases, the shape of the 
genital chamber is clear because the sclerotization immediately around it takes an 
obvious shape. The shape of the genital chamber was scored as 1) circular, 2) 
teardrop- or eye- shaped, or 3) elongate, in part parallel-sided. Tear-dropped shape is 
oval, rounded posteriorly and coming to a point anteriorly. Eye-shaped is oval, 
coming to a point at posterior and anterior ends. Both species of Symphoromyia in 
this sample are scored as having an elongate, parallel-sided genital chamber, however 
in S. plagens, the genital chamber is teardrop shaped. 
 
[#171]. Genital chamber size (not used in phylogenetic analysis). The size of the 
genital chamber was scored relative to the size of sternite nine. Ultimately, however, 
the size of the genital chamber did not differ in any striking ways and I place little 
value in a character such as this. The size of the genital chamber varied continuously 
and differences in scoring were somewhat subjective. Therefore, it was removed for 
the phylogenetic analysis. 
 
[#172]. Accessory gland posterior to genital chamber (prominence; not used in 
phylogenetic analysis). The accessory gland posterior to the genital chamber is 
 
 147 
 
present in most flies. For this reason, the accessory gland attached to the spermathecal 
ducts must be called the "spermathecal duct accessory gland." If only "accessory 
gland" of the female genitalia is used, it would undoubtedly be confused with this 
structure, with which most people are familiar. In most cases, such as among species 
of Rhagio, the accessory gland posterior to the genital chamber is inconspicuous and 
visible only after extensive staining. However in some cases, particularly in the 
Xylophagidae, the accessory gland retains dye easily and is prominent. The scoring of 
the accessory glands as 'prominent' or 'not prominent' was determined subjectively, by 
eye. Unfortunately, however, there were gray areas of 'prominence' between the two 
extreme states exemplified by Rhagio and the Xylophagidae. In intermediate 
conditions, the state could be scored either as prominent or not prominent, with only 
subtle differences between them. Typically, as the morphology of the female 
genitalia, and particularly of the accessory gland, become more familiar, it was 
natural that even more lightly stained structures were perceived as more 'prominent.' 
The character, therefore, was prone to experimental error in scoring and is not 
adequate for phylogenetic analysis. 
 
[#173]. Common duct of accessory gland posterior to genital chamber 
(presence). Typically, the accessory gland posterior to the genital chamber has a 
common duct that is divided distally, to form two ducts. The common duct of the 
posterior accessory gland is absent in Tabanus and Pelecorhynchus, however. The 
paired accessory gland ducts arise from membrane of sternite 9 separately in these 
taxa. 
 
 148 
 
 
[#174]. Common duct of accessory gland posterior to genital chamber (form; not 
used in phylogenetic analysis). The accessory gland posterior to the genital chamber 
may be difficult to see since it often does not stain. It is also easy damaged during 
dissection. Therefore, I had little confidence in scoring this and other characters of the 
posterior accessory gland for a fair number of taxa. It appears that there are generally 
two forms of the accessory gland common duct. It may either lead smoothly to 
distally paired ducts or may forms an irregular, inflated shape at the point at which 
the slender distal ducts arise. However, since I am unsure as to the extent of 
experimental error for this character (and it may be significant), I prefer not to use 
this character for phylogenetic analysis. 
 
[#175]. Common duct of accessory gland posterior to genital chamber (length; 
not used in phylogenetic analysis). The accessory gland posterior to the genital 
chamber may be difficult to see since it often does not stain. It is also easy damaged 
during dissection. Therefore, I had little confidence in scoring this and other 
characters of the posterior accessory gland for a fair number of taxa. It appears that 
the accessory gland common duct consistently varies in length. It may either be quite 
long (clearly longer than sternite 9) or short (as long as or shorter than sternite 9). 
However, since I am unsure as to the extent of experimental error for this character 
(and it may be significant), I prefer not to use this character for phylogenetic analysis. 
 
 
 149 
 
[#176]. Paired extensions of accessory gland posterior to genital chamber 
(presence; not used in phylogenetic analysis). The accessory gland posterior to the 
genital chamber may be difficult to see since it often does not stain. It is also easy 
damaged during dissection. Therefore, I had little confidence in scoring this and other 
characters of the posterior accessory gland for a fair number of taxa. It appears that 
the distal form of the accessory gland common duct varies. There may or may not be 
paired ducts distally. However, since I am unsure as to the extent of experimental 
error for this character (and it may be significant), I prefer not to use this character for 
phylogenetic analysis. 
 
[#177]. Paired extensions of accessory gland posterior to genital chamber 
(length; not used in phylogenetic analysis). The accessory gland posterior to the 
genital chamber may be difficult to see since it often does not stain. It is also easy 
damaged during dissection. Therefore, I had little confidence in scoring this and other 
characters of the posterior accessory gland for a fair number of taxa. Where present, 
the paired extensions of the accessory glands may vary in length. However, since I 
am unsure as to the extent of experimental error for this character (and it may be 
significant), I prefer not to use this character for phylogenetic analysis. 
 
[#178]. Ventral chamber duct. In Xylophagus lugens, there is a long a ventral duct, 
very similar in form to the dorsal accessory gland posterior of the genital chamber. 
This structure was not seen in any other of the sampled taxa. 
 
 
 150 
 
[#179]. Sternite 9, anterior end. The shape of the proximal end of sternite 9 was 
quite variable among the taxa sampled, with several autapomorphic conditions. Taxa 
exhibiting sternite 9 with tapered, pointed tip were designated as (1) pointed. Rhagio 
species and their allies exemplify this state. Austroleptis multimaculata is also scored 
as pointed, although the anterior end of sternite 9 is tapered, then has a narrow 
parallel-sided projection that is somewhat different than what is found in other taxa 
scored as having this state. When the anterior end of sternite 9 was short, smooth and 
rounded it was characterized as (2) rounded. Species of Symphoromyia, Spaniopsis, 
and several Ptiolina species were scored as having a rounded sternite 9. Spania nigra 
also has a short, broad sternite 9 anterior margin and was scored as rounded, however 
it is slightly more angular than in the other taxa scored as having the same state for 
this character. Atherix pachypus exhibits a peculiar form of sternite 9 in which the tip 
is elongated into a narrow stem then is capped with a swollen, knob-like 
protuberance. This was scored as (3) stem and knob. Bolbomyia nana is also scored 
as having a stem and knob shape although the stem is broader than that in Atherix 
pachypus. Species of Dasyomma exhibit a peculiar form of sternite 9 in which the 
proximal end is forked as a swallowtail. This was scored as (4) forked. Dichelacera 
marginata and Pseudoerrina jonesi is also scored as forked, although the fork tines 
are very short in both of these species. In many species of Chrysopilus, the anterior 
end of sternite 9 is narrowed severely so that the extension is thin, and parallel sided. 
The stem is considered narrow because its width is less that one-quarter the width of 
sternite 9 at its middle. The tip is rounded. This condition is scored as (5) narrow, 
thin, with rounded point. A similar state is found in Arthropeas americana, 
 
 151 
 
Coenomyia ferruginea, and Dialysis rufithorax. In these taxa, the tapering is equally 
abrupt but much less extensive, so that the distally end remains broad. This condition 
is scored as (6) narrowed and broadly rounded. In Glutops rossi, the anterior end of 
sternite 9 is parallel-sided, without any taper, with a fully rounded tip. This state is 
unique among the taxa scored here and is referred to as (7) longer than wide, parallel-
sided, tip rounded. Xylophagus lugens, Ptiolina mallochi, and P. zonata have a 
similar morphology, scored as (8) expanded distally, and then truncated abruptly. In 
this state, the anterior end of sternite 9 smoothly broadens and is capped by a straight 
transverse margin. Pelecorhynchus personatus, Vermileo vermileo, and Lampromyia 
canariensis each have autapomorphic conditions that are scored as separate states. In 
P. personatus, the anterior end of sternite 9 is long, evenly tapered to narrow, rounded 
point, with wide, dark blue-staining membrane margins. In V. vermileo, sternite 9 is a 
very simple, nearly absent rounded sclerite. The sternite 9 of L. canariensis, is 
irregularly formed into a bulbous sclerotized sac. On account of this character having 
more than 9 states, it cannot be included in Bayesian analyses. 
 
[#180]. Sternite 9, posterior end. The posterior end of sternite 9 (posterior of the 
genital chamber) has three typical forms. "free posteriorly, with two components (one 
medially positioned, other laterally positioned)" indicates that there is a vertical and 
horizontal component to sternite 9, posterior to the genital chamber. These 
components may be entirely fused or apparently branching. "Free posteriorly, in 
single projection" indicates that sternite 9 is does not join posterior to the genital 
chamber and has only a horizontal (or near horizontal) component. 
 
 152 
 
 
[#181]. Sternite 9, posteromedial form. Species of Arthroteles and Atherimorpha 
are distinct in that the medial components of sternite 9 posterior to the genital 
chamber meet centrally. There, the components are held against one another, in the 
vertical plane. In many Rhagio species, the medial components of sternite 9 posterior 
to the genital chamber are angled and perhaps even vertical at times, as they abut the 
opening of the accessory gland duct. The medial components in Rhagio species do 
not meet centrally as in Arthroteles and Atherimorpha, but they are very similar in 
form. The grouping of Atherimorpha and Arthroteles has been suspected, but never 
supported by an explicit synapomorphy (Stuckenberg, 1956; Nagatomi, 1982a). 
 
[#182]. Sternite 9, posterior margin. Symphoromyia and Ptiolina share a similar 
morphology at the posterior margin of the sternite 9. In these taxa, there is a narrow, 
crescent shaped emargination, directly posterior to the genital chamber. This 
character is inapplicable for taxa whose sternite 9 posterior margin is free. 
 
Larval characters 
I examined multiple species of Chrysopilus, Rhagio, and Symphoromyia, 
however, most of these were not determined to species. I was able to get a fair 
understanding of the diversity within each genus, however, and therefore attempted to 
develop genus-level (or higher) characters for all larvae.  All species within each of  
these genera, then, were scored identically for larval characters. This includes 
Ptiolina, although only a single larval specimen was examined for this genus.  
 
 153 
 
 
[#184]. Retractile head. The retractile head is a putative synapomorphy for the 
Tabanomorpha (Woodley, 1989) 
 
[#185]. Head capsule withdrawn into second segment. The tabanid and athericid 
larvae are distinction in that the head is telescoped within the second thoracic 
segment. Note 2.  
 
[#186]. Dorsal plate. The larval head capsule may be composed of a single, 
undivided plate or divided, forming a pair of metacephalic rods. 
 
[#187]. Head capsule length. The head capsule is noticeably lengthened in Atherix, 
Dasyomma, Glutops, Pelecorhynchus, and Tabanus. In these taxa, the head capsule 
(or 'dorsal shield') is more than 4.5 times longer than greatest its width. 
 
[#188]. Head capsule shape. The xylophagid taxa have a distinctive, strongly 
sclerotized cone-shaped head capsule that is a putative synapomorphy for this clade 
(James, 1981; Palmer and Yeates, 2000). 
 
[#189]. Mandibular brush. Sinclair (1992) notes that the presence of a larval 
mandibular brush is a synapomorphy for the Tabanomorpha. Note 3. 
 
 
 154 
 
In Ptiolina, there is no mandibular brush, although the cuticle is rough and distinctive 
at the point at which the mandibular brush arises in related taxa. Hence, I presume 
that the distinctive cuticle is homologous to the area containing the mandibular brush 
in related taxa such as Symphoromyia. The lack of a mandibular brush may be 
evidence of a different feeding behavior, as the mandibular brush is thought to play 
an important role in the predatory feeding habit, by anchoring the larvae within its 
prey as it feeds. However, Vermileo, which is a predator, also lacks a mandibular 
brush. 
 
[#190]. Mandibular brush association. Sinclair (1992) proposed that the association 
of the mandibular brush with an articulated rod (=Sclerotized stem (Webb, 1977a) 
was a synapomorphy uniting Tabanidae and Athericidae. He also proposed that the 
rod was held vertical in Athericidae, whereas in Tabanidae, the rod was held 
horizontal, however I found the latter character to be an exceptionally variable 
condition. Taxa that do not have a mandibular brush are not scored here. The rod is 
articulated and can be moved up and down in both Athericidae and Tabanidae. I 
found that the position of this rod was held in any number of angles, and was largely 
independent of family grouping. This is corroborated by illustrations of this character 
in the literature, where the rod is shown in a variety of positions for both Athericidae 
and Tabanidae. Teskey (1981a: 69) illustrates Tabanus marginalis Fabricius with 
articulated rod in the vertical position; however, Pechuman & Teskey (1981: 473) 
illustrate the articulated rod in Tabanus reinwardtii in the horizontal position. Teskey 
(1970a: 1132) illustrates the articulated rod in Glutops (Pelecorhynchidae) in the 
 
 155 
 
vertical position. Courtney et al. (1997: 108) have a line drawing of Hybomitra 
epistates Osten Sacken (Tabanidae) with the articulated rod clearly in the vertical 
position. The articulated rod, and its position, however, is not visible in SEM photos 
of Atherix sp., Glutops sp., and Tabanus sp. (Courtney et al., 1997). My observations 
of the articulated rod associated with mandibular brush in various taxa are consistent 
with what is shown in these SEM photographs. The articulated rod is not always 
easily observed. It is often reduced to a small sclerotized stem, as Webb notes 
(1977a).  
 
When the mandibular sclerites are dissected, the articulated rod may be in the 
vertical, oblique, or horizontal position. This is true regardless of the taxon observed, 
whether it is Pelecorhynchidae, Athericidae, or Tabanidae. To some extent, this is 
expected, given what is known about the articulation, musculature, and putative 
function of this structure. On the nature of the articulated rod, Webb notes (1977a: 
479): "I follow the speculation of Teskey (1969) that these spines are utilized like 
those of tabanid larvae in anchoring the head of the larva while it feeds within its 
prey.? This is stated again in Pechuman and Teskey (1981: 464): "Mandibles? 
linked with subdorsal brushes of spines that are erected to anchor head within host 
when mandibles strike downward." Mackerras & Fuller (1942) note that the 
mandibular brush, "when the mouthparts are contracted and when protruded, they are 
free to project up and backwards." This is also explained in Courtney et al. (1997: 
107). Thus, when the mandibular brush is erect and used to anchor the head, the 
articulated rod is in the vertical position. Otherwise, the articulated rod is in the 
 
 156 
 
horizontal position. The capacity for mandibular brush movement is similar among 
pelecorhynchids, athericids, and tabanids and this is consistent with the observation 
of the position of the articulated rod in dissected larvae. For this reason, I chose not to 
attempt to score it the angle of the articulated rod. 
 
The sister clade relationship of Athericidae and Tabanidae is well established on the 
basis of adult characters (Stuckenberg, 1973). Some of the character states most often 
used to indicate their common ancestry are thorax with postspiracular scale, female 
cercus single segmented, and male genitalia with aedeagal tines. In larval form, the 
position of articulated rod may be an invalid morphological character for 
corroborating current hypotheses of shared ancestry. However, I observed at least one 
unique condition shared between larval athericids and tabanids. In all of the 
specimens I examined, the first cephalic segment is entirely telescoped within the 
second cephalic segment. When the head is pulled anteriorly, the first segment, 
containing the head, is drawn out and is noticeably different from the other larval 
body segments. The first segment is elongate cylindrical and more slender than any of 
the posterior segments. In all other taxa when the head is pulled anteriorly, it becomes 
detached from the first cephalic segment. 
 
[#191]. Mandibular brush orientation (not used in phylogenetic analysis). In 
larvae of Chrysopilus, Rhagio, and Symphoromyia, the mandibular brush clearly run 
transversely, between the antenna and labrum. In the larvae of Atherix, Dasyomma, 
and Tabanus, the mandibular brush is oriented along the sagittal plane. This character 
 
 157 
 
may be non-independent with respect to the mandibular brush being associated with 
an articulated rod or fold of cuticle. For this reason, it is not included in the 
phylogenetic analysis. 
 
[#192]. Mandibular brush density. Where the mandibular brush runs transversely 
(where it is associated with a cuticular fold), the brush may consist of few, thick 
setae, as in Rhagio. Or the mandibular brush may consist of many thin setae, as in 
Chrysopilus and Symphoromyia. 
 
[#193]. Mandibular hook. The adoral surface of the mandibular hook may be scored 
for three separate states. The adoral surface may be smooth, as in species of Rhagio 
and Ptiolina. Alternatively, the adoral surface may either have an external groove 
open along its length (as in species of Atherix, Chrysopilus, and Symphoromyia) or an 
internal channel at its base, opening apically (as in species of Atherix, Glutops, and 
Pelecorhynchus). 
 
Pelecorhynchidae, Athericidae, and Tabanidae putatively share the apomorphic 
character of having the mandibular hook with an internal canal, opening subapically. 
However, there is significant variation in the distribution of this character state among 
these taxa. In the Tabanus sp. that I have examined, the mandibular hook has a deep 
canal, similar to those in all athericids, but is only closed the very base. For the most 
part, it is an open canal, forming a deep groove. The canal starts nearly dorsally 
(subdorsally), runs along the adoral side, and ends subapically, along the dorsal 
 
 158 
 
surface. From an anterior perspective, the groove ends dorsally. Dasyomma species 
(Athericidae) shows the mandibular hook with a deep grove along its entire length, 
including at its base. Also, Halucitherix (primitive Athericidae) has an external 
groove, open along its entire length along the mandibular hook. 
 
In Courtney et al. (1997: 120), SEM photographs show Glutops as having a deep 
open groove subdorsally which is a closed channel along approximately the basal 
third of the mandibular hook. Atherix sp. and Tabanus sp. are shown to have a long 
internal channel with a relatively small, deeply channeled opening subapically. 
 
The mandibular hooks when medially grooved, are reportedly appressed together to 
form a food canal (Schremmer, 1951; Tsacas, 1962; Sinclair, 1992; Courtney et al., 
1997). This is known as the suctorial type (Stuckenberg, 2001; 'promuscis-type' of 
Sinclair, 1992; 'Saugmandible' of Schremmer, 1951). Sinclair (1992) has suggested 
that this is the basal condition for Brachycera, associated with the transition to 
vertical mobility in the larval mandibles (Sinclair, 1992: 237-238). 
 
The evolution of mandibular hooks functioning as a feeding tube seems to require at 
least three integrated morphological features. One, the adoral surface of the 
mandibular hooks must complement on another in shape, so that they meet flush, 
centrally. Two, the groove on the adoral surface of the mandibular hook must match 
so that when the hooks are appressed together, a channel is formed. Three, there must 
 
 159 
 
be some development of musculature to support lateral movement of the hooks, 
allowing them to come together and stay together against counteracting forces. 
 
Tsacas (1962) reports having observed a feeding tube formed by the mandibular 
hooks being appressed medially in Vermileo. Initially, this seems difficult to believe, 
as the mouthparts of this larva are approximately 0.1 mm in length and are 
submerged, along with most of its head, within the prey as it feeds. However, the 
SEM images in Ludwig et al. (1996) are consistent with the feeding tube as illustrated 
and explained by Schremmer (1951). This is vastly different than what is seen in 
Vermileo comstocki, however. In V. comstocki, the mandibles are blunt, reduced in 
size, and do not meet medially. The canal or groove is difficult to see, if present in 
this species. In my observations, I have seen what may be a closed channel in one of 
two specimens of V. comstocki. 
 
In my opinion, the highly specialized feeding behavior of Vermileo and the 
inconsistency of this character within the genus (or at least among close relatives if 
Vermileo proves to be paraphyletic) discounts the likelihood that the mandibular hook 
feeding tube is necessarily associated with the vertical movement of the mandibles 
that defines brachyceran larvae. In most taxa within the Tabanomorpha, it appears 
that a groove in the mandibular hook is not a specialized condition for feeding. The 
hooks may have a medial groove simply to allow a passageway for food to pass to the 
hypopharanx when the mandibles are incidentally pushed together during feeding.  It 
doesn?t make sense that the mandibles are affixed together medially to form a feeding 
 
 160 
 
tube when, as Teskey (1970: 1131) writes, in most taxa the mandibles work in 
conjunction with the maxillae, as a unit, in a vertical manner, and may operate 
independent of the other pair (this is also mentioned in Courtney et al., 1997: 105).  
 
Regarding Vermileonidae, Nagatomi et al. (1998: 142) stated "it seems that the 
poison canal in Vermileonidae occurs independently from that in the 
Pelecorhynchidae, Athericidae, and Tabanidae,? an apparent misinterpretation of 
Ludwig's description of the mandibular hook of Vermileo vermileo. Tsacas (1962: 
214) and Ludwig (1996) observe that Vermileo vermileo and Vermileo sp., 
respectively, have an open groove (not a partially enclosed canal) along the adoral 
side of the mandibular hook. Sinclair (1992) doesn?t describe the mandibular hook in 
Vermileo specifically, aside from saying that the genus displays the groundplan larval 
condition. Since distantly related taxa such Xylophagus, Symphoromyia, and Thereva 
share having an adoral groove on the mandibular hook, one presumes that this is the 
case also for the Vermileo species that he examined. 
 
In at least some Chrysopilus species, for example, the groove on the adoral surface 
begins subdorsally, runs centrally, along the adoral surface, and ends subapically at a 
subdorsal position. The channel on each side, therefore, may meet at least along part 
of its length. However, along most of its length, the channel does not match its 
complement and cannot form a food canal (contrary to Tsacas, 1962). The plausibility 
of a mandibular hook food canal particularly suffers when considering aquatic 
 
 161 
 
environments. Would a food canal work in water for such aquatic species such as 
Chrysopilus tsacasi (Thomas, 1978b)?  
 
The mandibular hooks in Ptiolina are small, blunt, and without an external groove 
along its adoral surface. It appears unlikely that the hooks would be used in the same 
way as the sharp, serrated hooks in Rhagio, Atherix, Tabanus, and others. The 
mandibular hook of Rhagio has neither an external groove nor a partially closed 
channel. It is hollow, however. 
 
Stuckenberg (2001: 11) opined "the ?poison canal? was stated to be hypothetical so 
has no operational significance for cladistic analysis." I must respond by saying that 
while the function of the grooves/canal of the mandibular hook are not certain the 
corresponding structures (regardless of function) must not be rejected a priori as unfit 
for cladistic analysis. Shared presence of physical features with apparent positional, 
structural, and ontological homologies, regardless of the function of these features, 
may provide useful evidence for evaluating phylogenetic hypotheses. If the 
grooves/canal of the mandibular hook do not serve as a poison canal in the 
Pelecorhynchidae and related taxa, as Stuckenberg suggests, this may lend even more 
weight as a useful character to score for phylogenetic analyses since the character is 
less prone to functional and environmental restraints which may change and force 
change, independently of ancestral lineages. 
 
 
 162 
 
In Rhagio, the maxilla is closely associated with the mandibular hook. Ventrally, the 
lacinia envelops the hook so that an apparent channel is formed on the adoral side. 
This feature is most easily seen in dissected specimens that have not been cleared. 
This may assist in food uptake and compensate for not having a groove in the 
mandibular hook. Such a feature is not apparently present in Symphoromyia, 
however, in both this genus and Ptiolina, the maxilla are also closely associated with 
the mandibular hook. In Atherix, Dasyomma, and Tabanus, the mandibular hook is 
pronounced, exserted from the rest of the mouthparts, including the maxillae. 
 
[#194]. Sclerotization of maxilla (not used for phylogenetic analysis). The 
sclerotization of the maxilla was more prominent in some taxa (e.g., Symphoromyia) 
than in others (e.g., Vermileo). However, this feature is difficult to score consistently, 
as the amount of the sclerotization is apt to vary and the point at which it is scored or 
not scored as sclerotized is not clear. SEM images cannot help in determining the 
state of this character. For these reasons, it was not used in phylogenetic analyses. 
 
[#195]. Saw sclerite of basal mandibular sclerite. The saw sclerite is an 
autapomorphy for species of Rhagio. It is located in a ventral position, adoral to the 
basal mandibular sclerite. 
 
[#196]. Maxillary palp form. In species of Athericidae, maxillary palp has a 
distinctive form. The maxillary palp is sclerotized, thin and tubular, with the 
 
 163 
 
appearance of a cigarette. In other taxa, the maxillary palp is soft, not sclerotized, and 
its segments are poorly differentiated. 
 
[#197]. Maxillary palp segment number (not used in phylogenetic analysis). The 
maxillary palp is usually not sclerotized and it is difficult to identify segment breaks, 
if present. For this reason, it was not used in phylogenetic analysis. 
 
[#198]. Last segment of antenna. The terminal segment of the larval antenna may be 
split into two or more extensions. This character is most easily scored from SEM 
images. 
 
[#199]. Number of antennal segments (not used in phylogenetic analysis). The 
number of antennal segments may vary, the antenna is often unsclerotized and it is 
difficult to judge how many segments are present. For this reason, it was not included 
in phylogenetic analysis. 
 
[#200]. Unpaired salivary pump. In some larvae, there is a sac that is connected to 
the mouthparts via a long, central duct. It is difficult to follow to track the origin of 
this duct and verify to which mouthpart structure it is attached. The sac stains easily 
and is readily apparent in athericids and tabanids. I suspect that this is the 'huge, 
unpaired salivary pump" mentioned by Mainx (in Stuckenberg, 2001: 7). At the 
anterior margin of this pump is a pair of giant cells that Schremmer (1951) suspected 
were poison glands. The glands are not apparently present in Glutops, however, and 
 
 164 
 
this potentially conflicts with the scoring of Sinclair (1992) and Woodley (1989). 
Since pelecorhynchids have poison canals (interpreted from partially enclosed 
mandible hooks), it logically follows that this sac structure should be present in 
Glutops.  
 
Stuckenberg (2001) has refuted the contention that these are poison glands, arguing 
that in the original text, Schremmer (1951) only suspected that a pair of giant cells of 
salivary-gland nature represented poison glands. Furthermore, Stuckenberg (2001) 
pointed out, the gland ducts of these giant cells have not been followed to their 
anterior endpoint and therefore, an association with the mandibles has not been 
verified. This remains true, as I was unable to track the gland ducts directly to the 
mandibles and I could not see glandular cells with complete confidence, upon 
inspection with high magnification. These observations do not preclude the 
possibility that the sac and its associated elements form the apparatus for delivering 
poison into larval prey, however. 
 
[#201]. Posterior tentorial expansion. The posterior tentorial expansion has two 
lateral arms that run for approximately the length of the dorsal plate. Posteriorly, the 
expansions may be free, fused to one another, or fused to the dorsal plate. 
 
[#202]. Anterior margin of thorax. The anterior margin of the first thoracic sclerite 
of Ptiolina and Symphoromyia bears a distinctive ultrastructural pattern. In these taxa, 
 
 165 
 
the surface area immediately posterior to the head is scalloped or scaled. This is not 
found in any of the other sampled taxa. 
 
[#203]. Thoracic segments, ventrally. Athericidae, in particular, are distinct in 
having crocheted prolegs which are used to facilitate movement in rushing streams. 
Glutops, Pelecorhynchus, Symphoromyia, and the xylophagid taxa are smooth 
ventrally, whereas Chrysopilus, Ptiolina, and Rhagio have what are called 'creeping 
welts' on the ventral surface of the thoracic region. These welts are used to facilitate 
traction during movement. 
 
[#204]. Thick, waxy integument. The larvae of Pelecorhynchus and Glutops species 
have a tough, waxy covering that is unlike that found in other known tabanomorph 
larvae. Teskey (1970) noted "the shapes and relative sizes of the segments [of 
Glutops and Pelecorhynchus larvae] and the characteristics of their integuments are 
essentially identical."  
 
[#205]. Spiracular system (not used in phylogenetic analysis). The spiracular 
system may be apneustic (no spiracles present), propneustic (paired spiracles at front 
only), metapneustic (paired spiracles at back only), or amphipneustic (paired spiracles 
at front and back). However, this character was surprisingly hard to score because the 
spiracles are small and inconspicuous, and therefore are often difficult to find. I do 
not have a lot of confidence in the scoring this character and therefore do not use it in 
the phylogenetic analysis. 
 
 166 
 
 
[#206]. Hind segment sclerotization. The xylophagid taxa are unique among taxa 
sampled here in having the hind segment sclerotized. 
 
[#207]. Spiracular disc. The spiracular disc is the posterior face of the terminal 
segment which bears the terminal spiracles. Where present, it has a broad, open 
surface, as in Chrysopilus, Ptiolina, Rhagio, and Symphoromyia. 
 
[#208]. Hind segment form. The hind segment may be produced into elongate 
respiratory siphon; be smooth, elongate oval (from the dorsal/ventral perspective) or 
longitudinally striated, with an inflated appearance. 
 
 167 
 
 
Table 4. A list of the morphological characters and their state coding. Characters not 
used in the phylogenetic analyses are in italics. 
Character Character States 
#1. <Gender known>  1. male and female known  
2. only female known available  
3. only male known available  
 
#2. Clypeus  1. bulbous  
2. not bulbous  
 
#3. Scape  1. smaller than pedicel  
2. approximately the same size as pedicel  
3. clearly larger than pedicel  
 
#4. First flagellomere 
<laterally 
compressed?>  
1. laterally compressed  
2. rounded in cross section  
 
#5. <Antenna 
composition>  
 
1. antenna with many flagellomeres of similar shape and size, 
tapering distally  
2. first flagellomere of antenna enlarged bearing segmented 
stylus  
3. first flagellomere of antenna enlarged bearing stylus of single 
segment  
4. first flagellomere of antenna enlarged basally, bearing fused or 
distinct arista-like extension  
 
#6. Antennal 
flagellum <break 
after first 
flagellomere?>  
 
1. gradually tapered  
2. with abrupt change between the first flagellomere and those 
distal to it  
 
#7. Segment(s) distal 
of first flagellomere 
<in cases where 
abrupt change 
evident>  
 
1. segmented flagellomeres  
2. fused or distinct stylus of single segment  
3. fused or distinct arista  
 
#8. Arista 
<microsetose?>  
 
1. microsetose  
2. bare  
 
 
#9. Eyes 1. conspicuously microsetose  
 
 168 
 
Character Character States 
<microsetose>  
 
2. inconspicuously microsetose  
 
#10. Eyes in male 
<separation>  
 
1. holoptic  
2. dichoptic  
 
#11. Eyes in male 
<flattened dorsally>  
 
1. flattened dorsally  
2. not flattened dorsally  
 
#12. Eyes in male 
<facets>  
 
1. facets evenly distributed, of equal size  
2. facets evenly distributed, gradually smaller toward ventral 
margin  
3. facets split into upper and lower areas and smaller (more fine) 
ventrally  
 
#13. Parafacials in 
male <swollen?>  
 
1. not swollen  
2. swollen  
 
#14. Occiput <form>  
 
1. rounded, with smooth transition from dorsal to posterior part of 
head  
2. flattened, concave posteriorly  
 
#15. Head <width>  
 
1. wider than thorax  
2. approximately same width as thorax  
3. narrower than thorax  
 
External Mouthparts 
#16. Labellum <size>  
 
1. longer than palps  
2. about the same size as palps  
3. shorter than palps  
 
#17. Pseudotracheae 
<presence>  
 
1. present  
2. absent  
 
#18. Theca 
<elongate?>  
 
1. elongate  
2. short  
 
#19. Theca lateral 
sclerites <fused, 
adjacent with 
suture, touching or 
separated?>  
 
1. separate  
2. adjacent and touching, but mostly separated  
3. tightly adjacent, apparently fused with suture  
4. fused into single sclerite  
 
#20. Theca <sclerite 
composition>  
 
1. lateral sclerites separate  
2. lateral sclerites adjacent and or separated by medial suture  
3. fused into single sclerite, without medial suture  
 
 169 
 
Character Character States 
 
#21. Palps <segment 
number>  
 
1. one-segmented  
2. two-segmented  
3. three-segmented  
 
#22. Palps <relative 
length of each 
segment, if two>  
 
1. proximal segment longer than distal segment  
2. proximal and distal segments about the same length  
3. distal segment longer than proximal segment  
 
#23. Lateral ridge of 
clypeus <presence>  
 
1. absent  
2. present, subtending mouthparts basally  
 
#24. Oral margin 
<location>  
 
1. restricted to anterior part of head  
2. wraps around base to posterior portion of head  
 
#25. Stipes <position>  
 
1. apparently absent  
2. converging toward one another, directed medially  
3. attached to on in contact with tentorium at back side of head  
4. surrounded by membrane above theca, directed posteriorly  
 
#26. Cardo 
<presence>  
 
1. apparently absent  
2. not swollen  
3. swollen  
 
#27. Lacinia <length>  
 
1. shorter than palps  
2. longer than palps  
 
#28. Lacinia 
<serrated?>  
 
1. tip serrated  
2. tip not serrated  
 
#29. Mandibles 
<presence>  
 
1. absent  
2. present  
 
Internal mouthparts 
#30. Cibarial pump 
<length>  
 
1. short, as wide as long or wider  
2. long, clearly not as wide as long  
 
#31. Cornu <length>  
 
1. shorter than cibarial pump  
2. nearly as long as or longer than cibarial pump  
 
#32. Cornu 
<attachment>  
 
1. extending beyond pharyngeal pump  
2. apically fused to pharyngeal pump  
 
#33. Pharyngeal pump 
<width>  
1. narrow along most of length  
2. anteriorly broad  
 
 170 
 
Character Character States 
  
#34. Pharyngeal pump 
<presence of cup>  
 
1. forms cup-like structure  
2. mostly flat along its length  
 
#35. Pharyngeal pump 
<length>  
 
1. longer than length of cibarial pump  
2. approximately same length as cibarial pump  
3. approximately half length of cibarial pump  
4. short, reduced  
 
Thorax 
#36. Dorsum 
<vittate?>  
 
1. with vittae  
2. without vittae  
 
#37. Setae of dorsum 
<present>  
 
1. all of equal length  
2. acrostichal, dorsocentral, and intra-alar setae longer, 
distinguished from other setae of dorsum  
 
#38. Metallic- or 
scale-like thoracic 
setae, often with 
structural color 
<present absent>  
 
1. absent  
2. present  
 
#39. Proepimeron 
<reduced or 
absent?>  
 
1. absent  
2. reduced  
3. present  
 
#40. Proepimeron 
<setation>  
 
1. setose  
2. bare  
 
#41. Anepisternum 
<setation>  
 
1. bare  
2. with one or two setae  
3. setose on upper margin only  
4. setose on dorsal and anterior margins  
5. setose on dorsal and posterior margins  
6. setose throughout posterior half  
7. setose throughout upper half, including entire anterior and 
posterior margins  
8. setose throughout entire sclerite  
 
#42. Anepisternum 
<setation2>  
 
1. bare  
2. with one or two setae  
3. variously setose along margins or throughout dorsal or 
posterior half of sclerite  
4. setose throughout entire sclerite  
 
 171 
 
Character Character States 
 
#43. Anepisternum 
<setation3>  
 
1. bare or with one or two setae  
2. less than one-half of sclerite area with setae  
3. one-half or more of sclerite area with setae  
 
#44. Laterotergite 
<form>  
 
1. katatergite swollen, differentiated from anatergite  
2. katatergite and anatergite indistinguishable  
 
#45. Laterotergite 
<setation>  
 
1. bare  
2. setose  
 
#46. Laterotergite 
<arrangement of 
setae>  
 
1. present throughout laterotergite  
2. present mostly on katatergite  
3. present on katatergite only  
 
#47. Microsetation 
between base of 
halter and 
postspiracular 
sclerite  
 
1. absent  
2. present  
 
#48. Thoracic spiracle 
<flaps below 
laterotergite?>  
 
1. without flaps  
2. with flaps  
 
#49. Thoracic spiracle 
<lined with 
microsetae>  
 
1. not lined with microsetae  
2. lined with microsetae 
 
#50. Postspiracular 
scale <presence>  
 
1. absent  
2. present  
 
#51. Postspiracular 
sclerite <setae>  
 
1. without setae  
2. setose  
 
#52. Thoracic surface 
immediately 
posterior to 
postspiracular 
sclerite <presence of 
setae>  
 
1. bare  
2. setose  
 
#53. Proscutellum 
<lenticular swelling 
between the 
1. present  
2. absent  
 
 
 172 
 
Character Character States 
scutellum and 
mesoscutum>  
 
#54. Subscutellum 
<form>  
 
1. not bulbous  
2. bulbous  
 
#55. Subscutellum 
<setose?>  
 
1. bare  
2. setose on margins  
 
#56. Postmetacoxal 
bridge <presence>  
 
1. absent  
2. present  
 
#57. Postmetacoxal 
bridge <type>  
 
1. complete broad extension  
2. incomplete broad extension  
3. complete thin extension  
4. incomplete thin extension  
 
Wing 
#58. Wing <surface 
membrane>  
 
1. hyaline  
2. membrane lightly infuscated  
3. membrane darkly infuscated  
 
#59. Wing <presence 
of markings>  
 
1. with markings  
2. without markings  
 
#60. Wing <presence 
of pterostigma>  
 
1. with pterostigma  
2. without pterostigma  
 
#61. Lower calypter  
 
1. reduced  
2. present  
 
#62. Upper calypter 
<development>  
 
1. absent  
2. well developed, full and rounded  
3. underdeveloped, with a straight edge, appearing triangular in 
form  
 
#63. Upper calypter 
<shape in outline>  
 
1. with broad curvature, lobe-like, width twice length or less  
2. with reduced curvature, width more than twice length  
3. with narrow curvature, appearing stretched, width more than 
twice length  
 
#64. Alula 
<development>  
 
1. no curvature, reduced  
2. narrow curvature  
3. broad curvature  
 
 
 173 
 
Character Character States 
#65. Alula <shape in 
outline>  
 
1. curvature shifted distally so that lobe appears more triangular 
than round  
2. rounded evenly  
 
#66. Anal lobe 
<development>  
 
1. well developed  
2. reduced  
 
#67. Humeral 
crossvein (h) 
<strength>  
 
1. well developed  
2. weakly developed  
 
#68. Humeral 
crossvein <shape>  
 
1. bow-shaped  
2. straight, acute angle apically at wing margin  
3. straight, obtuse angle apically at wing margin  
4. straight, perpendicular to wing margin  
 
#69. Sc-r crossvein 
<presence>  
 
1. present, well developed  
2. present, weakly developed  
3. absent  
 
#70. Sc-r crossvein 
<location>  
 
1. proximal side of humeral crossvein (h), by less than length of h 
2. positioned distal to the humeral crossvein (h), by less than the 
length of h  
3. positioned distal to the humeral crossvein (h), by the 
approximate length of h  
4. positioned distal to the humeral crossvein (h), by more than the 
length of h  
5. located approximately midway between h and the origin of the 
radial-sector  
 
#71. Dorsal surface of 
R
1
 <setation>  
 
1. microsetose  
2. without microsetae  
 
#72. Ventral surface  
of R
1
 <setation>  
 
1. without microsetae  
2.microsetose  
 
#73. Wing veins R
1
 
and R
2+3
 <at wing 
margin>  
 
1. meet together at wing margin  
2. close together at wing margin (R
2+3
 clearly closer to R
1
 than to 
R4)  
3. separated at wing margin (R
2+3
 closer to R
1
 than to R
4
, but 
placed nearly between R
1
 and R
4
)  
4. widely separated at wing margin (R
2+3
 closer to R
4
 than to R
1
)  
 
#74. R
2+3
 curve 
<present>  
1. present, beneath pterostigma, as illustrated by Stuckenberg 
(2001:fig 11)  
 
 174 
 
Character Character States 
 2. absent  
 
#75. Wing vein R
2+3
 
<direction at 
margin>  
 
1. directed posteriorly, meeting wing at close angle  
2. directed toward wing margin, meeting margin abruptly  
3. directed posteriorly, but turned into the wing margin at 
terminus  
 
#76. Dorsal surface of 
R
2+3
 <setation>  
 
1. microsetose  
2. without microsetae  
 
#77. Ventral surface of 
R
2+3
 <setation>  
 
1. microsetose  
2. without microsetae  
 
#78. R
4
-R
5
 fork 
<presence>  
 
1. present  
2. absent  
 
#79. Base of R
4
-R
5
 
fork <location>  
 
 
1. distal of distal end of cell dm  
2. proximal or directly above distal end of cell dm  
#80. R
4
 at base 
<strongly curved?>  
 
1. strongly curved or angled  
2. relaxed, not strongly curved  
3. nearly straight  
 
#81. R
4
 <apical 
portion>  
 
1. straight or nearly straight apically  
2. curving towards the leading margin of the wing apically  
 
#82. <R
4
 and R
5
 
position at margin>  
 
1. R
5
 anterior to or ending at wing tip  
2. R
4
 and R
5
 encompass wing tip  
3. R
4
 ending at or posterior to wing tip  
 
#83. R
5
 <aligned with 
R
4+5
?>  
 
1. aligned with R
4+5
  
2. changing direction only slightly at fork with R
4+5
  
3. clearly changing direction at fork with R
4+5
  
 
#84. M
3
 wing vein 
<presence>  
 
1. present, reaching wing margin  
2. incompletely present, not reaching wing margin  
3. absent  
 
#85. Halter knob 
<size; begins at 
base of depression 
or oval form>  
 
1. 1/3 or shorter than length of stem  
2. between 1/3?1/2 length of stem  
3. approximately 1/2 length of stem  
4. between 1/2?2/3 length of stem  
5. 2/3 or longer than length of stem  
 
 
 175 
 
Character Character States 
#86. Halter stem <cell 
arrangement>  
 
1. with clear parallel rows of fine microsetae  
2. with rows or ripples, but not clearly parallel entire length at 
base of stem  
3. without any parallel rows of microsetae  
 
#87. Anterodorsal side 
of halter <setation>  
 
1. bare  
2. with thick row of setae running along trailing edge of stem  
3. with thick row of setae running along leading edge of stem and 
setae present in row along trailing edge of knob  
4. mostly bare with few setae at base  
5. mostly bare with few, fine setae running along trailing edge  
6. longish setae at base of halteres, otherwise bare  
 
#88. Ventroposterior 
side of halter 
<setation>  
 
1. bare  
2. microsetae running along trailing edge in row  
3. microsetae loosely arranged at base and near trailing edge  
4. microsetae loosely arranged on mostly basal half and running 
along trailing edge of knob in row  
5. microsetae mostly present in apical half of knob  
6. microsetae running along both edges  
 
Legs 
#89. Fore tibial spur  
 
1. absent  
2. one  
 
#90. Mid tibial spur  
 
1. absent  
2. one  
3. two  
 
#91. Hind tibial spur  
 
1. absent  
2. one  
3. two  
 
#92. Hind coxal 
tubercle <presence>  
 
1. absent  
2. present  
 
#93. Hind tibial 
macrochaetae 
<presence>  
 
1. absent  
2. present; small, easily overlooked, nearly flush with sclerite 
surface  
3. present; conspicuously present, obviously enlarged compared 
with other tibial setae  
 
#94. Hind tibial 
macrochaetae 
<form>  
1. short  
2. lengthened  
 
 
 176 
 
Character Character States 
 
#95. Hind tibia 
<presence of ventro-
apical swelling>  
 
1. without ventro-apical swelling  
2. with ventro-apical swelling  
 
#96. First hind 
metatarsus of male 
<swollen or not?>  
 
1. not swollen  
2. swollen  
 
Male Genitalia 
#97. Epandrial sclerite 
<form>  
 
1. wider than long  
2. longer than wide  
3. absent  
 
#98. Epandrial sclerite 
<anterior margin>  
(not scored where 
epandrial sclerite absent) 
 
1. strongly notched anteriorly  
2. modestly curved anteriorly  
3. not emarginate anteriorly  
 
#99. Epandrium 
<form, from 
posterior view>  
(not scored where 
epandrial sclerite absent) 
 
1. simple, not containing hypandrium ventrally  
2. wrapped laterally, surrounding cerci and hypoproct posteriorly 
and containing hypandrium ventrally  
#100. Cercus 
<attachment point>  
 
1. attached to subepandrial complex (subepandrial membrane or 
hypoproct)  
2. attached directly to epandrium via membrane  
 
#101. Subepandrial 
sclerite <presence>  
 
1. absent  
2. present as membrane  
3. present as sclerite  
 
#102. Subepandrial 
sclerite <form>  
 
1. undivided  
2. divided  
 
#103. Subepandrial 
sclerite <setation>  
 
1. bare  
2. setose  
 
#104. Hypoproct 
<form; outline>  
 
1. triangular (rounded posteriorly)  
2. sagittate  
3. two-pronged posteriorly  
4. rectangular; wider than long  
5. elongated; approximately twice as long as wide  
6. oval  
7. pentagonal  
 
 177 
 
Character Character States 
 
#105. Hypoproct 
<form, presence of 
lateral 
sclerotization?>  
 
1. lobes absent  
2. anterior and or posterior lobes present  
 
#106. Hypoproct 
<setae>  
 
1. setose  
2. tomentose, without setae  
 
#107. Cercus <leaf 
shaped or nearly 
square>  
 
1. squarish  
2. nearly leaf-shaped  
 
#108. Cerci 
<separation>  
 
1. directly adjacent to one another, separation distance one quarter 
width of cercus or less  
2. partially displaced from one another, separation distance 
approximately half the width of single cercus  
3. widely displaced from one another, separation distance greater 
than three quarters width of cercus  
 
#109. Cerci held 
<orientation>  
 
1. horizontal in relation to rest of abdomen  
2. at angle in relation to rest of abdomen  
3. vertical in relation to rest of abdomen  
 
#110. Cerci, in 
posterior view 
<form; curved or 
straight in posterior 
view>  
 
1. cupped, forming circular outline medially  
2. flat  
 
#111. Hypandrial 
sclerite  
 
1. fused entirely to gonocoxites  
2. separated partially from the gonocoxites by an incomplete 
suture  
3. separated from the gonocoxites by a complete suture  
 
#112. Gonocoxite 
<dorsal sinuous 
ridge presence>  
 
1. dorsal sinuous ridge present, leading to gonocoxal apodeme  
2. smooth dorsally, without sinuous ridge leading to gonocoxal 
apodeme  
 
#113. Gonocoxal 
apodemes 
<presence>  
 
1. present  
2. absent  
 
#114. Gonocoxal 
apodemes <length>  
1. short or long enough to reach anterior margin of hypandrium  
2. extending well beyond anterior margin of hypandrium  
 
 178 
 
Character Character States 
  
#115. Gonocoxal 
apodeme <origin>  
 
1. basal medial margin posterior to or at approximately the same 
transverse plane as basal lateral margin  
2. basal medial margin anterior to basal lateral margin  
 
#116. Parameral 
sheath <form>  
 
1. not developed into bulbous sac ventrally  
2. developed into bulbous sac ventrally, without distinct lobes  
3. developed into bulbous sac ventrally, with distinct lobes  
 
#117. Lateral 
ejaculatory 
processes 
<presence>  
 
1. absent  
2. present, integrated into sperm sac membrane  
3. present, not part of sperm sac posteriorly (homology uncertain) 
 
#118. Ejaculatory 
apodeme <length>  
 
1. reduced, nearly absent  
2. short, not reaching anterior margin of hypandrium  
3. moderately long, reaching anterior margin of hypandrium  
4. long, reaching beyond anterior margin of hypandrium  
 
#119. Ejaculatory 
apodeme <form>  
 
1. tubular  
2. laterally compressed  
3. compressed dorso-ventrally  
4. tripartite; dorsally compressed laterally and ventrally 
compressed dorso-ventrally  
5. umbrella-shaped anteriorly  
 
#120. Aedeagal tines 
<presence>  
 
1. absent  
2. present  
 
#121. Endoaedeagal 
process <presence>  
 
1. present  
2. absent  
 
#122. Endoaedeagal 
process <form, 
compressed?>  
 
1. strongly lateral compressed, as a butterknife  
2. rounded, narrowly conical  
3. dorso-ventrally flattened  
 
#123. Gonostylus 
<setation>  
 
1. heavily setose  
2. lightly setose or nearly bare  
 
Female abdomen and terminalia 
#124. Terminal 
abdominal segments  
 
1. evenly tapered  
2. narrowed and telescoped within wider basal abdominal 
segments  
 
#125. Tergite 7 1. much longer than wide  
 
 179 
 
Character Character States 
<shape>  
 
2. about as long as wide  
3. much wider than long  
 
#126. Intersegmental 
membrane between 
segments 7 and 8 
<presence>  
 
1. especially long  
2. short, as throughout abdomen  
 
#127. Tergite 9 
<length>  
 
1. not reduced  
2. reduced  
 
#128. Tergite 9 
<projections 
enveloping sternite 
9>  
 
1. normal, without such projections  
2. with narrow anteriorly-directed ventrolateral projections, 
enveloping sternite 9  
 
#129. Tergite 10 
<length>  
 
1. absent  
2. present, reduced (length less than half width)  
3. present, not reduced (length approximately equal to measured 
half width)  
4. present, not reduced (length more than half width)  
 
#130. Tergite 10 
<split?>  
 
1. entire  
2. split into two separate lateral sclerites  
 
#131. Sternite 8 
<divided?>  
 
1. sclerite entire, not divided into two segments  
2. sclerite divided into two segments, anterior segment long and 
wide, posterior segment rounded, cupped  
 
#132. Sternite 8 
cleavage <presence>  
 
1. present  
2. absent  
 
#133. Sternite 8 length 
<aspect ratio>  
 
1. wider than long  
2. as wide as long  
3. longer than wide  
4. elongated; more than twice as long as wide  
5. wider than long or as wide as long (triangular, ovoid, or nearly 
square)  
 
#134. Sternite 10 
<form>  
 
1. entire  
2. sclerotization weakened centrally, making it appear as if 
sclerite divided into two lateral components  
3. split into two sclerites  
 
#135. Sternite 10 1. triangular; nearly split into two sclerites  
 
 180 
 
Character Character States 
<shape>  
 
2. roughly pentagonal, pointed posteriorly  
3. roughly rectangular  
4. sagittate  
5. semi-circular  
6. ovoid  
 
#136. Sternite 10 
<position>  
 
1. nearly completely anterior to first cercus segment (~10% below 
basal cercus)  
2. posterior half below first cercus segment  
3. almost entirely underneath cercus segments  
 
#137. Cercus 
<segmentation>  
 
1. two-segmented  
2. one-segmented  
 
#138. Basal cercus 
<elongated or not>  
 
1. not elongated  
2. elongated (~3x longer than wide or more)  
 
#139. Basal cercus 
lobe <presence 
absence; gestalt>  
 
1. present  
2. absent  
 
#140. Basal cercus 
lobe <presence, 
extreme case>  
 
1. within range of continuous states between present and absent, 
not parallel-sided, not nearly as long as second cercus  
2. conspicuously present, extending posterior to point of 
attachment of second cercus, nearly parallel sided, 
approximately as long as second cercus  
 
#141. Basal cercus 
lobe <presence 
absence; strict 
scoring>  
 
1. present, area ventral to first cercus equal to or longer than 
width of second cercus at point of attachment  
2. absent, area ventral to first cercus shorter than width of second 
cercus at point of attachment  
 
#142. Second cercus 
segment <position 
on first cercus 
segment (where first 
segment not 
elongate)>  
 
1. dorsal (width clearly shorter above second cercus than width 
below)  
2. central (width approximately equal below and above second 
cercus)  
3. ventral (width clearly longer above second cercus than width 
below)  
 
#143. Basal lobe 
<presence absence; 
compound>  
 
1. absent; second cercus placed centrally or ventrally, ventral 
portion of basal cercus not expanded  
2. unclear; second cercus placed centrally or ventrally, ventral 
portion of basal cercus expanded  
3. present; second cercus placed dorsally, ventral portion of basal 
cercus expanded, extending mostly ventrally, rounded  
 
 181 
 
Character Character States 
4. present; second cercus placed dorsally, ventral portion of basal 
cercus expanded, extended mostly posteriorly and nearly 
parallel-sided  
 
#144. Ventral lobes of 
basal cercus 
<forming circle 
from posterior 
view?>  
 
1. curve ventrally towards one another to form a ring, visible in 
the posterior perspective  
2. do not curve ventrally towards one another to form a ring  
 
#145. Basal cerci 
<separation>  
 
1. adjacent dorsally  
2. separated from one another dorsally by approximately the 
width of the second cercal segment  
 
#146. Second cercus 
segment 
<elongated?>  
 
1. not elongated  
2. narrow, elongated (~3x longer than wide or more)  
 
#147. Cercus <apical 
sensory pits>  
 
1. with apical sensory pits  
2. without apical sensory pits  
 
#148. Tergite 8 <with 
ducts?>  
 
1. without any ducts  
2. with pair of ducts arising from posterior margin of sclerite, 
terminating in clump of sclerotized tissue  
 
#149. Spermathecae 
<number>  
 
1. one  
2. two  
3. three  
4. four  
 
#150. Spermathecae 
<shape>  
 
1. clubbed  
2. spherical  
3. swollen  
4. oval  
5. elongate oval  
6. eye-shaped  
7. pear-shaped  
8. elongated and misshapen, with what appear to be trachea 
attached distally  
 
#151. Spermathecae 
<sclerotization>  
 
1. not sclerotized  
2. lightly sclerotized  
3. sclerotized  
 
#152. Glandular hairs 1. absent  
 
 182 
 
Character Character States 
at distal tip of 
spermathecae 
<presence>  
 
2. present  
 
#153. Spermathecal 
ducts <length>  
 
1. no more than three times the length of sternite 9  
2. more than three times but less than five times the length of 
sternite 9  
3. longer than five times the length of sternite 9, but not so long 
as to be difficult to measure  
4. very long, folding upon themselves many times  
 
#154. Spermathecal 
duct accessory 
glands <presence>  
 
1. absent  
2. present  
 
#155. Spermathecal 
duct accessory 
glands <origin>  
 
1. arise at approximately the proximal third of the spermathecal 
ducts  
2. arise at approximately halfway along the length of the 
spermathecal ducts  
3. arise at approximately the distal third of the spermathecal ducts 
4. arise at the base of each spermatheca  
 
#156. Circular ridge 
marking the distal 
terminus of the 
ejection apparatus 
 
1. absent  
2. present  
 
#157. Ring of more 
heavy sclerotization 
near base of 
spermathecal ducts  
 
1. absent  
2. present  
 
#158. Spermathecal 
ducts <swelling 
presence>  
 
1. without swelling halfway between genital chamber and 
spermathecae  
2. with swelling halfway between genital chamber and 
spermathecae  
 
#159. Ejection 
apparatus 
<presence of 
furrows>  
 
1. furrows present  
2. furrows absent  
 
#160. Ejection 
apparatus <furrows 
at base, form>  
1. ultrastructural surface ringed  
2. ultrastructural surface furrowed at an angle  
 
 
 183 
 
Character Character States 
 
#161. Spermathecal 
ducts <insertion>  
 
1. extended smoothly into junction duct tubing  
2. visibly inserted within junction duct tubing  
 
#162. Ejection 
apparatus 
<sclerotized?>  
 
1. sclerotized  
2. not sclerotized  
 
#163. Junction and 
base of 
spermathecae 
<thickened?>  
 
1. junction and base of spermathecal ducts not thickened  
2. junction not thickened, base of spermathecal ducts thickened  
3. junction and base of spermathecal ducts thickened  
 
#164. Common 
spermathecal duct 
<presence>  
 
1. present  
2. absent  
 
#165. Common 
spermathecal duct 
<type; specially 
modified?>  
 
1. duct, without any special modifications  
2. enlarged pilose tapering  
 
#166. Common 
spermathecal duct 
<enlarged?>  
 
1. thinner than individual ducts (distally)  
2. not thickened (approximate thickness of individual duct)  
3. thickened (approximate summation of duct volumes)  
4. thickened (larger than summation of duct volumes)  
 
#167. Common 
spermathecal duct 
<anterior end, at 
junction with 
spermathecal ducts>  
 
1. without apical transverse ridge and suture at junction of 
spermathecal ducts  
2. with apical transverse ridge and suture at junction of 
spermathecal ducts  
 
#168. Common 
spermathecal duct 
<length, distance 
from duct junction 
to spermathecal 
opening>  
 
1. short, shorter than longest diameter of genital chamber  
2. moderate, about the same length as the longest diameter of 
genital chamber  
3. long, clearly longer than longest diameter of genital chamber  
 
#169. Genital chamber 
boundary <defined 
by membrane or 
sclerotization?>  
 
1. membranous; sclerotization of sternite 9 laterally contained  
2. tightly defined by medial sclerotization of sternite 9  
 
 
 184 
 
Character Character States 
#170. Genital chamber 
<shape>  
 
1. circular  
2. teardrop- or eye- shaped  
3. elongate, parallel-sided in part  
 
#171. Genital 
chamber <size>  
 
1. reduced  
2. small, occupying fraction of sternite 9 area  
3. moderately sized  
4. elongate, occupying most of sternite 9 area  
 
#172. Accessory gland 
posterior to genital 
chamber 
<prominent?>  
 
1. inconspicuous, easily overlooked even after staining  
2. prominent, retains dye easily  
 
#173. Accessory gland 
posterior to genital 
chamber <common 
duct present?>  
 
1. common duct present  
2. common duct absent  
 
#174. Accessory gland 
posterior to genital 
chamber <common 
duct form>  
 
1. distal end of common duct bulbous, knob-like  
2. distal end of common duct evenly tapered  
 
#175. Accessory gland 
posterior to genital 
chamber <length of 
common duct>  
 
1. long; common duct clearly longer than sternite 9  
2. short; common duct as long or shorter than sternite 9  
 
#176. Accessory gland 
posterior to genital 
chamber <presence 
of paired 
extensions?>  
 
1. with paired extensions posteriorly  
2. without paired extensions posteriorly  
 
#177. Accessory gland 
posterior to genital 
chamber <length of 
paired ducts>  
 
1. paired ducts long, clearly longer than common duct  
2. paired ducts short, as short or shorter than common duct  
 
#178. Ventral chamber 
duct  
 
1. absent  
2. present  
 
#179. Sternite 9 
anterior end 
1. tapered to a point  
2. as wide or wider than long, broadly rounded at tip  
 
 185 
 
Character Character States 
<anterior end, 
shape>  
 
3. stem and knob  
4. forked  
5. narrow, parallel-sided, round at tip  
6. narrowed, but broadly rounded tip  
7. longer than wide, parallel-sided, rounded tip  
8. expanding distally, then truncated abruptly  
9. autapomorphic state represented by Pelecorhynchus personatus 
10. autapomorphic state represented by Vermileo vermileo  
11. autapomorphic state represented by Lampromyia canariensis  
 
#180. Sternite 9 
<posterior end, split 
into two 
components?>  
 
1. fused posteriorly  
2. free posteriorly, in single projection  
3. free posteriorly, with two components (one medially 
positioned, other laterally positioned)  
 
#181. Posteromedial 
component of 
sternite 9 <medial 
projection>  
 
1. held horizontal  
2. held vertical medially  
 
#182. Sternite 9 
posterior margin 
<crescent shaped 
medially?>  
 
1. not crescent-shaped  
2. crescent-shaped  
 
Larva 
#183. Larva <known 
or not>  
 
1. known  
2. unknown not available  
 
#184. Head 
<retractile>  
 
1. retractile  
2. not retractile  
 
#185. Head capsule 
<withdrawn into 
second segment?>  
 
1. not folded within second segment  
2. withdrawn into second segment  
 
#186. Head capsule 
(dorsal plate) 
<internal portion>  
 
1. composed of a single, undivided plate  
2. divided, forming a pair of metacephalic rods  
 
#187. Head capsule 
<length>  
 
1. less than 4.5 times longer than greatest width  
2. more than 4.5 times longer than greatest width  
 
#188. Head capsule 
<shape>  
1. not cone-shaped  
2. cone shaped  
 
 186 
 
Character Character States 
  
#189. Mandibular 
brush <presence>  
 
1. absent  
2. present  
 
#190. Mandibular 
brush <association>  
 
1. associated with simple fold of cuticle  
2. associated with articulated rod  
 
#191. Mandibular 
brush <orientation>  
 
1. oriented along sagittal plane  
2. oriented along transverse plane  
 
#192. Mandibular 
brush <brush 
density>  
 
1. consists of less than 25 setae (per brush)  
2. consists of more than 25 setae (per brush)  
 
#193. Mandibular 
hook <external 
surface>  
 
1. without groove or canal  
2. with external groove on adoral surface  
3. canal with apical opening  
 
#194. Maxilla 
<sclerotized?>  
 
1. not sclerotized  
2. sclerotized  
 
#195. Saw sclerite of 
basal mandibular 
sclerite <presence>  
 
1. absent  
2. present  
 
#196. Maxillary palp  
 
1. segments sclerotized, cylindrical  
2. soft, segments poorly differentiated  
 
#197. Maxillary palp 
segment number  
 
1. two  
2. three  
 
#198. Antenna 
<bifurcating?>  
 
1. last segment multifurcated  
2. last segment bifurcated  
3. last segment entire  
 
#199. Antenna <no. of 
segments>  
 
1. one-segmented  
2. two-segmented  
3. three-segmented  
 
#200. Central duct 
leading to unpaired 
salivary pump   
 
1. present  
2. absent  
 
#201. Posterior 
tentorial expansion  
1. fused to DP (dorsal plate)  
2. free  
 
 187 
 
Character Character States 
 3. fused to each other  
 
#202. Anterior 
segment <with 
scalloping>  
 
1. ruffled  
2. not ruffled  
 
#203. Thoracic 
segments <welts, 
prolegs?>  
 
1. with creeping welts ventrally  
2. with crocheted prolegs ventrally  
3. smooth ventrally  
 
#204. Thick waxy 
covering <presence>  
 
1. absent  
2. present  
 
#205. Spiracular 
system  
 
1. apneustic <no spiracles present>  
2. propneustic <paired spiracles at front only>  
3. metapneustic <paired spiracles at back only>  
4. amphipneustic <paired spiracles at front and back>  
 
#206. Hind segment 
<sclerotized?>  
 
1. not sclerotized posteriorly  
2. sclerotized posteriorly  
 
#207. Spiracular disc 
<presence>  
 
1. absent  
2. present  
 
#208. Hind segment 
<form>  
 
1. produced into elongate respiratory siphon  
2. smooth, elongate oval (from dorsal ventral perspective)  
3. longitudinally striated, with inflated appearance  
4. soft, wrinkly, with a pair of setose projections  
 
 
 
Maximum Parsimony Analysis of Morphological Data 
MP heuristic searches found 24 most parsimonious trees. Fig. 4 shows the strict 
consensus of these trees (tree length = 657), with bootstrap support values above the 
branches, where the support is greater than 50% [statistics of one tree: consistency 
index (CI) = 0.3202, homoplasy index (HI) = 0.6798, CI excluding uninformative 
 
 188 
 
characters = 0.3121, HI excluding uninformative characters = 0.6879, retention index 
(RI) = 0.6341, rescaled consistency index (RC) = 0.2030].  
 
 
Figure 3. Strict consensus of 24 most parsimonious trees generated from the 
morphological matrix. Bootstrap values ? 50% are noted above supported 
branches (TBR, nreps = 1000 / addseq reps = 10). 
 
 189 
 
 
 190 
 
 
The most parsimonious trees recover the same major groups as in the molecular 
analyses. The Tabanomorpha are shown as monophyletic, as is the Tabanoidea, 
Vermileonidae, Rhagioninae, Chrysopilinae, and Spaniinae. There is apparently more 
resolution in how they are related to one another, however. The Tabanoidea is 
recovered as the basal clade of Tabanomorpha, with Vermileonidae sister to the 
Rhagionidae in its broadest sense, including Austroleptis, Bolbomyia, Rhagioninae, 
Chrysopilinae, and Spaniinae). Austroleptis multimaculata and Bolbomyia are sister 
taxa and along with Alloleptis tersus, form the sister group to Rhagioninae, 
Chrysopilinae, and Spaniinae. Rhagioninae is sister to the clade formed by 
Chrysopilinae and Spaniinae.  
 
The monophyly of the outgroup Xylophagomorpha, is supported by the elongated 
basal segment of the female cercus and several thoracic setal characters 
(postspiracular sclerite with setae, subscutellum setose on margins) in addition to the 
well known synapomorphy of having a highly sclerotized, cone-shaped larval head. 
  
The monophyly of Tabanomorpha is supported by the retractile larval head. The 
Tabanoidea is recovered, including Pseudoerrina jonesi as the basal taxon of the 
pelecorhynchid lineage, which is sister to the Athericidae and Tabanidae of the 
dataset. Tabanoidea is supported by the presence of a postspiracular scale (Fig. 177; 
lacking in some specimens of Glutops rossi), a bulbous subscutellum (lacking in 
Suragina concinna), and exceptionally long gonocoxal apodemes (Fig. 164; however 
 
 191 
 
absent in Glutops spp. and Pseudoerrina jonesi). Larval characters that may be used 
to define the Tabanoidea include an elongate head capsule (character 187; Fig. 197), 
and the mandibular brush associated with an articulated rod (character 190; Fig. 196). 
The monophyly of the Pelecorhynchidae is supported by the presence of paired ducts 
arising from the posterior margin of tergite 8 (Fig. 168). This is a new character, and 
the first adult synapomorphy developed for the family. The ventral process of the 
female first cercus segment in these taxa is distinctive and has been recognized as a 
possible synapomorphy for the group (Woodley, 1989; Figs. 158A, 166, 171A), 
however an explicit definition of this character is problematic as other tabanomorph 
taxa also have elongated first cercal segments (see discussion in Appendix A, 
regarding characters 140-143). The thick, waxy outer layer of the larva also supports 
this grouping, although the larva of Pseudoerrina species is not known. Additional 
characters that may support a close relationship between Glutops and Pelecorhynchus 
are the bare laterotergite, non-emarginate anterior margin of the epandrial sclerite. 
The sister group relationship between Tabanidae and Athericidae is supported 
principally by larval characters such as the presence of a salivary pump and the first 
thoracic segment withdrawn into the second thoracic segment. This sister group 
relationship is also supported by the shared presence of aedeagal tines in the male 
(independently derived in Bolbomyia and Arthroceras; Fig. 16), a single-segmented 
cercus (also present, but unrelated in Austroleptis), and a narrow ridge marking the 
distal end of the ejection apparatus of the spermathecal duct in the female (character 
56). The monophyly of the Athericidae is supported by wing veins R
1
 and R
2+3
 which 
autapomorphically meet together at the wing margin. Athericidae are also unique in 
 
 192 
 
having the larval maxillary palp fully sclerotized and cylindrical (Fig. 195) and 
having crocheted larval prolegs. 
 
The Vermileonidae are highly autapomorphic in many ways, particularly in regards to 
the male and female genitalia, to the degree that homology assessments for many 
characters were difficult for this taxon. The common spermathecal duct in 
vermileonids, for example, is highly modified into an enlarged, pilose structure that 
tapers distally. The wing is distinct in lacking an alula and anal lobe. 
  
Increased sampling within Rhagionidae, in comparison with the molecular analysis, 
allows for a better understanding of the relationships present in this group. I have 
some reservation in the placement of all taxa, however, that are based on a single 
gender (Litoleptis alaskensis, Sierramyia chiapasensis, Stylospania lancifera) or 
entirely scored from the literature (Alloleptis tersus).  
 
The sister group relationship between Bolbomyia nana and Austroleptis 
multimaculata is supported by the shared loss of wing vein M
3
. Both of these taxa 
also lack setae on the laterotergite, a condition shared by Alloleptis which is 
recovered as sister to Bolbomyia and Austroleptis. The lack of laterotergite setae, 
however, is present in other species of Pelecorhynchidae and Spaniinae. The lack of 
setae on the laterotergite appears particularly associated with small size. 
 
 
 193 
 
Taxa representing the chrysopilines (Chrysopilus), rhagionines (Rhagio), and 
spaniines (Ptiolina and Symphoromyia) have several larval features in common that 
provide evidence for their recency of common ancestry and support the monophyly of 
the Rhagionidae. The larval mandibular brush is associated with a simple fold of 
cuticle (hypothesized by Sinclair, 1992 as plesiomorphic; Fig. 186), a broad 
spiracular disc that is folded laterally, and a longitudinally striated, inflated larval 
hind segment (Fig. 187). Larvae of most taxa within this group (and possible sister 
taxa such as Alloleptis, Austroleptis and Bolbomyia) are not known, however. Based 
on the morphological analysis, Arthrocerinae belongs in the Rhagionidae although the 
larva of Arthroceras is also not known. 
 
The monophyly of the Rhagioninae is supported by the (butter-knife) form of the 
endoaedeagal process, which is strongly laterally compressed (Fig. 107). Rhagio and 
its relatives are united by having lateral ejaculatory processes that are integrated into 
the sperm sac membrane. Arthroteles and Atherimorpha are united by having 
conspicuously present hind tibial macrochaetae and in the female genitalia, the 
posteromedial component of sternite nine is held in a horizontal plane. The South 
American Atherimorpha species are shown as more closely related to Arthroteles 
bombyliiformis (South Africa) than to Atherimorpha vernalis (Australia). This is 
supported by the shared presence of two setal sizes of the thorax. Stuckenberg has 
speculated that the South African Atherimorpha may be more closely related to 
Arthroteles than the other members of the group, however the fact that South 
American members of Atherimorpha may show this affinity is a new revelation 
 
 194 
 
(Stuckenberg in Nagatomi & Nagatomi, 1990). Several larval features support the 
monophyly of the genus Rhagio. These include basal larval mandible with saw 
sclerite (Fig. 193) and mandibular brush with fewer than 25 setae (Fig. 186B).  These 
synapomorphies may be more inclusive once the larvae of other putatively related 
taxa within the Rhagioninae are found and described. 
 
Evidence supporting the common ancestry of the following clades is the shared 
presence of spermathecal duct accessory glands. This character is present within 
species of Arthroceras, Chrysopilus, Ptiolina, Schizella, Spania, Spaniopsis, and 
Symphoromyia (Figs. 40B, 62A, 64, 93A, 94-96, 126, 138, 149A, 150, 152-153). 
Female Litoleptis were not available, but this genus is placed in the Spaniinae 
tentatively on the basis of other characters. Within the Chrysopilinae, Stylospania 
furcicornis and Stylospania lancifera are recovered as sister taxa and form a 
monophyletic group along with Chrysopilus. These taxa form the basal lineage of the 
clade that composes Chrysopilinae, Arthrocerinae, and Spaniinae. The chrysopilines 
are supported by two synapomorphies; the presence of a reduced proepimeron sclerite 
(Fig. 175) and the presence of metallic- or scale-like thoracic setae (Fig. 112). All of 
these genera also tend to have pilose aristae and the gonostyli of the male genitalia 
tend to be heavily setose (Figs. 61, 117). The shared presence of a parameral sheath 
developed into ventral lobes in the male hypandrium supports the placement of 
Arthroceras pollinosum (Arthrocerinae) sister to the Spaniinae. The subfamily 
Spaniinae is defined by the shared presence of an anterior-facing lateral process of 
tergite 9, enveloping sternite 9 in the female (Figs. 94, 125A, 149-151; the state of 
 
 195 
 
Litoleptis is unknown). The larvae of Symphoromyia and Ptiolina have an unusually 
scalloped first thoracic segment that is probably an additional synapomorphy for this 
clade (Figs. 186C, 190). Several features of the female terminalia Ptiolina, Spania, 
and Spaniopsis support their close relationship. These characters include a laterally 
displaced first cercus segment (Figs. 92, 93A, 125A, 139; Symphoromyia may also 
have this character, see Fig. 149); the genital chamber tightly defined by medial 
sclerotization; a wide, broadly rounded anterior sternite 9 area; and a posteriorly 
fused sternite 9 (Figs. 93-94, 125, 139A). These taxa also have stylate antennae (Figs. 
73-76, 122, 127-130; see also Litoleptis spp., Fig. 70) and lack laterotergite setae. 
Ptiolina is recovered as a paraphyletic group, formed in a grade of two separate 
lineages. The P. zonata and P. mallochi clade are recovered as more closely related to 
Spaniopsis spp., Spania nigra, and Litoleptis alaskensis. Characters that support this 
relationship are the shared presence of a bare anepisternum (as S.  nigra and L. 
alaskensis, but not Spaniopsis spp.), the R
2+3
 vein directed toward wing margin 
anteriorly (a character that was not used in phylogenetic analysis), the R
5
 anterior to 
wing tip (Figs. 71, 81-82, 123, 132-135), the absence of an endoaedeagal process, a 
shortened spermathecal common duct (Figs. 94, 125, 139) and a tear-dropped shape 
of the genital chamber (Figs. 94, 125; as S. nigra, but not Spaniopsis spp.; Litoleptis 
alaskensis unscored for all female characters). Eliminating characters of the wing for 
phylogenetic analysis does not change this relationship (Ptiolina remains 
paraphyletic). The sister group relationship of Spania and Spaniopsis is supported by 
the reduction of female tergite 9 as a very narrow sclerite (Figs. 125A, 139).  
 
 196 
 
Discussion 
Monophyly of the Rhagionidae 
Rhagionidae were recently redefined (Stuckenberg 2001). The most important 
putative synapomorphies proposed for Rhagionidae sensu Stuckenberg include tergite 
7 of the female either as long as wide or much longer than wide (character 125); a 
wide intersegmental membrane between segments 7 and 8 of the female (character 
126); basal segments of the female cerci closely adjacent dorsally (character 145); 
and basal segment of the female cerci with posteroventral lobes which curve ventrally 
towards one another to form a ring, as seen in posterior view (character 144). 
Stuckenberg lists other putative synapomorphies, but I have found these to be 
subjective and variable. Unfortunately the most consistent characters used to define 
the Rhagionidae sensu Stuckenberg are not unambiguous with respect to related taxa 
in lower Brachycera (see below). The autapomorphies listed for the family in 
Stuckenberg (2001: 16) are more appropriately considered suites of diagnostic 
characters that are currently used to define Rhagionidae. 
 
Both putative synapomorphies relating to the female abdomen (tergite 7 square or 
elongate (character 125); presence of wide intersegmental membrane between 
segments 7 and 8 (character 126)) fail to distinguish Xylophagidae (sensu Woodley, 
1989) from Rhagionidae sensu Stuckenberg. There is also some internal disagreement 
of the characters within Rhagionidae. A wide intersegmental membrane between 
segments 7 and 8 of the female is not present in Bolbomyia Loew, for instance. 
 
 197 
 
Similarly, while most Rhagionidae sensu Stuckenberg have a square or elongate 
tergite 7 of the female abdomen, this is not exhibited in Symphoromyia Frauenfeld. 
 
The proximity of the female cerci to one another dorsally is not a discrete character 
and therefore its scoring may be subject to interpretation (see Nagatomi & Iwata 
1976). Nonetheless, when present, the basal female cerci are adjacent dorsally in 
many lower brachyceran taxa (e.g. xylophagids, coenomyiids) and the Rhagionidae 
sensu Stuckenberg are not exclusively defined by this character. 
  
The monophyly of Rhagionidae in its broad sense, including Austroleptis and 
Bolbomyia is supported by the morphological phylogenetic analysis, although no 
single character defines the group. This concept conforms to Woodley (1989) and 
differs from Stuckenberg (2001) and Nagatomi (1982) in the exclusion of the 
pelecorhynchid genera from the family. It can be argued that family-level recognition 
of Austroleptis species (Stuckenberg, 2001) may be justified. Another new family 
concept, Spaniidae sensu Stuckenberg (2001), however is unwarranted. This group is 
neither monophyletic nor located outside of the clade containing other rhagionid 
groups. Evidence in favor of retaining the Rhagionidae as a single taxonomic unit is 
revealed in the special similarity of larval terminal segments shared among rhagionid 
genera Chrysopilus, Ptiolina, Rhagio, and Symphoromyia. The larvae of Alloleptis, 
Austroleptis and Bolbomyia are not known, however.  
 
 198 
 
Pelecorhynchidae 
The monophyly and position of the Pelecorhynchidae as recognized by Sinclair 
(1992), Sinclair et al. (1993), and Woodley (1989) are confirmed by the phylogenetic 
analysis of morphological characters. This may be somewhat surprising given the 
divergent morphology of Pelecorhynchus, particularly in regards to the male 
genitalia, when compared with Glutops (Figs 156-157, 164-165). Pelecorhynchus was 
originally placed in the Tabanidae (Hardy, 1920b) while Glutops has been variously 
placed in the Rhagionidae (Nagatomi, 1977; Kovalev, 1981; Nagatomi, 1982a; 
Nagatomi, 1984; Stuckenberg, 2001) or in its own group (Krivosheina, 1971). The 
male genitalia of Pelecorhynchus resemble species of Tabanidae in several ways 
(e.g., elongated gonocoxal apodemes (character 114), fused hypandrial sclerite 
(character 111), sinuate R
4
 wing vein that diverges from R
5
 and toward the leading 
margin of the wing in a conspicuous manner (character 81)), seeming to suggest that 
this genus has strong affinities to Tabanidae and may actually be more closely related 
to this family than to Glutops. The male genitalia of Glutops is bulbous laterally in a 
way that is similar to Pelecorhynchus and Pseudoerinna, but is otherwise primitive 
(Figs. 157, 164). The male genitalia of Glutops recalls Rhagio in that it apparently 
lacks noticeably derived features.  
 
Several morphological features that suggest the pelecorhynchids are a monophyletic 
group are distinctive, but ultimately unscorable ?gestalt? characteristics. This includes 
the look of the eyes and swollen gena, and the roundedness of the male genitalia. 
Fortunately, there are other consistent features that may be scored as discrete 
 
 199 
 
characters suitable for phylogenetic analysis and used as evidence of monophyly for 
the family. The monophyly of the Pelecorhynchidae is supported by the presence of 
paired ducts arising from the posterior margin of tergite 8 (character 148; Fig. 168). 
This is a new character, and the first adult synapomorphy developed for the family. 
The ventral process of the female first cercus segment in these taxa is distinctive and 
has been recognized as a possible synapomorphy for the group (Woodley, 1989; 
character 140; Figs. 158A, 166, 171A), however an explicit definition of this 
character is problematic as other tabanomorph taxa also have elongated first cercal 
segments (see discussion regarding characters 140-143). The thick, waxy outer layer 
of the larva (character 204) also supports this grouping, although the larva of 
Pseudoerrina species is not known. Additional characters that may support a close 
relationship between Glutops and Pelecorhynchus are the bare laterotergite (character 
45) and non-emarginate anterior margin of the epandrial sclerite (character 98). 
Definition of Chrysopilinae, Arthrocerinae, and Spaniinae 
Three subfamilies of Rhagionidae are united by the presence of an accessory gland 
stemming from each of the spermathecal ducts (character 154). These structures, 
which are delicate and faint even after staining, have been overlooked by previous 
workers (Nagatomi, 1952; Nagatomi and Iwata, 1976; Webb, 1977b; Nagatomi, 
1986). The Spaniinae, within this clade, are defined by the shared presence of 
ventrolateral arms, extending posteriorly from tergite 9, surrounding and fusing to 
sternite 9 laterally (character 128). Both of these features are consistent in resolving 
the groups, just as recovered in the phylogenetic analysis. However, in the 
morphological analysis, Ptiolina is paraphyletic as currently recognized (Fig. 5).  
 
 200 
 
 
Particularly in most Chrysopilus specimens, the basal cercal lobes of the female 
clearly arc towards one another ventrally to form a circle, in line with the anus 
(character 144). Stuckenberg (2001) used this character as a synapomorphy of 
Rhagionidae. Rhagio species also exhibit this feature. I have found that when females 
die, however, the basal lobes of the cerci may dry and be fixed in a number of 
positions. Specimens may have the basal lobes appressed together medially or the 
basal lobes may be directed parallel or slightly laterally, creating space between the 
lobes. This variation, along with the imprecise nature of judging when the ?ring? 
formed by the cerci is circular, creates difficulties in scoring this character. In my 
opinion, however, the basal cercal lobes of Arthroteles Bezzi, Atherimorpha White 
and Symphoromyia (Rhagionidae sensu Stuckenberg) females should not be scored 
differently from the Spaniopsis and Ptiolina Zetterstedt species (Spaniidae sensu 
Stuckenberg) that I have seen. 
 
The close association of Schizella and Chrysopilus is self-evident. Species of both 
genera have derived, slightly flattened, often metallic-colored setae on the 
mesonotum (character 38) and the females of these genera are virtually 
indistinguishable. The possibility certainly exists that Schizella species represent an 
autapomorphic species group within the Chrysopilus lineage, distinguished only by 
the unusual antenna exhibited in the male (Fig. 113). Although there is some reason 
to believe Schizella species are derived forms of Chrysopilus, it is premature to 
synonymize the genera without more information. This is certainly the case in light of 
 
 201 
 
this result, which shows a monophyletic Chrysopilus. Examination of the larva of 
Schizella would be particularly instructive to help answer the question of Chrysopilus 
monophyly. 
Ptiolina and Omphalophora 
The genus Ptiolina Zetterstedt as currently recognized (Nagatomi, 1982a; Nartshuk, 
1995) was found to be a grade of two lineages in the morphological phylogenetic 
analysis (Fig. 5). This was somewhat of a surprising result given the close 
morphological similarity of Ptiolina and Omphalophora. Below I discuss the history 
of the Ptiolina and Omphalophora and morphological observations made during this 
study. Although some species remain undetermined regarding their generic 
designation (see list of included species for Ptiolina), the recognition of 
Omphalophora appears warranted. 
 
In 1900, Becker established Omphalophora oculata, a new genus and species from 
West Siberia. Frey added Omphalophora lapponica Frey to the genus in 1911 and 
another species was added in 1918, when Szil?dy transferred Chrysopilus arctica 
Frey to the genus. In 1982, Nagatomi examined all three types of these species and 
determined that Chrysopilus arctica Frey was erroneously placed in Omphalophora 
by Szil?dy (1934) and was a true Chrysopilus species, in agreement with Narchuk 
(1969). For the Omphalophora species, Nagatomi noted that wing vein R
2+3
 is 
straight in its apical portion, wing cell sc is wider at wing margin than r1, wing vein 
R
5
 at wing margin beyond wing tip and the anal cell is open. These features were 
believed to vary within Ptiolina, however, and he synonymized Omphalophora with 
 
 202 
 
Ptiolina. Majer (1988) did not recognize Nagatomi?s contribution and followed 
Szil?dy (1934), keeping all three species in Omphalophora. Narchuk (1969) indicated 
that Omphalophora differed from Ptiolina by its larger size and having a reduced or 
absent hind tibial spur. In addition to O. lapponica and O. oculata, Ptiolina grandis 
Frey and P. uralensis Becker shared these qualities and were consequently placed in 
Omphalophora. Later, however, Nartshuk (1995) adhered to a strict interpretation of 
Omphalophora, recognizing only a single species in the genus O. lapponica (Frey), 
which she then synonymized with Ptiolina oculata (Becker). In so doing, she 
assigned Omphalophora as a junior synonym of Ptiolina. Nagatomi also treated 
Omphalophora as a synonym of Ptiolina (Nagatomi, 1982a). 
 
I examined the type of Omphalophora lapponica (Frey) and dissected the female 
terminalia. Examination of its morphology reveals that Omphalophora Becker is a 
valid concept, defined by a suite of characters that have not been fully analyzed 
previously. The female genitalia are especially important in demonstrating important 
differences between O. lapponica (Frey) and its allies, from the remaining Ptiolina 
species. These differences are consistent with other differences evident in the male 
genitalic, wing, thoracic, and to some degree, antennal morphologies. It seemed as if, 
everywhere I looked, there were consistent differences between Omphalophora and 
Ptiolina and the morphological evidence supporting the monophyly of each of these 
groups is quite strong.  
 
 
 203 
 
Omphalophora and Ptiolina have also have divergent, but consistent morphologies in 
the length and form of tergites 7 and 9, the degree of fusion between tergite 9 and 
sternite 9, sternite form, size and shape of the genital chamber, spermathecal duct 
length, and spermathecal duct accessory gland position.  
 
In Omphalophora, the 7
th
 tergite is clearly longer than wide, whereas in Ptiolina, this 
tergite is clearly wider than long. Comparisons of the female terminalia of 
Omphalophora and Ptiolina are made with reference to Figs 81 and 95, which show 
the genitalia of these taxa, respectively. Tergite 9 in Omphalophora is bulbous 
laterally and tapers posteriorly, with a length that is greater than half its width (Fig. 
81). In Ptiolina, tergite 9 is rectangular and narrow; its length is less than half its 
width (Fig. 95). Also in Ptiolina, the ventrolateral arms of tergite 9 are easily 
distinguished, forming a modest s-curve (as observed in the dorsal/ventral 
perspective). The distal (anterior) tip of these ventrolateral arms is fused to sternite 9 
to varying degrees (e.g., lightly fused as in P.  zonata Hardy & McGuire or firmly 
attached as in P. mallochi Hardy & McGuire), but for the most part, the arms of 
tergite 9 are free from sternite 9 posteriorly. There is no such separation in 
Omphalophora, where the ventrolateral arms are bound to sternite 9 by a thick 
membrane along their entire length. The form of sternite 9 itself differs significantly 
between Omphalophora and Ptiolina. In Omphalophora, sternite 9 is narrow 
posteriorly and then broadens as it extends anteriorly beyond the ventrolateral arms of 
tergite 9 and is broadly rounded apically (anteriorly). The genital chamber, formed at 
the base of the common spermathecal duct, is narrow, with lateral margins that are 
 
 204 
 
nearly parallel, in line with a lengthened common spermathecal duct. In Ptiolina, 
sternite 9 is widest posteriorly, within the ventrolateral arms of tergite 9, and extends 
anteriorly as a narrow, nearly parallel-sided process which is flat truncated at its 
anterior apex. Female genitalia of Omphalophora also have partially sclerotized lobes 
located in the membrane between the ninth tergite and ninth sternite. The origin and 
homology of these structures are unclear. They are absent in species of Ptiolina. 
 
The genital chamber occupies a larger area and the margins of which are clearly oval 
(not nearly parallel sided as in Omphalophora). The common spermathecal duct is 
reduced to a short length (less than the length of the genital chamber) in Ptiolina and 
the spermathecal ducts, themselves, are less than three times the length of sternite 9. 
Spermathecal duct accessory glands arise at or near the base of the sclerotized 
spermathecae (Fig. 96). In Omphalophora, the common spermathecal duct is 
lengthened to at least the same length as the genital chamber, or longer, and the 
spermathecal ducts are at least three times the length of sternite 9 or greater. The 
spermathecal duct accessory glands arise at approximately two-thirds to four-fifths 
the distal length (from sternite 9) leading to the spermathecae (Figs. 81, 95). The duct 
is wider and thicker between the accessory gland and the sclerotized spermatheca, 
suggesting that it may be an unsclerotized proximal expansion of the spermatheca. 
This is true for all taxa (in Ptiolina, see in particular, P. mallochi Hardy & McGuire; 
it is readily apparent in all Omphalophora). Spermathecal form appears to vary on a 
species level and does not distinguish between Omphalophora and Ptiolina; 
spermathecae may be oval, egg-shaped, or spherical, regardless of grouping. 
 
 205 
 
Spermathecal sclerotization is generally well developed in Ptiolina sensu lato; any 
variation appears to be on a individual-level basis. The degree of lateral displacement 
between the basal cercal segments is also variable at the species level and is 
inadequate for distinguishing between Omphalophora and Ptiolina (not to mention 
difficult to score objectively on account of its continuous nature). 
 
Omphalophora also exhibit several distinctive characters in the male genitalia that 
instantly distinguish it from Ptiolina (Figs 77-80, 89-94). The most pronounced 
feature is that each gonostylus comes to a sharp point apically (Figs 79-80), as 
opposed to in Ptiolina, where the gonostylus is rounded apically (Figs 92-94). 
Omphalophora also have an elongate aedeagal sheath, posterior of the gonocoxites 
medially, where it narrows. In Ptiolina, the aedeagal sheath tends to be shorter. 
Similarly, the gonocoxal apodemes tend to be longer in Omphalophora than in 
Ptiolina. This feature is less striking and therefore, perhaps more difficult to 
distinguish without directly comparing samples. However, in all of the samples I have 
examined, the gonocoxal apodemes of Omphalophora reach the anterior margin of 
the hypandrium when examined in a direct dorsal view; in Ptiolina, the gonocoxal 
apodemes end well short of this.  
 
The epandrium is clearly different in Omphalophora, where the subepandrial sclerite 
is either as wide as long or squared and nearly oval (as in P. fasciata) (Figs. 77-78). 
Whereas in Ptiolina, the subepandrial sclerite is narrow rectangular, approximately 
three times as wide as long (Figs. 89-91). The hypoproct is tomentose in 
 
 206 
 
Omphalophora; setose in Ptiolina. The epandrial sclerite is also much more firmly 
secured to the subepandrial sclerite, and the cerci more firmly attached to the 
hypoproct, (in a single plane) in Omphalophora. In Ptiolina, the epandrial and 
subepandrial sclerites tend to rest at a perpendicular angle, loosely attached, after 
being cleared and placed in glycerol (Fig. 89B). Separation between cerci seems to 
vary at the species level, independent of higher level patterns and the form of the 
epandrial sclerite (e.g., notching or curvature of the posterior and/or anterior margin) 
does not appear to resolve differences at the subgeneric level. While a medial line of 
increased sclerotization of the epandrial sclerite (as evidenced by P. lapponica and 
especially P. majuscula) is more common in Omphalophora (Figs. 77B, 78), it is an 
unreliable character to differentiate Omphalophora and Ptiolina. 
 
In Omphalophora species (at that time, represented by O. oculata Becker and O. 
lapponica (Frey)), Nagatomi (1982: 56) noted that wing vein R
2+3
 is straight in its 
apical portion, wing cell sc is wider at wing margin than r1, ?wing vein R
5
 at wing 
margin beyond wing tip? and the anal cell is open. Now that more species are added 
to Omphalophora based on male and female genitalic characters, the distribution of 
wing character states has shifted. Wing vein R
2+3
, for example, is not always straight 
in its apical portion in Omphalophora (see P. fasciata Loew and P. nigripilosa Hardy 
& McGuire; Figs. 75A, 76A) and the anal cell is not always open at the margin (see 
P. fasciata Loew, Fig. 75A; however all Ptiolina available for examination have 
closed anal cells). However, at least one wing character consistently separates 
Omphalophora and Ptiolina. In Omphalophora, wing veins R
4
 and R
5
 contain the 
 
 207 
 
wingtip (Figs. 80, 81A) whereas in Ptiolina, R
5
 is anterior to the wingtip (Figs. 76B, 
88). In many Omphalophora, R
2+3
 is longer than the length of R
5
, however this is 
character is less reliable (P. nigripilosa is not obviously longer). In Ptiolina, wing 
veins R
2+3
 and R
5
 are either approximately equal in length, or R
5
 is longer. I have 
found that the relative length of wing cells sc and r
1
 is not a discrete and reliable 
feature distinguishing Omphalophora and Ptiolina.  
 
Perhaps because of their accessibility, the antennae often receive special attention in 
taxonomic treatments, and this is certainly the case for Omphalophora and Ptiolina. 
Species within these groups have obvious differences in antennal morphology. The 
first flagellomere in P. (Omphalophora) majuscula, for example, is rounded, enlarged 
and conical at its base then smoothly tapered into an extended stylus, somewhat 
reminiscent of the antenna of species in Litoleptis Chillcott and Spaniopsis White (Fig 
73B; compare with Litoleptis, Fig. 70, and Spaniopsis, particularly Figs. 127B and 
128B). In many Ptiolina (e.g., P. edeta, P. zonata, others), the first flagellomere is 
enlarged but flattened laterally, with a clear break between the stylus and the first 
flagellomere (Figs. 74, 84B). However, enough overlap of antennal morphologies 
exists between Omphalophora and Ptiolina to break down the reliability of this 
character system. Generally, the first flagellomere of Omphalophora tends to be sub-
globular oval or conical (three-dimensional) while in Ptiolina, it is laterally 
compressed (flattened). Differences in the way the specimen is collected and 
preserved may mask these differences however, and it is risky to base determinations 
on this alone (especially based on photos or illustrations from the lateral 
 
 208 
 
perspective!). The first flagellomere may be smoothly integrated into the style, or 
there may be a break between style and first flagellomere, without regard to 
subgeneric affiliation. Similarly, first flagellomere profile form (from a lateral 
perspective) and antennal style length may vary independently, on the species level. I 
could not detect differences between male and female specimens, in particular, above 
differences that exist naturally between individuals within the same species. 
 
Finally, I considered thoracic characters, body size, tibial spurs, and other features 
that may add to the evidence supporting two phylogenetic lineages within the genus 
Ptiolina. In the anepisternum of Omphalophora species is setose near the posterior 
margin whereas in Ptiolina, this sclerite is completely bare; lending further evidence 
of phylogenetic distance between the two groups.  
 
Ptiolina male eyes are clearly split into upper and lower regions, where facets are 
larger above and smaller ventrally. Omphalophora males also have facets that are 
larger above and smaller below, however, there is a smooth transition and an abrupt 
demarcation where facet size changes is lacking. 
 
Body size was an important consideration for distinguishing Omphalophora and 
Ptiolina in the past (Hardy and McGuire, 1947). This was probably attributable to the 
fact that the original species placed in Omphalophora are particularly large (P. 
oculata probably remains the largest in genus Ptiolina), which called attention to 
them. Their size, however, was largely independent of ancestry, as it would seem now 
 
 209 
 
in light of further examination. P. (Omphalophora) fasciata and P. (Omphalophora) 
nigripilosa, for example, are comparable in size to some of the smallest species in the 
Ptiolina. Tibial spurs have also been identified as a character that may assist 
separating Omphalophora and Ptiolina (Narchuk, 1969). I could not find any 
meaningful differences of size, number, or placement of tibial spurs, however.  

Table 5. Determination of Ptiolina species into Ptiolina and Omphalophora subgroups. 
Current 
Genus 
Original 
species 
Original Author Year Species name 
status 
Valid species Biotic 
Region
Subgenus 
determination
Basis of determination 
Ptiolina alberta Leonard 1931 valid alberta NA indet. not clear from description; although likely Ptiolina 
Ptiolina arctica Becker 1900 junior  
synonym  
grandis PA Ptiolina synonymy 
Ptiolina arctica Malloch 1923 junior  
synonym  
mallochi NA Ptiolina examination of holotype 
Ptiolina attenuata Nagatomi 1986 valid attenuata OR Ptiolina (Nagatomi, 1986) 
Ptiolina augusta Curran 1931 valid augusta NA indet. not clear from description; although likely Ptiolina 
Ptiolina calamodytes Schiner 1868 junior  
synonym  
cinereofasciata PA Omphalophora synonymy 
Ptiolina cinereofasciata (Schummel) 1837 valid cinereofasciata PA Omphalophora synonymy 
Ptiolina dudai Lindner 1942 valid dudai PA indet. specimen not available 
Ptiolina edeta (Walker) 1849 valid edeta NA Ptiolina examination of determined material 
Ptiolina fasciata Loew 1869 valid fasciata NA Omphalophora examination of determined material 
Ptiolina fulva Becker 1900 junior  
synonym  
cinereofasciata PA Omphalophora examination of determined material 
Ptiolina grandis Frey 1918 valid grandis PA Ptiolina examination of determined material 
Ptiolina grisea (Stroble) 1892 junior  
synonym  
paradoxa PA indet. specimen not available 
Ptiolina grisea Curran 1931 junior  
synonym  
edeta NA Ptiolina examination of determined material 
Ptiolina lapidaria Nowiczki 1868 junior  
synonym  
paradoxa PA indet. specimen not available 
Ptiolina lapponica (Frey) 1911 junior  
synonym  
oculata PA Omphalophora examination of holotype 
Ptiolina latifrons Nagatomi 1986 valid latifrons OR indet. specimen unavailable; description inadequate 
Ptiolina longipilosa Nagatomi 1986 valid longipilosa OR indet. specimen unavailable; description inadequate 
Ptiolina majuscula Loew 1869 valid majuscula NA Omphalophora examination of determined material 
Ptiolina mallochi Hardy & McGuire 1947 valid mallochi NA Ptiolina examination of determined material 
 
 212 
 
Current 
Genus 
Original 
species 
Original Author Year Species name 
status 
Valid species Biotic 
Region
Subgenus 
determination
Basis of determination 
Ptiolina nervosa Nagatomi 1986 valid nervosa OR indet. specimen unavailable; description inadequate 
Ptiolina nigra Zetterstedt 1842 junior  
synonym  
obscura PA Ptiolina examination of determined material 
Ptiolina nigrina Wahlberg 1854 junior  
synonym  
obscura PA Ptiolina synonymy 
Ptiolina nigripes Zetterstedt 1859 Junior 
synonym  
obscura PA Ptiolina synonymy 
Ptiolina nigripilosa Hardy & McGuire 1947 valid nigripilosa NA Omphalophora examination of holotype 
Ptiolina nitida Wahlberg 1854 valid nitida PA Ptiolina examination of determined material 
Ptiolina nitidifrons Hardy & McGuire 1947 valid nitidifrons NA Ptiolina examination of holotype 
Ptiolina obscura (Fallen) 1814 valid obscura PA Ptiolina examination of determined material 
Ptiolina obsoleta Leonard in Curran 1931 valid obsoleta NA Ptiolina examination of determined material 
Ptiolina oculata (Becker) 1900 valid oculata PA Omphalophora synonymy 
Ptiolina paradoxa (Jaennicke) 1866 valid paradoxa PA indet. specimen not available 
Ptiolina pelliticornis Becker 1900 valid pelliticornis PA indet. specimen not available 
Ptiolina phragmitophila Schiner 1868 junior  
synonym  
cinereofasciata PA Omphalophora synonymy 
Ptiolina shimai Nagatomi 1985 valid shimai OR Ptiolina (Nagatomi, 1985) 
Ptiolina sphaeralis Nagatomi 1986 valid sphaeralis OR indet. specimen unavailable; description inadequate 
Ptiolina tristis (Schummel) 1837 junior  
synonym  
obscura PA Ptiolina synonymy 
Ptiolina tristis (Walker) 1949 junior  
synonym  
obscura PA Ptiolina synonymy 
Ptiolina uralensis Becker 1921 junior  
synonym  
oculata PA Omphalophora synonymy 
Ptiolina vicina Hardy & McGuire 1947 valid vicina NA Ptiolina examination of determined material 
Ptiolina wodzickii Frauenfeld 1867 junior  
synonym  
paradoxa PA indet. specimen not available 
Ptiolina zonata Hardy & McGuire 1947 valid zonata NA Ptiolina examination of determined material 
 
 
 213 
 
 
Rhagioninae 
Rhagio and Rhagina 
All characters of Rhagina are also seen in Rhagio species, except for the exceptional 
wing characters seen in some species of Rhagina. Yang et al. note that Rhagina male 
lacks the subepandrial sclerite (tergite 10), whereas in Rhagio, it is present (Yang et 
al., 1997: 187), however, I find the male genitalia indistinguishable; both lack the 
subepandrial sclerite. The wing in Rhagina incurvatus de Meijere is distinctive, 
however, there appears to be a morphological grade in the genus, especially as one 
examines the wing of Rhagina sinensis Yang and Nagatomi (1997: 186) which has a 
sinuous R
2+3
 wing vein, but not distinctively so, and not far removed from venation 
found in some R. hirtus (Say) specimens and R. dichomaticus Chillcott. Another 
distinctive feature of Rhagina emphasized by Nagatomi (1982) and Yang et al. (1997) 
is a prominent ventro-apical ?hump? on the hind femur. This is a variable character in 
both Rhagio and Rhagina, however. Although most commonly absent in Rhagio, I 
have seen the ?hump? in an undescribed Rhagio species from Laos. Similarly, I have a 
Rhagina specimen which lacks such a hind tibial hump. Yang et al. indicate that the 
presence or absence of such a hump does not necessarily determine the genus (Yang 
et al., 1997: 115).  
Atherimorpha 
Stuckenberg suggested that South African members of the genus Atherimorpha may 
be more closely related to Arthroteles than to its nominal congeners. The results of 
this study may at first, seem to confirm this idea, since Atherimorpha is recovered 
 
 214 
 
paraphyletic with respect to Arthroteles. However, South African Atherimorpha 
species were not included in the phylogenetic analysis. In the current analysis 
Arthroteles bombyliiformis is recovered as sister to two South American 
Atherimorpha species. While it is possible that the South American and South 
African Atherimorpha species form a clade that is sister to Arthroteles, I strongly 
doubt it. I have examined extensive collections of Atherimorpha from South America 
and Australia, as well as a synoptic collection of South African Atherimorpha species. 
The South American members are very diverse and in some cases, highly divergent 
(e.g., A. albohirta) however I believe that Atherimorpha is monophyletic on the basis 
of consistent similarity among species throughout their distribution. I expect that as 
more Australian and South African Atherimorpha species are scored and included in 
the analysis, a more accurate resolution of the relationships will be recovered and the 
monophyly of Atherimorpha will be supported. The odd result here is most likely an 
effect of poor taxon sampling of the genus.  
Bolbomyia and Austroleptis 
Austroleptis and Bolbomyia are grouped in the most parsimonious arrangement, 
consistent with Grimaldi and Cumming (1999). Nonetheless, the shared loss of the 
third medial vein does not seem like a particularly compelling synapomorphy. 
Alloleptis tersus is located as the sister of these genera, but there are no clear 
synapomorphies supporting this placement. 
 
 
 215 
 
Molecular Treatment 
Introduction 
The taxonomic classification of the Rhagionidae has been unstable for decades 
because there are few morphological characters that can be used to support 
hypotheses of relationship among its members. When new morphological 
synapomorphies are proposed (such as the presence of aedeagal tines (Sinclair et al., 
1993)), few additional sources of evidence are available to corroborate or refute such 
ideas (Stuckenberg, 2001). One potentially very powerful and independent source of 
evidence comes from molecular data.   
 
The Tabanomorpha has already been studied in a molecular context to some degree 
(Wiegmann et al., 2000). No previous molecular study, however, has tested the 
monophyly of the Rhagionidae in light of several recent, conflicting hypotheses for 
the group (Sinclair, 1992; Stuckenberg, 2001; Woodley, 1989; Table 11). At issue is 
the position of three key taxon groups; Austroleptis, Bolbomyia, and Spaniinae 
(recognized at the family level by Stuckenberg, 2001).  
 
Austroleptis Hardy is a highly autapomorphic genus that shares few character states 
with its relatives in Lower Brachycera and has puzzled morphological taxonomists 
for decades. Most have placed the genus in Tabanomorpha but its location within this 
group has been disputed (Hardy, 1920; Steyskal, 1953; Hardy, 1955; Nagatomi, 
1982a, 1984; Woodley, 1989; Stuckenberg, 2001). Austroleptis is traditionally 
considered a primitive member of the Rhagionidae (Hardy, 1920; Nagatomi, 1982a). 
 
 216 
 
Chillcott (1963) suggested that within Rhagionidae, Austroleptis had affinities to 
Bolbomyia, Litoleptis, Archicera (= Spania), and Hilarimorpha. Grimaldi & 
Cumming (1999) used similarities in wing morphology to argue that Austroleptis is 
most closely related only to Bolbomyia and Litoleptis (either within or outside of 
Rhagionidae). Stuckenberg preferred to use the derived condition of the genus as 
evidence for supporting its own, family-level recognition (Stuckenberg in Nagatomi, 
1982a; Stuckenberg, 2001). It is not certain, however, that Austroleptis belongs in the 
Tabanomorpha. Ecological information, such as larval feeding habits, has been used 
as a surrogate to direct morphological evidence as a basis for its proposed placement 
in the Xylophagomorpha (Colless & McAlpine, 1991; Sinclair et al., 1993).  
 
Similarly, Bolbomyia Loew has been variously been placed within Tabanomorpha or 
outside of it (Xylophagidae) due to the ambiguity of the morphological evidence. 
Bolbomyia has been placed in the Rhagionidae on account of an elongated 
intersegmental region in the female abdomen (Nagatomi, 1982a; Stuckenberg, 2001); 
together with Austroleptis and Litoleptis (inside or outside of Rhagionidae) because it 
lacks wing vein M
3
 (Grimaldi & Cumming, 1999); or as sister to Athericidae and 
Tabanidae on account of having aedeagal tines in the male genitalia (Sinclair et al., 
1993). James (1965) located Bolbomyia among the Xylophagidae because of its 
flattened clypeus (James, 1965).  
 
The spaniine group of Rhagionidae (Hennig, 1973; Nagatomi, 1982a) includes 
Ptiolina, Spania, Spaniopsis, and Litoleptis. These taxa have been placed together on 
 
 217 
 
the basis of their antennal form (short, stylate), short female abdominal 
intersegmental length, and the wide separation of the female cercus segments 
(Nagatomi, 1982). Recently, this group was raised to family level status 
(Stuckenberg, 2001). Stuckenberg (2001) did not suggest a sister taxon to the group, 
however, and the monophyly of the group and its placement with respect to the 
Rhagionidae have not been tested in a phylogenetic context. Its proper status, 
therefore, remains somewhat uncertain.  
Materials and Methods 
Taxon Sampling 
Complete taxon sampling for molecular analysis of the Rhagionidae is difficult 
because so many of the genera of this family are rare. Nonetheless, ingroup taxa 
include representatives from all subfamilies of the genera in Rhagionidae recognized 
by Nagatomi (1982) and Stuckenberg (2001). Outgroup taxa included representatives 
from all families within Tabanomorpha, as well as a diversity of genera within the 
Xylophagomorpha and two genera from the Stratiomyiomorpha. 
 
The breadth of taxon sampling was determined on the basis of availability of 
specimens for study and the importance of taxa for testing specific hypotheses of 
relationship including the monophyly and position of Pelecorhynchidae and 
Spaniidae, the monophyly of Chrysopilus, Rhagio, and Atherimorpha, as well as the 
position of Vermileonidae, Bolbomyia, and Austroleptis. Table 3 shows the species 
used in the molecular analysis, their source, locality, and recent family designations. 
Higher level classification follows Woodley (1989). 
 
 218 
 
 
Newly determined sequences were combined with sequences from the literature, 
available from GenBank. Twenty 28S rDNA sequences representing twenty species 
in fifteen genera of Tabanomorpha have been published previously (Wiegmann et al., 
2000). Of these, seven species in four genera are currently placed in Rhagionidae. 
Brian Wiegmann (North Carolina State University, Raleigh, NC, USA) supplied 
additional, unpublished sequences of Exeretonevra angustifrons Hardy, Xylophagus 
abdominalis Loew, and Heterostomus sp. The sequence identified as P. fasciata 
Loew (Wiegmann et al., 2000; GenBank accession numbers AF238554, AF238530, 
AF238508) showed very strong affinity with Chrysopilus species and not with the P. 
fasciata identified and used in this study (results not shown). Because a voucher for 
this specimen was not available to confirm identification, GenBank accessions 
AF238554, AF238530, and AF238508 were not included in this study. Sequences 
from Wiegmann et al. (2000) were shorter than those presented here, therefore 
identical sequences determined from the same species for this study were included in 
favor of the shorter sequences. A summary of species, source, and GenBank 
accession numbers is shown in Table 3. 
 
 219 
 
Table 6. Taxon sampling for molecular analyses. Classification consistent with 
Woodley (1989) / Stuckenberg (2001). 
Classification Species Locality Source GenBank 
Accession 
Numbers 
STRATIOMYOMORPHA 
Stratiomyiidae  Pachygaster leachii England GenBank 
AF238548, AF238524, AF238502 
Pantophthalmidae  Pantophthalmus sp. Costa Rica GenBank 
AF238547, AF238523, AF238501 
XYLOPHAGOMORPHA    
 
Arthropeas magnum Saskatchewan GenBank 
AF238549, AF238525, AF238503 
Dialysis elongata North Carolina GenBank 
AF238551, AF238527, AF238505 
Exeretonevra 
angustifrons 
Australia B. Wiegmann, unpubl. 
 
Heterostomus sp. Costa Rica B. Wiegmann, unpubl. 
 
Coenomyia ferruginea Tennessee GenBank 
AF238550, AF238526, AF238504 
Xylophagidae  
Xylophagus 
abdominalis 
North Carolina B. Wiegmann, unpubl. 
 
TABANOMORPHA 
Atherix variegata Wisconsin GenBank 
AF238565, AF238541, AF238517 
Dasyomma sp. 1  Chile New sequence 
 
Athericidae 
Dasyomma sp. 2  Chile New sequence 
 
Chrysops sp. Maryland New sequence 
 
Tabanus atratus North Carolina GenBank 
AF238568, AF238544, AF238519 
Tabanus rufofrater Georgia GenBank 
AF238561, AF238537, AF238513 
Tabanidae 
Tabanus sp. Maryland New sequence 
 
Atherimorpha atrifemur Chile New sequence 
 
Atherimorpha sp. 1  Chile New sequence 
 
Atherimorpha sp. 2  Tasmania New sequence 
 
Atherimorpha sp. 3  Tasmania New sequence 
 
Atherimorpha vernalis  Tasmania New sequence 
 
Bolbomyia nana Quebec New sequence 
 
Chrysopilus quadratus Maryland New sequence 
 
Chrysopilus thoracicus Maryland New sequence 
 
Chrysopilus sp. 1  Ecuador New sequence 
 
Chrysopilus sp. 2  Queensland, 
Australia 
New sequence 
 
Chrysopilus sp. 3  Tasmania New sequence 
 
Chrysopilus sp. 3  Tasmania New sequence 
 
Rhagio hirtus Illinois GenBank 
AF238509 ,AF238532, AF238556 
Rhagionidae 
Rhagio mystaceus Illinois GenBank 
AF238510 , AF238531, AF238555
Glutops rossi Washington GenBank 
AF238570, AF238546, AF238521 
Pelecorhynchidae/ 
Rhagionidae Pelecorhynchus 
personatus 
Australia GenBank 
AF238569, AF238545, AF238520 
Austroleptis collessi  Tasmania New sequence 
 
Austroleptis 
multimaculata 
Tasmania New sequence 
 
Rhagionidae/ 
Austroleptidae 
Austroleptis rhyphoides Tasmania New sequence 
 
Ptiolina fasciata Saskatchewan New sequence 
 
Spaniopsis clelandi Tasmania New sequence 
 
Spaniopsis longicornis Victoria, Australia New sequence  
 
Symphoromyia atripes Illinois GenBank 
AF238559, AF238535 
Symphoromyia hirta Illinois GenBank 
AF238558, AF238534, AF238512 
Rhagionidae/ Spaniidae 
Symphoromyia sp. Alaska New sequence  
TABANOMORPHA/  
VERMILEOMORPHA 
Leptynoma hessei South Africa Genbank 
AF238552, AF238528, AF238506 
Vermileo opacus California GenBank 
AF238553, AF238529, AF238507 
Vermileonidae 
Vermileo sp. Israel New sequence 
 
 
 
 220 
 
Extraction 
DNA was extracted from specimens stored in 85-100% ethyl alcohol with the 
Nucleon Phytopure resin-based extraction kit, using the protocol provided for small 
samples (Amersham Pharmacia, Uppsala, Sweden) or with the DNEasy Plant mini kit 
(Qiagen, Valencia, CA, USA). Quality of the extracted DNA was assessed via 
agarose gel electrophoresis and ethidium bromide staining.  
Amplification 
Amplification of 28S rRNA was performed using three primer pairs (rc28C-28E, 
rc28D-28K, and rc28Q-28Z) for the polymerase chain reaction (PCR). In cases where 
amplification of these regions failed, internal primers (28P, rc28P, 28H, rc28H, 28X, 
and/or rc28X) were used to amplify smaller fragments (Table 2; Fig. 1). The PCR 
was done using a Biometra PCR machine with the following program: 95?C initial 
denature step of 3 minutes followed by the amplification cycle of 95?C for 20 
seconds, 54?C for 20 seconds, and 75? for 1 minute and 10 seconds. The cycle was 
repeated 30 times. After 10 minutes at 75?C, the products were cooled to 4?C. The 
resulting PCR products were purified using a modified polyethylene glycol (PEG) 
precipitation (Morgan and Soltis 1993). An equal volume of 20% weight:volume 
PEG 8000, 2.5 M NaCl was added to each PCR product, vortexed briefly and spun at 
16,000g for 15 minutes. The solution was removed and the resulting DNA pellet was 
washed once with 80% cold ethanol. The solution was spun at 16,000g for 10 minutes 
and the ethanol was removed. The pellet was then air-dried and resuspended in 25 ?l 
de-ionized water. The PEG-purified PCR product was quantified via agarose gel 
electrophoresis and ethidium bromide staining for subsequent sequencing reactions. 
 
 221 
 
Table 7. 28S rDNA primers (from Wiegmann et al., 2000). 
Primer Sequence 3?- 5? 
rc28C CCG AAG TTT CCC TCA GGA TAG C 
28P GGC TTA CGC CAA ACA CTT CTA GGC 
rc28P TGG TAT GCG TAG AAG TGT TTG GC 
28E CCT TAT CCC GAA GTT ACG 
rc28D CCG CAG CTG GTC TCC AAG 
28H GGT TTC GCT GGA TAG TAG 
rc28H CTA CTA TCC AGC GAA ACC 
28K GAA GAG CCG ACA TCG AAG 
rc28Q GGA CAT TGC CAG GTA GGG AGT T 
28X CGG ATA CGA CCT TAG AGG CG 
rc28X  CGC CTC TAA GGT CGT ATC CG 
28Z GCA AAG GAT AAG CTT CAG TGG 
 
 
 222 
 
 
Figure 4. Primers used for sequencing of 28S rDNA gene. 
Sequencing 
Sequencing reactions were performed in 7 ?l final volume (0.5 ?l PEG-purified PCR 
product, 3.5 ?l de-ionized water, 1.0 ?l 2.5 ?M primer, 1.5 ?l 5X buffer [400 mM 
Tris pH 9.0, 10 mM MgCl
2
], 0.5 ?l BigDye Terminator Ready Reaction Mix v2 
[Perkin Elmer Biosystems, Foster City, CA]), cycled and purified according to the 
manufacturer?s protocols and resolved either using an ABI 377 slab gel sequencer 
performed by the University of Maryland Center for Agricultural Biotechnology or an 
ABI 3100 capillary sequencer. The resulting sequences were blasted against Genbank 
to confirm their identity. Sequenc fragments were edited and compiled using the 
computer program Sequencher 3.1.1 (Gene Codes Corporation, Ann Arbor, MI). 
Opposite strands were confirmed for all templates.  
Sequence Alignment 
The aligned 28S data matrix was 3201 base pairs in length. The average base 
frequencies of the entire matrix were A= 30.22%, C=17.33%, G=23.20%, T=29.26% 
and base composition did not significantly differ from these proportions among taxa 
(?
2
 = 52.542 (df=126), P = 1.0).  
 
 
 223 
 
Many new sequences generated for this study were approximately 55bp longer at 
their 3? ends than the sequences acquired from GenBank, after primer ends had been 
discarded and the sequences aligned. Taxa that were shorter by comparison include 
Arthropeas magnum (55bp), Atherix variegata (46bp), Chrysopilus quadratus (46bp), 
Coenomyia ferruginea (55bp), Dialysis elongata (55bp), Exeretonevra angustifrons 
(55bp), G. rossi (55bp), Heterostomus sp. (55bp), Leptynoma hessei (55bp), 
Pelecorhynchus personatus (55bp), Rhagio hirtus (46bp), Rhagio mystaceus (46bp), 
Symphoromyia hirta (46bp), Tabanus atratus (55bp), Tabanus rufofrater (55bp), 
Vermileo opacus (55bp), Xylophagus abdominalis (55bp), Pachygaster leachii 
(55bp), and Pantophthalmus sp. (60bp). The region of 28S between primer pairs 
rc28C and 28E could not be amplified for Atherimorpha sp. 2 [Cradle Mountain 
Lodge, Tasmania, Australia], so the first 1134 bp of the aligned sequence are missing. 
The sequences acquired from GenBank were also partially incomplete in other areas. 
Chrysopilus thoracicus and Symphoromyia atripes were missing the first 503bp and 
537bp of aligned sequence, respectively. The primer region between base pairs 442-
557 of the missing set (116 bases, containing the starting ends of primers rc28P and 
28P) was missing for many of the GenBank sequences including Glutops rossi, 
Leptynoma hessei, Pelecorhynchus personatus, Tabanus atratus, Vermileo opacus, 
Pachygaster leachii, and Pantophthalmus sp. Parts of this region was also missing for 
Chrysopilus quadratus (base pairs 480-537; a total of 58bp), Coenomyia ferruginea 
(504-524; 21bp), and Dialysis elongata (base pairs 464-551; 88 bp). The region about 
site 1770 of the aligned sequence is another area of relatively small regions of 
missing sequence for Chrysops sp. (new sequence; 1735-1789; 55bp), G. rossi (1735-
 
 224 
 
1805; 71bp), Heterostomus sp. (new sequence from B. Wiegmann; 1766-1798; 33 
bp), L. hessei (1735-1849; 115bp), P. personatus (1734-1806; 73bp), Tabanus 
rufofrater (1735-1805; 71bp), V. opacus (1735-1811; 77bp), P. leachii (1721-1805; 
85bp), and Pantophthalmus sp. (1735-1805; 71bp). The Tabanus atratus sequence 
also lacks a large part of sequence containing this area (1735-2293; 259bp). The 
Atherix variegata sequence ends at site 2744. E. angustifrons, G. rossi, and V. opacus 
ends are slightly shorter (ending at base pair 3078, 3071, and 3055), respectively than 
all of the other sequences which are cut at the 5? end at base pair 3102 of the aligned 
sequence. 
 
The alignment is a process of determining homology between sequence sites and can 
have a large influence on the ultimate phylogeny produced. The structure of the 28S 
sequences is well known to have areas that are highly conserved and areas that are 
more free to vary due to the secondary structure of the molecule, consisting of stems 
and loops, respectively. Most of the 28S rDNA molecule is highly conserved and 
therefore, easily aligned. Within regions of the alignment that are subject to higher 
rates of evolution, phylogenetic signal is retained to varying degrees or lost entirely. 
Regions where phylogenetic signal is retained may be important to include so that 
tips of the branches may be more accurately resolved among closely related taxa, 
even though across other parts of the taxon sample the molecule may be saturated. At 
what point does saturation undermine homology assessment for phylogenetic 
analysis? 
 
 
 225 
 
Generally, researchers invoke their own authority for deciding such matters. In some 
cases, any perceived uncertainties in the alignment process are excised from the 
phylogenetic analysis. Presumably, there is enough phylogenetic information within 
the more conserved regions to yield a biologically relevant phylogenetic tree. But are 
the results of these analyses compromised by the loss of phylogenetically useful 
information? May the baby thrown out with the bathwater using this approach? 
 
On the other hand, allowing saturated positions in the analysis are certain to introduce 
additional noise, or even worse, may incorporate investigator bias that overrides true 
phylogenetic signal.   
 
Where is the balance between maximizing phylogenetic information and minimizing 
noise and bias? An explicit approach was taken here to investigate this further, in a 
two step process. The first is to classify regions of the sequence according to the 
degree of saturation and difficulty of alignment. The second step is to evaluate the 
utility of these regions for phylogenetic analysis.  
 
Ideally, the first step of this process would employ a model to quantify the amount of 
saturation of a given site, across the data set. Such a model would take several factors 
into account. First, it would detect accelerated rates of evolution within the molecule, 
presumably correlated with secondary structure loops. This may be inferred from 
secondary structure modeling or by other means, such as estimating evolutionary 
rates of specific sites given a tree, or manually designated as a region of alignment 
 
 226 
 
ambiguity by the user. Once loop regions are identified, these are excluded and a 
maximum likelihood search is carried out to generate a phylogeny that may be used 
as the underlying topology for the homoplasy calculation.  This would be an 
important component of the model since there is a correlation between character 
homoplasy and difficulty of homology assessment (i.e., alignment). The alignment of 
a constant site, for example, is trivial yet as steps are added to the site, the alignment 
becomes increasingly difficult. Therefore, some measure of homoplasy would be 
factored into the model as an indication of homology assessment uncertainty due to 
homoplasy. 
 
Another aspect affecting alignment is the frequency of gaps. Alignment uncertainty is 
often associated with highly gapped characters and this uncertainty is not reflected in 
the homoplasy index. For example, characters of a saturated insertion present in two 
species will have a consistency index value of 1. Therefore, a correction must be 
incorporated into the model that accounts for the uncertainty of homology assessment 
due to the insertion of gaps.  
 
The reliability of homology assessment may also be affected by neighboring sites. 
Sites that involve gaps or other uncertainties may invoke alternative alignments that 
affect the stability of homology assessment of neighboring sites. A character that has 
few gaps and a moderate level of homoplasy, for example, may intuitively have 
differing levels of certainty of homology depending on whether or not its neighbors 
 
 227 
 
on each side are constant across the taxon set or are contained within a secondary 
structure loop.  
 
The interaction and relative weights of model components for determining relative 
alignment uncertainty in an objective manner is a difficult proposition and a work in 
progress. Incorporating the secondary structure of ribosomal DNA into the alignment 
may be alternative to devising such a model. Mapping secondary structure is an 
arduous process however, that does not guarantee an improved alignment. 
 
With a model approach in mind, characters were classified according to the degree of 
saturation and difficulty of alignment. This was done using the following 
methodology and criteria. First, sequences were arranged in congeneric clusters of no 
particular order and the base pairs were aligned by eye. Regions exhibiting increased 
rates of evolution were identified by eye and removed for a preliminary phylogenetic 
analysis using maximum likelihood. This preliminary analysis produced a 
biologically reasonable topology that was used as a guide to place close related taxa 
adjacent to one another in the alignment. The alignment was further refined by eye. 
The taxa were then arranged in alphabetical order and another pass of alignment 
refinements were made by eye. A second preliminary phylogenetic analysis was 
undertaken and the taxa were placed near closest relatives in the alignment, as 
indicated by the resulting topology.  
 
 
 228 
 
Characters within regions of the molecule that exhibited accelerated rates of evolution 
were then classified as ?lowambig,? ?modambig,? ?highambig,?or ?unalign? based on 
the degree of uncertainty associated with their alignment, as determined by the ease 
of alignment between increasingly distantly related taxa. Areas exhibiting variation 
that was easy to align within closely related groups and across most of the taxon set 
designated as ?lowambig.? Areas of the sequence that were easily aligned between 
related groups, but more challenging to align between more distantly related groups 
were categorized as ?modambig.? Areas alignable only among close relatives were 
designated as ?highambig? sites. Characters that were unalignable even among closely 
related taxa were classified as ?unalign? and discarded for all analyses. All of these 
classifications of alignment ambiguity were done by eye, as an objective model has 
not been developed for such a process. Classifications were compared against one 
another for consistency across the alignment, and necessary modifications were made, 
as appropriate, to maintain adherence to the classification criteria listed above. The 
full alignment including all regions is archived in TreeBase (www.treebase.org). 
 
The next step was to evaluate the utility of these regions for phylogenetic analysis. 
First, the effects of differing alignment decisions were explored by creating eight 
separate, successively more conservative matrices from the complete data set. 
Unambiguously aligned regions were included for all analyses. The data matrices 
created were as follows:  
1) Only ?unalign? regions excluded; all other regions included (Alignment Set 
1 (AS#1)).  
 
 229 
 
2) ?Ends? and ?unalign? regions excluded; all other regions included (AS#2). 
3) ?Highambig? and ?unalign? regions excluded; all other regions included 
(AS#3). 
4) ?Ends,? ?Highambig,? and ?unalign? regions excluded; all other regions 
included (AS#4). 
5) ?Modambig,? ?Highambig? and ?unalign? regions excluded; all other regions 
included (AS#5). 
6) ?Ends,? ?Modambig,? ?Highambig? and ?unalign? regions excluded; all other 
regions included (AS#6). 
7) ?LowAmbig? ?Modambig,? ?Highambig? and ?unalign? regions excluded; 
all other regions included (AS#7). 
8) ?LowAmbig? ?Ends,? ?Modambig,? ?Highambig? and ?unalign? regions 
excluded; all other regions included (AS#8). 
 
For each alignment set, a maximum parsimony heuristic search (tbr, nreps=1000) was 
conducted to generate one or more equally most parsimonious trees (MPTs). A strict 
and majority rule consensus tree was then calculated from this collection of trees for 
each alignment set (if there was only one MPT, this step was unnecessary). 
 
A clear manner to decide upon the best alignment set is not immediately obvious. 
Since the number of characters varies between alignment sets, parsimony scores 
cannot be compared directly. Bootstrap values for these datasets, however, may be 
informative. A bootstrap analysis was conducted for each alignment set and a 
 
 230 
 
bootstrap (majority rule) consensus tree was created. Bootstrap consensus trees and 
most parsimonious consensus trees were then compared for each alignment set. The 
fidelity in recovering bootstrap-supported most parsimonious trees may be an 
indication of the degree of internal consistency within the data. To quantify this 
consistency in a way that could be compared between alignment sets, the number of 
nodes present in both the bootstrap support consensus tree and the most parsimonious 
strict consensus tree was counted for each alignment set. Nodes present in both the 
bootstrap support consensus and most parsimonious majority rule (MR) tree were 
also counted for each alignment set as a secondary measure. The alignment set with 
the highest overlap between bootstrap and most parsimonious consensus trees, as 
counted by the number of shared nodes, was determined as the most internally 
consistent data set and chosen as the data set for all subsequent phylogenetic analyses. 
Phylogenetic Analysis 
Phylogenetic analyses were performed using two computer programs, on both PC and 
Macintosh operating systems. PAUP* 4.0b10 (Swofford 2002) was used for both 
maximum parsimony (MP) and maximum likelihood (ML) phylogenetic analyses.  
 
For parsimony analyses, characters were unordered and assigned equal weights at all 
sites (Fitch, 1971). Heuristic search replicates (n = 1000) were performed with 
random-taxon-addition, tree bisection reconnection (TBR) branch swapping, steepest 
descent and ?MulTrees? options in effect. Scores of all MP trees were verified to 
avoid artifactually inflated sets of MP trees. 
 
 
 231 
 
Modeltest 3.06 (Posada and Crandall, 1996) was used to select among 56 nested 
models of sequence evolution. Under the Akaike information criterion (AIC), the best 
fitting model was found to be the general-time-reversible model with invariant sites 
and gamma distributed rates for variable sites (GTR+I+? ). An iterative series of 
searches were then performed to re-estimate the parameters until the parameters 
stabilized. A maximum likelihood heuristic search was conducted, using this model 
and the parameters estimated by Modeltest, with five random taxon addition 
sequences, tree-bisection reconnection (TBR) branch swapping, and steepest descent 
option active. The resulting tree provided the basis for re-estimating the parameters. 
The ML search and re-estimation of parameters based on the most likely tree was 
carried out again, as the parameters converged on identical values. These parameter 
values were used for ML analyses. 
 
In order to gauge the internal consistency of the data, bootstrap analyses (Felsenstein 
1985) were carried out. Parsimony bootstrap analyses were performed with 1000 
repetitions, each with 30 random-taxon-addition TBR heuristic searches. ML 
bootstrap analyses were performed with 100 repetitions, each with three random-
taxon-addition NNI heuristic searches.  
Tests of alternative hypotheses using Maximum Parsimony 
Hypotheses regarding the monophyly and placement of the Pelecorhynchidae, 
Vermileonidae, Bolbomyia, Austroleptis, and Rhagionidae were tested in light of the 
molecular evidence. This was done by constraining nodes to conform to hypothesized 
arrangements. Heuristic searches were then carried out, enforcing the constrained 
 
 232 
 
nodes. The length of resulting most parsimonious trees constrained to fit each 
hypothesis was then compared with the most parsimonious unconstrained trees. 
Hypotheses were evaluated by the number of additional steps they incur on the data; 
greater additional steps indicate less concordance with the molecular data.  The 
Kishino-Hasegawa and Wilcoxon sign-rank tests were also applied using PAUP* to 
compare constrained and unconstrained trees to determine if the hypotheses predicted 
signficantly different arrangements than what was found given the molecular data. 
Results 
Alignment region classification 
Most of the 28S rDNA molecule is highly conserved and unambiguously aligned. The 
sizes of the regions subject to higher rates of evolution are listed in Table XX, by 
alignment classification type.  
Table 3. Alignment region classifications. 
Alignment classification Total bp 
Unambiguously aligned 2779 
?Lowambig? 39 
?Modambig? 123 
?Highambig? 47 
?Ends? 56 
?Unalign? 190 
 
 
 233 
 
Effects of alignment 
The results of most parsimonious tree searches for each of the eight alignment sets are 
summarized in Table 4. The average pairwise genetic distances between each species 
and the rest of the sampled taxa, for the entire character matrix excluding the 
unalignable regions are shown in Table 6. The tabanids and stratiomyiomorph 
outgroups Pachygaster leachii and Pantophthalmus sp. are the most divergent (Table 
7). The deviation from the mean of the average pairwise differences did not appear to 
change as the number of characters varied (no statistics applied however). From the 
bootstrap analyses, AS#4 was determined as the most internally consistent data set 
(Table 8). 
 
 
 
 
Table 8. Relationships among the major taxa, evident in the Maximum Parsimony majority rule consensus trees of different alignment 
sets (AS). 
excluded 
character 
sets of AS 
No. of  
MPTs 
Tabanomorpha 
monophyly/ 
paraphyly 
Sister group 
of 
Vermileonidae
Sister group of 
Austroleptis 
Sister group of  
Bolbomyia 
No. of resolved strict 
consensus nodes 
1. unalign  6 paraphyletic Rhagionidae, 
sensu lato 
Rhagioninae Austroleptis + 
Rhagioninae  
37 
2. Ends, 
unalign  
6 paraphyletic Rhagionidae, 
sensu lato 
Rhagioninae Austroleptis + 
Rhagioninae 
37 
3. 
Highambig 
unalign  
30 monophyletic Austroleptis + 
Rhagioninae 
Rhagioninae Vermileonidae + 
(Austroleptis + 
Rhagioninae) 
31 
4. Ends, 
highambig 
unalign  
6 monophyletic Bolbomyia Rhagioninae Vermileonidae 36 
5. 
Modambig, 
highambig, 
unalign  
54 paraphyletic Rhagioninae (Spaniinae + 
Chrysopilinae) + 
(Bolbomyia+ 
(Rhagioninae + 
Vermileonidae)) 
Rhagioninae + 
Vermileonidae 
30 
6. Ends, 
modambig, 
highambig, 
unalign  
21 paraphyletic Bolbomyia (Rhagioninae + 
(Spaniinae + 
Chrysopilinae)) + 
(Bolbomyia+ 
Vermileonidae) 
Vermileonidae 32 
 
 235 
 
excluded 
character 
sets of AS 
No. of  
MPTs 
Tabanomorpha 
monophyly/ 
paraphyly 
Sister group 
of 
Vermileonidae
Sister group of 
Austroleptis 
Sister group of  
Bolbomyia 
No. of resolved strict 
consensus nodes 
7. 
Lowambig, 
modambig, 
highambig, 
unalign  
36 monophyletic Rhagioninae + 
(Spaniinae + 
Chrysopilinae) 
rest of 
Tabanomorpha 
Tabanoidea 36 
8. Ends, 
lowambig, 
modambig, 
highambig, 
unalign  
27 monophyletic Rhagioninae + 
(Spaniinae + 
Chrysopilinae) 
rest of 
Tabanomorpha 
Tabanoidea 27 
 
Table 9. Relationships among the major taxa, evident in the Maximum Likelihood trees of different alignment sets (AS). 
Alignment sets Sequence 
length (bp) 
Tabanomorpha 
monophyly/ 
paraphyly 
Sister group of 
Vermileonidae 
Sister group 
of Austroleptis
Sister group of  
Bolbomyia 
1. Only unalign regions excluded 2944 monophyletic Rhagionidae, 
sensu lato 
Rhagioninae + 
(Chrysopilinae 
+ Spaniinae) 
Austroleptis 
(Rhagioninae + 
(Chrysopilinae + 
Spaniinae)) 
2. Ends and only unalign excluded 2888 monophyletic Rhagionidae, 
sensu lato 
Rhagioninae + 
(Chrysopilinae 
+ Spaniinae) 
Austroleptis 
(Rhagioninae + 
(Chrysopilinae + 
Spaniinae)) 
3. Highambig and unalign regions 2873 monophyletic rest of Rhagioninae + Tabanoidea 
 
 236 
 
Alignment sets Sequence 
length (bp) 
Tabanomorpha 
monophyly/ 
paraphyly 
Sister group of 
Vermileonidae 
Sister group 
of Austroleptis
Sister group of  
Bolbomyia 
excluded Tabanomorpha (Chrysopilinae 
+ Spaniinae) 
4. Ends, highambig and unalign 
regions excluded 
2817 monophyletic rest of 
Tabanomorpha 
Rhagioninae + 
(Chrysopilinae 
+ Spaniinae) 
Tabanoidea 
5. Modambig, highambig, and 
unalign regions excluded 
2774 paraphyletic Bolbomyia 
(together, they 
Xylophagidae Vermileonidae 
6. Ends, modambig, highambig, and 
unalign regions excluded 
2718 paraphyletic Bolbomyia 
(together, they 
Xylophagidae Vermileonidae 
7. Lowambig, modambig, highambig, 
and unalign regions excluded 
2735 monophyletic Bolbomyia + 
Tabanoidea 
rest of 
Tabanomorpha
Tabanoidea 
8. Ends, lowambig, modambig, 
highambig, and unalign regions 
excluded 
2679 monophyletic Bolbomyia + 
Tabanoidea 
rest of 
Tabanomorpha
Tabanoidea 
 
 
 
Table 10. Average pairwise genetic distance between each species and the rest of the 
sampled set. All characters included except the ?unalign? character set (AS #1).  
The mean of all pairwise distances is 0. 06139372. 
Species Avg vs. all other 
taxa 
Deviation from 
mean 
Arthropeas magnum 0.05553619 (0.0058575) 
Atherimorpha atrifemur 0.054439048 (0.0069547) 
Atherimorpha sp. 1 (South
America) 
0.055517143 (0.0058766) 
Atherimorpha sp. 2 (Australia) 0.052862619 (0.0085311) 
Atherimorpha sp. 3 (Australia) 0.050461429 (0.0109323) 
Atherimorpha vernalis  0.052245714 (0.0091480) 
Atherix variegata 0.078396905 0.0170032 
Austroleptis multimaculata  0.060189048 (0.0012047) 
Austroleptis rhyphoides 0.053920952 (0.0074728) 
Austroleptis collessi  0.056150000 (0.0052437) 
Bolbomyia nana 0.052441190 (0.0089525) 
Chrysopilus quadratus 0.051426667 (0.0099671) 
Chrysopilus thoracicus 0.058680000 (0.0027137) 
Chrysopilus sp. 1 (Ecuador) 0.059182857 (0.0022109) 
Chrysopilus sp. 2 (Australia) 0.050049048 (0.0113447) 
Chrysopilus sp. 3 (Tasmania) 0.050351429 (0.0110423) 
Chrysopilus sp. 3 (Tasmania) 0.050632857 (0.0107609) 
Chrysops sp. 0.075362619 0.0139689 
Coenomyia ferruginea 0.061550952 0.0001572 
Dasyomma sp. 1 (Chile) 0.077999524 0.0166058 
Dasyomma sp. 2 (Chile) 0.077398810 0.0160051 
Dialysis elongata 0.059441905 (0.0019518) 
Exeretonevra angustifrons 0.073409762 0.0120160 
Glutops rossi 0.060331429 (0.0010623) 
Heterostomus sp. 0.056274048 (0.0051197) 
Leptynoma hessei 0.064375000 0.0029813 
Pelecorhynchus personatus 0.071864762 0.0104710 
Ptiolina fasciata 0.050122857 (0.0112709) 
Rhagio hirtus 0.057822381 (0.0035713) 
Rhagio mystaceus 0.059472619 (0.0019211) 
Spaniopsis clelandi 0.054455476 (0.0069382) 
Spaniopsis longicornis 0.053835238 (0.0075585) 
Symphoromyia atripes 0.056689762 (0.0047040) 
Symphoromyia hirta 0.055975238 (0.0054185) 
 
 238 
 
Species Avg vs. all other 
taxa 
Deviation from 
mean 
Symphoromyia sp. (Alaska) 0.051126429 (0.0102673) 
Tabanus atratus 0.077501905 0.0161082 
Tabanus rufofrater 0.089475238 0.0280815 
Tabanus sp. 0.082094762 0.0207010 
Vermileo opacus 0.053200238 (0.0081935) 
Vermileo sp. 0.056690476 (0.0047032) 
Xylophagus abdominalis 0.067588333 0.0061946 
Pachygaster leachii 0.080747619 0.0193539 
Pantophthalmus sp. 0.072639524 0.0112458 
 
Table 11. Average pairwise genetic distance between each clade and the rest of the 
sampled set.  All characters included except the ?unalign? character set (AS #1). 
 
Clades Avg pairwise distance  
vs. all other taxa 
 
Avg. deviation from the mean 
Outgroup 0.076693571 0.01529985 
Xylophagidae 0.062300198 0.00090648 
Tabanoidea 0.073573615 0.01217989 
Vermileonidae 0.058088571 (0.00330515) 
Rhagioninae 0.054688707 (0.00670501) 
(Chrysopilinae 
+ Spaniinae) 0.053543988 (0.00784973) 
Bolbomyia 0.052441190 (0.00895253) 
Austroleptis 0.056753333 (0.00464039) 
 
 
 239 
 
Table 12. Overlap between most parsimonious consensus trees and bootstrap 
consensus trees. The highest overlap between all compatible majority rules 
consensus trees and bootstrap consensus trees occurs in alignment sets 3, 4, and 6. 
The highest overlap between strict consensus and bootstrap consensus trees 
occurs in alignment set #4. 
Alignment sets (ASs) Sequence 
length 
No. of 
MPTs 
No. of Strict / MR 
consensus nodes 
consistent with 
bootstrap (all 
compatible) consensus 
tree 
Clades with > 
50% bootstrap 
values in strict 
consensus 
(bootstrap value) 
1. Only unalign 
regions excluded 
2944 26 33 / 36  - 
2. Ends and only 
unalign excluded 
2888 6 33 / 36 - 
3. Highambig and 
unalign regions 
excluded 
2873 30 31 / 38 Monophyly of 
Tabanomorpha (68)
4. Ends, highambig 
and unalign regions 
excluded 
2817 6 36 / 38  Monophyly of 
Tabanomorpha (65)
5. Modambig, 
highambig, and 
unalign regions 
excluded 
2774 54 30 / 36 - 
6. Ends, modambig, 
highambig, and 
unalign regions 
excluded 
2718 21 32 / 38  - 
7. Lowambig, 
modambig, 
highambig, and 
unalign regions 
excluded 
2735 36 32 / 36  Monophyly of 
Tabanomorpha (69)
8. Ends, lowambig, 
modambig, 
highambig, and 
unalign regions 
excluded 
2679 27 35 / 37 Monophyly of 
Tabanomorpha (71)
 
 
 240 
 
Maximum Parsimony Analysis of Molecular Dataset 
Within Tabanoidea, the Pelecorhynchidae (G. rossi and P. personatus), Athericidae 
(A. variegata and Dasyomma spp.), and Tabanidae (Tabanus spp. and Chrysops sp.) 
are shown to be monophyletic, with Athericidae and Tabanidae sister taxa (Fig. 2). 
Vermileo is shown to be paraphyletic with respect to Leptynoma hessei in all 
analyses. In the Rhagioninae, both Rhagio and Atherimorpha are monophyletic, and 
within Atherimorpha, there are two monophyletic clades, consisting of South 
American and Australian species, respectively. Within Chrysopilinae, species of 
Chrysopilus are formed by monophyletic New World and Australian clades. 
Chrysopilinae is sister to the Spaniinae. Within Spaniinae, Ptiolina is recovered basal 
to sister genera Spaniopsis and Symphoromyia. Bolbomyia is recovered as sister to the 
vermileonids in the strict consensus. However, this relationship is only supported by a 
20% bootstrap value. The internal relationships of each of these eight clades, where 
resolved, is consistent throughout all analyses, except for the Xylophagomorpha 
where there are varying internal relationships across the trees and the entire group is 
sometimes shown as paraphyletic (in alignment sets 7 and 8). 
 
For all figures, the current taxonomic placement of each species is indicated in 
parenthesis next to the taxon name. A summary of the taxonomic abbreviations is 
presented in Table 10. The most parsimonious tree found for AS #4 is presented in 
Fig. 2. 
 
 
 241 
 
Eight clades of the ingroup (exclusive of the stratiomyiomorphs Pachygaster leachii 
and Pantophthalmus sp.) were consistently recovered in all analyses (regardless of 
alignment), although the particular relationships between these clades were not. To 
facilitate reporting of the results and subsequent discussion, I will refer to these eight 
clades that are of interest. These are the Tabanoidea, Xylophagomorpha, 
Vermileonidae, Rhagioninae, Spaniinae, Bolbomyia, and Austroleptis (Table 9).   
Table 13. Clades of interest. 
Clade of interest Taxonomic rank of clade Taxa included in clade 
Xylophagomorpha Infraorder Xylophagidae 
Tabanoidea Superfamily Athericidae, Tabanidae, 
Pelecorhynchidae 
Vermileonidae Family Vermileonidae (e.g., 
Vermileo and Leptynoma) 
Rhagioninae Subfamily Rhagio, Atherimorpha 
(and related genera not 
represented in the 
molecular data set) 
Spaniinae Subfamily Ptiolina, Spania, 
Spaniopsis, Symphoromyia 
Chrysopilinae Subfamily Chrysopilus (and related 
genera not represented in 
the molecular data set) 
Austroleptis Genus (family?) Austroleptis  
Bolbomyia Genus (family?) Bolbomyia 
 
 
Since Xylophagomorpha consists of a single family, the Xylophagidae, I may use 
these two terms interchangeably in reference to the same clade. The 
Xylophagomorpha in this sample includes Arthropeas magnum, Coenomyia 
ferruginea, Dialysis elongata, Exeretonevra angustifrons, Heterostomus sp., and 
Xylophagus abdominalis.  
 
 242 
 
 
Hennig (1973) and others (McAlpine, 1981) have called the Tabanoidea the group 
that I consider the infraorder Tabanomorpha (Woodley, 1989). I consider the 
Tabanoidea in a new sense, as a superfamily that includes the families 
Pelecorhynchidae, Tabanidae, and Athericidae. In this sample, the tabanoid species 
are Atherix variegata, Chrysops sp., all species of Dasyomma, Glutops rossi, 
Pelecorhynchus personatus, and all species of Tabanus.  
 
The subfamily Rhagioninae is recognized in a new sense here and includes only close 
relatives of Rhagio, and their relatives. For this sample, the rhagionines are 
represented by Rhagio and Atherimorpha.  
 
The subfamily Spaniinae is also recognized in a new sense here. Spaniinae is defined 
by a synapomorphy present in the female terminalia which I will discuss below. In 
this sample, the Spaniinae includes Ptiolina, Symphoromyia, Spania and Spaniopsis.  
 
The final two clades of interest are each represented by single genus, Bolbomyia and 
Austroleptis.  
 
The average pairwise genetic distance between each of these clades and the rest of the 
sampled taxa is shown in Table 7. 
 
 243 
 
Table 14. Taxonomic abbreviations. 
Abbreviation Family Infraorder 
Ath Athericidae  Tabanomorpha 
Pel/Rhag Pelecorhynchidae sensu Woodley, 1989 or 
Rhagionidae sensu Stuckenberg 2001 
Tabanomorpha 
Rhag Rhagionidae Tabanomorpha 
Tab Tabanidae Tabanomorpha 
Rhag/Aust Rhagionidae sensu Woodley, 1989 or 
Austroleptidae sensu Stuckenberg 2001 
Tabanomorpha 
Rhag/Span Rhagionidae sensu Woodley, 1989 or 
Spaniidae sensu Stuckenberg 2001 
Tabanomorpha 
STR-Pant Pantophthalmidae Stratiomyomorpha 
STR-Strat Stratiomyiidae Stratiomyomorpha 
Verm Vermileonidae Tabanomorpha or 
Vermileonomorpha 
XYL Xylophagidae Xylophagomorpha 
 
 
 244 
 
 
Figure 5. Maximum Parsimony analysis of 28S rDNA data (AS #4).  Strict consensus 
of 6 most parsimonious trees. Bootstrap values ? 50% are noted above supported 
branches (TBR, nreps = 1000 / addseq reps = 10). 
 
 
 245 
 
Maximum Likelihood Analysis of Molecular Dataset 
In all of the trees, the eight major groups of interest (Table 9) were recovered as in 
MP analysis.  The ML tree for AS # 4 is given in Fig. 3. The relationships within 
these groups are the same for all analyses. Within Xylophagidae, there are two clades. 
D. elongata is the sister to A. magnum + C. ferruginea in one clade and X. 
abdominalis is sister to E. angustifrons + Heterostomus sp. in the other. Each of the 
three tabanoid families is recovered as monophyletic, as are all genera. The 
Pelecorhynchidae are basal to Athericidae + Tabanidae. Within the Vermileonidae, 
Vermileo is recovered as paraphyletic with respect to L. hessei in all analyses. 
Rhagioninae retain a monophyletic Rhagio and Atherimorpha in the ML analyses, 
with Atherimorpha composed of a monophyletic South American and Australian 
clade, respectively. In Chrysopilinae, Chrysopilus spp. are divided into two clades, 
composed of Australian and New World species, respectively. The Chrysopilinae is 
sister to the Spaniinae, which show P. fasciata sister to Spaniopsis spp. + 
Symphoromyia spp. 
 
The way in which each of these eight lineages are related to one another is unclear, 
except in regards to the relationship of Rhagioninae + (Chrysopilinae + Spaniinae) 
which is recovered in all analyses (regardless of alignment). This provides evidence 
in support of a monophyletic Rhagionidae, exclusive of Bolbomyia and Austroleptis. 
In the alignment chosen, Vermileonidae are sister to the rest of the Tabanomorpha, 
which is divided into two lineages. One is composed of Bolbomyia + Tabanoidea. 
 
 246 
 
The other recovers Austroleptis sister to Rhagionidae (Rhagioninae, Chrysopilinae, 
and Spaniinae).  
 
 247 
 
 
 
Figure 6. Maximum Likelihood analysis of 28S rDNA data (AS #4). Bootstrap values 
? 50% are noted above supported branches (NNI, nreps = 50 / addseq reps = 3).  
 
 248 
 
Discussion 
Exploring the effects of alignment 
Ultimately, the most desired sequence alignment is maximally phylogenetically 
informative yet minimally affected by noise and bias from non-phylogenetic sources. 
While this is self-evident in a theoretical framework, realizing such an ideal requires 
difficult practical decisions when working with a large, complex molecular dataset. 
 
Perhaps the only justifiable alignment for use in phylogenetic analysis is one that 
excludes all sites where the homology is not certain. However, although rates may 
vary within a gene, such as is apparent for 28S rDNA, areas experiencing faster rates 
have not necessarily reached saturation for all taxa. Saturation in parts of the 
alignment may not preclude information for determining relationships among closer 
groups, where saturation has not yet been reached. Deciding at what point portions of 
the dataset ought to be removed on the basis of alignment is not clear in all cases. 
This may be important particularly for ribosomal DNA genes, where inherent 
evolutionary constraints associated with the secondary structure of the molecule 
produce regions of highly conserved sequence side-by-side with regions that are more 
free to vary. The molecule in its entirety may carry phylogenetic information, but the 
signal present in conserved regions alone may not be strong enough to be recognized 
amid natural phylogenetic noise. Furthermore, different methods of phylogenetic 
inference may have differing abilities to tolerate noise within a sample, so that the 
point at which a site should be excluded on the basis of uncertain homology 
 
 249 
 
assessment may also differ, for example, depending on whether Maximum Parsimony 
or Maximum Likelihood methods are used. 
Evaluating the Alignment Sets. 
It is hard to evaluate the various alignment sets objectively, as the phylogenetic signal 
of the data set is heterogeneous in unpredictable ways, leading to uneven 
contributions of signal. Even though the datasets have been divided into increasingly 
smaller sets (from AS#1-8), the amount of signal does not necessarily taper at the 
smallest part of this range, in a corresponding manner. The 'noise' of the data may 
also similarly vary among the alignment sets, according to which sites are included, 
rather than the overall size of the data set. It can be argued, however, when the 
internal consistency of the data is higher, noise in the data will be minimized. 
Bootstrapping is a measure of the internal consistency of the data and may therefore 
provide a way in which two data matrices (developed from the same data set, but 
different due to different alignment decisions) may be evaluated for phylogenetic 
utility. The true alignment, like the true tree, is not known so gauging phylogenetic 
accuracy of one alignment over the other is a tricky proposition. The effects of 
various alignment decisions cannot be simply ignored, however, particularly with a 
variable-rate molecule such as large subunit rDNA. Homology assessment in 
molecular data is a very important step in developing robust phylogenetic hypotheses. 
A modest new method for choosing one alignment over the other has been presented 
here as a first attempt to both acknowledge the problem and treat it in an objective 
manner. Certainly this is an aspect of molecular systematics that needs much more 
attention. 
 
 250 
 
 
Assessment of the Most Parsimonious Trees 
Although it may not be immediately clear which of the 8 alignment sets most 
accurately reflects the common evolutionary history of the taxon set, several common 
themes become apparent in the parsimony analysis.  
 
First, the composition of most nodes of the tree is maintained throughout all 
alignments. The formation and resolution of the Rhagioninae, Chrysopilinae, 
Spaniinae, Tabanoidea, and Vermileonidae is consistent throughout all 143 most 
parsimonious trees, totaled among all of the alignments. There is also consistency of 
alignment sets in regards to the basal nodes of the Tabanomorpha lineage. Alignment 
schemes did not affect support levels for these nodes; in all analyses, there is weak 
support at the base of the Tabanomorpha lineage. The relationships among the major 
tabanomorph clades, including Austroleptis and Bolbomyia at best, remain weakly 
supported. The sister group of each of the major groups remains uncertain. 
 
Pairwise distances are susceptible to the same flaw that allows for long branch 
attraction; that is, the actual amount of sequence evolution may be underestimated 
because changes may be overwritten and unrecognized (e.g., a change from A to C, 
back to A). It is instructive to note, however, that Tabanoidea is comparatively quite 
distant from the rest of the taxa (Table 7). Comparatively long branches of 
Tabanoidea and the stratiomyiomorph outgroup taxa may induce maximum 
parsimony to misinterpret random overlap of sequence between these groups as 
synapomorphic change. As a result, in a few analyses (not shown), Tabanoidea is 
 
 251 
 
pulled from the rest of the tabanomorph taxa, to a position sister to the 
Xylophagomorpha, rendering Tabanomorpha paraphyletic. This reconstruction is not 
recovered in any of the bootstrap consensus trees. 
Assessment of the Most Likely Trees 
Although there is no overlap between the four most likely trees and any of the most 
parsimonious trees, produced by all of the alignment data sets, similar patterns 
emerge in each of the analyses. 
 
As in the MP analysis, the formation and resolution of Rhagioninae, Chrysopilinae, 
Spaniinae, Tabanoidea, and Vermileonidae remain consistent throughout. Similarly, 
support levels for nodes at the base of the Tabanomorpha lineage were exceptionally 
low; this area of the tree is particularly unstable. Unlike in the MP analyses, however, 
Rhagioninae is consistently recovered as sister taxa to the clade formed by 
Chrysopilinae and Spaniinae. Although recovered in all ML analyses, this 
relationship (Rhagioninae + (Chrysopilinae + Spaniinae) has low bootstrap support 
(55%). 
Monophyly of Rhagionidae 
Recent concepts of the Rhagionidae are rejected by the molecular evidence. A 
monophyletic Rhagionidae sensu Nagatomi (1982) invokes an additional 31 steps to 
the most parsimonious tree (Table 10). Similarly, the monophyly of Rhagionidae 
sensu Stuckenberg (2001) requires that 30 steps be added to the most parsimonious 
arrangement, given the data. Other concepts of the subfamily Rhagioninae 
(Nagatomi, 1982a; Stuckenberg, 2001) are also inconsistent with the molecular 
 
 252 
 
results. As these are compound hypotheses, it is somewhat easier to discuss why such 
little support is afforded to them by examining their components. This is done in the 
following sections. 
 
 
Table 15. Hypotheses tested in light of molecular evidence. 
Concept Authors Constraint Requisite 
increase in steps 
(MP) 
Wilcoxon 
Signed-Rank 
Test  
(p value) 
Kishino- 
Hasegawa 
Test 
(p value) 
Monophyly and Position of Pelecorhynchidae 
Rhagionidae includes 
Glutops 
Nagatomi, 
1982a 
Atherimorpha+ Austroleptis + 
Bolbomyia + Chrysopilus + 
Glutops + Ptiolina + Rhagio 
+  Symphoromyia + 
Spaniopsis  
31 0.0016* 0.0016* 
Rhagionidae includes 
Glutops and 
Pelecorhynchus 
Stuckenberg, 
2001 
Atherimorpha+ Bolbomyia + 
Chrysopilus + Glutops + 
Pelecorhynchus +  Rhagio +  
Symphoromyia  
30 0.0031* 0.0049* 
Glutops and 
Pelecorhynchus 
(with Pseudoerinna) 
form monophyletic 
unit 
Woodley, 1989; 
Sinclair, 1992; 
Stuckenberg, 
2001 
Glutops + Pelecorhynchus 0 - - 
Position of Vermileonidae 
Vermileonomorpha Nagatomi, 
1977;  et al. 
Vermileonidae  + 
(Xylophagidae + 
Tabanomorpha) 
13 0.0906 0.0906 
Vermileonidae, basal 
clade of 
Tabanomorpha 
Nagatomi, 
1977; Sinclair 
et al., 1993 
(Vermileonidae + (other  
tabanomorph taxa) 
0 - - 
 
 254 
 
Concept Authors Constraint Requisite 
increase in steps 
(MP) 
Wilcoxon 
Signed-Rank 
Test  
(p value) 
Kishino- 
Hasegawa 
Test 
(p value) 
Position of Bolbomyia 
Bolbomyia within 
Tabanoidea 
Sinclair et al., 
1993 
Pelecorynchidae (Bolbomyia 
+ (Athericidae + Tabanidae)) 
20 0.0230* 0.0235* 
Bolbomyia sister to 
Athericidae and 
Tabanidae 
Sinclair et al., 
1993 
Bolbomyia + (Athericidae + 
Tabanidae) 
20 0.0604 0.0610 
Bolbomyia in 
Xylophagomorpha 
James, 1965 Bolbomyia + Xylophagidae 8 0.1167 0.1167 
Unnamed concept 
defined by 
lacking M
3
 vein 
Grimaldi and 
Cumming, 1999 
Bolbomyia + Austroleptis 3 0.6744 0.7745 
Position of Austroleptis 
Austroleptinae Nagatomi, 
1982a 
(Austroleptis + (Atherimorpha 
+ Bolbomyia + Chrysopilus + 
Ptiolina + Rhagio + 
Spaniopsis + Symphoromyia)) 
5 0.4658 0.4659 
Austroleptis 
placement in or sister 
to Xylophagidae 
Sinclair et al., 
1994 
Austroleptis + Xylophagidae 5 0.3980 0.3981 
Position and monophyly of Spaniidae 
Spaniinae / 
Spaniidae (exclusive 
of Symphoromyia) 
Nagatomi, 
1982a; 
Stuckenberg, 
2001 
 
Spaniopsis + Ptiolina 3 0.7142 0.7099 
 
 255 
 
Concept Authors Constraint Requisite 
increase in steps 
(MP) 
Wilcoxon 
Signed-Rank 
Test  
(p value) 
Kishino- 
Hasegawa 
Test 
(p value) 
Composition of Rhagioninae 
Rhagioninae  Nagatomi, 
1982a 
Atherimorpha + Bolbomyia + 
Chrysopilus + Rhagio +  
Symphoromyia 
22 0.0470* 0.0428* 
Rhagioninae Stuckenberg, 
2001; et al. 
Atherimorpha + Chrysopilus 
+ Rhagio + Symphoromyia 
17 0.1113 0.1066 
Chrysopilus within 
Rhagioninae 
Nagatomi, 
1982a; et al. 
Chrysopilus + (Rhagio and 
Atherimorpha) 
10 0.2102 0.2114 
Chrysopilus sister to 
Rhagio 
- Chrysopilus + Rhagio 19 0.0502 0.0512 
 
 
 256
Monophyly and composition of Pelecorhynchidae 
Various hypotheses have been presented for the composition and placement of 
Glutops and Pelecorhynchus (Mackerras and Fuller, 1942; Teskey, 1970; 
Krivosheina, 1971; Nagatomi, 1982a; Woodley, 1989; Sinclair, 1992; Wiegmann et 
al., 2000; Stuckenberg, 2001; among others). Perhaps it should not be too surprising 
that the pelecorhynchid genera are recovered here as a monophyletic unit, sister to the 
Athericidae and Tabanidae, confirming the results of Wiegmann et al. (2000). The 
Wiegmann et al. study (2000) however used a smaller fragment of the molecule and 
examined the question only in terms of a single MP analysis. Here, the relationship is 
confirmed by both ML and MP analyses, in a way that may refute contentions that a 
biased alignment may have affected the results. Support values for this relationship 
are similar across alignment sets and similar to those found by Wiegmann et al. 
(2000).  
 
These results appear to contradict the hypotheses that Glutops belongs in its own 
genus (Krivosheina, 1971); that Glutops is a part of Rhagionidae, exclusive of 
Pelecorhynchus (Nagatomi, 1982a); and that Glutops and Pelecorhynchus form a 
monophyletic group within the rhagionid lineage (Stuckenberg, 2001).  
Monophyly and composition of Chrysopilinae and Spaniinae 
In all analyses, Chrysopilus was placed as sister to Ptiolina, Spaniopsis, and 
Symphoromyia. The bootstrap values for this placement ranged from 62-100% in ML 
and 56-99% in MP. The lowest of these bootstrap values were recovered in 
alignments that generally had weak tree support throughout (AS #5 in ML, and AS #5 
 
 257
and AS#6 in MP). Ptiolina and Spaniopsis have been recognized as members of the 
Spaniinae (Nagatomi, 1982a; others) or as Spaniidae (Stuckenberg, 2001). However, 
the molecular data makes this pairing paraphyletic with respect to Symphoromyia in 
all analyses. Although only three additional steps are required to force Ptiolina and 
Spaniopsis together in the MP analysis (using AS # 4), bootstraps supporting a 
Symphoromyia and Spaniopsis sister group relationship are typically above 50% 
(appx. 73-78% in ML, 53% - 81% in MP). Symphoromyia has not been grouped with 
Ptiolina and Spaniopsis in the modern era although the genus was originally 
established by Frauenfeld to clarify the distinction of species that had been 
mistakenly placed in Ptiolina. Symphoromyia differs from Spaniopsis and Ptiolina in 
several aspects, most conspicuously in the form of the antenna, but these genera share 
a special resemblance in their compact habitus, which is difficult to describe. The 
same cannot be said about Chrysopilus, which has never been mistaken for any of the 
typical spaniine genera. Yet forcing Chrysopilus into the Rhagioninae, its traditional 
affiliation, adds 10 additional steps to the tree. Given this result, it is expected that 
morphological evidence may be found to support the relationship between 
Chrysopilus and the spaniine genera (including Symphoromyia). The morphological 
analysis that follows will discuss this further and provide new evidence that may 
independently support sister group relationship between Chrysopilinae and Spaniinae. 
Bolbomyia 
The position of Bolbomyia is unstable among the basal branches of Tabanomorpha. It 
may be recovered as sister to any of the major tabanomorph groups and none of these 
 
 258
relationships are strongly supported. Several hypotheses regarding its placement, 
however, apparently may be rejected.  
 
The first is that Bolbomyia is sister to the Athericidae and Tabanidae (Sinclair et al., 
1993). When the data are constrained to this relationship, an additional 20 steps are 
forced upon the most parsimonious arrangement. This suggests that aedeagal tines of 
the male genitalia may be homoplasious and the tines in Bolbomyia are not 
homologous to those in Athericidae and Tabanidae, as Sinclair et al. have asserted 
(1993).  
 
Loew (1850) originally placed Bolbomyia in the Xylophagidae, a position that James 
(1965) retained in the Catalog of the Diptera of America North of America. Although 
the Xylophagidae concept of Loew and James were from different each other, and 
different from the Xylophagidae of today, Bolbomyia does not have a bulbous 
clypeus, a typical tabanomorph character and speculation may remain whether or not 
this genus belongs within the Tabanomorpha. When Bolbomyia is constrained as a 
xylophagomorph, eight steps are added to the most parsimonious tree. This is a 
clearly a less parsimonious arrangement for Bolbomyia than any of the possible 
positions in Tabanomorpha (outside of established groups), even though a specific 
sister group for Bolbomyia has not been recognized. Forcing Bolbomyia as sister to 
Austroleptis 3 steps are added.  
 
 259
Austroleptis 
The phylogenetic signal of Austroleptis is similar to that found in Bolbomyia. The 
divergence of this genus from other taxa in the sampled set appears quite ancient, 
where the resolving power of the 28S rDNA is weak. It is recovered within 
Tabanomorpha in all analyses, and this is most likely its correct location, however on 
account of its poorly resolved status with respect to specific sister groups, alternate 
hypotheses aren't as easily rejected. Five additional steps are required to place this 
genus within the Xylophagomorpha; the same may be said for Nagatomi's concept of 
Austroleptinae, a subfamily sister to the rest of the Rhagionidae sensu lato 
(Nagatomi, 1982a). The suggestion that Austroleptis and Bolbomyia form a 
monophyletic entity (with other taxa, not included here), on the basis of the shared 
loss of the third medial vein (Grimaldi and Cumming, 1999) is not recovered in the 
most parsimonious arrangement, yet forcing such a sister group relationship adds only 
3 steps to the tree. This seems like a reflection of the relatively poor resolving power 
of 28S rDNA for both of these taxa rather than the plausibility of this hypothesis, 
however. 
Vermileonidae 
The monophyly of Vermileonidae has never been disputed, on account of many 
morphological and behavioral synapomorphies. The placement of this family, 
however, is unclear and a number of authorities have called for its removal from 
Tabanomorpha (Nagatomi, 1977, 1991; Griffiths, 1994; Stuckenberg, 2001). 
Although its placement with respect to other tabanomorph taxa remains uncertain, the 
Vermileonidae do not appear to represent a separate infraorder. By forcing this 
 
 260
arrangement, 13 steps are added to the most parsimonious arrangement. Placing it 
anywhere within the Tabanomorpha, outside of the Tabanoidea, Rhagioninae, 
Chrysopilinae, and Spaniinae, adds one or no steps to the tree. Within Vermileonidae, 
Vermileo is shown as paraphyletic with respect to L. hessei, with strong support. 
Bootstrap values for the sister group relationship between Vermileo sp. (Israel) and L. 
hessei are 100% in all analyses, for all alignments. Although this may be a surprising 
result, Vermileo is an old, well established genus, the morphology of the larval 
mouthparts confirms that Vermileo of the New World (as V. opacus) exhibit striking 
differences from their congeners in the Old World (noted in Appendix A). 
 
 
 
 261
Combined Analysis 
Introduction 
The strength of an hypothesis may be determined by the quality of its support, from 
all sources of information that are available. There are many potential pitfalls for both 
morphological and molecular methods that in each case, may lead to positively 
misleading conclusions. Therefore, while considering independent lines of evidence 
separately is essential, the integration of this knowledge into a single synoptic 
hypothesis is the ultimate aim of phylogenetic systematics.  
 
There has been some controversy regarding the combined approach. One may argue 
that by combining data sets, the 'total evidence' available is used and the analysis will 
consequently render a most conservative tree with the greatest descriptive and 
explanatory power. However, combining data sets of heterogeneous phylogenetic 
signal may give misleading results (de Queiroz et al., 1995). At the heart of this 
debate is whether or not there are separate classes of evidence. That is, can evidence 
be inherently different and therefore incapable of being compared simultaneously? Or 
are multiple lines of phylogenetic evidence simply the documentation of a single 
series of events from different perspectives? De Quieros et al. (1995) give an 
overview of this discussion.  
 
If data sets are shown to be congruent (i.e., homogeneous) and to have the same 
underlying assumptions, the data may be combined (Bull et al., 1993). In such cases, 
combining data sets, theoretically, will enhance the detection of phylogenetic signal 
 
 262
and generate a more robust phylogenetic hypothesis. Even so, it is always good to 
keep a keen watch for imbalances. Regier and Shultz (1997) note that morphological 
characters may have undue influence when combining data sets given that typological 
approaches may lower the rate of homoplasy and thereby artificially increase 
phylogenetic signal.  
 
This section evaluates the morphological and molecular contributions developed in 
previous sections in a combined analysis using three techniques. The first two use a 
concatenated morphological and molecular data set for simultaneous analysis using 
MP and BI. The final combined approach is a type of meta-analysis, which uses the 
trees generated independently by the morphological and molecular data, to give an 
overall "supertree" representation of the combined phylogenetic hypothesis. 
Materials and Methods 
Taxon Sampling 
For the combined analyses, the intersection of taxa in morphological and molecular 
datasets was sampled for the combined analyses. The molecular data set excluded 
?ends?, ?highambig?, and ?unalign? character sets (AS #4). In some cases, the genera 
in the combined analysis are a chimera of two different species (see Table 15). This 
was allowed only in cases where the genus was sampled once in the dataset and the 
monophyly of the species contained within the genus was irrefutable. In the supertree 
analysis, the dataset is the matrix representation of trees found in separate 
morphological (MP) and molecular (ML) analyses, so that all taxa are included.  
 
 263
Table 16. Combined analysis dataset; morphological and molecular taxon overlap. 
Morphological 
characters 
Molecular characters Recent Family 
Placement 
Atherimorpha atrifemur Atherimorpha atrifemur Rhagionidae 
Atherimorpha vernalis  Atherimorpha vernalis  Rhagionidae 
Atherix pachypus  Atherix variegata Athericidae 
Austroleptis 
multimaculata  
Austroleptis 
multimaculata  
Rhagionidae/ 
Austroleptidae 
Bolbomyia nana  Bolbomyia nana  Rhagionidae 
Chrysopilus quadratus  Chrysopilus quadratus  Rhagionidae 
Chrysopilus thoracicus  Chrysopilus thoracicus  Rhagionidae 
Dasyomma atratulum  Dasyomma sp. Athericidae 
Glutops rossi  Glutops rossi  Rhagionidae/ 
Pelecorhynchidae/ 
Glutopidae 
Pelecorhynchus 
personatus  
Pelecorhynchus 
personatus  
Rhagionidae/ 
Pelecorhynchidae 
Ptiolina majuscula  Ptiolina fasciata  Rhagionidae/ 
Spaniidae 
Rhagio hirtus  Rhagio hirtus  Rhagionidae 
Rhagio mystaceus  Rhagio mystaceus  Rhagionidae 
Spaniopsis longicornis  Spaniopsis longicornis  Rhagionidae/ 
Spaniidae 
Spaniopsis clelandi  Spaniopsis clelandi  Rhagionidae/ 
Spaniidae 
Symphoromyia hirta  Symphoromyia hirta  Rhagionidae 
Tabanus atratus  Tabanus atratus  Tabanidae 
Vermileo vermileo  Vermileo sp.  Vermileonidae 
Arthropeas americana  Arthropeas magnum Xylophagidae 
Coenomyia ferruginea  Coenomyia ferruginea  Xylophagidae/ 
Coenomyiidae 
Dialysis rufithorax  Dialysis elongata  Xylophagidae 
Xylophagus lugens  Xylophagus abdominalis  Xylophagidae 
 
 
 
 264
Combined Phylogenetic Analysis 
The combined analyses were carried out in two parts. First, to gauge the contribution 
of signal from each of the two types of character data, separate analyses were 
performed, using only the morphological or molecular characters. The ?pruned 
morphological? data set refers to the matrix that contains the combined analysis taxon 
set but includes only morphological characters. The ?pruned molecular? data set 
refers to the matrix that contains the combined analysis taxon set but includes only 
molecular characters. The data was then combined and a full analysis of the data was 
carried out using MP and BI. A test was also performed to determine if the molecular 
and morphological data sets are deemed combinable by the partition homogeneity 
test.  
 
MP analyses were carried out using PAUP* (Swofford, 2001). All characters were 
treated unordered, 1000 heuristic search replicates were performed with random-
taxon-addition, tree bisection reconnection (TBR) branch swapping, steepest descent 
and ?MulTrees? options in effect. Scores of all MP trees were verified to avoid 
artifactually inflated sets of MP trees. In order to gauge the internal consistency of the 
data, bootstrap analyses were carried out for each of these analyses. MP bootstrap 
analyses were performed with 500 repetitions, each with 5 random-taxon-addition 
TBR heuristic searches.  
 
For the BI analysis, each run was set for four million generations, with sampling 
every 200
th
 generation for a total of 20,000 tree samples. Tree scores were plotted 
 
 265
against generation number to assess at what point stationarity was reached. All of the 
generations previous to the highest point stationarity are considered burn-in and 
discarded. The posterior distribution of trees sampled from the Markov chain was 
summarized using the ?sumt? command. 
 
Weighting was done in both the MP and BI phylogenetic analyses so that the high 
number of molecular characters did not override the signal of the morphological data. 
Due to the acquisition bias (Lewis, 2001) of the morphological dataset, however, the 
number of constant characters of the sequence data was large and therefore, the 
average per character signal of the molecular dataset was much less than in the 
morphology matrix. Simply weighting each set by the inverse of the total character 
number would de-emphasize the signal in the molecular dataset. Therefore, characters 
of each dataset were weighted by the inverse of the number of phylogenetically 
informative characters. 
Supertree analysis 
The software program RadCon (Thorley and Page, 2000) was used to convert the set 
of trees generated by the separate and complete MP (morphology) and ML (sequence 
data) analyses into its matrix representation. The matrix was then analyzed using MP 
in PAUP*, consistent with the Matrix Representation using Parsimony (MRP) 
consensus method (Baum, 1992; Ragan, 1992; Sanderson et al., 1998). Two MRP 
analyses were carried out. The first analysis used two trees; the strict consensus of the 
trees from the molecular and morphological analyses, respectively. The second MRP 
 
 266
analysis used the most likely tree(s) from the molecular data and all of the most 
parsimonious trees found for the morphological data. 
Results 
Maximum Parsimony Analysis of Combined Data 
Subset Taxon Sampling of Morphological Data  
MP heuristic search of the pruned morphological dataset used for the combined 
analysis gave a single most parsimonious tree (Fig. 5). The reduced sampling 
produced a somewhat different topology than that from the full morphological 
dataset. A monophyletic Tabanomorpha is recovered, consisting of two clades. 
Vermileo sp. is recovered as sister to Austroleptis + Tabanoidea. In the other clade, 
Bolbomyia nana is sister to the remaining rhagionid taxa. In this analysis, Chrysopilus 
is sister to Rhagio spp., with 51% bootstrap support, to the exclusion of 
Atherimorpha. The Spaniinae are recovered sister to the group formed by 
Atherimorpha, Chrysopilus, and Rhagio.  
 
 
 267
 
Figure 7. Maximum Parsimony analysis of pruned morphological matrix. Bootstrap 
values ? 50% are noted above supported branches (TBR, nreps = 500 / addseq 
reps = 5). The terminal named ?Dialysis elongata / A. rufithorax? includes 
molecular information generated from the sequence of D. elongata and 
morphological information generated from specimens of D. rufithorax. The same 
is true for other terminals that have a combination of two names (e.g. ?Ptiolina 
fasciata / P. majuscula?); these taxa are composed of molecular and 
morphological components generated from these species, respectively. 
 
 268
 
Subset Taxon Sampling of Molecular Data  
MP heuristic search of the pruned morphological dataset used for the combined 
analysis also gave a single most parsimonious tree (Fig. 6). As in the morphological 
analysis, reduced sampling had a noticeable effect on the topology of the reduced  
molecular dataset. Again, a monophyletic Tabanomorpha is recovered, however, 
Austroleptis is the basal clade of the group. A comb-like topology of the major groups 
is recovered in the order from most basal to the most derived as: Austroleptis, 
Vermileo, Rhagioninae, Tabanoidea, and Bolbomyia nana + (Chrysopilinae + 
Spaniinae). Bootstrap support of greater than 50% is found for all of the major 
subgroups (Rhagioninae (84%), Tabanoidea (99%), Chrysopilinae + Spaniinae 
(78%)), Chrysopilinae (100%), and Spaniinae (90%), although Pelecorhynchus 
personatus is placed sister to the Tabanidae + Athericidae, to the exclusion of Glutops 
rossi. The paraphyly of Pelecorhynchidae is supported by a bootstrap value of 67% 
for P. personatus (Tabanus sp. (Dasyomma sp. + Atherix sp.)).  
 
 269
 
Figure 8. Maximum Parsimony analysis of pruned molecular matrix. Bootstrap values 
? 50% are noted above supported branches (TBR, nreps = 500 / addseq reps = 5). 
The terminal named ?Dialysis elongata / A. rufithorax? includes molecular 
information generated from the sequence of D. elongata and morphological 
information generated from specimens of D. rufithorax. The same is true for other 
terminals that have a combination of two names (e.g. ?Ptiolina fasciata / P. 
majuscula?); these taxa are composed of molecular and morphological 
components generated from these species, respectively. 
 
 270
 
Partition Homogeneity Test 
A partition homogeneity test showed that the data are combinable (P value = 
0.057778). 
 
Combined Sample of Morphology and Molecular Data  
The MP heuristic search found a single most parsimonious tree (Fig. 7). Character 
sets are downweighted by number of parsimony informative characters (1/351 
molecular, 1/127 morphological). Tree length is 3.49639.  
 
The most parsimonious tree of the combined morphology and molecular sample has 
the same topology as the most parsimonious tree from the morphology maximum 
parsimony analysis. The Tabanoidea is recovered as the basal clade of 
Tabanomorpha, with Vermileonidae sister to the Rhagionidae in its broadest sense, 
including Rhagioninae, Chrysopilinae, Spaniinae, and Austroleptis and Bolbomyia. 
Austroleptis multimaculata and Bolbomyia nana are sister taxa and form the sister 
group to the clade formed by Rhagioninae, Chrysopilinae and Spaniinae. Rhagioninae 
is recovered as the sister taxon of Chrysopilinae plus Spaniinae. The relationships 
within Spaniinae are also preserved as in the morphological analysis. The difference 
between the results of the complete morphology and the combined analysis (aside 
from the difference in taxon sampling and character number) is the bootstrap support 
values for the various clades. In the combined analysis, bootstrap values are higher 
across the board. The Tabanomorpha are shown as monophyletic with a 99% 
 
 271
bootstrap, as is the Tabanoidea (94% bootstrap), Rhagioninae (86% bootstrap), 
Chrysopilinae and Spaniinae (78% bootstrap), and Spaniinae (78% bootstrap). 
Bootstrap values within these groups are also higher in the combined analysis. In 
regards to the morphological/molecular conflict as to the sister group of the 
Spaniopsis group (incl. Spania and perhaps Litoleptis), the combined analysis 
supports the morphological placement of Ptiolina + Spaniopsis. Although bootstrap 
values for the Ptiolina + Spaniopsis clade are less than 50% (48%), they are greater 
than the bootstrap values for a Symphoromyia + Spaniopsis clade (42%), and this is 
the relationship recovered in the most parsimonious reconstruction. 
 
 
 272
 
Figure 9. Maximum Parsimony analysis of combined matrix. Character sets 
downweighted by the number of parsimony informative characters within each set 
(1/351 molecular, 1/127 morphological). Bootstrap values ? 50% are noted above 
supported branches (TBR, nreps = 500 / addseq reps = 5).  
 
 273
Bayes Analysis of Combined Dataset 
Subset Taxon Sampling of Morphological Data  
Mr. Bayes uses the Mkv model (Lewis, 2001) to model morphological data, which is 
essentially a Jukes-Cantor model modified to allow for a greater number of character 
states (ref). The Bayes tree of the morphological data ?pruned? subset gives a novel 
topology with weak posterior probabilities (Fig. 8). A monophyletic Tabanomorpha is 
recovered. At the base of the Tabanomorpha, Vermileo sp. is recovered as sister to 
Tabanoidea. The Spaniinae is maintained as a monophyletic group that is sister to a 
diverse clade that includes Bolbomyia, Austroleptis, Chrysopilinae, and a polyphyletic 
Rhagioninae. Sister to the Atherimorpha is Bolbomyia + Austroleptis. Sister to this, 
within the other subgroup, is a clade formed by Chrysopilus and Rhagio. The 
posterior probabilities of this diverse clade (and within it) are weak, however.  
 
 274
 
Figure 10. Bayesian tree for pruned morphology matrix. Posterior probabilities are 
noted above branches. 
 
 
 275
Subset Taxon Sampling of Molecular Data  
Bayes analysis of the molecular data ?pruned? subset yields yet another topology 
within the Tabanomorpha (Fig. 9). Within a monophyletic Tabanomorpha, Vermileo 
is the basal taxon. The tabanoid clade is located sister to the Rhagionidae in its 
broadest sense. In this construction, Bolbomyia is sister to Austroleptis + 
Rhagioninae, which in turn, is sister to the clade formed by monophyletic subfamilies 
Spaniinae and Chrysopilinae. Within the Spaniinae, Ptiolina is recovered as the basal 
taxon and there is a 93% posterior probability for the sister group relationship 
between Symphoromyia and Spaniopsis.  
 
 276
 
Figure 11. Bayesian tree generated from molecular characters of data matrix  
composed of taxa appropriate for a combined molecular and morphological 
analysis. Posterior probabilities are noted above branches. 
 
 
 277
Combined Analysis of Morphology and Molecular Data  
Bayes analysis of the molecular data ?pruned? subset yields yet another topology 
within the Tabanomorpha (Fig. 10). Within a monophyletic Tabanomorpha, Vermileo 
is the basal taxon, the tabanoid clade is then located sister to the Rhagionidae in the 
broadest sense. In this construction, Bolbomyia is sister to Austroleptis + Rhagioninae 
and together, these taxa are sister to the clade formed by Spaniinae and 
Chrysopilinae. Within the Spaniinae, Ptiolina is recovered as the basal group and 
there is a 93% posterior probability for the sister group relationship between 
Symphoromyia and Spaniopsis. Since posterior probabilities are known to often be 
overestimates of the actual support of the data (Douady et al., 2003), another 
representation of this result may be given in the form of a consensus tree restricted to 
posterior probability values of greater than 80%. This tree shows a monophyletic 
Rhagionidae sensu lato, without resolution of its basal lineages. 
 
 278
 
Figure 12. Bayesian tree for combined morphology and molecular data. Posterior 
probabilities are noted above branches. 
Supertrees Analysis  
Matrix Representation using Parsimony (MRP) 
The supertree analysis is different from the two previous combined analyses by 
including all taxa, of both molecular and morphological datasets. This yields a tree 
 
 279
with 71 terminal taxa. The first MRP analysis combines two trees, each the strict 
consensus from the MP analysis of morphological data and the ML analysis of 
molecular data, respectively. This may be considered the most conservative 
interpretation of the combined results. The second MRP analysis allows all of the 
most likely (3) and most parsimonious (24) trees from the molecular and 
morphological analyses, respectively.  
 
The MRP analysis of the strict consensus trees generated from molecular and 
morphological datasets yields 55733 trees. The strict consensus of these trees (Fig. 
11) shows Xylophagomorpha fully resolved, sister to Tabanomorpha. Vermileonidae 
are sister to the Rhagionidae, sensu lato, whose basal clade is composed of Alloleptis 
tersus, Bolbomyia nana, and Austroleptis spp.  The remainder of the tree shows 
Rhagioninae sister to Chrysopilinae plus Spaniinae, which are sister to one another. 
Within the Rhagioninae, Atherimorpha comprises two unresolved clades, one of 
which contains Arthroteles bombyliiformis. Rhagio is recovered as a monophyletic 
group, sister to a clade formed of Desmomyia thereviformis, and Rhagina incurvatus 
plus Sierramyia chiapasensis. Chrysopilinae is composed of Chrysopilus, Schizella 
furcicornis, and Stylospania lancifera, which are mostly unresolved. Chrysopilus may 
be paraphyletic with respect to S. furcicornis, S. lancifera, or both.  Arthroceras 
pollinosum is sister to the Spaniinae and forms its own group, the Arthrocerinae. 
Within the Spaniinae, Ptiolina is again recovered as paraphyletic and sister to 
Spaniopsis (L. alaskensis + S.  nigra), as in the morphological analysis. 
 
 280
 
Figure 13. Strict consensus of 958299 trees from MRP analysis.  
 
 281
Discussion. 
Combined analysis 
The combined maximum parsimony analysis produced essentially the same topology 
as recovered using morphological data and the basal tabanomorph topology is 
consistent with the molecular tree recovered by Wiegmann et al. (2000).  
More taxa or more characters? Both! 
A coincidental result of this study involves the effects of taxon sampling (LeCointre 
et al., 1993; Hillis, 1998; Poe, 1998; Rosenberg and Kumar, 2001; Zwickl and Hillis, 
2002; Hillis et al., 2003). Data sets with the full complement of taxa in the 
morphological and molecular analysis converge on similar topologies, and these 
resemble the topology of the combined analysis where, although the taxon sampling 
is reduced, there are an increased number of characters. When the taxon sampling is 
reduced for strictly morphological or molecular data sets, however, phylogenetic error 
becomes apparent. The trees produced by these sets are unusual. This result is 
consistent with those who have argued that both the number of characters and the 
number of taxa sampled are important for phylogenetic accuracy (Swofford et al., 
1996). The results may also suggest that, at least in this case, the problems associated 
with a limited taxon sample may be overcome somewhat by adding more characters. 
Similarly, it appears that phylogenetic accuracy is improved with greater taxon 
sampling.  
 
 
 282
Bayes analysis 
The Bayes analysis of the combined, full data set gives a result that is similar to the 
most likely tree when only the molecular data are considered. Vermileonidae are 
placed at the base of the Tabanomorpha, although the posterior probability of this 
placement is relatively low. The bayesian combined tree differs from the trees 
generated in other analyses in the placement of Austroleptis and Bolbomyia however 
these relationships also have low support. Since posterior probabilities are known to 
inflate support values (Douady et al., 2003), I prefer to collapse nodes with posterior 
probability values of less than 80%. When this is done, areas of the tree that have 
weak throughout the tree collapse. The monophyly of Rhagionidae, including 
Austroleptis and Bolbomyia is retained. Although the Bayes tree is a reasonable 
estimate of the relationships, I tend to give this tree less weight because the model on 
which the morphological data is based is very simple and may not accurately 
represent the evolution in this part of the data set. When nodes of low probability are 
collapsed, the tree is consistent with all other analyses. 
Supertree analysis 
Supertree analysis is an alternative method of deriving comprehensive topologies 
from various sources of character data. The input of supertree analysis, however, is in 
the form of trees and not the character data itself. This makes supertrees analysis 
vulnerable to confounding effects that may result in phylogenetic error, for a variety 
of reasons (Gatesy et al., 2002). The two biggest problems are that one, the input trees 
may be derived from different studies, employing different levels of phylogenetic 
rigor. A related issue is that the levels of support in each of the branches of the trees 
 
 283
are not incorporated into the supertree matrix. The second problem is that in many 
supertrees analyses, there is a redundancy of character data, resulting in nodes that are 
supported by multiple synapomorphies from different data sets, but ultimately are 
simply repetitions of the same character state.  
 
The supertree analysis here avoids both of those problems. Each of the two input trees 
was generated by systematic phylogenetic analysis of equal rigor. Furthermore, the 
input trees were derived from completely independent data sets.  
 
Gatesy et al. (2002) recommended that large trees be constructed using a supermatrix 
of all characters rather than employing supertree analysis from all available trees. The 
combined analysis in this study confirms that a ?supermatrix? may have improved 
resolving power over smaller matrices, analyzed alone. But what if taxon sampling is 
an issue? Many of the taxa included in the morphological data set were unavailable 
for molecular study. Therefore, the combined analysis was limited in scope. 
 
There may be some justification for forming a supertree by simply grafting richly 
sampled morphological branches onto the molecular tree, where common ancestry 
may be assumed. However, I prefer to use the explicit methods of MRP. 
 
The results of the supertree analysis were remarkable in that the recovered topology is 
consistent with the one found by the combined analysis.  
 
 284
Conclusions of the Phylogenetic Analysis. 
Morphological and molecular data, when analyzed separately and in combination, 
yield similar hypotheses of the evolution within the Tabanomorpha. Tabanoidea 
(Pelecorhynchidae, Athericidae, and Tabanidae), Vermileonidae, Rhagioninae 
(Atherimorpha, Desmomyia, Rhagina, and Rhagio), Chrysopilinae (Chrysopilus, 
Schizella, and Stylospania), Arthrocerinae (Arthroceras), and Spaniinae (Ptiolina, 
Spania, Spaniopsis, and Symphoromyia) are recovered consistently with high 
bootstrap support. Sierramyia and Litoleptis probably belong to Rhagioninae and 
Spaniinae, respectively, however limited taxon sampling prevents confirmation of 
this (Sierramyia male needed for examination of aedeagal form; Litoleptis female 
needed for confirmation of tergite 10 and spermathecal duct morphologies). 
 
These data conflict with the most recent treatment of Tabanomorpha (Stuckenberg, 
2001), particularly in regards to the composition of Rhagionidae. It has been shown 
here that Glutops, Pelecorhynchus, and Pseudoerinna do not belong in the 
Rhagionidae.  
 
The morphology of Bolbomyia (consistent with 28S rDNA sequence data) suggests 
that it is distantly related to the rest of the rhagionid taxa and it may be placed in its 
own family, Bolbomyiidae.  
 
Over the course of the long history of the concept of the Rhagionidae, the family was 
always considered (whether explicitly or not) to include the single common ancestor 
 
 285
of Chrysopilus and Rhagio, and all of its descendants. Evidence is presented here to 
show that Chrysopilus is more closely related to the spaniine group than it is to 
Rhagio. Thus, if Rhagionidae are defined on such terms today, descendents of the 
single common ancestor of Chrysopilus and Rhagio most certainly include the 
spaniine taxa. On this basis, Spaniinae should remain a subfamily within the 
Rhagionidae.  
 
The evolution of blood feeding within Rhagionidae remains unclear. The molecular 
data suggest that blood feeding arose once within the Rhagionidae, in the ancestor of 
sister taxa Symphoromyia and Spaniopsis. The morphological data suggest, however, 
that with the consideration of additional genera, Symphoromyia and Spaniopsis are 
more distantly related; Spaniopsis forms a clade with Litoleptis alaskensis and Spania 
nigra that is sister to two separate lineages of Ptiolina. In interpreting the evolution of 
this behavior from a morphological perspective, one must conclude that blood feeding 
arose separately, once in each of the Symphoromyia and Spaniopsis lineages, 
respectively. The combined maximum parsimony analysis shows that the two 
hypotheses are equally parsimonious, while the bayesian inference of the combined 
data set is consistent with the single origin hypothesis. However, it is important to 
keep in mind that the combined sets do not include L. alaskensis and S. nigra, the 
putative sister taxa to Spaniopsis. Until greater sampling is carried out in this area of 
the tree, our understanding of blood feeding in Rhagionidae will remain clouded. 
 
 
 286
The position of Alloleptis tersus remains unclear. It does not appear to be a member 
of either Rhagioninae or Spaniinae, nor does it bear any apparent synapomorphies 
with any other known groups. This may change as specimens of this species become 
available for examination. However, presently, this genus must be considered 
incertae sedis within Tabanomorpha. 
 
Consistent with other analyses that have examined the effect of taxon sampling, it 
appears that as more taxa are added to the matrix, phylogenetic accuracy is increased. 
Although the phylogeny of the major groups within Tabanomorpha is a difficult 
problem and remains unsettled to some degree, the results here indicate that efforts to 
amplify new sequences may be well rewarded. Arthroceras is of particular interest in 
this regard. More Ptiolina material will also be helpful to resolve the question of 
monophyly for the genus, as well as indicate with greater certainty its sister group. 
 
Similarly, the addition of characters will certainly help resolve lesser known groups 
such as Alloleptis, Litoleptis, and Sierramyia. A determined effort must be made to 
collect more of this material. The importance of larval characters especially must be 
emphasized. Although larvae are rarely found, even for common genera, a successful 
systematic program to collect larval Austroleptis, Arthroceras, Bolbomyia, and 
Atherimorpha, in particular, seems possible. Fairly reliable localities are known for 
each of these genera, yet a thorough, painstaking search for these larvae has not been 
carried out. These efforts may at least generate information regarding the first instar 
 
 287
larva, since egg-laying may be induced from freshly collected, live females (J. 
Cumming, pers. comm.). 
 
Based on the results of this phylogenetic analysis, I will propose a re-classification of 
the group. This reclassification is presented as part of Chapter 3.  
Chapter 2 Notes. 
Note 1. The first character [#1] is used to score the availability of male and female 
specimens in order to help avoid data-entry mistakes. In DELTA, certain character 
states may be programmed to control other characters. The character scored as having 
state 'male only', for example, renders all female characters inapplicable.  DELTA 
blocks data-entry for inapplicable characters, as programmed by the user so that, as in 
this example, female character states cannot be scored for species that are without 
females available for examination. It is a bookkeeping character only and is removed 
from all analyses.  
Note 2. The family of Oreoleptis, a new undescribed genus, is not clear to the authors 
who have collected larvae and reared adults (Szloty, Pritchard, and Sinclair, pers. 
comm.). I have not seen the adult, but I have read a brief description of its 
peculiarities. It has aedeagal tines as Athericidae + Tabanidae, but lacks a 
postspiracular scale and the female first cercus is two segmented. I have examined the 
larva and it is athericid in nearly all respects. All Athericidae that I have examined 
have elongate cylindrical palps that look like cigarettes, unlike those in any other 
larvae. They also uniquely have elongate prolegs armed with crochets apically, used 
 
 288
to adhere to its substrate. However, I have examined one Halucitherix that was 
already partially dissected so I cannot be sure that it does not have a telescoped first 
cephalic segment, as currently known Athericidae and Tabanidae. It appears to have a 
first segment similar to rhagionids. 
 
Note 3. I use the term ?mandibular brush? as Woodley (1989) and Sinclair (1992), 
however, other terms have been used for this structure (?cephalic brush? (Cameron, 
1934; Teskey, 1970; Webb; 1977), ?setaceous region? (Cook, 1949), ?spinose area? 
(Mackerras and Fuller, 1942), ?borstenfeld? (Schremmer, 1951), ?champ de??pines 
(Tsacas, 1962), ?bristle area? (Roberts, 1969)). In the tabanoids (Pelecorhynchidae, 
Athericidae, Tabanidae) the brush is associated with the mandibles by way of an 
articulated rod. As the mandibles are adducted downward, the articulated rod is lifted 
(Pechuman and Teskey (1981: 464). It appears to work similarly in the rhagionids, as 
illustrated in Chrysopilus auratus (Tsacas, 1962: 176). In the Rhagionidae, however, 
the brush is associated with cuticular tissue. Inherent in the term ?mandibular brush? 
may be an implicit assumption of functional association between the brush and 
mandibles. While the homology between the brush in tabanoids and the brush in 
rhagionids is not necessarily certain, the brush does seem functionally associated with 
the mandibles in all cases. Since this functional association is not immediately 
obvious, and the brush is a very conspicuous element of the larval head, the term 
?cephalic brush? may be an appealing term, particularly in reference to Rhagionidae. 
However, the brush may be completely concealed by the cephalic sclerite, when the 
mandibles are abducted (as in SEM of Pelecorhynchus, Fig. <will be in next draft>). 
 
 289
In this case, the scoring may be affected solely by the position of the mandibles 
(cephalic brush apparently absent). Therefore, in my opinion, the term ?mandibular 
brush? is most appropriate. 
 
 
 290
Chapter 3. Definition of Rhagionid Genera and Related Taxa, and their Family 
Groups.  
Introduction 
This chapter is dedicated to updating current knowledge of the Rhagionidae and 
related taxa. The genera of Austroleptidae, Bolbomyiidae, and Rhagionidae are 
defined and re-classified based on characters and relationships established in the 
previous chapter.  
Key to Adult Genera of Rhagionidae.   
The following key includes leads for Athericidae, Tabanidae, Vermileonidae, and 
Xylophagidae; members of which may be mistaken for Rhagionidae. The key also 
includes Pelecorhynchidae (Glutops Burgess, Pelecorhynchus Macquart, and 
Pseudoerrina Shiraki) and Austroleptidae (Austroleptis Hardy), and Bolbomyiidae 
(Bolbomyia Loew) which have been placed in Rhagionidae by recent authors 
(Nagatomi, 1982a; Woodley, 1989; Stuckenberg, 2001, etc.). Neorhagio Lindner is an 
unavailable name and is replaced by Sierramyia Kerr (new genus). Solomomyia 
Nagatomi is considered a junior synonym of Chrysopilus Macquart, Archicera 
Szil?dy a junior synonym of Spania Meigen, and Spatulina Szil?dy a junior synonym 
of Ptiolina Zetterstedt. Omphalophora Becker sensu Kerr is treated herein. 
1. Clypeus flat or recessed................................................................................... 23 
 Clypeus bulbous or at least slightly swollen ..................................................... 2 
2(1). Postspiracular sclerite with scale or slender ridge........................................... 19 
 Postspiracular sclerite smooth and flat .............................................................. 3 
 
 291
3(2). Hind tibial spur absent; laterotergite bare; wing vein M
3
 usually absent or 
reduced .......................................................................................................... 4 
 Hind tibial spur one; laterotergite setose or bare; wing vein M
3
 present, 
reaching wing margin .................................................................................... 6 
 Hind tibial spur two; laterotergite setose; wing vein M
3
 present, reaching wing 
margin ...........................................................................................................13  
4(3). Lacking mid tibial spurs; discal cell absent; scape clearly smaller than pedicel; 
eyes in male dichoptic; dorsal surface of wing vein R
1
 bare; Alaska, Chile, 
Japan .................................................................................  Litoleptis Chillcott 
 Mid leg with two tibial spurs; discal cell present; scape and pedicel 
approximately the same size; eyes in male holoptic; dorsal surface of wing 
vein R
1
 microsetose; Australian, Holarctic.................................................... 5 
5(4). Body greater than 4 mm in length; length of wing vein R
2+3
 shorter than length 
of R
5
; CuA
1
 arises from discal cell; Australian .................. Spaniopsis White 
 Body less than 4 mm in length; length of wing vein R
2+3
 about the same length 
as R
5
; CuA
1
 arises from basal medial cell or at juncture between basal 
medial and discal cells; Palaearctic ........................................ Spania Meigen 
6(3). Laterotergite setose............................................................................................ 9 
 Laterotergite bare................................................................................................7  
7(6). Eyes pilose; one mid tibial spur; body length less than 4 mm; Celebes (female 
unknown)                                      ..................Alloleptis Nagatomi & Saigusa  
 
 292
 Eyes without conspicuous microsetae; two mid tibial spurs; body length 
greater than 4 mm; Holarctic ..........................................................................7  
8(7). Anepisternum bare; wing vein R
5
 anterior to or ending at wing tip; wing vein 
R2+3 about as long or shorter  than R5; spermathecal duct no more than 
three times the length of sternite 9; female tergite 9 short, length 
approximately one half width or less; spermathecal duct accessory glands 
arise from the base of the spermathecae; Holarctic ..........................................  
.......................................................................................... Ptiolina Zetterstedt  
 Anepisternum setose; wing veins R
4
 and R
5
 encompass wing tip; wing vein 
R2+3 clearly longer than R5; spermathecal duct more than three times but 
less than five times the length of sternite 9; female tergite 9 longer, length 
greater than one half width; spermathecal duct accessory glands arise at 
approximately the distal third of the spermathecal ducts; Holarctic                               
.................................................................................... Omphalophora Becker 
9(6). Macrochaetae of hind tibia absent; palp two-segmented; scale-like thoracic 
setae with structural color never present; ventral parameral sheath of male 
genitalia with paired lobes ..........................................................................  12 
 Macrochaetae of hind tibia present; palp one-segmented; scale-like thoracic 
setae with structural color often present; ventral parameral sheath of male 
genitalia without paired lobes.......................................................................10 
10(9). First antennal flagellomere elongate, stylate; anepisternum bare, hind coxal 
tubercle present; Philippines ...............................................  Stylospania Frey  
 
 293
 First antennal flagellomere ovoid, bearing long arista-like extension apically 
(in female; in male, various forms, but not stylate); anepisternum setose; 
hind coxal tubercle absent; Cosmopolitan....................................................11  
11(10) First antennal flagellomere of male highly modified, bifurcate; male dichoptic; 
Philippines ............................................................................. Schizella Bezzi 
 First antennal flagellomere of male as in female; males almost always 
holoptic; Cosmopolitan ............................................... Chrysopilus Macquart 
12(9). First antennal flagellomere kidney-shaped, with dorso-apical arista; mandibles 
present; proscutellum usually present; hypandrium fused entirely to 
gonocoxites; Holarctic .........................................  Symphoromyia Frauenfeld 
 Antenna with many similarly sized antennal flagellomeres, tapering distally; 
mandibles absent; proscutellum absent; hypandrium separated partially from 
the gonocoxites by an incomplete suture; Nearctic ..... Arthroceras Williston 
13(3). Palp two-segmented; first antennal flagellomere enlarged bearing segmented 
stylus; setae of laterotergite arranged in row(s)........................................... 18 
 Palp one-segmented; first antennal flagellomere enlarged basally, fused with 
distinct arista-like apical extension; setae of laterotergite in arranged in tuft14  
14(13). Fore tibial spur one; dorsal surface of R
1
 without microsetae; spermathecal 
duct no more than three times the length of sternite 9; female tergite 7 much 
wider than long ............................................................  VERMILEONIDAE  
 
 294
 Fore tibial spur absent; dorsal surface of R
1
 microsetose; spermathecal duct 
longer than five times the length of sternite 9, but not so long as to be 
difficult to measure; female tergite 7 much longer than wide ......................15  
15(2). Alula completely reduced, without any curvature; wing vein R2+3 setulose; 
proepimeron bare; Mexico   ................................................. Sierramyia Kerr 
 Alula present, with narrow or broad curvature; wing vein R2+3 bare; 
proepimeron setose; Holarctic, including Mesoamerica  ............................ 16 
16(3). First hind metatarsus of male swollen; scape clearly larger than pedicel; India 
and China .......................................................................  Desmomyia Brunetti 
 First hind metatarsus of male not swollen; scape about the same size as pedicel                              
............................................  Rhagio Fabricius (including Rhagina Malloch)  
17(13). Proboscis short, fleshy; alula with narrow curvature; female tergite 7 about 
as long as wide; cardo not swollen; South Africa, Australia, and South 
America.........................................................................  Atherimorpha White 
 Proboscis elongate, sclerotized, specialized for nectarivory; alula with broad 
curvature; female tergite 7 much longer than wide; cardo swollen (as 
protuberance arising from base of palp); South Africa 
 ........................................................................................... Arthroteles Bezzi 
18(2). Tuft of setae posterior to postspiracular sclerite present; female cercus two-
segmented .....................................................  PELECORHYNCHIDAE 20 
 Tuft of setae posterior to postspiracular sclerite absent; female cercus usually 
one-segmented ..............................................................................................19 
 
 295
19(18). Wing veins R
1
 and R
2+3
 meet together or close together at wing margin; 
lower calypter undeveloped; first abdominal sclerite without median notch 
or suture; Cosmopolitan...................................................... ATHERICIDAE 
 Wing veins R
1
 and R
2+3
 close together at wing margin; lower calypter 
developed; first abdominal sclerite with median notch or suture; 
Cosmopolitan........................................................................... TABANIDAE 
20(18). Laterotergite setose; eyes conspicuously microsetose; short fore tibial spur 
present; Holarctic.......................................................... Pseudoerrina Shiraki  
 Laterotergite bare; eyes inconspicuously microsetose; fore tibial spur lacking; 
Cosmopolitan................................................................................................21  
21(20). Hind tibial spur one; wing veins R
2+3
 and R
4
 nearly straight; subscutellum 
enlarged, noticeably bulbous; Holarctic, including Mesoamerica 
.............................................................................................. Glutops Burgess 
 Hind tibial spur two; wing veins R
2+3
 and R
4
 sinuous, arcing toward anterior 
wing margin; subscutellum not enlarged; Australia and South America 
 ..............................................................................Pelecorhynchus Macquart 
22(1). Postspiracular scale present; lower calypter developed; first abdominal sclerite 
with median notch or suture; mandibles present; occiput flattened, concave 
posteriorly .................................................................................TABANIDAE  
 Postspiracular scale absent; lower calypter undeveloped; first abdominal 
sclerite without median notch or suture; mandibles absent; occiput rounded 
to various degrees .........................................................................................23  
 
 296
23(22). Postspiracular sclerite setose .........................................  XYLOPHAGIDAE  
 Postspiracular sclerite without setae ................................................................24  
24(23). Laterotergite setose; wing vein M
3
 present; anal lobe reduced; eyes in male 
evenly distributed, of equal size; Holarctic 
.......................................................Vermileo Linnaeus VERMILEONIDAE  
 Laterotergite bare; wing vein M
3
 absent; anal lobe well developed; eyes in 
male split into upper and lower areas and smaller in lower area..................25 
25(24). Lacking mid and hind tibial spurs; palp one-segmented; first antennal 
flagellomere elongate, stylate; wing veins R
1
 and R
2+3
 widely separated at 
wing margin; Alaska, Chile, Japan ...................................  Litoleptis Chillcott 
 Mid and hind leg with two tibial spurs; palp two-segmented; first antennal 
flagellomere bearing segmented stylus; wing veins R
1
 and R
2+3
 close 
together at wing margin; Australian, Neotropical, Holarctic....................... 26 
26(25). Fore tibial spur absent; female cercus one-segmented; female tergite 7 much 
wider than long; three spermathecae; subscutellum bulbous; theca elongate; 
pseudotrachae present; eyes in male flattened dorsally; male hypandrium 
fused entirely to gonocoxites; aedeagal tines absent; Australia and South 
America....................................... Austroleptis Hardy AUSTROLEPTIDAE 
 Fore tibial spur present; female cercus two-segmented; female tergite 7 much 
longer than wide; two spermathecae; subscutellum inconspicuous; theca 
compact; pseudotrachae absent; eyes in male rounded dorsally; hypandrium 
 
 297
separated from the gonocoxites by a complete suture; aedeagal tines present; 
North America and Kamchatka .........  Bolbomyia Loew BOLBOMYIIDAE 
 
 
 298
Austroleptidae 
Genus AUSTROLEPTIS Hardy 
Figures 17-24, 178. 
 
AUSTROLEPTIS Hardy, 1920: 126.  Type species Austroleptis rhyphoides Hardy, 
1920, by original designation. 
 
DIAGNOSIS.  
Austroleptis is unique among lower brachyceran taxa in having the cornu apically 
setulose and by having sternite 8 of the female terminalia laterally divided into two 
segments. In the male genitalia, the presence of paired sclerotized lobes arising 
ventrally, near the center of the gonocoxites, is another autapomorphic development. 
Austroleptis is also characterized by the combination of having a recessed clypeus, 
female cercus one-segmented, wing vein M
3
 missing, and male genitalia without 
gonocoxal apodemes. Although Austroleptis is not unique among lower brachycerans 
in having each these four features, the phylogenetic placement of the genus suggests 
that most, if not all, of these character states are each independently derived and 
represent additional autapomorphies for the genus. Within Tabanomorpha, 
Austroleptis is unique in having a recessed clypeus and is the only taxon outside of 
Athericidae and Tabanidae that has one-segmented female cercus.  Nagatomi & Iwata 
(1976: 43) and Nagatomi & Nagatomi (1987:140) state that Austroleptis is peculiar 
among the Tabanomorpha in having sternite 9; they were mistakenly referring to the 
posterior sclerite of sternite 8. All lower brachyceran flies retain sternite 9 (also 
 
 299
known as the genital fork, vaginal apodeme, or furca), in some form. Austroleptis is 
restricted to the southern hemisphere, in South America and Australia. 
 
Austroleptis is a small to moderately sized fly (3.1 to 7.7 mm) of black, brown, brown 
and black, orangish or yellowish coloration. There is sexual dimorphism in the 
coloration; males are usually black or darker, whereas females often have at least 
some light brown or orange, if not entirely yellowish. All Australian Austroleptis 
have spotted wings, whereas South American Austroleptis wings are hyaline 
(although I have seen one undescribed Austroleptis species from Chile, Malleco 
Province, with infuscate wing veins). Antenna with basal flagellomere enlarged, oval, 
laterally compressed, bearing 2 to 4 distal flagellomeres; eyes in male holoptic; 
laterotergite bare; tibial spur formula 0:2:2 (spurs very short); and tibia without 
macrochaetae. Due to its unusual combination of character states (listed above) and 
its restricted distribution, Austroleptis is unlikely to be confused with related Diptera. 
In South America, it is most readily distinguished from Atherimorpha by its recessed 
clypeus, bare laterotergite and the absence of wing vein M
3
 and from Litoleptis by the 
multisegmented flagellomere, spurs on mid and hind tibia, and the presence of the 
discal cell. In Australia, it is most readily distinguished from Spaniopsis by its 
multisegmented flagellomere and the presence of hind tibial spurs.  
 
DESCRIPTION.  
Head. Clypeus not bulbous. Scape approximately the same size as pedicel. 
Flagellomeres 3-5; first flagellomere enlarged, oval, laterally compressed, bearing 
setae; distal flagellomeres robust, cylindrical, short (except terminal flagellomere 
 
 300
which is more elongated). Eyes inconspicuously setulose; in female, dichoptic; in 
male, holoptic, flattened dorsally, ommatidia split into dorsal and ventral areas and 
smaller ventrally. Labella with pseudotrachae, longer or shorter than palps. Theca 
elongate, lateral sclerites tightly adjacent, apparently fused with suture. Palps two-
segmented; proximal and distal segments about the same length. Stipes surrounded by 
membrane above theca, directed posteriorly (very reduced). Cardo not swollen. 
Lacinia longer than palps, lacinia apex not serrated. Mandibles absent. Cibarial pump 
short, as wide as long or wider. Cornu shorter than cibarial pump. Pharyngeal pump 
narrow along most of length, mostly flat; longer than length of cibarial pump. Thorax. 
Mesonotum with vittae. Dorsocentral bristles absent, all dorsal setae of equal length 
or present (as in A. collessi). Anepisternum bare. Laterotergite bare. Postspiracular 
scale absent. Proscutellum present. Subscutellum enlarged, noticeably bulbous. Wing 
hyaline or lightly infuscate, membrane with or without markings, with or without 
pseudostigma. Lower calypter reduced. Upper calypter well developed, with broad 
curvature, lobe-like, width twice length or less. Costa extends to wingtip or past 
wingtip (to at least R
5
). Humeral crossvein (h) well developed. Sc-r crossvein present, 
weakly developed, positioned distal to the humeral crossvein (h), by less than the 
length of h. Dorsal side of R
1
 setulose, ventral side with or without setulae; wing 
veins R
4
 and R
5
 with or without setulae; all cells and other wing veins bare. R
2+3
 
sinuous, apical third ultimately bends toward leading edge of wing margin (creating 
concave flex anteriorly); shorter than R
5
. Base of R
4
-R
5
 fork proximal or directly 
above distal end of cell dm. R
4
 at base relaxed, not strongly curved, nearly straight 
apically. R
5
 anterior to or ending at wing tip, or R
4
 and R
5
 encompass wing tip. R
5
 
 
 301
anterior to, posterior to, or ending at wing tip; clearly longer than R
4+5
 (r-m to R
4
 
origin). Origin of CuA
1
 at discal cell. M
3
 wingvein absent. CuA
2
 reaches wing 
margin, about 2/3 length of posterior vein of bm cell. M
3
 wing vein absent. Alula with 
broad curvature, rounded evenly. Anal lobe well developed. Anal cell closed. Halter 
knob between 1/3?1/2 length of stem. Tibial spur formula 0:2:2. Mid and hind tibial 
spurs short. Hind coxal tubercle absent. Hind tibial macrochaetae absent. 
Postmetacoxal bridge present, as an incomplete, thin extension. Abdomen. Terminal 
abdominal segments 5?10 evenly tapered from segments 1?4. In female, tergite 7 
much wider than long. Intersegmental membrane between segments 7 and 8 short, as 
throughout abdomen. Sternite 8 sclerite elongated; more than twice as long as wide; 
divided into two segments, anterior segment long and wide, posterior segment 
rounded, cupped. Male terminalia with epandrium simple, not containing hypandrium 
ventrally. Epandrial sclerite wider than long, modestly curved anteriorly. 
Subepandrial sclerite absent. Hypoproct triangular (rounded posteriorly), flattened; 
anterior margin entire, with an even amount of sclerotization; appearing posteriorly 
lobed, with paired region of increased sclerotization; setose. Cercus directly adjacent 
to epandrial sclerite; directly adjacent to one another, separation distance one quarter 
width of cercus or less; held horizontal in relation to rest of abdomen; in posterior 
view flat. Hypandrium fused entirely to gonocoxites. Gonocoxite smooth dorsally, 
without sinuous ridge leading to gonocoxal apodeme; ventrally, with paired, 
sclerotized, lobe-like processes. Gonocoxal apodemes absent. Lateral ejaculatory 
processes absent. Ejaculatory apodeme moderately long, reaching anterior margin of 
hypandrium. Ejaculatory apodeme laterally compressed, umbraculate (umbrella-
 
 302
shaped) anteriorly. Aedeagal tines absent. Endoaedeagal process absent. Female 
terminalia with three spermathecae, clubbed, lightly to well sclerotized. Spermathecal 
ducts more than three times but less than five times the length of sternite 9,  inflated 
at base of spermathecae. Spermathecal duct accessory glands absent. Ejection 
apparatus of spermathecal ducts sclerotized, with surface ringed furrows. Common 
spermathecal duct thickened, about the same length as the longest diameter of genital 
chamber. Genital chamber circular, small, occupying fraction of sternite 9 area. 
Accessory gland posterior to genital chamber inconspicuous, easily overlooked even 
after staining. Sternite 9 anterior end tapered to a point; posterior end with broad 
lateral extensions, free, held in vertical plane. Tergite 10 present. Tergite 10 narrow, 
split into two separate lateral sclerites; short (length less than half width). Sternite 10 
roughly pentagonal, pointed posteriorly; almost entirely underneath cercus segments. 
Cercus one-segmented; separated from one another dorsally by approximately the 
width of the second cercal segment; without apical sensory pits. Larva. The larvae are 
unknown. 
LITERATURE. Paramonov (1962) provides a key to species for the Australian 
fauna. Nagatomi and Nagatomi (1987) provide a key to species for the Neotropical 
fauna. 
List of included species 
Species Author, reference Type country 
Austroleptis atrata Nagatomi & Nagatomi, 1987: 141 Chile 
Austroleptis atriceps Malloch, 1932: 203 Chile 
Austroleptis breviflagella Nagatomi & Nagatomi, 1987: 148 Chile 
Austroleptis collessi Paramonov, 1962: 138 Australia 
Austroleptis fulviceps Malloch, 1932: 202 Chile 
Austroleptis multimaculata Hardy, 1920: 128 Australia 
 
 303
Species Author, reference Type country 
Austroleptis penai Nagatomi & Nagatomi, 1987: 153 Chile 
Austroleptis rhyphoides Hardy, 1920: 127 Australia 
 
 
 
 304
Bolbomyiidae status revised 
Genus BOLBOMYIA Loew 
Figures 16c, 25-31, 179. 
 
BOLBOMYIA Loew, 1850: 304.  Type species Bolbomyia nana Loew, by 
subsequent monotypy (Loew, 1862: 188). 
 MISGOMYIA Coquillett, 1908: 145.  Type-species obscura Coquillett, by 
original designation. 
 CEKENDIA Szil?dy, 1934: 264 (as subgenus of Ptiolina Zetterstedt, 1842).  
Type species Ptiolina (Cekendia) wuorentausi Szil?dy, by 
monotypy. 
 CECHENIA Frey, 1954: 9.  Unjustified emendation. 
 
DIAGNOSIS.  
The best autapomorphy defining Bolbomyia species is the female terminalia, which 
have only two spermathecae and whose ducts lead directly to the genital chamber and 
attach to this structure independently, without joining into a common duct. Aedeagal 
tines are present in the male genitalia of Bolbomyia. Although aedeagal tines are also 
found in Arthroceras and in members of Athericidae and Tabanidae, the tines in 
Bolbomyia are likely independently derived and may represent another 
autapomorphic character state.  
 
 
 305
Species of Bolbomyia are small (2.3 to 3.5 mm), brown or black in color, with lightly 
infuscate wings, restricted to the north temperate region of North America and eastern 
Asia (Kamchatka). Males holoptic; antenna with basal flagellomere enlarged, 
elongate oval or subconical, laterally compressed, bearing 2 to 3 distal flagellomeres; 
laterotergite bare; wing vein M
3
 absent; tibial spur formula 1:2:2; and tibia without 
macrochaetae. The fore tibial spur will separate Bolbomyia from nearly all other 
small brachycerans. Litoleptis is similar in size and appearance, but lacks tibial spurs 
on all tibia, bears a single elongate antennal flagellomere, and lacks the discal medial 
cell of the wing. Ptiolina is larger and more robust, and may be distinguished from 
Bolbomyia by the presence of wing vein M
3
, by having only one hind tibial spur, and 
antenna with enlarged first flagellomere with a single-segmented style.  
 
DESCRIPTION.  
Clypeus not bulbous. Scape approximately the same size as pedicel. Flagellomeres 3 
to 4; first flagellomere enlarged, elongate oval or subconical, laterally compressed, 
bearing weak setae; distal flagellomeres cylindrical, short (except terminal 
flagellomere which is more elongated). Eyes inconspicuously setulose; in female, 
dichoptic; in male, holoptic, not flattened dorsally, ommatidia split into dorsal and 
ventral areas and smaller ventrally. Labella lacking pseudotrachae, shorter than palps. 
Theca short and stout; lateral sclerites adjacent and touching, but mostly separated. 
Palps two-segmented; distal segment longer than or about the same length as 
proximal segment. Stipes surrounded by membrane above theca, directed posteriorly. 
Cardo not swollen. Lacinia shorter than palps, tip not serrated. Mandibles absent. 
Cibarial pump short, as wide as long or wider. Cornu nearly as long as or longer than 
 
 306
cibarial pump. Pharyngeal pump narrow along most of length, mostly flat, longer than 
length of cibarial pump (cibarial pump very short). Thorax. Mesonotum without 
vittae. Dorsocentral setae not longer than other mesonotal setae. Anepisternum bare 
or bearing 1-2 setulae. Laterotergite bare. Postspiracular scale absent. Proscutellum 
present or absent. Subscutellum not enlarged nor lengthened; inconspicuous. Wing 
membrane lightly infuscate, without markings, pseudostigma absent or lightly 
present. Lower calypter reduced. Upper calypter well developed. Upper calypter with 
broad curvature, lobe-like, width twice length or less. Costa extends past wingtip, to 
approximately R
5
. Humeral crossvein (h) well developed. Sc-r crossvein present, 
weakly developed or absent; positioned distal to the humeral crossvein (h), by less 
than the length of h. Dorsal side of R
1
 setulose, ventral side bare. Dorsal side of R
2+3
 
with or without setulae, as ventral side. All other wing cells and veins bare. R
2+3
 
sinuous, apical third of R
2+3
 ultimately bends toward wingtip (creating convex flex 
anteriorly); longer than R
5
, but less than twice as long. Base of R
4
-R
5
 fork distal of 
distal end of cell dm. R
4
 at base relaxed, not strongly curved; nearly straight apically. 
R
5
 anterior to, posterior to, or ending at wing tip; clearly longer than R
4+5
 (r-m to R
4
 
origin). M
3
 wing vein absent. M-cu crossvein absent. Origin of CuA
1
 at discal cell. 
CuA
2
 greater than 2/3 length of posterior vein of bm cell.  Alula with narrow or broad 
curvature, rounded evenly. Anal lobe well developed. Anal cell open.  Halter knob 
2/3 or longer than length of stem. Tibial spur formula 1:2:2. Hind coxal tubercle 
present. Hind tibial macrochaetae absent. Postmetacoxal bridge present as incomplete 
thin extension. Abdomen. Terminal abdominal segments 5?10 evenly tapered from 
segments 1?4. In female, tergite 7 much longer than wide. Intersegmental membrane 
 
 307
between segments 7 and 8 short, as throughout abdomen. Sternite 8 cleavage 
superficial, open broadly; longer than wide or as long as wide. Male terminalia with 
epandrium simple, not containing hypandrium ventrally. Epandrial sclerite wider than 
long, modestly curved anteriorly. Subepandrial sclerite absent. Hypoproct triangular 
(rounded posteriorly); rounded, virtually encircling cerci; not fused with cercus; 
anterior margin entire, with an even amount of sclerotization; posterior margin entire, 
with an even amount of sclerotization; tomentose, without setae. Cercus displaced 
away from epandrial sclerite; partially displaced from one another, separation 
distance approximately half the width of single cercus; held horizontal in relation to 
rest of abdomen;  in posterior view cupped, forming circular outline medially. 
Hypandrium separated from the gonocoxites by a complete suture. Gonocoxite with 
sinuous dorsal ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short or 
long enough to reach anterior margin of hypandrium. Parameral sheath not developed 
into bulbous sac or separate lobes. Lateral ejaculatory processes absent. Ejaculatory 
apodeme moderately long to long, reaching anterior margin of hypandrium, or 
somewhat beyond this. Ejaculatory apodeme tripartite; dorsally compressed laterally 
and ventrally, compressed dorso-ventrally. Aedeagal tines present. Endoaedeagal 
process present. Female terminalia with two spermathecae; spherical or elongate oval; 
sclerotization light to none. Spermathecal ducts less than three times the length of 
sternite 9. Spermathecal duct accessory glands absent. Circular ridge present at distal 
end of reinforced base of spermathecal ducts. Ejection apparatus of spermathecal 
ducts thickened, sclerotized, with ringed surface furrows; swelling present halfway 
between genital chamber and spermathecae. Common spermathecal duct absent. 
 
 308
Genital chamber oval, moderately sized. Accessory gland posterior to genital 
chamber inconspicuous, easily overlooked even after staining. Sternite 9 anterior end 
narrow with rounded tip; posterior end with narrow lateral extensions, free, held in 
vertical plane. Tergite 10 short (length less than half width). Sternite 10 roughly 
pentagonal, pointed posteriorly; posterior half below first cercus segment. Cercus 
two-segmented. First segment of cercus not elongate, with ventral process. Ventral 
lobes of first segment of cercus not curving ventrally towards one another to form a 
ring (sometimes slightly arched medially, but forming narrow elliptical opening 
only). Basal cerci adjacent dorsally. Second cercus segment not elongated, with apical 
sensory pits. Larva. The larvae are unknown. 
LITERATURE. Webb (1987a) provides a key for the extant species of the world. 
List of included species 
Species (synonyms indented) Author, reference Type country 
?Bolbomyia loewi Meunier, 1902: 96 Baltic Region 
(Eocene/ Oligocene) 
Bolbomyia melanderi Chillcott, 1963: 1189 USA 
Bolbomyia nana Loew, 1862: 188 USA 
Misgomyia obscura Coquillett, 1908: 146 USA 
Ptiolina mitis Curran, 1931: 249 USA 
Bolbomyia andiscalcella Webb, 1969: 286 USA 
Bolbomyia  wuorentausi (as Cekendia; Szil?dy, 1934: 
264) 
Russia 
Bolbomyia macgillisi Chillcott 1961: 634 USA 
 
 309
Rhagionidae 
Genus ALLOLEPTIS Nagatomi and Saigusa 
ALLOLEPTIS Nagatomi and Saigusa, 1982: 40.  Type species Alloleptis tersus 
Nagatomi and Saigusa, by monotypy. 
 
DIAGNOSIS.  
Alloleptis is the only rhagionid genus with a tibial spur formula of 0:1:1.  It is a small 
fly (body and wing length 3.8 mm), with antenna similar to those found in Bolbomyia 
(first flagellomere enlarged, bearing a short, two-segmented stylus). The antennal 
flagellomeres were lost prior to measurement and no illustration exists, however.  
Alloleptis may be distinguished immediately from Bolbomyia by having 
conspicuously setulose eyes and wing vein M
3
 present.  Other distinctive features 
include bare laterotergite, wing veins CuA
2
 and A
1
 join well before the wing margin 
(CuA
2
+A
1
 long), and the wings are darkly infuscate. 
 
DESCRIPTION.  
No specimens available for examination; character state scoring based on Nagatomi 
(1982, 1984). Head. Clypeus slightly bulbous. Scape slightly smaller than pedicel. 
First flagellomere enlarged bearing two-segmented stylus. Eyes conspicuously 
setulose; in male, holoptic. Palps one-segmented or two-segmented (the original 
description reads "probably two-segmented (if so, basal segment short)." The 
illustrations of the head (Nagatomi, 1982a: 56), however, show palp one-segmented). 
Mandibles absent. Thorax. Mesonotum and scutellum dark, with long, erect setae. 
 
 310
Wing membrane infuscate, without markings, without pseudostigma. Lower calypter 
reduced. Upper calypter well developed, with broad curvature, lobe-like, width twice 
length or less. Costa apparently extends past wingtip or at least to R
5
. R
2+3
 nearly 
straight; longer than R
5
, but less than twice as long. Base of R
4
-R
5
 fork proximal or 
directly above distal end of cell dm. R
4
 at base relaxed, not strongly curved, nearly 
straight apically. R
5
 clearly longer than R
4+5
 (r-m to R
4
 origin); ending at wingtip. M
3
 
wing vein present. Wing cell m3 parallel-sided at margin. Origin of CuA
1
 at crossvein 
separating discal and basal medial cells. Length of CuA
2
 v. posterior vein of bm cell 
less than 1/2 length of posterior vein of bm cell. Alula with broad curvature, rounded 
evenly. Anal lobe well developed. Anal cell closed (CuA
2
 + A
1
 vein as long or longer 
than  CuA
2
). Laterotergite bare. Subscutellum absent. Tibial spur formula 0:1:1. 
Abdomen is ?comparatively long and narrow? (Nagatomi, 1982a: 41). Male genitalia 
with epandrium simple, not containing hypandrium ventrally. Epandrial sclerite wider 
than long, modestly curved anteriorly. Subepandrial sclerite absent. Hypoproct 
tomentose, without setae. Cercus base held underneath epandrial sclerite. Cerci 
directly adjacent to one another, separation distance one quarter width of cercus or 
less. Hypandrium fused entirely to gonocoxites. Gonocoxite smooth dorsally, without 
sinuous ridge leading to gonocoxal apodeme. Gonocoxal apodemes very short. 
Parameral sheath narrow. Lateral ejaculatory processes present, integrated into sperm 
sac membrane. Ejaculatory apodeme moderately long, reaching slightly beyond 
margin, anteriorly; rod-shaped. Aedeagal tines absent. Endoaedeagal process present. 
Female are not known. Larva. The larvae are unknown.  
 
 311
LITERATURE. Illustrations of wing and head in Nagatomi (1982); male genitalia 
illustrated in Nagatomi (1984). 
 
List of included species 
Species (synonyms indented) Author, reference Type country 
Alloleptis tersus Nagatomi & Saigusa, 1982: 41 Celebes 
 
Genus ARTHROCERAS Williston 
Figures 16D, 32-41, 180. 
 
ARTHROCERAS Williston, 1886: 107.  Type species Arthroceras pollinosum 
Williston, by subsequent designation (Coquillett, 1910: 510). 
NOTE 1. 
 USSURIELLA Paramonov, 1929: 181.  Type species Ussuriella gadi 
Paramonov, by monotypy. 
 PSEUDOCOENOMYIA ?uchi, 1943: 493.  Type species Pseudocoenomyia 
sinensis ?uchi, by original designation. 
 
DIAGNOSIS.  
No obvious autapomorphic feature is known that defines Arthroceras conclusively. 
However, it retains the following unique combination of primitive and derived states: 
males are unique among orthorrhaphous flies in that their genitalia have both well-
developed lateral ejaculatory processes and aedeagal tines. The spermathecal ducts of 
females are unique in that they are visibly inserted within the tubing that forms their 
 
 312
common junction, near the genital chamber. The female terminalia also have 
spermathecal duct accessory glands.  
 
Arthroceras species are mid-sized to large (4.5 to 13 mm), black, gray, or often 
yellowish-colored flies that have a fairly long, tapering antenna consisting of 5-8 
flagellomeres, mandibles absent, laterotergite setose, tibial spur formula 0:2:1, wing 
vein M
3
 present; hind tibial macrochaetae absent; female tergite 9 without 
ventrolateral arms; female spermathecal ducts with accessory glands. Arthroceras are 
restricted to the Holarctic Region. They are distinguished from all other Rhagionidae 
by the form of their antenna, which is composed of 5-8 similarly sized flagellomeres, 
tapering distally (Figs. 12- 13). They may be distinguished from Glutops Burgess 
(Pelecorhynchidae) by their setose laterotergite and parafacials not swollen and from 
Pseudoerrina Shiraki (Pelecorhynchidae) by lacking conspicuously setulose eyes and 
fore tibial spurs.  
DESCRIPTION.  
Head. Clypeus bulbous. Scape smaller than or subequal to pedicel. First flagellomere 
slightly enlarged, round in cross section. Antenna with 5-8 flagellomeres of similar 
shape, tapering distally; terminal flagellomere usually more elongate. Eyes 
inconspicuously setulose; in female, dichoptic; in male, holoptic, ommatidia split into 
dorsal and ventral areas and smaller ventrally, not strongly flattened dorsally. Labella 
with pseudotrachae, longer or shorter than palps. Theca short and stout, divided into 
two separate, lateral sclerites. Palps two-segmented; distal segment longer than 
proximal segment. Stipes convergent toward one another medially. Lacinia present, 
 
 313
shorter than palps, tip not serrated. Mandibles absent. Cibarial pump long, clearly not 
as wide as long. Cornu nearly as long as or longer than cibarial pump. Pharyngeal 
pump moderately broad anteriorly, mostly flat along its length, approximately same 
length as cibarial pump (excluding cornu). Thorax. Mesonotum with or without 
vittae. Dorsocentral bristles absent; all dorsal setae of equal length. Anepisternum 
setulose on dorsal and posterior margins. Laterotergite setose. Proscutellum absent. 
Subscutellum mostly flat or slightly bulbous. Wing hyaline, without markings, or 
membrane lightly infuscate. Lower calypter reduced. Upper calypter well developed, 
with broad curvature, lobe-like, width twice length or less. Costa extends to wingtip 
or just past wingtip, to R
5
. Humeral crossvein (h) well developed. Sc-r crossvein 
present, very weakly developed, positioned distal to the humeral crossvein (h), by the 
approximate length of h. Dorsal side of R
1 
setulose, ventral side bare. All other wing 
veins without setulae. Wing veins R
1
 and R
2+3
 separated at wing margin. R
2+3
 
sinuous, apical third of R
2+3
 ultimately bends toward wingtip (creating convex flex 
anteriorly). Length of R
2+3
 longer than R
5
, but less than twice as long. R
4+5
 aligned 
with R
5
. R
4
 and R
5
 contain wing tip. Base of R
4
-R
5
 fork proximal or directly above 
distal end of cell dm. R
4
 nearly straight apically. R
4
 and R
5
 encompass wing tip. R
5
 
aligned with R
4+5
. R
5
 clearly longer than R
4+5
 (r-m to R
4
 origin). R-m crossvein at 
proximal side of central one-third of discal cell (or more centrally). CuA
1
 origin at 
bm. CuA
2
 greater than 1/2 length of posterior vein of bm cell, less than 2/3 length of 
posterior vein of bm cell. M
3
 wing vein present. Alula full, rounded, with broad 
curvature. Anal lobe well developed. Anal cell open. Halter knob between 1/2?2/3 
length of stem. Tibial spur formula 0:2:1. Hind coxal tubercle absent. Hind tibial 
 
 314
macrochaetae absent. Postmetacoxal bridge absent. Epandrial sclerite wider than 
long. Abdomen. Abdominal segments evenly tapered. In female, last 3 abdominal 
segments telescoping; tergite 7 much wider than long; intersegmental membrane 
between segments 7 and 8 especially long; sternite 8 as wide as long or wider than 
long. Male terminalia with epandrium simple, not containing hypandrium ventrally. 
Epandrial sclerite wider than long, strongly notched anteriorly. Subepandrial sclerite 
absent. Hypoproct triangular (rounded posteriorly), setose. Cercus base held 
underneath epandrial sclerite. Cerci directly adjacent, separation distance one quarter 
width of cercus or less. Cerci, in posterior view cupped, forming circular outline 
medially. Hypandrium separated partially from the gonocoxites by an incomplete 
suture. Gonocoxite with sinuous dorsal ridge, leading to gonocoxal apodeme. 
Gonocoxal apodemes short or long enough to reach anterior margin of hypandrium. 
Parameral sheath forming separate, distinct lobes ventrally. Lateral ejaculatory 
processes present, not part of sperm sac posteriorly. Ejaculatory apodeme moderately 
long, reaching anterior margin of hypandrium, rod-shaped or laterally compressed. 
Aedeagal tines present. Endoaedeagal process present. Female terminalia with three 
spermathecae, spherical, moderately to well sclerotized. Spermathecal ducts longer 
than five times the length of sternite 9, but not so long as to be difficult to measure; 
not inflated at base of spermathecae. Spermathecal duct accessory glands present, 
arise at approximately halfway along the length of the spermathecal ducts. Ejection 
apparatus of spermathecal ducts thickened, sclerotized, with furrows. Common 
spermathecal duct thickened, about the same length as the longest diameter of genital 
chamber. Genital chamber oval, moderately sized. Accessory gland posterior to 
 
 315
genital chamber prominent, retains dye easily; with paired extensions posteriorly. 
Sternite 9 anterior end pointed, posterior end with broad extensions posteriorly that 
are held in horizontal plane. Tergite 10 partially split, short (length less than half 
width). Sternite 10 roughly pentagonal, pointed posteriorly, posterior half below first 
cercus segment. Cercus two-segmented. First segment of cercus not elongate, with or 
without ventral process. Ventral lobes of first segment of cercus not curving ventrally 
towards one another to form a ring. Basal cerci adjacent dorsally. Second cercus 
segment not elongated with or without apical sensory pits. Larva. The larvae are 
unknown. 
LITERATURE. Key to Arthroceras of the world in Nagatomi (1966). Key to 
Nearctic species in Webb (1987). 
List of included species 
Species (synonyms indented) Author, reference Type country 
Arthroceras fulvicorne Nagatomi, 1966: 46 Canada 
Arthroceras fulvicorne 
nigricapite 
Nagatomi, 1966: 49 USA 
Arthroceras fulvicorne 
subsolanum 
Nagatomi, 1966: 49 USA 
Arthroceras subaquilum Nagatomi, 1966: 59 USA 
Arthroceras gadi (as Ussuriella; Paramonov, 
1929: 181) 
Russia 
Arthroceras japonicum Nagatomi, 1954: 13 Japan 
Arthroceras leptis (as Arthropeas; Osten-Sacken, 
1878: 223) 
USA 
Arthroceras pollinosum Williston, 1886: 108 USA 
Leptis pruinosus Bigot, 1887: 115 USA 
Arthroceras rubrifrons Nagatomi, 1966: 56 Japan 
Arthroceras sinense (as Pseudocoenomyia; Ouchi, 
1943: 493) 
China 
 
 
 316
Genus ARTHROTELES Bezzi 
Figures 42-48, 181. 
 
ARTHROTELES Bezzi, 1926: 321.  Type-species Arthroteles bombyliiformis 
Bezzi, 1926, by original designation. 
DIAGNOSIS.  
The most striking autapomorphy for this genus is the elongate, sclerotized proboscis, 
which is adapted for nectar feeding. At the base of the mouthparts, the cardo is 
swollen distinctively.   
 
Species of Arthroteles are moderately sized (5 to 7.5 mm) flies of gray to dark gray 
coloration, having an antenna that bears seven to eight tapering flagellomeres (first 
flagellomere much larger than all others); eyes in male holoptic (with the exception of 
A. longipalpus Nagatomi & Nagatomi); laterotergite setose; tibial spur formula 0:2:2; 
wing vein M
3
 present; short macrochaetae on all tibiae; female tergite 9 without 
ventrolateral arms; and female spermathecal ducts without accessory glands. 
Arthroteles is most similar to Atherimorpha in general form, but may be distinguished 
from this and all other related flies by the form of its mouthparts. It also differs from 
Atherimorpha in having hind coxal tubercles and short macrochaetae on all tibia.  
 
DESCRIPTION.  
Head. Clypeus bulbous, produced anteriorly. Scape approximately the same size as or 
slightly larger than pedicel. Flagellomeres 7-8, cylindrical; first flagellomere larger 
than other flagellomeres; terminal flagellomere more elongate than flagellomeres of 
 
 317
equal girth. Eyes inconspicuously setulose; in male, eyes holoptic or dichoptic (in A. 
longipalpus Nagatomi and Nagatomi only), flattened dorsally, ommatidia evenly 
distributed, of equal size (in A. longipalpus only) or ommatidia split into dorsal and 
ventral areas and smaller ventrally. Labella reduced, very short, with few 
pseudotrachae. Hypopharynx, labium, and labrum very elongate; theca lateral 
sclerites adjacent, apparently fused with suture. Palps two-segmented; distal segment 
longer than proximal segment. Lateral ridge of oral margin absent. Stipes surrounded 
by membrane above theca, directed posteriorly. Cardo swollen. Lacinia longer than 
palps; tip not serrated. Mandibles absent. Cibarial pump long, clearly not as wide as 
long. Cornu shorter than cibarial pump. Pharyngeal pump narrow along most of 
length, mostly flat along its length, longer than length of cibarial pump. Thorax. 
Mesonotum with vittae. Dorsocentral bristles absent, all dorsal setae of equal length. 
Anepisternum bare. Laterotergite, katatergite, and anatergite indistinguishable. 
Laterotergite setose, in rows, mostly on katatergite. Proscutellum present. 
Subscutellum not enlarged nor lengthened; inconspicuous. Wing hyaline or lightly 
infuscate, without markings. Pseudostigma absent. Lower calypter reduced. Upper 
calypter well developed, with broad curvature, lobe-like, width twice length or less. 
Costa extends past wingtip (to at least R
5
). Humeral crossvein (h) well developed. Sc-
r crossvein present, weakly developed, positioned distal to the humeral crossvein (h), 
by the approximate length of h. Dorsal side of R
1
 setulose, ventral side bare. All other 
wing veins without setulae. R
2+3
 nearly straight, apical third of R
2+3
 ultimately bends 
slightly toward wingtip (creating concave flex posteriorly) (although very nearly 
straight). Length of R
2+3
 longer than R
5
, but less than twice as long. Base of R
4
-R
5
 
 
 318
fork proximal or directly above distal end of cell dm. R
4
 at base strongly curved or 
angled; nearly straight apically. R
4
 and R
5
 encompass wing tip. R
5
 clearly longer than 
R
4+5
 (r-m to R
4
 origin). M
3
 wing vein present. Wing cell m3 convergent at margin. 
Origin of CuA
1
 at bm cell. CuA
2
 about 2/3 length of posterior vein of bm cell. Alula 
with broad curvature, rounded evenly. Anal lobe well developed. Anal cell open. 
Halter knob between 1/3?1/2 length of stem. Tibial spur formula 0:2:2. Hind coxal 
tubercle present. Hind tibial macrochaetae present, short. Abdomen. Terminal 
abdominal segments 5?10 evenly tapered from segments 1?4. In female, tergite 7 
much longer than wide, intersegmental membrane between segments 7 and 8 
especially long. Sternite 8 length wider than long or as wide as long. Male terminalia 
with epandrium simple, not containing hypandrium ventrally. Epandrial sclerite wider 
than long, strongly notched anteriorly. Subepandrial sclerite absent. Hypoproct 
triangular, tomentose, without setae. Cercus base held underneath epandrial sclerite, 
directly adjacent to one another, separation distance one quarter width of cercus or 
less. Cerci, in posterior view cupped, forming circular outline medially. Hypandrium 
separated from the gonocoxites by a complete suture. Gonocoxite with sinuous dorsal 
ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short or long enough to 
reach anterior margin of hypandrium. Parameral sheath bulbous, with shallowly 
paired swellings ventrally. Lateral ejaculatory processes present, not part of sperm sac 
posteriorly. Ejaculatory apodeme moderately long, reaching anterior margin of 
hypandrium, or long, reaching beyond anterior margin of hypandrium. Ejaculatory 
apodeme laterally compressed. Aedeagal tines absent. Endoaedeagal process present. 
Female terminalia with three spermathecae, swollen in shape, lightly sclerotized or 
 
 319
not sclerotized. Spermathecal ducts no more than three times the length of sternite 9, 
not inflated at base of spermathecae. Spermathecal duct accessory glands absent. 
Spermathecal ducts near junction sclerotized, thickened, with surface furrows in 
rings. Spermathecal duct junction not thickened. Common spermathecal duct 
thickened, with apical transverse ridge and suture at junction of spermathecal ducts; 
long, clearly longer than longest diameter of genital chamber. Genital chamber oval, 
moderately sized. Accessory gland posterior to genital chamber inconspicuous, easily 
overlooked even after staining. Sternite 9 anterior end pointed; posterior end with 
broad lateral extensions, joined medially with seam, in vertical plane. Tergite 10 
entire, short (length less than half width) (however, elongate in Nagatomi & Iwata, 
1976). Sternite 10 entire, roughly pentagonal, pointed posteriorly; posterior half 
below first cercus segment. Cercus two-segmented. First segment of cercus not 
elongate, with ventral process. Ventral lobes of first segment of cercus not curving 
ventrally towards one another to form a ring. Basal cerci adjacent dorsally. Second 
cercus segment not elongated. Cercus with apical sensory pits. Larva. The larvae are 
unknown. 
LITERATURE. Illustrations of mouthparts, antenna, wing, male and female 
genitalia, and dichotomous key to species is given by Stuckenberg (1956b). A new 
species is illustrated by Nagatomi and Nagatomi (1990a). 
List of included species 
Species (synonyms indented) Author, reference Type country 
Arthroteles bombyliiformis Bezzi, 1926: 322 South Africa 
Arthroteles cinerea Stuckenberg, 1956: 329 South Africa 
Arthroteles longipalpis Nagatomi & Nagatomi, 1990: 312 South Africa 
Arthroteles orophila Stuckenberg, 1956: 327 South Africa 
 
 320
 
Genus ATHERIMORPHA White 
Figures 49-57. 
ATHERIMORPHA White, 1915: 41.  Type species Atherimorpha vernalis White, 
1915, by monotypy. 
 BICALCAR Lindner, 1923: 4.  Type species Chrysopila obscuripennis 
(Loew), 1873, by monotypy [misidentification = 
Atherimorpha]. NOTE 2. 
 THEREVIRHAGIO Lindner, 1925: 20.  Type species Therevirhagio 
setosiradiatus Lindner, 1925, by monotypy. 
 PHILIPPOLEPTIS Malloch, 1931: 276 (as subgenus). Type species Leptis 
praefica Philippi, by original designation. 
 ARTHERIMORPHA Nagatomi, 1982a: 44 (lapsus). 
 NEORHAGIO Lindner, 1924: 75. Type species Leptis setosa Philippi, 1865, 
by monotypy. NEW SYNONYMY. NOTE 3. 
 
DIAGNOSIS.  
The phylogenetic analysis in the previous chapter indicated that South American 
Atherimorpha may be more closely related to the African genus Arthroteles than to 
other Atherimorpha. African species of Atherimorpha need to be studied further to 
verify this result. 
 
 
 321
The distinctive form of the antenna may provide evidence for the monophyly of the 
species of Atherimorpha. While the first flagellomere may vary in shape (from 
subglobose to onion-shaped to conical) and the total number of flagellomeres may 
also vary (3 to 7), the first flagellomere is always enlarged compared to the other 
flagellomeres, which are narrow and rod-like in form. Species of the genus 
Atherimorpha are small to moderately sized (4.4 to 11.4 mm) flies of coloration that 
varies from entirely gray or black and gray (as all Australian members of this genus) 
to brown, brown and yellow, entirely yellow, or orangish. Eyes in male holoptic or 
dichoptic, laterotergite setose, tibial spur formula 0:2:2, macrochaetae present on hind 
tibiae, and wing vein M
3
 present. 
 
Atherimorpha is very Rhagio-like in form and behavior, but can be distinguished 
from this genus by its antenna and by having two-segmented palps. Atherimorpha is 
most similar to Arthroteles in the form of its antenna, but it differs in that the first 
flagellomere is larger in comparison to the other flagellomeres and the other 
flagellomeres are narrower. Atherimorpha may be distinguished immediately from 
Arthroteles by its fleshy proboscis. Atherimorpha is distributed in South America, 
South Africa, and Australia and among these faunas, South America is the richest in 
terms of species number and morphological form and color, although many of these 
remain undescribed.  
 
DESCRIPTION.  
 
 322
Head. Clypeus bulbous. Scape approximately the same size as pedicel or clearly 
larger than pedicel (as in A. albohirta and A. praefica). Flagellomeres 3 to 7; first 
flagellomere clearly larger than other flagellomeres, round in cross section or laterally 
compressed; round, conical, subglobose to globose, or fusiform; terminal 
flagellomere more elongate than other cylindrical flagellomeres. Eyes 
inconspicuously setulose; in female, dichoptic; in male, holoptic or dichoptic, not 
flattened dorsally, ommatidia evenly distributed, of equal size. Labella with 
pseudotrachae, longer or shorter than palps. Theca short and stout, lateral theca 
sclerites adjacent and touching, but mostly separated. Palps two-segmented; distal 
segment longer than proximal segment. Stipes surrounded by membrane above theca, 
directed posteriorly. Cardo not swollen. Lacinia shorter than palps, tip not serrated. 
Mandibles absent. Cibarial pump long, clearly not as wide as long. Cornu shorter than 
cibarial pump. Pharyngeal pump narrow along most of length, mostly flat along its 
length, approximately same length as cibarial pump. Thorax. Mesonotum with vittae. 
Dorsocentral bristles present or absent. Anepisternum bare. Laterotergite setose, in 
row(s), mostly on katatergite. Postspiracular scale absent. Proscutellum absent. 
Subscutellum not enlarged nor lengthened; inconspicuous. Wing hyaline or lightly or 
darkly infuscate; without markings. Pseudostigma present or absent. Lower calypter 
reduced. Upper calypter well developed, with reduced curvature, width more than 
twice length. Costa extends to wingtip (between R
4
 and R
5
). Humeral crossvein (h) 
well developed. Sc-r crossvein present, weakly developed; positioned distal to the 
humeral crossvein (h), by less than the length of h, by the approximate length of h, or 
by more than the length of h. Dorsal side of R
1
 setulose, ventral side with or without 
 
 323
setulae. Other wing veins and cells variously setulose or bare. R
2+3
 sinuous, apical 
third of R
2+3
 ultimately bends toward wingtip (creating convex flex anteriorly). 
Length of R
2+3
 longer than R
5
, but less than twice as long. Base of R
4
-R
5
 fork 
proximal or directly above distal end of cell dm. R
4
 at base strongly curved or angled, 
nearly straight apically. R
4
 and R
5
 encompass wing tip. R
5
 aligned with R
4+5
. R
5
 
clearly longer than R
4+5
 (r-m to R
4
 origin). Origin of CuA
1
 at bm cell. M
3
 wing vein 
present. Wing cell m3 parallel-sided at margin. CuA
2
 greater than 2/3 length of 
posterior vein of bm cell. Alula with narrow or broad curvature, rounded evenly. Anal 
lobe well developed. Anal cell open or closed. Halter knob between 1/2?2/3 length of 
stem. Tibial spur formula 0:2:2. Hind coxal tubercle absent. Hind tibial macrochaetae 
enlarged. First hind metatarsus of male not swollen. Postmetacoxal bridge absent. 
Abdomen. Terminal abdominal segments 5?10 evenly tapered from segments 1?4. In 
female, tergite 7 about as long as wide, intersegmental membrane between segments 
7 and 8 especially long, sternite 8 length wider than long or as wide as long 
(triangular, ovoid, or nearly square). Male terminalia with epandrium simple, not 
containing hypandrium ventrally. Epandrial sclerite usually wider than long (longer 
than wide in A. albohirta and A. praefica), strongly notched anteriorly. Subepandrial 
sclerite absent. Hypoproct margins entire, setose. Cercus base held underneath 
epandrial sclerite. Cerci directly adjacent to one another, separation distance one 
quarter width of cercus or less, held horizontal or at angle in relation to rest of 
abdomen; in posterior view, flat. Hypandrium separated from the gonocoxites by a 
complete suture. Gonocoxite with sinuous dorsal ridge, leading to gonocoxal 
apodeme. Gonocoxal apodemes short or long enough to reach anterior margin of 
 
 324
hypandrium. Parameral sheath bulbous, with shallowly paired swellings ventrally. 
Lateral ejaculatory processes present, not part of sperm sac posteriorly. Ejaculatory 
apodeme moderately long, reaching anterior margin of hypandrium, laterally 
compressed. Aedeagal tines absent. Endoaedeagal process present. Female terminalia 
with three spermathecae, swollen, not sclerotized. Spermathecal ducts no more than 
three times the length of sternite 9, not inflated at base of spermathecae. Spermathecal 
duct accessory glands absent. Spermathecal ducts near junction thickened, but not 
sclerotized, without surface furrows. Spermathecal duct junction thickened. Common 
spermathecal duct thickened, about the same length as the longest diameter of genital 
chamber. Genital chamber oval, moderately sized. Accessory gland posterior to 
genital chamber inconspicuous, easily overlooked even after staining. Sternite 9 
anterior end pointed, posterior end with broad lateral extensions that meet medially, 
in vertical plane. Tergite 10 entire. Sternite 10 roughly rectangular, posterior half 
below first cercus segment. Cercus two-segmented. First segment of cercus not 
elongate, with ventral process. Ventral lobes of first segment of cercus not curving 
ventrally towards one another to form a ring. Basal cerci adjacent dorsally. Second 
cercus segment not elongated. Cercus with apical sensory pits.  Larva. The larvae are 
unknown. 
LITERATURE. A revision of the African fauna, including a key to species, is 
given by Nagatomi and Nagatomi (1990b). Malloch (1932a) gives a key to the 
Neotropical species. Paramonov (1962) treats the Australian species, and provides a 
key. 
 
 325
List of included species 
Species (synonyms indented) Author, reference Type country 
Atherimorpha agathae Paramonov, 1962: 167 Australia 
Atherimorpha albipennis Bezzi, 1926: 318 South Africa 
Atherimorpha albohirta Malloch, 1932: 208 Argentina 
Atherimorpha alisae Paramonov, 1962: 158 Australia 
Atherimorpha angustifrons Nagatomi & Nagatomi, 1990: 44 South Africa 
Atherimorpha atrifemur Malloch, 1932: 210 Chile 
Atherimorpha bevisi Stuckenberg, 1956: 143 South Africa 
Atherimorpha claripennis (as Leptis; Philippi, 1865: 772) Chile 
Atherimorpha commoni Paramonov, 1962: 164 Australia 
Atherimorpha corpulenta Paramonov, 1962: 166 Australia 
Atherimorpha crassitibia Nagatomi & Nagatomi, 1990: 52 South Africa 
Atherimorpha edgari Paramonov, 1962: 164 Australia 
Atherimorpha edwardsi Malloch, 1932: 212 Chile 
?Atherimorpha festuca Jell & Duncan, 1986: 181 Australia 
Atherimorpha flavicorpus Nagatomi & Nagatomi, 1990: 57 South Africa 
Atherimorpha flavofasciata Paramonov, 1962: 161 Australia 
Atherimorpha flavolateralis Malloch, 1932: 213 Argentina 
Atherimorpha fulva Hardy, 1920: 121 Australia 
Atherimorpha fusca Malloch, 1932: 213 Chile 
Atherimorpha fuscicoxa Malloch, 1932: 215 Chile 
Atherimorpha gracilipennis Nagatomi & Nagatomi, 1990: 59 South Africa 
Atherimorpha grisea (as Leptis; Philippi, 1865: 774) Chile 
Atherimorpha hirtula Bigot, 1887: 116 Chile 
Atherimorpha imitans Malloch, 1932: 211 Chile 
Atherimorpha infuscata Paramonov, 1962: 162 Australia 
Atherimorpha irwini Nagatomi & Nagatomi, 1990: 59 South Africa 
Atherimorpha latipennis Stuckenberg, 1956: 144 South Africa 
Atherimorpha longicornu Nagatomi & Nagatomi, 1990: 59 South Africa 
Atherimorpha lugens (as Leptis; Philippi, 1865: 773) Chile 
Atherimorpha mcalpinei Paramonov, 1962: 167 Australia 
Atherimorpha mensaemontis Stuckenberg, 1961: 116 South Africa 
Atherimorpha montana Hardy, 1927: 125 Australia 
Atherimorpha nemoralis (as Leptis; Philippi, 1865: 772) Chile 
Atherimorpha nigrata (as Leptis; Philippi, 1865: 772) Chile 
Atherimorpha norrisi Paramonov, 1962: 162 Australia 
Atherimorpha occidens Hardy, 1927: 126 Australia 
Atherimorpha ornata Nagatomi & Nagatomi, 1990: 74 South Africa 
Atherimorpha praefica (as Leptis; Philippi, 1865: 772) Chile 
Leptis setosus (Philippi, 1865: 773) Syn. Nov. Chile 
Psilocephala macrochaeta (Bigot, 1889: 325) Chile 
Psilocephala pilosa (Bigot, 1889: 326) Chile 
Atherimorpha pusilla Paramonov, 1962: 165 Australia 
Atherimorpha rieki Paramonov, 1962: 157 Australia 
 
 326
Species (synonyms indented) Author, reference Type country 
Atherimorpha scutellaris Malloch, 1932: 214 Chile 
Atherimorpha setosiradiata (as Therevirhagio; Lindner, 1925: 20) Australia 
Atherimorpha stuckenbergi Nagatomi & Nagatomi, 1990: 79 South Africa 
Atherimorpha subannulata (as Leptis; Philippi, 1865: 771) Chile 
Atherimorpha tonnoiri Paramonov, 1962: 159 Australia 
Atherimorpha triangularis Malloch, 1932: 234 Chile 
Atherimorpha uptoni Paramonov, 1962: 163 Australia 
Atherimorpha vernalis White, 1914: 42 Australia 
Atherimorpha victoriana Paramonov, 1962: 160 Australia 
Atherimorpha villosissima Paramonov, 1962: 158 Australia 
 
Genus CHRYSOPILUS Macquart 
Figures 58-65, 175, 186A. 
 
CHRYSOPILUS Macquart, 1826: 403.  Type species Musca diadema Linnaeus 
1767, by designation of Westwood, 1840: 134 
(misidentification) = Rhagio  aureus Meigen, 1804. 
 LEPTIPALPUS Rondani, 1850: 183. Type species Tabanus brasiliensis 
Rondani 1850, by monotypy. 
 HELIOMYIA Doleschall, 1857: 402.  Type species Heliomyia ferruginea 
Doleschall, 1857 [=Leptis ferruginosus Wiedemann, 1819], by 
monotypy.  
 MACELLOPALPUS Bigot, 1886: xlviii.  Type species Macellopalpus 
flaveolus Bigot, 1886 [=Leptis ferruginosus Wiedemann, 
1819], by monotypy. 
 ?PALEOCHRYSOPILA Meunier, 1892: lxxxiii. Type species Chrysopilus 
nagatomii Evenhuis, 1994: 292 (= Chrysopilus meunieri Kerr, 
present work). NOTE 4. 
 
 327
 POPPIUSIELLA Frey, 1918: 30 (as subgenus). Type species Chrysopilus 
arctica, by original designation. NOTE 5. 
 ACHRYSOPILUS Szil?dy, 1934: 255 (as subgenus; no type species given; 
name invalid by article 13.3 of the code). 
 SAPPOROMYIA Szil?dy, 1934: 233. Type species Leptis basalis Matsumura, 
1915, by monotypy. 
 CHRYSOPILODES Frey, 1954: 15 (as subgenus). Type species Chrysopilus 
boettcheri Frey, 1954, by monotypy. 
 VARIOPILUS Frey, 1954: 22 (as subgenus). Type species Chrysopilus 
aequicellulatus Frey, 1954, by original designation. 
 SOLOMOMYIA Nagatomi, 1982a: 50. Type species Solomomyia gressitti 
Nagatomi, 1982a, by original designation. Correct original 
spelling by present revision. New synonymy. NOTE 6.  
 SOLOMYIA Nagatomi, 1982a: 68. Incorrect original spelling. 
 CHRYSOPILA, CHRYSOPYLA, errors for Chrysopilus Macquart. 
  
DIAGNOSIS.  
The monophyly of the species of Chrysopilus is uncertain, due to the recognition of 
Schizella, which shares all of the potential autapomorphies of the genus. These 
synapomorphies include thoracic setae that are slightly flattened, with a metallic 
sheen, often with structural color present and a reduced, bare proepimeron. 
Chrysopilus may also be paraphyletic with respect to Stylospania.  
 
 
 328
Chrysopilus species are delicate to fairly robust flies, variably sized (3.7 to 19 mm), 
usually with long, thin legs; black, gray, brown, or orange-brown; often with colored 
setae on thorax and/or abdomen that adds to color pattern. Wings are hyaline or 
infuscate, with or without markings; male holoptic (males dichoptic in a few African 
species; eyes separated in female); first flagellomere subcircular, laterally 
compressed, with terminal arista; mandibles absent; laterotergite setose; wing vein M
3
 
present; tibial spur formula 0:2:1; hind tibia with short macrochaetae; tergite 9 
without ventrolateral arms; female spermathecal ducts with accessory glands. 
Chrysopilus is cosmopolitan, found on all continents except Antarctica, throughout 
the tropics, up to near 4000m in Bolivia (pers. obs.); as far north as the Arctic circle, 
and as far south as Chilo? Island, Chile; in Africa, most Chrysopilus species are 
confined to humid montane forest (Stuckenberg, 1996). In the northern hemisphere, 
Chrysopilus species are most commonly confused with species of Rhagio, but may be 
distinguished by having a single hind tibial spur; arista with microsetae longer than 
width of arista; and a reduced, bare proepimeron. In the Philippines and possibly its 
surrounding area, Chrysopilus may be distinguished from Schizella and Stylospania 
solely by its antenna, which has the arista arising from the first flagellomere centrally 
(not ventrally), and is the same in both sexes (the female of Stylospania is unknown). 
In the southern hemisphere, Chrysopilus is distinguished from Atherimorpha by 
having a single hind tibial spur and aristate antenna. 
 
DESCRIPTION.  
 
 329
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere laterally compressed, rounded and slightly enlarged, bearing fused 
arista-like extension. Eyes inconspicuously setulose; in female, dichoptic; in male, 
holoptic or dichoptic, ommatidia evenly distributed, of equal size, or ommatidia split 
into dorsal and ventral areas and smaller ventrally, not flattened dorsally. Labella with 
pseudotrachae, longer or shorter than palps. Theca short and stout, divided into lateral 
sclerites that are tightly adjacent, apparently fused with suture. Palps one-segmented. 
Stipes surrounded by membrane above theca, directed posteriorly (very reduced). 
Lacinia present, shorter than palps, tip not serrated. Mandibles absent. Cibarial pump 
long, clearly not as wide as long. Cornu shorter than cibarial pump. Pharyngeal pump 
narrow and flat along most of length, approximately half the length of cibarial pump. 
Thorax. Mesonotum with or without vittae. Dorsocentral bristles absent; all dorsal 
setae of equal length. Anepisternum setulose on dorsal margin only or setulose on 
dorsal and posterior margins. Laterotergite setose, throughout laterotergite. Metallic- 
or scale-like thoracic setae, often with structural color, present. Proscutellum absent. 
Subscutellum bulbous or not bulbous. Wing hyaline or infuscate, with or without 
markings; pseudostigma present or absent. Lower calypter reduced. Upper calypter 
variously developed. Costa extends to wingtip or past wingtip. Humeral crossvein (h) 
well developed. Sc-r crossvein present, weakly developed, or absent; positioned distal 
to the humeral crossvein (h), by less than the length of h. Dorsal side of R
1
 setulose, 
ventral side bare. All other wing veins bare. Wing veins R
1
 and R
2+3
 close together at 
wing margin. R
2+3
 sinuous, apical third ultimately bends toward leading edge of wing 
margin (creating concave flex anteriorly), length of R
2+3
 about the same length as R
5
, 
 
 330
or longer. Base of R
4
-R
5
 proximal of, directly above, or distal of distal end of cell dm. 
R
4
 at base usually strongly curved, leading directly to wing margin or with short 
proximal offshoot at point of curvature near R
5
; along most of its length, nearly 
straight or lightly sinuous. R
4
 and R
5
 contain wing tip or R
4
 ending at wing tip. R
5
 
longer or shorter than R
4+5
 (r-m to R
4
 origin). R-m crossvein proximal to one-third of 
discal cell. Origin of CuA
1
 at bm cell. CuA
2
 greater than 2/3 length of posterior vein 
of bm cell. M
3
 wing vein present. Alula full, rounded, with broad curvature. Anal 
lobe well developed. Anal cell closed. Halter knob between 1/3?1/2 length of stem. 
Tibial spur formula 0:2:1. Hind coxal tubercle absent. Hind tibial macrochaetae 
present, short. Postmetacoxal bridge present, reaching internally to base of 
metasternal furcum as incomplete thin extension. Abdomen. Abdominal segments 
evenly tapered. In female, last 3 abdominal segments telescoping; tergite 7 much 
longer than wide; intersegmental membrane between segments 7 and 8 especially 
long; sternite 8 as wide as long or wider than long. Male genitalia with epandrium 
simple, not containing hypandrium ventrally. Epandrial sclerite wider than long, 
strongly notched anteriorly. Subepandrial sclerite present, divided medially, without 
setae. Hypoproct tomentose, without setae. Cerci directly adjacent to one another, 
separation distance one quarter width of cercus or less. Cerci, in posterior view 
flattened or lightly rounded. Hypandrium fused entirely to gonocoxites. Gonocoxite 
with or without dorsal sinuous ridge, leading to gonocoxal apodeme. Gonocoxal 
apodemes short or long enough to reach anterior margin of hypandrium. Parameral 
sheath bulbous, without paired swellings ventrally. Lateral ejaculatory processes 
present, not part of sperm sac posteriorly. Ejaculatory apodeme moderately long to 
 
 331
long, reaching at least anterior margin of hypandrium. Ejaculatory apodeme rod-
shaped or laterally compressed (often upside-down v-shaped in profile). Aedeagal 
tines absent. Endoaedeagal process absent. Female terminalia with three 
spermathecae, clubbed or swollen, lightly to moderately sclerotized. Spermathecal 
ducts no more than three times the length of sternite 9, not inflated at base of 
spermathecae, without swelling halfway between genital chamber and spermathecae. 
Spermathecal duct accessory glands present or absent; where present, arise at 
approximately the distal third of the spermathecal ducts or at the base of each 
spermatheca. Ejection apparatus of spermathecal ducts lightly sclerotized, not 
thickened, with surface furrows. Common spermathecal duct thickened, moderately 
long, about the same length as the longest diameter of genital chamber. Genital 
chamber elongate, occupying most of sternite 9 area. Accessory gland posterior to 
genital chamber inconspicuous, easily overlooked even after staining. Sternite 9 
anterior end broadly paddle-shaped; posterior end with broad extensions posteriorly, 
held in vertical plane. Tergite 10 partially split or split into two separate lateral 
sclerites, short (length less than half width). Sternite 10 split into two sclerites, almost 
entirely underneath cercus segments. Cercus two-segmented. First segment of cercus 
not elongate, with or without ventral process. Ventral lobes of first segment of cercus 
curve ventrally towards one another to form a ring, visible in the posterior 
perspective. Basal cerci adjacent dorsally. Second cercus segment not elongated, with 
or without apical sensory pits. Larva. Body with 11 segments (not counting head). 
Thoracic segments with creeping welts ventrally. Head capsule not folded within 
second segment. Head capsule composed of a single, undivided plate (dorsal plate). 
 
 332
Head capsule less than 4.5 times longer than greatest width (1.5 width/ 4.5 length). 
Mandibular brush present, associated with simple fold of cuticle. Mandibular hook 
with external groove on adoral surface, smooth, without microsetae. Labral teeth 
developed, sclerotized, in two rows, converging anteriorly (teeth separated by central 
depression). Maxilla not sclerotized. Saw sclerite of mandibular base absent. 
Maxillary palp soft, segments poorly differentiated. Maxillary palp segment number 
three. Antenna last segment entire (as nub). Antenna three-segmented. Unpaired 
salivary pump absent. Posterior tentorial expansion free, with thin extension produced 
dorsally.  
LITERATURE. Hardy (1949) gives to key to species of North America. Narchuk 
(1969) gives key to eastern European species. Yang et al. (1997) give key to species 
of China. Nagatomi (1978) gives key to species of Japan. For keys of Chrysopilus of 
Africa, see Stuckenberg (1965, 1996).  
List of included species 
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus aequalis (as Leptis; Walker, 1848: 216) Australia 
Chrysopilus aequicellulatus (as Variopilus; Frey, 1954: 22) Argentina 
Chrysopilus alaskaensis Hardy, 1949: 147 USA 
Chrysopi lus albicornis de Meijere, 1914: 28 Java 
Chrysopilus albobasalis Brunetti, 1920: 140 India 
Chrysopilus albopictus Brunetti, 1909: 428 India 
Chrysopilus alpicola (as Chrysopila; Pokorny, 1886: 
194) 
Switzerland 
Chrysopilus alternatus Brunetti, 1920: 144 India 
Chrysopilus amamiensis Nagatomi, 1968: 33 Japan 
Chrysopilus americanus (as Chrysopila; Schiner, 1868: 
197) 
?South America? 
Chrysopilus amulus Kerr, new name. Mexico 
Chrysopila latifrons  Williston, 1901: 266. 
Preoccupied by Chrysopilus 
Italy 
 
 333
Species (synonyms indented) Species reference Species Type 
Country 
latifrons Bezzi, 1898: 32. 
Chrysopilus amurensis Soboleva, 1986: 112 Russia 
Chrysopilus andersoni Leonard, 1930: 131 USA 
Chrysopilus andicola Lindner, 1924: 71 Bolivia 
Chrysopilus andringitrensis Stuckenberg, 1965: 116 Madagascar 
Chrysopilus androgynus Paramonov, 1962: 123 Australia 
?Chrysopilus anglicus Cockerell, 1921: 471 England 
(Eocene/Oligocene) 
Chrysopilus angustifacies Hardy, 1949: 148 USA 
Chrysopilus angustifrons Frey, 1954: 18 Burma 
Chrysopilus ankaratrae Stuckenberg, 1965: 126 Madagascar 
Chrysopilus anthracinus Bigot, 1887: 105 USA 
Chrysopilus antipoda Bigot, 1887: 105 Australia 
Chrysopilus antipodes Nagatomi & Evenhuis, 1989: 
298. error  
Australia 
Chrysopilus antongilensis Stuckenberg, 1965: 150 Madagascar 
Chrysopilus apicalis Wulp, 1882: 119 Guadeloupe 
Chrysopilus apicimaculatus Yang & Yang, 1991: 274 China 
Chrysopilus arctica (as Chrysopila; Frey, 1918: 29) Russia 
Chrysopilus arctiventris James, 1936: 343 USA 
Chrysopilus argenteofasciatus (as Chrysopila; Bromley, 1931: 
9) 
Guyana 
Chrysopilus argenteus Paramonov, 1962: 123 Australia 
Chrysopilus argyrophorus (as Chrysopila; Schiner, 1868: 
199) 
not given; South 
America 
Chrysopilus ater (as Chrysopila; Williston, 1896: 
304 
St. Vincent 
Chrysopilus aterrimus (as Chrysopila; Williston, 1901: 
264) 
Mexico 
Chrysopilus atricornis Stuckenberg, 1965: 149 Madagascar 
Chrysopilus asiaticus Lindner, 1923: 9 Central Asia 
Chrysopilus auratus (as Atherix; Fabricius, 1805: 73) Denmark 
Chrysopilus aureus (as Rhagio; Meigen, 1804: 309) not given; Europe 
Rhagio diadema Fabricius, 1775: 762. 
misidentification 
"Lipsiae hortis" 
Leptis vitripennis Meigen, 1820: 101 not given; Europe 
Rhagio aurulans Meigen, 1820: 101 not given; Europe 
Rhagio luridus Meigen, 1820: 101 not given; Europe 
Chrysopilus aureus 
meridionalis 
Bezzi, 1898: 32 Italy 
Chrysopilus austriacus Kerr, new name Austria 
Chrysopilus unicolor  Becker, 1922: 71. Preoccupied by 
Chrysopilus unicolor Brunetti, 
1909: 432. 
Austria 
 
 334
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus aymara Lindner, 1924: 72 Peru 
Chrysopilus azurinus Frey, 1954: 19 Philippines 
Chrysopilus basalis Walker, 1860: 285 Mexico 
Chrysopilus basifasciatus Paramonov, 1962: 124 Australia 
Chrysopilus basiflavus Yang & Yang, 1992: 355 China 
Chrysopilus basilaris (as Leptis; Say, 1823: 36) USA 
Chrysopilus batak Kerr, new name Sumatra 
Chrysopilus tomentosus   
 
de Meijere, 1924: 13. 
Preoccupied by Chrysopilus 
tomentosus  Bigot, 1887: 104. 
Sumatra 
Chrysopilus beameri Hardy 1949: 152 USA 
Chrysopilus bequaerti Curran,, 1931: 3 Cuba 
Chrysopilus betsileorum Stuckenberg, 1965: 118 Madagascar 
Chrysopilus binoculatus Edwards, 1915: 397 Indonesia 
Chrysopilus binotatus Loew, 1871: 62 Greece 
Chrysopilus birmanensis Brunetti, 1920: 137 Burma 
Chrysopilus bisectus Oldroyd, 1939: 17 Uganda 
Chrysopilus bistriatipennis Brunetti, 1927: 300 Malaya 
Chrysopilus boettcheri Frey, 1954: 15 Philippines 
Chrysopilus brunneifrons Kert?sz, 1902: 147 Peru 
Chrysopilus caducus (as Leptis; Wiedemann, 1828: 
579) 
Brazil 
Chrysopilus terminalis Macquart, 1846: 234 ?Columbia, 
Venezuela? 
Chrysopilus calchaqui Coscaron & Coscaron, 1995: 267 Argentina 
Chrysopilus calopterus (as Leptis; Schiner, 1868: 579) Brazil 
Chrysopilus choui Yang & Yang, 1989: 243 China 
Chrysopilus chrysopiliformis (as Atherix; Lindner, 1924: 69) Bolivia 
Chrysopilus clarapex Frey, 1954: 18 Burma 
Chrysopilus claricinctus Lindner, 1923: 10 Central Asia 
Chrysopilus clarus (as Leptis; Walker, 1852: 164) Brazil 
Chrysopilus clemendoti Stuckenberg, 1965: 131 Madagascar 
Chrysopilus cochinensis Brunetti, 1920: 136 India 
Chrysopilus coeruleothorax Lindner, 1925: 22 Fiji 
Chrysopilus cognatus Stuckenberg, 1965: 163 Madagascar 
Chrysopilus collesi Paramonov, 1962: 127 Australia 
Chrysopilus commoni Paramonov, 1962: 128 Australia 
Chrysopilus connexus Johnson, 1912: 108 USA 
Chrysopilus consanguineus Schiner, 1868: 197 Brazil 
Chrysopilus correctus Osten Sacken, 1882: 101 Philippines 
Chrysopilus cricosphaerota Speiser, 1914: 4 Cameroun 
Chrysopilus cubensis Curran, 1931: 5 Cuba 
Chrysopilus dauricus Frey, 1954: 22 Russia 
Chrysopilus davisi Johnson, 1912: 4 USA 
 
 335
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus decisus (as Leptis; Walker, 1857: 15) Malaya 
Chrysopilus decoratus de Meijere, 1911: 290 Java 
Chrysopilus depressiconus Frey, 1954: 20 Burma 
Chrysopilus dilatus Cresson, 1919: 177 USA 
Chrysopilus diplostigma Bezzi, 1917: 120 Philippines 
Chrysopilus ditissimis Bezzi, 1912: 451 Japan 
Chrysopilus apyros S?guy, 1948: 154 Japan 
Chrysopilus dives Loew, 1871: 62 Russia 
Chrysopilus divisus Hardy, 1949: 152 USA 
Chrysopilus donato Curran, 1931: 6 Panama 
Chrysopilus edgari Paramonov 1962: 130 Australia 
Chrysopilus egregius de Meijere,, 1919: 22 Sumatra 
Chrysopilus elegans Schiner, 1868: 198 Colombia 
Chrysopilus erythrophthalmus Loew, 1840: 3 Poland 
Leptis hyalipennis von Roser, 1840: 52 not given; Europe 
Chrysopilus facetticus Paramonov, 1962: 125 Australia 
Chrysopilus fasciatus (as Leptis; Say, 1823: 37) USA 
Leptis par Walker, 1848: 215 USA 
Chrysopilus fascipennis (as Macellopalpus; Brunetti) 
1920: 123 
India 
Chrysopilus fasciventris Curran, 1931: 7 Panama 
Chrysopilus fenestratus (as Chrysophilus; Bezzi, 1912: 
448) 
Taiwan 
Chrysopilus sanjodokeana Matsumura, 1916: 348 Japan 
Chrysopilus ferruginosus (as Leptis; Wiedemann) 1819: 4 Indonesia 
Heliomyia ferruginea Doleschall, 1857: 402 Indonesia 
Macellopalpus flaveolus Bigot, 1886 Papua New Guinea 
Chrysopilus frater Brunetti, 1909: 431 Burma 
Macellopalpus fulvidus Brunetti, 1909: 424 India 
Chrysopilus ferruginosus 
dimidiatus 
Frey, 1954: 18 Vietnam 
Chrysopilus ferruginosus 
philippinus 
Frey, 1954: 18 Philippines 
Chrysopilus ferruginosus 
burmanicus 
Frey, 1954: 19 Burma 
Chrysopilus fimbriatus Stuckenberg, 1997: 238 South Africa 
Chrysopilus flaveolus (as Leptis; Meigen, 1820: 100) ?Alpen? 
Chrysopilus flavibarbus Adams, 1904: 438 USA 
Chrysopilus cameroni Curran, 1926: 170 USA 
Chrysopilus aldrichi James, 1936: 343 USA 
Chrysopilus flaviscutellus Yang & Yang, 1989: 290 China 
Chrysopilus flavopilosus Brunetti, 1920: 138 India 
Chrysopilus flavopunctatus Brunetti, 1909: 213 India 
Chrysopilus foedus Loew, 1861: 317 USA 
 
 336
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus fulvidus Bigot, 1891: 370 Ivory Coast 
Chrysopilus fuscicinctus Brunetti, 1927: 299 Malaya 
Chrysopilus fuscipes Bigot, 1887: 103 France 
Chrysopilus gansuensis Yang & Yang, 1991: 95 China 
Chrysopilus gemmiferus Frey, 1954: 17 Laos 
Chrysopilus georgianus Hardy, 1949: 154 USA 
Chrysopilus gilvipennis Edwards, 1919: 30 Sumatra 
Chrysopilus golbachi Coscaron & Coscaron, 1995: 263 Argentina 
Chrysopilus grandis Yang & Yang, 1993: 3 China 
Chrysopilus gratiosus Paramonov, 1962: 131 Australia 
Chrysopilus gravelyi Brunetti, 1920: 137 India 
Chrysopilus griffithi Johnson, 1897: 119 USA 
Chrysopilus griseipennis Bezzi, 1912: 451 Formosa 
Chrysopilus griveaudi Stuckenberg, 1965: 113 Madagascar 
Chrysopilus guangxiensis Yang & Yang, 1992: 354 China 
Chrysopilus guianicus Curran, 1931: 5 Guyana 
Chrysopilus guttipennis Walker, 1861: 282 Indonesia 
Chrysopilus guttulatus de Meijere, 1914: 31 Indonesia 
Chrysopilus fenestratus de Meijere, 1913: 321. 
Preoccupied by Chrysopilus 
fenestratus Bezzi, 1912: 448. 
Indonesia 
Chrysopilus hakusanus Nagatomi, 1978: 446 Japan 
Chrysopilus hardyi Nagatomi & Evenhuis, 1989: 297 Australia 
Chrysopilus fascipennis Hardy, 1933: 408. Preoccupied 
by Chrysopilus fascipennis 
(Brunetti, 1920: 123). 
Australia 
Chrysopilus helvolus (as Leptis; Meigen, 1820: 100) Switzerland 
Chrysopilus heroicus Paramonov, 1962: 126 Australia 
Chrysopilus howei Paramonov, 1962: 119 Lord Howe Island 
Chrysopilus huashanus Yang & Yang, 1989: 244 China 
Chrysopilus hubeiensis Yang & Yang, 1991: 274 China 
Chrysopilus humeralis Brunetti, 1912: 466 India 
Chrysopilus humilis Loew, 1874: 379 USA 
Chrysopilus hybridus Lindner, 1924: 73 Peru 
Chrysopilus iani Paramonov, 1962: 128 Australia 
Chrysopilus illustris Frey, 1954: 17 Burma 
Chrysopilus imitator Paramonov, 1962: 129 Australia 
Chrysopilus impar Walker, 1861: 282 Indonesia 
Chrysopilus incidens Curran, 1927: 95 Zaire 
Chrysopilus indris Stuckenberg, 1965: 151 Madagascar 
Chrysopilus inka Lindner, 1924: 73 Peru 
Chrysopilus insularis Schiner, 1868: 199 Nicobar Islands 
Chrysopilus intermedius Paramonov, 1962: 131 Australia 
Chrysopilus intermedius Bezzi, 1895: 42 Italy 
 
 337
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus invalidus Williston, 1901: 265 Mexico 
Chrysopilus irroratus Schiner, 1868: 198 ?South America? 
Chrysopilus itoi Nagatomi, 1958: 36 Japan 
Chrysopilus ivontakae Stuckenberg, 1965: 161 Madagascar 
Chrysopilus jamaicensis (as Chrysopila; Johnson, 1894: 
273) 
Jamaica 
Chrysopilus keiseri Stuckenberg, 1965: 158 Madagascar 
Chrysopilus kimoroensis Stuckenberg, 1965: 129 Madagascar 
Chrysopilus kincaidi Hardy, 1949: 156 USA 
Chrysopilus komurae Matsumura, 1911: 68 Russia 
Chrysopilus kyotoensis Frey, 1954: 23 Japan 
Chrysopilus laetus Zetterstedt, 1842: 224 Sweden 
Chrysopilus lateralis Oldroyd, 1939: 18 Uganda 
Chrysopilus latifrons Bezzi, 1898: 32 Italy 
Chrysopilus latipennis Stuckenberg, 1965: 156 Madagascar 
Chrysopilus latistigma Curran, 1931: 7 Panama 
Chrysopilus latus Brunetti, 1920: 143 India 
Chrysopilus lemur Stuckenberg, 1965: 124 Madagascar 
Chrysopilus leonardi Curran, 1931: 4 Puerto Rico 
Chrysopilus leptiformis Kert?sz, 1902: 148 Peru 
Chrysopilus lii Yang, Yang, and Nagatomi, 
1997: 140 
China 
Chrysopilus lilianae Soboleva, 1986: 114 Russia 
Chrysopilus lineatus Lindner, 1929: 267 Brazil 
Chrysopilus lokobiensis Stuckenberg, 1965: 144 Madagascar 
Chrysopilus longipalpis Hardy, 1949: 157 USA 
Chrysopilus lucifer Walker, 1852: 164 Colombia 
Chrysopilus lucimaculatus Yang & Yang, 1992: 355 China 
Chrysopilus luctuosus (as Chrysophilus; Brunetti, 1909: 
430) 
India 
Chrysopilus luculentus Nagatomi, 1968: 41 Japan 
Chrysopilus ludens Loew, 1861: 34 Cuba 
Chrysopilus lugubrinus de Meijere, 1924: 12 Sumatra 
Chrysopilus lupinus Osten Sacken, 1881: 420 Sumatra 
Chrysopilus luteolus (as Leptis; Fall?n, 1814: 10) Sweden 
Chrysopilus mackerrasi Paramonov, 1962: 120 Australia 
Chrysopilus macu laris Curran, 1931: 6 Puerto Rico 
Chrysopilus maculipennis (as Chrysophilus; Walker, 1857: 
118) 
Borneo 
Chrysopilus madecassus Stuckenberg, 1965: 134 Madagascar 
Chrysopilus madecassus 
merinanus 
Stuckenberg, 1965: 137 Madagascar 
Chrysopilus maerens Loew, 1873: 36 Not specified (Korfu 
or Herkulesbad?) 
 
 338
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus magnipennis Brunetti, 1909: 213 Sumatra 
Chrysopilus malaisei Frey, 1954: 21 Burma 
Chrysopilus marmoratus Brunetti, 1909: 429 India 
Chrysopilus mawambus Kerr, new name Africa (?Mawambi-
Ukaika?) 
Chrysopilus obscuripes Brunetti, 1927: 298. Preoccupied 
by Chrysopilus obscuripes 
Speiser, 1923: 98. 
Africa (?Mawambi-
Ukaika?) 
Chrysopilus mcalpinei Paramonov, 1962: 121 Australia 
?Chrysopilus meunieri Kerr, new name Baltic Region 
(Eocene/ Oligocene) 
Chrysopilus nagatomii Evenhuis 1994: 292. Preoccupied 
by Chrysopilus nagatomii Yang 
& Yang, 1991: 273. 
Baltic Region 
(Eocene/ Oligocene) 
Chrysopilus megacephalus Stuckenberg, 1965: 142 Madagascar 
Chrysopilus mexicanus Bellardi, 1861: 196 Mexico 
Chrysopilus modestus Loew, 1872: 58 USA 
Chrysopilus mojiangensis Yang & Yang, 1989: 281 China 
Chrysopilus montanorum Paramonov, 1962: 132 Australia 
Chrysopilus moramangensis Stuckenberg, 1965: 160 Madagascar 
Chrysopilus morimotoi Nagatomi, 1968: 44 Japan 
Chrysopilus mundus Stuckenberg, 1965: 130 Madagascar 
Chrysopilus mutabilis Stuckenberg, 1965: 147 Madagascar 
Chrysopilus nagatomii Yang & Yang, 1991: 273 China 
Chrysopilus nanus Williston, 1901: 265 Mexico 
Chrysopilus neimongolicus Yang & Yang, 1990: 289 China 
Chrysopilus nemoris Stuckenberg, 1965: 153 Madagascar 
Chrysopilus niger Bellardi, 1862: 27 Mexico 
Chrysopilus nigricauda Beling, 1873: 547 Austria 
Chrysopilus nigrifacies Nagatomi, 1968: 44 Japan 
Chrysopilus nigrimaculatus Yang & Yang, 1991: 92 China 
Chrysopilus nigrimarginatus Yang & Yang, 1990: 281 China 
Chrysopilus nigripalpis Bezzi, 1912: 448 Formosa 
Chrysopilus nigrocinctus Brunetti, 1927: 297 Malaya 
Chrysopilus ningminganus Yang & Yang, 1993: 51 China 
Chrysopilus nitidiventris Tonnoir, 1927: 105 New Zealand 
Chrysopilus niveofarinosus Frey, 1954: 19 Philippines 
Chrysopilus nobilipennis Frey, 1954: 16 Philippines 
Chrysopilus norrisi Paramonov, 1962: 124 Australia 
Chrysopilus nubecula (as Leptis; Fall?n, 1814: 9) Sweden 
Leptis auricollis Meigen, 1820: 103 Europe (?Harz?) 
Chrysopilus nudus Cresson, 1919: 177 USA 
Chrysopilus obscuralatus Yang & Yang, 1989: 245 China 
Chrysopilus ningxianus Yang & Yang, 1991: 94 China 
 
 339
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus obscuratus de Meijere, 1914: 1630 Java 
Chrysopilus obscuribarbus (as Chrysopila; Loew) 1869: 99 'Central Asia' 
Chrysopilus obscuripes Speiser, 1923: 98 Malaya 
Chrysopilus occidentalis Kerr, new name USA 
Chrysopilus lucifer Adams, 1904: 437. Preoccupied 
by Chrysopilus lucifer Walker, 
1852: 164. 
USA 
Chrysopilus okutanii Nagatomi, 1968: 49 Japan 
Chrysopilus opacifrons de Meijere, 1911: 288 Java 
Chrysopilus opalescens Brunetti, 1920: 134 Ceylon 
Chrysopilus opalizans de Meijere, 1913: 49 Indonesia 
Chrysopilus ornatipennis (as Chrysophilus; Brunetti, 1909: 
212) 
India 
Chrysopilus ornatus (as Leptis; Say, 1823: 34) USA 
Chrysopilus pallipes (as Chrysopila; Loew, 1869: 54) Greece 
Chrysopilus pallipilosus Yang & Yang, 1992: 354 China 
Chrysopilus palparis Loew, 1869: 50 Greece 
Chrysopilus panamensis Curran, 1931: 2 Panama 
Chrysopilus parvus Yang, Yang, and Nagatomi, 
1997: 161 
China 
Chrysopilus pele Kerr, new name Brazil 
Chrysopilus fascipennis Bromley, 1931: 8. Preoccupied 
by Chrysopilus fascipennis 
(Brunetti, 1920: 123). 
Brazil 
Chrysopilus peruanus Kert?sz, 1902: 149 Peru 
Chrysopilus philippii Lindner, 1924: 70 Peru 
Chrysopilus pilosus Leonard, 1930: 152 USA 
Chrysopilus pingquanus Yang, Yang, and Nagatomi, 
1997: 163 
China 
Chrysopilus pingxianganus Yang & Yang, 1992: 353 China 
Chrysopilus plebeius Williston, 1901: 264 Mexico 
Chrysopilus poecilapterus (as Chrysophilus; Bezzi, 1912: 
450) 
Taiwan 
Chrysopilus stigmaticus Statz, 1940: 129 Germany 
Chrysopilus praetiosus Loew, 1869: 55 Greece 
Chrysopilus propinquus Kert?sz, 1902: 146 Mexico 
Chrysopilus proximus (as Leptis; Walker, 1848: 214) USA 
Leptis propinquus Walker, 1848: 215 USA 
Chrysopilus puella Williston, 1901: 265 Mexico 
Chrysopilus pullus Loew, 1869: 43 Germany 
Chrysopilus pusilla (as Atherix; Macquart 1855: 88) Australia 
Chrysopilus quadratus (as Leptis; Say, 1823: 35) USA 
Leptis fumipennis Say, 1823: 37 USA 
Leptis reflexus Walker, 1848: 216 USA 
 
 340
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus dispar Wulp, 1867: 143 USA 
Chrysopilus flavidus Bigot, 1887: 104 USA 
Leptipalpis limbipennis Bigot, 1887: 107 USA 
Leptipalpis obscuripennis Bigot, 1887: 107 USA 
Chrysopilus rhagiodes Bromley, 1931: 8 Panama 
Chrysopilus rotundipennis Loew, 1861: 317 USA 
Chrysopilus rufipes Macquart, 1850: 103. NOTE 7 Australia 
Chrysopilus ruiliensis Yang & Yang, 1990: 280 China 
Chrysopilus saffranus (as Leptipalpis; Bigot, 1887: 108) Chile 
Chrysopilus sauteri Bezzi 1907: 564 Taiwan 
Chrysopilus matsumurai Nagatomi, 1968: 42 Japan 
Leptis basalis Matsumura, 1915: 39 Japan 
Chrysopilus schnusei Lindner, 1924: 74 Peru 
Chrysopilus segmentatus Brunetti, 1909: 430 Nepal 
Chrysopilus sericeus Bromley, 1931: 9 Guyana 
Chrysopilus shaanxiensis Yang & Yang, 1989: 244 China 
Chrysopilus shananus (as Chrysophilus; Frey, 1954: 21) Burma 
Chrysopilus shibuyai Nagatomi, 1968: 51 Japan 
Chrysopilus siculus Loew, 1869: 49 Italy 
Chrysopilus sigillatus Lindner, 1930: 65 Costa Rica 
Chrysopilus silvaticus Nagatomi, 1968: 53 Japan 
Chrysopilus silvicola Nagatomi, 1968: 54 Japan 
Chrysopilus similis Brunetti, 1920: 138 Ceylon 
Chrysopilus simillimus de Meijere, 1914: 29 Java 
Chrysopilus simplex de Meijere, 1904: 97 Java 
Chrysopilus smaragdinus Kert?sz, 1902: 145 Peru 
Chrysopilus sogai Stuckenberg, 1965: 120 Madagascar 
Chrysopilus sordidus Brunetti, 1920: 143 India 
Chrysopilus chlorophthalmus Loew, 1840: 4 not given; Europe 
Chrysopilus splendidus Meigen, 1820: 102 Germany 
Chrysopilus squamithorax Brunetti, 1927: 297 Malaya 
Chrysopilus stigma Brunetti, 1909: 432 Burma 
Chrysopilus stigmatias (as Leptipalpus; Bigot, 1887: 
106) 
USA 
?Chrysopilus stigmaticus Cockerel, 1921: 471 England 
(Eocene/Oligocene) 
Chrysopilus strigipennis de Meijere, 1914: 26 Java 
Chrysopilus stylatus Walker, 1864: 208 Indonesia 
Chrysopilus subaquilis Nagatomi, 1968: 56 Japan 
Chrysopilus sucini Stuckenberg, 1965: 138 Madagascar 
Chrysopilus suomianus (as Achrysopilus; Szil?dy, 1934: 
256) 
Finland 
Chrysopilus superbus Stuckenberg, 1965: 140 Madagascar 
Chrysopilus tanakai Nagatomi, 1978: 451 Japan 
 
 341
Species (synonyms indented) Species reference Species Type 
Country 
Chrysopilus tasmaniensis White, 1914: 40 Australia 
Chrysopilus tenggeranus Frey, 1934: 308 Java 
Chrysopilus testaceipes Bigot, 1887: 105 USA 
Chrysopilus bellus Adams, 1904: 438 USA 
Chrysopilus testaceus Loew, 1858: 367 South Africa 
Chrysopilus rhodesiensis Bruggen, 1960: 297 Zimbabwe 
Chrysopilus thoracicus (as Leptis; Fabricius, 1805: 70) USA 
Chrysopilus tomentosus Bigot, 1887: 104 USA 
Chrysopilus tonnoiri Paramonov, 1962: 126 Australia 
Chrysopilus torrentium Thomas, 1978: 311 France 
Chrysopilus trifasciatus Walker, 1860: 284 Mexico 
Chrysopilus trimaculatus Yang & Yang, 1989: 245 China 
Chrysopilus tsacasi Thomas, 1979: 136 Morocco 
Chrysopilus tuckerei Bezzi, 1929: 320 South Africa 
Chrysopilus turkestanus Lindner, 1931: 85 Turkestan 
Chrysopilus ugensis Nagatomi, 1968: 59 Japan 
Chrysopilus ungaranensis de Meijere, 1911: 291 Java 
Chrysopilus unicolor Brunetti, 1909: 432 India 
Chrysopilus unicus Curran, 1931: 3 Panama 
Chrysopilus vacillans Walker, 1858: 89 Indonesia 
Chrysopilus vadoni Stuckenberg, 1965: 165 Madagascar 
Chrysopilus valdivianus Philippi, 1865: 774 Chile 
Chrysopilus varius Kert?sz, 1902: 150 Peru 
Chrysopilus velutinus Loew, 1861: 316 USA 
Chrysopilus vespertinus Stuckenberg, 1965: 145 Madagascar 
Chrysopilus villosissimus Paramonov, 1962: 129 Australia 
Chrysopilus virtuosus Nagatomi, 1958: 33 Japan 
Chrysopilus waigiensis (as Leptipalpus; Bigot, 1887: 
108) 
Indonesia 
Chrysopilus wirthi Stuckenberg, 1997: 241 South Africa 
Chrysopilus xanthocromus Yang & Yang, 1990: 280 China 
Chrysopilus xanthopus Hardy, 1949: 163 USA 
Chrysopilus xizangensis Yang & Yang, 1991: 93 China 
Chrysopilus yerburyi Brunetti, 1920: 139 Ceylon 
Chrysopilus yezonis Nagatomi, 1968: 61 Japan 
Chrysopilus yunnanensis Yang & Yang, 1990: 279 China 
Chrysopilus zanjensis Stuckenberg, 1965: 154 Madagascar 
 
nomina dubia Author, reference 
Musca asiliformis Preyssler, 1791: 99 
Leptis cristatus Fabricius, 1775: 782 
 
 
 342
Genus DESMOMYIA Brunetti 
Figures 66-68, 182. 
DESMOMYIA Brunetti, 1912: 462.  Type species Desmomyia thereviformis 
Brunetti, by original designation. 
DIAGNOSIS.  
Desmomyia is characterized by having the antennal scape elongated, clearly longer 
than the pedicel; and the male hind first tarsomere enlarged. Both of these character 
states are autapomorphies for the genus.  
 
Species of Desmomyia are mid-sized flies (5.0 to 6.7 mm) of gray, black, or brownish 
coloration; legs with some yellow or concolorous dark brown to black; wings lightly 
infuscate, with light markings; male holoptic (eyes widely separated in female); 
laterotergite setose, wing vein M
3
 present, tibial spur formula 0:2:2, and hind tibia 
with or without short macrochaetae. Desmomyia is restricted to India and China. It is 
most likely to be confused with Rhagio, which is very similar in general appearance, 
and overlaps Desmomyia in its geographic distribution. The males of Desmomyia are 
distinguished by the autapomorphic characters of the antenna and hindleg mentioned 
above and may also be separated from nearly all Rhagio males by having pronounced, 
swollen parafacials. Females may be separated reliably from Rhagio by having the 
scape longer than pedicel. Desmomyia is distinguished from Chrysopilus by having 
two hind tibial spurs and by lacking thoracic setae that have a metallic sheen, in 
addition to the autapomorphic character states mentioned above. 
  
 
 343
DESCRIPTION.  
Head. Clypeus bulbous. Scape clearly larger than pedicel. First flagellomere laterally 
compressed, enlarged basally, bearing fused or distinct arista-like extension. Eyes 
inconspicuously setulose; in male, ommatidia evenly distributed, of equal size, or 
ommatidia split into dorsal and ventral areas and smaller ventrally, holoptic, not 
flattened dorsally. Parafacials in male swollen. Labella with pseudotrachae, longer or 
about the same size as palps. Theca short and stout. Palps one-segmented. Thorax. 
Mesonotum with or without vittae. Dorsocentral bristles absent, all dorsal setae of 
equal length. Anepisternum setulose on dorsal and posterior margins. Laterotergite 
katatergite swollen, differentiated from anatergite. Laterotergite setose, on katatergite 
only. Postspiracular scale absent. Proscutellum absent. Subscutellum not enlarged nor 
lengthened; inconspicuous. Wing hyaline, without markings. Pseudostigma present. 
Lower calypter reduced. Upper calypter well developed, with broad curvature, lobe-
like, width twice length or less. Costa extends to wingtip. Humeral crossvein (h) well 
developed. Sc-r crossvein present, weakly developed; positioned distal to the humeral 
crossvein (h), by the approximate length of h. Dorsal side of R
1
 setulose, ventral side 
of R
1
 bare. All other wing veins without setulae. Apical third of R
2+3
 ultimately bends 
toward leading edge of wing margin (creating concave flex anteriorly). Length of 
R
2+3
 about the same length as R
5
. Base of R
4
-R
5
 fork proximal to or directly above 
distal end of cell dm. R
4
 at base relaxed, not strongly curved; nearly straight apically. 
R
4
 and R
5
 encompass wing tip. R
5
 clearly longer than R
4+5
 (r-m to R
4
 origin). R-m 
crossvein proximal to one-third of discal cell. Origin of CuA
1
 at bm cell. M
3
 wing 
vein present. Wing cell m3 convergent at margin. Length of CuA
2
 v. posterior vein of 
 
 344
bm cell less than 1/2 length of posterior vein of bm cell. Alula narrow curvature, 
rounded evenly. Anal lobe well developed. Anal cell closed. Halter knob 
approximately 1/2 length of stem. Tibial spur formula 0:2:1. Hind coxal tubercle 
absent. Hind tibial macrochaetae present, short. First hind metatarsus of male 
swollen. Abdomen. Terminal Abdomen. Abdominal segments 5?10 evenly tapered 
from segments 1?4. In female, tergite 7 much longer than wide; intersegmental 
membrane between segments 7 and 8 especially long. Sternite 8 sclerite entire, not 
divided into two segments, length wider than long or as wide as long. Male terminalia 
with epandrium simple, not containing hypandrium ventrally. Epandrial sclerite wider 
than long, modestly curved anteriorly. Subepandrial sclerite absent. Hypoproct 
tomentose, setae absent. Cercus base held underneath epandrial sclerite. Cerci 
partially displaced from one another, separation distance approximately half the width 
of single cercus. Cerci, in posterior view flat, held in horizontal orientation. 
Hypandrium separated partially from the gonocoxites by an incomplete suture. 
Gonocoxite with sinuous dorsal ridge, leading to gonocoxal apodeme. Gonocoxal 
apodemes short or long enough to reach anterior margin of hypandrium. Parameral 
sheath bulbous, without paired swellings ventrally. Lateral ejaculatory processes 
present, integrated into sperm sac membrane. Ejaculatory apodeme moderately long, 
reaching anterior margin of hypandrium or long, reaching beyond anterior margin of 
hypandrium; laterally compressed. Aedeagal tines absent. Endoaedeagal process 
present. Female terminalia with three spermathecae, elliptical, lightly sclerotized. 
Spermathecal ducts longer than five times the length of sternite 9, but not so long as 
to be difficult to measure; not inflated at base of spermathecae. Spermathecal duct 
 
 345
accessory glands absent. Ejection apparatus of spermathecal ducts not sclerotized, 
without surface furrows. Spermathecal duct junction thickened. Common 
spermathecal duct thickened; short, shorter than longest diameter of genital chamber. 
Genital chamber oval, moderately sized. Accessory gland posterior to genital 
chamber inconspicuous, easily overlooked even after staining. Sternite 9 anterior end 
pointed; posterior end with broad lateral extensions which are joined medially with 
seam, in the vertical plane. Tergite 10 not greatly reduced. Sternite 10 entire, roughly 
pentagonal, pointed posteriorly; posterior half below first cercus segment. Cercus 
two-segmented. First segment of cercus not elongate, without ventral process. Ventral 
portions of first segment of cercus do not curve ventrally, towards one another, to 
form a ring. Basal cerci adjacent dorsally. Second cercus segment not elongated, 
without apical sensory pits. Larva. The larvae are unknown. 
LITERATURE. Yang et al. (1997) illustrate the species from China. 
List of included species 
Species (synonyms indented) Author, reference Type country 
Desmomyia sinensis Yang & Yang, 1997: 181 China 
Desmomyia thereviformis Brunetti, 1912: 462 India 
 
Genus LITOLEPTIS Chillcott 
Figures 69-72. 
LITOLEPTIS Chillcott, 1963: 1186.  Type species Litoleptis alaskensis Chillcott, by 
original designation. NOTE 8. 
DIAGNOSIS.  
 
 346
Several autapomorphic features support the monophyly of Litoleptis. The wing lacks 
the medial crossvein (so that the discal cell is absent); all tibiae are without spurs; the 
male aedeagus is very weakly developed, lacking an aedeagal apodeme; and the 
gonocoxal apodemes are also absent. Unfortunately, the female terminalia have not 
been described in the literature and I have been unable to examine any females 
personally, so these characters remain unknown and its phylogenetic placement 
remains uncertain.  
 
Litoleptis is composed of species that are small (1.8 to 2.7 mm); black in color; with 
hyaline wings; dichoptic eyes in the male; antenna bearing a single, elongated, 
tapering flagellomere; bare laterotergite; discal cell and wing vein M
3
 absent; and 
tibial spur formula of 0:0:0. This genus also shows an unusual disjunct distribution; 
Litoleptis species are found in Alaska, Chile, and the Philippines. This is among the 
most distinctive of rhagioniform Diptera, however it is most likely to be confused 
with Hilarimorpha, an asiloid either placed in its own family, Hilarimorphidae, or 
within Bombyliidae. Litoleptis differs from Hilarimorpha in having a pulvilliform 
empodium; antenna bearing a single, tapering flagellomere; and wing veins R
5
 and 
R
4+5
 subequal in length. Litoleptis may also be confused with Spania on account of its 
similar size and the shape of its antenna. Litoleptis may be distinguished from other 
small flies within Tabanomorpha, such as Bolbomyia and Spania, by the absence of 
the discal cell of the wing, dorsal surface of wing membrane bare, lack of tibial spurs, 
and eyes in male dichoptic.  
 
 
 347
DESCRIPTION.  
Head. Clypeus slightly bulbous. Pedicel clearly larger than scape. First flagellomere 
laterally compressed or rounded in cross section. First flagellomere of antenna 
elongate as long tapering segment or oval and enlarged near base, with straight, 
tapering stylus. Eyes inconspicuously setulose; in male, dichoptic (but not widely 
separated), not flattened dorsally, ommatidia split into dorsal and ventral areas and 
smaller ventrally. Labella without pseudotrachae, about the same size as palps. Theca 
short and stout. Palps one-segmented. Tentorium low on the face. Mandibles absent. 
Cibarial pump long, narrow (clearly not as wide as long). Thorax. Mesonotum 
without vittae. Anepisternum bare. Laterotergite bare. Postspiracular scale absent. 
Proscutellum absent. Subscutellum not enlarged nor lengthened; inconspicuous. Wing 
hyaline, without markings, without pseudostigma. Lower calypter reduced. Upper 
calypter triangular in form, underdeveloped. Upper calypter with broad curvature, 
lobe-like, width twice length or less. Costa stops before wingtip or extends past 
wingtip. Humeral crossvein (h) weakly developed. Sc-r crossvein present, well 
developed, positioned at proximal side of humeral crossvein (h), by less than length 
of h. All wing veins and cells bare. R
2+3
 sinuous, apical third of R
2+3
 ultimately bends 
toward leading edge of wing margin (creating concave flex anteriorly); longer than 
R
5
, but less than twice as long.Base of R
4
-R
5
 fork distal of distal end of cell dm. R
4
 at 
base nearly straight entire length. R
5
 anterior to, posterior to, or ending at wing tip; 
about as long as R
4+5
 (r-m to R
4
 origin). M
3
 wing vein absent. M-cu crossvein absent. 
Discal cell absent. CuA
2
 greater than 2/3 length of posterior vein of bm cell. CuA
2
 
does not reach wing margin. Alula reduced, without curvature or with narrow 
 
 348
curvature. Alula full, rounded evenly. Anal lobe well developed. Halter knob between 
1/2?2/3 length of stem. Tibial spur formula 0:0:0. Hind coxal tubercle absent. Hind 
tibial macrochaetae absent. Postmetacoxal bridge absent. Abdomen. Male terminalia 
with epandrium simple, not containing hypandrium ventrally. Epandrial sclerite wider 
than long; strongly notched anteriorly. Subepandrial sclerite absent. Hypoproct 
triangular (rounded posteriorly); flattened, distinct from sclerites above cerci; with or 
without setae. Cercus displaced away from epandrial sclerite; widely displaced from 
one another, separation distance greater than three quarters width of cercus; held 
vertical in relation to rest of abdomen; in posterior view flat. Hypandrium fused 
entirely to gonocoxites; in L. alaskensis, with broad ventral processes separated by a 
central notch. Gonocoxite smooth dorsally, without sinuous ridge leading to 
gonocoxal apodeme. Gonocoxal apodemes absent. Parameral sheath not developed 
into bulbous sac or separate lobes. Lateral ejaculatory processes absent. Ejaculatory 
apodeme reduced, nearly absent. Ejaculatory apodeme laterally compressed. 
Aedeagal tines absent. Endoaedeagal process absent. Female unavailable and remains 
undescribed. Larva. The larvae are unknown. 
LITERATURE. Chillcott (1963) illustrated the head, wing, and male genitalia of 
L. alaskensis.  Hennig (1972) compared the three species currently described in the 
genus, and illustrated the head, wing and male genitalia of L. chilensis.  
List of included species 
Species (synonyms indented) Author, reference Type country 
Litoleptis alaskensis Chillcott, 1963: 1187 USA 
Litoleptis chilensis Hennig, 1972: 6 Chile 
Litoleptis orientalis (as Hilarimorpha; Frey, 1954: 25) Philippines 
 
 
 349
Genus OMPHALOPHORA Becker 
Figures 73, 74A, 75, 76A, 77-83. 
OMPHALOPHORA Becker 1900: 12. Type species Omphalophora oculata Becker, 
1900: 13, by monotypy. 
  
DIAGNOSIS.  
The monophyly of Omphalophora is supported by the unique form of the female 
tergite 9, which is distinctively bulbous and tapers posteriorly. Partially sclerotized 
lobes located in the membrane between the ninth tergite and ninth sternite may also 
be a synapomorphy for the group. Omphalophora and Ptiolina are very similar in 
their antennal form and general habitus, but phylogenetic analysis reveals that they 
form a paraphyletic grade when grouped together. Therefore, the recognition of 
Omphalophora Becker is merited. 
 
Omphalophora species are delicate to fairly robust flies, small to moderately sized 
(3.0 to 10.0 mm) that are entirely black or brown in color. Wings are hyaline or 
infuscate near wing veins; male holoptic, eyes separated in female; antenna with 
unsegmented terminal stylus, round or lightly laterally compressed in cross section; 
mandibles absent; laterotergite bare; wing vein M
3
 present; tibial spur formula 0:2:1; 
hind tibia without macrochaetae; tergite 9 with ventrolateral arms, extending 
posteriorly, surrounding and fusing to sternite 9 laterally; female spermathecal ducts 
with accessory glands. Omphalophora is restricted to the Holarctic Region. 
Omphalophora is easily confused with Ptiolina. It may be distinguished from this 
 
 350
genus by having setose anepisternum, wing veins R
4+5
 containing the wing tip, the 
female 7
th
 tergite clearly longer than wide (in Ptiolina, this tergite is clearly wider 
than long), and in the male, the subepandrial sclerite is as wide as long or nearly oval 
and the gonostylus comes to a sharp point apically. Species of Omphalophora may be 
confused with Symphoromyia but is immediately distinguished from this genus by 
having the scape approximately the same size as the pedicel, first flagellomere longer 
than wide (elongated anteriorly, not reniform), and bare laterotergite. Omphalophora 
may be distinguished from Bolbomyia by the unsegmented style, by lacking fore 
tibial spurs, and by having wing vein M
3
 present. Omphalophora is usually 
significantly larger than Spania and also differs by having a hind tibial spur. 
 
DESCRIPTION.  
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere lightly compressed laterally or rounded in cross section (may vary from 
specimen to specimen), bearing stylus of single segment. Eyes inconspicuously 
setulose; in female, dichoptic; in male, holoptic, not strongly flattened dorsally, 
ommatidia evenly distributed, gradually smaller toward ventral margin. Labella with 
pseudotrachae, length variable. Theca short and stout, with medial suture. Palps one- 
or two-segmented. When two-segmented, distal palp segment longer than proximal 
segment. Stipes surrounded by membrane above theca, directed posteriorly. Lacinia 
present, shorter than palps, not serrated at tip. Mandibles absent. Cibarial pump short, 
as long as wide or slightly longer than wide. Cornu shorter than cibarial pump. 
Pharyngeal pump narrow along most of length, mostly flat, approximately same 
 
 351
length as cibarial pump. Thorax. Mesonotum usually with vittae, but may be without. 
Dorsocentral bristles absent; all dorsal setae of equal length. Anepisternum haired 
throughout posterior half. Laterotergite bare. Proscutellum usually present. 
Subscutellum not enlarged nor lengthened; inconspicuous. Wing hyaline or lightly 
infuscate, without markings. Costa extends to R
5
, at wingtip. Lower calypter reduced. 
Upper calypter well developed, with broad curvature, lobe-like, width twice length or 
less. Humeral crossvein (h) well developed. Sc-r crossvein present, weakly 
developed, positioned distal to the humeral crossvein (h), by approximate length of h 
or less. Dorsal side of R
1
 setulose, ventral side bare. All other wing veins without 
setulae. Wing veins R
1
 and R
2+3
 separated at wing margin. R
2+3
 sinuous, apical third 
of R
2+3
 ultimately bends toward leading edge of wing margin, creating concave flex 
anteriorly. Base of R
4
-R
5
 fork proximal or directly above distal end of cell dm. R
4
 
nearly straight apically or curving slightly towards anterior margin. R
4
 and R
5
 
encompass wing tip. R
5
 clearly longer than R
4+5
 (r-m to R
4
 origin). R-m crossvein at 
proximal one-fifth to near halfway of discal cell. M
3
 wing vein present. M-cu 
crossvein present. Origin of CuA
1
 at bm cell. CuA
2
 greater than 1/2 length, less than 
2/3 length of posterior vein of bm cell. Anal lobe well developed. Alula full, rounded 
with broad curvature. Anal cell closed. Halter knob between 1/2?2/3 length of stem. 
Tibial spur formula 0:2:1. Hind tibial spur short. Hind coxal tubercle absent. Hind 
tibial macrochaetae absent (long delicate setae sometimes present). Postmetacoxal 
bridge absent. Abdomen. Abdominal segments evenly tapered. In female, last 3 
abdominal segments telescoped; tergite 7 much wider than long; intersegmental 
membrane between segments 7 and 8 especially long; sternite 8 as wide as long or 
 
 352
wider than long. Male terminalia with epandrium simple, not containing hypandrium 
ventrally. Epandrial sclerite wider than long, modestly curved anteriorly. 
Subepandrial sclerite present, undivided, narrow (clearly wider than long), setose. 
Hypoproct present, setose. Cercus base held underneath epandrial sclerite, or directly 
adjacent to epandrial sclerite. Cerci displaced from one another, separation distance 
greater than three quarters width of cercus. Cerci, in posterior view flat or cupped, 
forming circular outline medially. Hypandrium fused entirely to gonocoxites. 
Gonocoxite with sinuous dorsal ridge, leading to gonocoxal apodeme. Gonocoxal 
apodemes short or long enough to reach anterior margin of hypandrium. Parameral 
sheath forming separate, distinct lobes ventrally. Lateral ejaculatory processes 
present, not part of sperm sac posteriorly. Ejaculatory apodeme moderately long, 
reaching anterior margin of hypandrium. Ejaculatory apodeme tubular. Aedeagal 
tines absent. Endoaedeagal process present; short, blunt, rounded. Female terminalia 
with tergite 9 entire, with narrow anteriorly-directed ventrolateral projections, 
enveloping sternite 9. Spermathecae three, clubbed, sclerotized. Spermathecal ducts 
more than three times but less than five times the length of sternite 9, not inflated at 
base of spermathecae. Spermathecal duct accessory glands present, arise at 
approximately the distal third of the spermathecal ducts. Spermathecal ducts near 
junction with common duct not sclerotized. Common spermathecal duct thickened; of 
moderate length, about the same length as the longest diameter of genital chamber. 
Genital chamber oval, moderately sized. Accessory gland posterior to genital 
chamber inconspicuous, easily overlooked even after staining. Accessory gland 
posterior to genital chamber common duct present with short paired extensions 
 
 353
posteriorly. Sternite 9 anterior end narrowly paddle-shaped, posterior end with broad 
extensions posteriorly, joined together in horizontal plane centrally, held in vertical 
plane laterally. Tergite 10 entire; short, length less than half width. Sternite 10 
sclerotization weakened centrally, making it appear as if sclerite divided into two 
lateral components. Sternite 10 roughly rectangular or ovoid, pointed posteriorly. 
Cercus two-segmented. First segment of cercus not elongate, without ventral process. 
Ventral lobes of first segment of cercus not curving ventrally towards one another to 
form a ring. Basal cerci separated from one another dorsally by approximately the 
width of the second cercal segment. Second cercus segment narrow, elongated (~3x 
longer than wide or more), with apical sensory pits. Larva. Unknown, probably very 
similar to Ptiolina.  
 
LITERATURE.  
Omphalophora species descriptions and treatments are almost entirely contained 
within the body of work covering the genus Ptiolina. Hardy and McGuire (1947) 
provide a key to North American species. Narchuk (1969) gives key to eastern 
European species. 
List of included species 
Species (synonyms indented) Author, reference Type country 
Omphalophora cinereofasciata (as Leptis; Schummel, 1837: 109) ?Sanderberg? 
Europe 
Ptiolina phragmitophila Schiner, 1868: 910 Austria 
Ptiolina calamodytes Schiner, 1868: 911 Slovakia 
Ptiolina fulva Becker, 1900: 110 Siberia 
Omphalophora fasciata (as Ptiolina; Loew, 1869: 164) Canada 
Omphalophora majuscula (as Ptiolina; Loew, 1869: 165) Canada 
Omphalophora nigripilosa (as Ptiolina; Hardy & McGuire, USA 
 
 354
Species (synonyms indented) Author, reference Type country 
1947: 9) 
Omphalophora oculata Becker, 1900: 106 Russia 
Omphalophora lapponica Frey, 1911: 16 Finland 
 
Genus PTIOLINA Zetterstedt 
Figures 74B, 76B, 84-96, 183-184, 188-192. 
PTIOLINA Zetterstedt, 1842: 226.  Type species Leptis obscura Fall?n, by 
subsequent designation of Frauenfeld, 1867: 497. NOTE 9.  
 EURYTION Jaennicke, 1867: 99.  Type species Eurytion paradoxus 
Jaennicke, 1867: 99, by monotypy. 
 TYOLINA Walker, 1848: 220.  Misspelling. 
 SPATULINA Szil?dy, 1942: 625. Type species Spatulina engeli Szil?dy, by 
monotypy. NOTE 10. 
DIAGNOSIS.  
The monophyly of Ptiolina is supported by the unique form of the female tergite 9 
which is rectangular and narrow, with ventrolateral arms that are easily distinguished 
from sternite 9.  
 
Ptiolina species are delicate to fairly robust flies, small to moderately sized (3.0 to 
10.0 mm), that are entirely black or brown in color. Wings are hyaline or infuscate 
near wing veins; male holoptic, eyes separated in female; antenna with unsegmented 
terminal stylus, round or laterally compressed in cross section; mandibles absent; 
laterotergite bare; wing vein M
3
 present; tibial spur formula 0:2:1; hind tibia without 
macrochaetae; tergite 9 with ventrolateral arms, extending posteriorly, surrounding 
 
 355
and fusing to sternite 9 laterally; female spermathecal ducts with accessory glands. 
Ptiolina is restricted to the Holarctic Region and although it is fairly distinctive for 
this region, it may be confused with Symphoromyia. Ptiolina may be immediately 
distinguished from this genus by having the scape approximately the same size as the 
pedicel, first flagellomere longer than wide (elongated anteriorly, not reniform), and 
bare laterotergite. Ptiolina may be distinguished from Bolbomyia by the unsegmented 
style, by lacking fore tibial spurs, and by having wing vein M
3
 present. Ptiolina is 
usually significantly larger than Spania and also differs by having a hind tibial spur. 
 
DESCRIPTION.  
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere enlarged, laterally compressed, bearing stylus of single segment. Eyes 
inconspicuously setulose; in female, dichoptic; in male, holoptic, not strongly 
flattened dorsally, ommatidia split into dorsal and ventral areas and smaller ventrally. 
Labella with pseudotrachae, length variable. Theca short and stout, with medial 
suture. Palps one- or two-segmented. When two-segmented, distal palp segment 
longer than proximal segment. Stipes surrounded by membrane above theca, directed 
posteriorly. Lacinia present, shorter than palps, not serrated at tip. Mandibles absent. 
Cibarial pump short, as long as wide or slightly longer than wide. Cornu shorter than 
cibarial pump. Pharyngeal pump narrow along most of length, mostly flat, 
approximately same length as cibarial pump. Thorax. Mesonotum with or without 
vittae. Dorsocentral bristles absent; all dorsal setae of equal length. Anepisternum 
bare. Laterotergite bare. Proscutellum present or absent. Subscutellum not enlarged 
 
 356
nor lengthened; inconspicuous. Wing hyaline or lightly infuscate, without markings. 
Costa extends to R
5
, at wingtip (in P. nitida, it extends past wingtip). Lower calypter 
reduced. Upper calypter well developed, with broad curvature, lobe-like, width twice 
length or less. Humeral crossvein (h) well developed. Sc-r crossvein present, weakly 
developed, positioned distal to the humeral crossvein (h), by approximate length of h 
or less. Dorsal side of R
1
 setulose, ventral side bare. All other wing veins without 
setulae. Wing veins R
1
 and R
2+3
 separated at wing margin. R
2+3
 sinuous, apical third 
of R
2+3
 ultimately bends toward leading edge of wing margin, creating concave flex 
anteriorly. Length of R
2+3
 shorter than or about the same length as R
5
. Base of R
4
-R
5
 
fork proximal or directly above distal end of cell dm. R
4
 nearly straight apically or 
curving slightly towards anterior margin. R
5
 ending anterior to or at wing tip, clearly 
longer than R
4+5
 (r-m to R
4
 origin). R-m crossvein at proximal one-fifth to near 
halfway of discal cell. M
3
 wing vein present. M-cu crossvein present. Origin of CuA
1
 
at bm cell. CuA
2
 greater than 1/2 length, less than 2/3 length of posterior vein of bm 
cell. Anal lobe well developed. Alula full, rounded with broad curvature. Anal cell 
closed. Halter knob between 1/2?2/3 length of stem. Tibial spur formula 0:2:1. Hind 
tibial spur short. Hind coxal tubercle absent. Hind tibial macrochaetae absent (long 
delicate setae sometimes present). Postmetacoxal bridge absent. Abdomen. 
Abdominal segments evenly tapered. In female, last 3 abdominal segments 
telescoped; tergite 7 much wider than long; intersegmental membrane between 
segments 7 and 8 especially long; sternite 8 as wide as long or wider than long. Male 
terminalia with epandrium simple, not containing hypandrium ventrally. Epandrial 
sclerite wider than long, strongly notched anteriorly. Subepandrial sclerite present, 
 
 357
undivided, narrow (clearly wider than long), setose. Hypoproct present, setose. 
Cercus base held underneath epandrial sclerite, or directly adjacent to epandrial 
sclerite. Cerci displaced from one another, separation distance greater than three 
quarters width of cercus. Cerci, in posterior view cupped, forming circular outline 
medially. Hypandrium fused entirely to gonocoxites. Gonocoxite with sinuous dorsal 
ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short or long enough to 
reach anterior margin of hypandrium. Parameral sheath forming separate, distinct 
lobes ventrally. Lateral ejaculatory processes present, not part of sperm sac 
posteriorly. Ejaculatory apodeme short or moderately long, not reaching anterior 
margin of hypandrium. Ejaculatory apodeme rod-shaped or laterally compressed. 
Aedeagal tines absent. Endoaedeagal process absent or present (when present, 
reduced). Female terminalia with tergite 9 entire, with narrow anteriorly-directed 
ventrolateral projections, enveloping sternite 9. Spermathecae three, clubbed, 
sclerotized. Spermathecal ducts no more than three times the length of sternite 9, not 
inflated at base of spermathecae. Spermathecal duct accessory glands present, arise at 
the base of each spermatheca. Spermathecal ducts near junction with common duct 
sclerotized, thickened, with furrows. Common spermathecal duct thickened; short, 
shorter than longest diameter of genital chamber. Genital chamber oval, moderately 
sized. Accessory gland posterior to genital chamber inconspicuous, easily overlooked 
even after staining. Accessory gland posterior to genital chamber common duct 
present with short paired extensions posteriorly. Sternite 9 anterior end narrowly 
paddle-shaped, posterior end with broad extensions posteriorly, joined together in 
horizontal plane centrally, held in vertical plane laterally. Tergite 10 entire; short, 
 
 358
length less than half width. Sternite 10 sclerotization weakened centrally, making it 
appear as if sclerite divided into two lateral components. Sternite 10 roughly 
pentagonal, pointed posteriorly (very broad, wider than long; nearly rectangular). 
Cercus two-segmented. First segment of cercus not elongate, without ventral process. 
Ventral lobes of first segment of cercus not curving ventrally towards one another to 
form a ring. Basal cerci separated from one another dorsally by approximately the 
width of the second cercal segment. Second cercus segment narrow, elongated (~3x 
longer than wide or more), with apical sensory pits. Larva. Body with 11 segments. 
Thoracic segments with creeping welts ventrally. Head capsule not folded within 
second segment. Head capsule  composed of a single, undivided plate (dorsal plate). 
Head capsule less than 4.5 times longer than greatest width (1.5 width/5.0 length). 
Mandibular brush absent. Mandibular hook without groove or canal, smooth, without 
microsetae. Labral teeth very well developed, heavily sclerotized, in two rows, 
converging anteriorly. Maxilla sclerotized (strongly toothed, with three teeth). Saw 
sclerite of mandibular base absent. Maxillary palp soft, segments poorly 
differentiated, three-segmented. Antenna one-segmented. Antenna last segment 
entire, dome-shaped, short. Unpaired salivary pump absent. Posterior tentorial 
expansion fused to each other posteriorly, with thin extension produced anteriorly.  
 
LITERATURE.  
Hardy and McGuire (1947) provide a key to North American species. Narchuk (1969) 
gives key to eastern European species. 
 
 359
List of included species 
Species tentatively placed in the genus are indicated by an asterisk (*). These species 
require inspection to verify their placement. 
 
Species (synonyms indented) Author, reference Type country 
Ptiolina alberta* Leonard in Curran, 1931: 259 Canada 
Ptiolina attenuata Nagatomi, 1986: 311 Japan 
Ptiolina augusta* Curran, 1931: 249 Canada 
Omphalophora cinereofasciata (as Leptis; Schummel, 1837: 109) ?Sanderberg? 
Europe 
Ptiolina phragmitophila Schiner, 1868: 910 Austria 
Ptiolina calamodytes Schiner, 1868: 911 Slovakia 
Ptiolina fulva Becker, 1900: 110 Siberia 
Ptiolina dudai* Lindner, 1942: 240 Austria 
Ptiolina edeta (as Spania; Walker, 1849: 489) Canada 
Atherix vidua Walker, 1849: 1153 Canada 
Ptiolina grisea Curran, 1931: 251 USA 
Omphalophora fasciata (as Ptiolina; Loew, 1869: 164) Canada 
Ptiolina grandis Frey, 1918: 31 Russia 
Ptiolina arctica Becker, 1921: 62 Russia 
Ptiolina latifrons* Nagatomi, 1986: 312 Taiwan 
Ptiolina longipilosa* Nagatomi, 1986: 314 Japan 
Omphalophora majuscula (as Ptiolina; Loew, 1869: 165) Canada 
Ptiolina mallochi Hardy & McGuire, 1947: 8 USA 
Ptiolina arctica Malloch, 1923: 181. Preoccupied by 
Ptiolina arctica Becker, 1921: 62. 
USA 
Ptiolina nervosa* Nagatomi, 1986: 317 Japan 
Omphalophora nigripilosa (as Ptiolina; Hardy & McGuire, 
1947: 9) 
USA 
Ptiolina nitida Wahlberg, 1854: 215 Norway 
Ptiolina nitidifrons Hardy & McGuire, 1947: 10 USA 
Ptiolina obscura (as Leptis; Fall?n, 1814: 11) Sweden 
Ptiolina nigra Zetterstedt, 1842: 227 Sweden 
Ptiolina nigrina Wahlberg, 1854: 215 Sweden 
Ptiolina nigripes Zetterstedt, 1859: 4975 Sweden 
Tyolina tristis Walker, 1949: 220 Great Britain 
Ptiolina obsoleta (as Ptiolina; Leonard in Curran, 
1931: 250) 
USA 
Omphalophora oculata Becker, 1900: 106 Russia 
Omphalophora lapponica Frey, 1911: 16 Finland 
Ptiolina uralensis Becker 1921: 62 Russia 
Ptiolina paradoxa* (as Eurytion; Jaennicke, 1866: 99) Switzerland 
Ptiolina lapidaria Nowiczki, 1868: 74 Poland 
Ptiolina wodzickii Frauenfeld, 1867: 497 not given; Europe 
Spania grisea Strobl, 1892: 124 Austria 
 
 360
Species (synonyms indented) Author, reference Type country 
Ptiolina pelliticornis* Becker, 1900: 113 Italy 
Ptiolina shimai Nagatomi, 1985: 211 Nepal 
Ptiolina sphaeralis* Nagatomi, 1986: 320 Japan 
Ptiolina vicina Hardy & McGuire, 1947: 12 USA 
Ptiolina zonata Hardy & McGuire, 1947: 13 USA 
 
Genus RHAGIO Fabricius 
Figures 97-111, 176, 193. 
RHAGIO Fabricius, 1775: 761.  Type species Musca scolopacea Linnaeus, 1758, by 
subsequent designation of Latreille, 1810: 443. 
 LEPTIS Fabricius, 1805: 69 (unjustified emendation for Rhagio Fabricius, 
1775).  Type-species Musca scolopacea Linnaeus, 1758, 
automatic. 
 ?PALAEOHILARIMORPHA Meunier, 1902: 400. Type species 
Palaeohilarimorpha bifurcata Meunier, 1902, by monotypy. 
 RHAGINA Malloch, 1932: 117. Type species Leptis incurvata de Meijere, 
1911. New synonymy. NOTE 11. 
 RHAGIONELLA Szil?dy, 1934: 239 (as subgenus). Type species Nemotelus 
maculatus De Geer, 1776, by original designation. 
 RHAGIELLA Szil?dy, 1934: 240 (as subgenus). Type species Rhagio lineola 
Fabricius, 1794, by original designation. 
DIAGNOSIS.  
The monophyly of the species of Rhagio is supported by a unique feature found in the 
larva. All Rhagio larvae have a saw sclerite attached ventrally to the basal mandibular 
sclerite. The autapomorphic nature of this character state is evident among 
 
 361
tabanomorph larvae. However, the larva of many Rhagio species remain undescribed 
and the larva of putatively closely related genera such as Desmomyia and 
Atherimorpha, among many other rhagionid taxa, are also not known.  
 
Rhagio species are delicate to fairly robust flies, variably sized (4.2 to 18 mm); black, 
gray, brown, orange-brown, yellow or yellow and black. Wings are hyaline or 
infuscate, with or without markings; male holoptic or dichoptic, eyes separated in 
female; first flagellomere subcircular, laterally compressed, with terminal arista 
arising ventrally or from central position; mandibles absent; laterotergite setose; wing 
vein M
3
 present; tibial spur formula 0:2:2; hind tibia with short macrochaetae when 
present; tergite 9 without ventrolateral arms; female spermathecal ducts without 
accessory glands. Rhagio is distributed throughout the Holarctic Region. Rhagio 
species are most commonly confused with species of Chrysopilus, but may be 
distinguished by having two hind tibial spurs; an arista that is nearly bare; and a 
prominently setose proepimeron. In India, and perhaps in surrounding areas, Rhagio 
may be distinguished from local Desmomyia by having the scape approximately the 
same size as pedicel and in the male, first tarsomere not enlarged. Rhagio is very 
similar in form to Atherimorpha, although their distributions are not sympatric. 
Rhagio may be distinguished immediately from Atherimorpha by having an aristate 
antenna. 
  
DESCRIPTION.  
 
 362
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere. First flagellomere of antenna enlarged; basally rounded in cross section 
or laterally compressed, bearing fused or arista-like extension. Eyes inconspicuously 
setulose; in male, holoptic or dichoptic (dichoptic in female), not flattened dorsally; 
ommatidia evenly distributed of equal size or ommatidia larger dorsally with smooth 
transition to slightly smaller ommatidia ventrally. Labella with pseudotrachae, length 
variable. Theca short and stout, lateral thecal sclerites separated. Palps one-
segmented, often with constriction, making it appear that there are two palp segments. 
Stipes convergent toward one another medially or surrounded by membrane above 
theca, directed posteriorly. Cardo not swollen. Lacinia shorter than palps; tip not 
serrated. Mandibles absent. Cibarial pump long, clearly not as wide as long. Cornu 
nearly as long as or longer than cibarial pump. Pharyngeal pump narrow along most 
of length, mostly flat along its length, approximately same length as cibarial pump. 
Thorax. Mesonotum with or without vittae. Dorsocentral bristles absent, all dorsal 
setae of equal length. Anepisternum bare (R. maculatus De Geer, R. dichomaticus 
Chillcott), setulose on dorsal margin only, or throughout posterior half of sclerite. 
Laterotergite setose, on katatergite only. Postspiracular scale absent. Proscutellum 
present or absent. Subscutellum not enlarged nor lengthened; inconspicuous. Wing 
hyaline or lightly infuscate; with or without markings. Wing with or without 
pseudostigma. Lower calypter reduced. Upper calypter well developed, with broad 
curvature, lobe-like, width twice length or less. Costa ends before or approximately at 
wingtip Humeral crossvein (h) well developed. Sc-r crossvein present, well 
developed, positioned distal to the humeral crossvein (h), by more than the length of 
 
 363
h. Dorsal side of R
1
 setulose, ventral side bare. All other wing cells and veins bare. 
R
2+3
 nearly straight or sinuous; longer than but less than twice as long as R
5
; apical 
third ultimately bends either toward leading edge of wing margin (creating concave 
flex anteriorly) or toward wingtip (creating convex flex anteriorly). Base of R
4
-R
5
 
fork proximal of, directly above, or distal of distal end of cell dm. R
4
 at base strongly 
curved or angled; leads directly to wing margin or with short proximal offshoot at 
point of curvature near base; nearly straight or sinuous apically (as in R. tuberculatus 
(Yang et al., 1997: 245)); anterior to, ending at, or posterior to wing tip. R
5
 clearly 
longer than R
4+5
 or about as long as R
4+5
 (r-m to R
4
 origin).  M
3
 wing vein present. 
Wing cell m3 parallel-sided at margin. Origin of CuA
1
 at bm cell. CuA
2
 greater than 
1/2 length of posterior vein of bm cell and greater or less than 2/3 length of posterior 
vein of bm cell. Alula with narrow or broad curvature, rounded evenly. Anal lobe 
well developed. Anal cell open or closed. Halter knob approximately 1/2 length of 
stem. Tibial spur formula 0:2:2. Hind coxal tubercle absent or present. Hind femora 
with or without ventro-apical swelling. Hind tibial macrochaetae absent or present; 
when present, short. First hind metatarsus of male not swollen. Postmetacoxal bridge 
present. Postmetacoxal bridge reaching to internal base of metasternal furcum as 
incomplete thin extension. Abdomen. Terminal Abdomen. Abdominal segments 5?10 
evenly tapered from segments 1?4. In female, tergite 7 much longer than wide; 
intersegmental membrane between segments 7 and 8 especially long; sternite 8 length 
variable, wider than long to much longer than wide. Male terminalia with epandrium 
simple, not containing hypandrium ventrally. Epandrial sclerite wider than long, 
strongly notched anteriorly. Subepandrial sclerite absent. Hypoproct with or without 
 
 364
setae. Cercus attached to hypoproct, displaced away from epandrial sclerite; partially 
displaced from one another, separation distance approximately half the width of 
single cercus. Cerci, in posterior view flat. Hypandrium separated from the 
gonocoxites by a complete or incomplete suture. Gonocoxite with sinuous dorsal 
ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short or long enough to 
reach anterior margin of hypandrium. Parameral sheath bulbous, without paired 
swellings ventrally. Lateral ejaculatory processes present, integrated into sperm sac 
membrane. Ejaculatory apodeme long, reaching beyond anterior margin of 
hypandrium; laterally compressed. Aedeagal tines absent. Endoaedeagal process 
present, very reduced (as in R. plumbeus), or apparently absent (as in R. 
punctipennis). Female terminalia with three spermathecae, spherical or elliptical, 
lightly sclerotized or without sclerotization. Spermathecal ducts longer than five 
times the length of sternite 9, but not so long as to be difficult to measure. 
Spermathecal duct accessory glands absent. Ejection apparatus of spermathecal ducts 
thickened, lightly sclerotized, surface furrows that run at an angle. Spermathecal duct 
junction thickened. Common spermathecal duct thickened; short, shorter than longest 
diameter of genital chamber. Genital chamber oval, moderately sized. Accessory 
gland posterior to genital chamber inconspicuous, easily overlooked even after 
staining. Sternite 9 anterior end pointed; posterior end with broad lateral extensions, 
free, held in horizontal plane. Tergite 10 length aproximately equal to half measured 
width, or longer. Sternite 10 entire, roughly pentagonal, pointed posteriorly; posterior 
half below first cercus segment. Cercus two-segmented. First segment of cercus not 
elongate, with or without ventral process. Ventral lobes of first segment of cercus 
 
 365
curve ventrally towards one another to form a ring, visible in the posterior 
perspective. Basal cerci adjacent dorsally. Second cercus segment not elongated, with 
or without apical sensory pits. Larva. Body with 11 segments, amphipneustic. 
Thoracic segments with creeping welts ventrally. Head capsule not folded within 
second segment, composed of a single, undivided plate (dorsal plate); less than 4.5 
times longer than greatest width (2 width/ 5.5 length); not cone-shaped. Mandibular 
brush present, associated with simple fold of cuticle. Mandibular hook canal with 
apical opening. Hook serrate, transversely smooth. Stiff microsetae pointing 
anteriorly on dorsal ridge of mandibular hook absent. Labral teeth developed, 
sclerotized; in single row. Maxilla sclerotized (and thrice toothed, as in Ptiolina). Saw 
sclerite of mandibular base present. Maxillary palp soft, segments poorly 
differentiated. Maxillary palp segment number three. Antenna last segment entire 
(nub). Antenna three-segmented. Unpaired salivary pump absent. Posterior tentorial 
expansion free, with thin extension produced dorsally. 
 
LITERATURE.  
Chillcott (1965) gives a key to the species of Eastern North America. Narchuk (1969) 
provides a key to the species of Russia. Yang et al. (1997) give key to species of 
China. 
List of included species 
Species (synonyms indented) Author, reference Type country 
Rhagio albicornis (as Leptis; Say, 1823: 38) USA 
Xylophagus fasciatus Say, 1829: 155 Canada 
Leptis boscii Macquart, 1840: 30 USA 
Rhagio albipilosus Becker, 1921: 47 Turkey 
 
 366
Species (synonyms indented) Author, reference Type country 
Rhagio albus Yang, Yang, and Nagatomi, 1997: 
191 
China 
Rhagio algericus (as Leptis; Becker, 1906: 282) Algeria 
Rhagio alumnus Walker, 1852: 163 ?South America? 
Rhagio amurensis Makarkin, 1992: 255 Russia 
Rhagio annulatus (as Nemotelus; De Geer, 1776: 
164) 
 
Leptis conjugens (as Leptis; Ruthe, 1831: 1214) not given, prob. 
Sweden 
Rhagio apiciflavus Yang & Yang, 1991: 275 China 
Rhagio apicipennis (as Leptis; Brunetti, 1909: 423) India 
Rhagio arcuatus (as Leptis; de Meijere, 1911: 292) India 
Rhagio ardea Fabricius, 1794: 275 ?Europe? 
Rhagio asticta Yang & Yang, 1994: 32 China 
Rhagio balcanicus (as Leptis; Strobl, 1902: 475) Yugoslavia 
Rhagio basiflavus Yang & Yang, 1993: 48 China 
Rhagio basimaculatus Yang & Yang, 1993: 48 China 
Rhagio beckeri Lindner, 1923: 7 France 
? Rhagio bifurcatus (as Palaeohilarimorpha; Meunier, 
1902: 400) 
Baltic Region 
(Eocene/Oligocene) 
Rhagio biroi Szil?dy, 1934: 8. NOTE 12 India 
Rhagio bisectus Yang, Yang, and Nagatomi, 1997: 
200 
China 
Rhagio bitaeniatus (as Leptis; Bellardi, 1862: 26) Mexico 
Rhagio brunneipennis Leonard, 1930: 92 USA 
Rhagio calcaratus Statz, 1940: 128 Germany 
Rhagio californicus Leonard, 1930: 93 USA 
Rhagio cavannae (as Leptis; Bezzi, 1898: 28) Italy 
Rhagio centrimaculatus Yang & Yang, 1993: 47 China 
Rhagio chillcotti James, 1965: 333 Canada 
Rhagio choui Yang & Yang, 1997: 205 China 
Rhagio chrysopilaeformis (as Leptis; Bezzi, 1898: 31) Italy 
Rhagio chrysostigma (as Leptis; Loew, 1857: 33) Yugoslavia 
Rhagio cinerascens (as Leptis; von R?der, 1884: 2) Italy 
Rhagio cinereus (as Leptis; Bellardi 1861: 95) Mexico 
Rhagio cingulatus (as Leptis; Loew, 1856: 28) Russia 
Rhagio cingulatus 
canescens 
Szil?dy, 1934: 243 France 
Rhagio conspicuus Meigen, 1804: 299 Russia 
Leptis janotae Nowicki, 1867: 349 Czechoslovakia 
Leptis conspicuus 
alpinus 
Loew, 1869: 35 Austria 
Leptis conspicuus 
florentinus 
Loew, 1869: 34 Italy 
Leptis marchalii Pierre, 1889: 5 France 
 
 367
Species (synonyms indented) Author, reference Type country 
Rhagio corsicanus Becker, 1910: 640 France 
Rhagio costalis Matsumura, 1911: 68 Russia 
Rhagio costatus (as Leptis; Loew, 1826: 187) USA 
Rhagio costimaculata Matsumura, 1916: 330 Japan 
Rhagio dichromaticus Chillcott, 1965: 788 USA 
Rhagio difficilis Becker, 1921: 54 Greece 
Rhagio dimidiatus (as Leptis; Loew, 1863: 10) USA 
Leptis albibarbis Bigot, 1887: 114 USA 
Leptis pleuralis Adams, 1904: 441 USA 
Leptis flavoniger Coquillett, 1904: 20 USA 
Rhagio discoidalis (as Leptis; Brunetti, 1912: 463) India 
Rhagio elenae Soboleva, 1991: 96 Russia 
Rhagio ephippium (as Leptis; Zetterstedt, 1842: 219) Sweden 
Rhagio expansus James, 1964: 564 USA 
Rhagio franciscanus James, 1964: 565 USA 
? Rhagio expassus (as Leptis; Meunier, 1910: 69) Baltic Region (Eocene/ 
Oligocene) 
? Rhagio exporrectus (as Leptis; Meunier, 1910: 70) Baltic Region (Eocene/ 
Oligocene) 
? Rhagio expositus (as Leptis; Meunier, 1910: 71) Baltic Region (Eocene/ 
Oligocene) 
? Rhagio exsanguis (as Leptis; Meunier, 1910: 70) Baltic Region (Eocene/ 
Oligocene) 
? Rhagio fascinatoris (as Leptis; Meunier, 1910: 71) Baltic Region (Eocene/ 
Oligocene) 
? Rhagio ferus (as Leptis; Meunier, 1910: 72) Baltic Region (Eocene/ 
Oligocene) 
Rhagio filius (as Atherix; Walker, 1848: 219) USA 
Rhagio flavicornis (as Leptis; Macquart, 1826: 402) Japan 
Rhagio flavimedia (as Leptis; Coquillett, 1898: 307) Japan 
Rhagio floridensis Chillcott, 1965: 789 USA 
Rhagio formosus Bezzi, 1912: 445 Taiwan 
?Rhagio fossitius Melander, 1949: 29 USA (Miocene) 
Rhagio freyae Lindner, 1923: 8 Germany 
Rhagio funebris Meigen, 1820: 98 not given; Europe 
Rhagio fuscipennis (as Leptis; Meigen, 1820: 93) Austria 
Rhagio gansuensis Yang & Yang, 1997: 207 China 
Rhagio gracilis (as Leptis; Johnson, 1912: 3) USA 
Rhagio graeculus (as Leptis; Loew, 1869: 32) Greece 
Rhagio grandis Szil?dy, 1934: 248 France 
Rhagio guadarramensis Czerny & Strobl, 1909: 166 Spain 
Rhagio guangxiensis Yang & Yang, 1993: 46 China 
Rhagio guizhouensis Yang & Yang, 1992: 587 China 
Rhagio hainanensis Yang, Yang, and Nagatomi, 1997: 
212 
China 
 
 368
Species (synonyms indented) Author, reference Type country 
Rhagio hangzhouensis Yang & Yang, 1989: 291 China 
Rhagio huashanensis Yang & Yang, 1997: 215 China 
Rhagio idaeus Bezzi, 1908: 390 Greece 
Rhagio immaculatus (as Leptis; Meigen, 1804: 301) not given; Europe 
Rhagio immaculatus 
hahnleae 
Lindner, 1923: 9 Germany 
Rhagio incisus (as Leptis; Loew, 1872: 59) USA 
Rhagio insularis Becker, 1921: 55 Greece 
Rhagio iriomotensis Nagatomi & Nagatomi, 1990: 38 Japan 
Rhagio itoi Nagatomi, 1952: 11 Japan 
Rhagio japonicus Matsumura, 1916: 346 Japan 
Rhagio javanus Lindner, 1925: 21 Java 
Rhagio jinxiuensis Yang & Yang, 1993: 50 China 
Rhagio karafutonis Matsumura, 1916: 343 Japan 
Rhagio latipennis (as Leptis; Loew, 1856: 19) Germany 
Rhagio libanonicus Szil?dy, 1934: 252 Lebanon 
Rhagio lineola Fabricius, 1794: 275 Denmark 
Rhagio albifrons Meigen, 1804: 300 Austria 
Leptis lineola monticola Egger, 1860: 348 Austria 
Leptis albifrons 
monticola 
Egger, 1860: 668 Austria 
Leptis lineola andalusica Strobl, 1909: 166 Spain 
Leptis lineola polonica Szil?dy, 1934: 241 Poland 
Rhagio longshengensis Yang & Yang, 1993: 50 China 
Rhagio longzhouensis Yang & Yang, 1993: 49 China 
Rhagio luteus Soboleva, 1984: 122 Russia 
Rhagio lutifaciatus Okada, 1941: 256 Japan 
Rhagio maculatus (as Nemotelus; De Geer, 1776: 
165) 
not given; Europe 
Rhagio annulatus Meigen, 1804: 299 not given; Europe 
Rhagio nigrofasciatus Meigen, 1804: 301 not given; Europe 
Leptis distigma Meigen, 1820: 93 Germany 
Leptis stigmaticus Zetterstedt, 1849: 218 Sweden 
Leptis bimaculatus Gobert, 1877: 1 not given; Europe 
Leptis maculatus 
obscurus 
Strobl, 1898: 20 Bosnia 
Rhagio maculatus 
dalmaticus 
Szil?dy, 1934: 240 Croatia 
Rhagio maculatus 
macedonicus 
Szil?dy, 1934: 240 Macedonia 
Rhagio maculifer (as Leptis; Bigot, 1887: 113) USA 
Leptis hoodiana Bigot, 1887: 115 USA 
Rhagio maculifer 
concavus 
Leonard, 1930: 94 USA 
Rhagio maculipennis (as Leptis; Loew, 1854: 1) Turkey 
 
 369
Species (synonyms indented) Author, reference Type country 
Rhagio maolanus Yang & Yang, 1993: 280 China 
Rhagio matsumurae Lindner, 1923: 11 Japan 
Rhagio medeae Iacob, 1971: 353 Romania 
Rhagio meridionalis Yang & Yang, 1993: 2 China 
Rhagio miyonis Nagatomi, 1952: 7 Japan 
Rhagio mongolicus Lindner, 1923: 6 Mongolia 
Rhagio montanus Becker, 1921: 54 Italy 
Rhagio montanus 
striatus 
Lindner, 1934: 244 Poland 
Rhagio montivagus Edwards, 1919: 29 Sumatra 
Rhagio morulus Nagatomi, 1971: 39 Japan 
Rhagio mystaceus (as Leptis; Macquart, 1840: 30) not given; North 
America 
Rhagio naganensis Nagatomi, 1952: 9 Japan 
Rhagio nagatomii Yang & Yang, 1997: 227 China 
Rhagio niger (as Leptis; Wiedemann in Meigen, 
1820: 98) 
Portugal 
Rhagio niger trajani Szil?dy, 1934: 243 Romania 
Rhagio notatus (as Leptis; Meigen, 1820: 95) not given; Europe 
Rhagio notatus nigricans Lindner, 1934: 248 Czechoslovakia 
Rhagio ochraceus (as Leptis; Loew, 1862: 187) USA 
Rhagio olgae Soboleva, 1991: 94 Russia 
Rhagio olsufjevi Soboleva, 1989: 401 Russia 
Rhagio orestes Chillcott, 1965: 792 USA 
Rhagio pallidipennis Becker, 1921: 55 Greece 
Rhagio pallidistigma (as Leptis; de Meijere, 1924: 14) Sumatra 
Rhagio palpalis (as Leptis; Adams, 1904: 442) USA 
Rhagio costatus limbatus Leonard, 1930: 96 USA 
Rhagio perdicaceus Frey, 1954: 11 China 
Rhagio petrovae Soboleva, 1989: 403 Russia 
Rhagio philippinensis Frey, 1954: 11 Philippines 
Rhagio pilosus Yang, Yang, and Nagatomi, 1997: 
229 
China 
Rhagio plumbeus (as Leptis; Say, 1823: 39) USA 
Rhagio poecilopterus Bezzi, 1908: 389 Japan 
Rhagio politaeniatus (as Leptis; Bellardi, 1862: 27) Mexico 
Rhagio pollinosus Leonard, 1930: 116 USA 
Rhagio pseudasticta Yang & Yang, 1994: 32 China 
Rhagio puellaris Nagatomi, 1971: 276 Japan 
Rhagio pullata (as Chrysopila; Coquillett, 1898: 
307) 
Japan 
Rhagio punctipennis (as Leptis; Say, 1823: 34) USA 
Rhagio rolandi Becker, 1921: 48 France 
Rhagio rondanii Bezzi, 1908: 389 Italy 
Rhagio sabahensis Nagatomi & Nagatomi, 1990: 41 Malaysia 
 
 370
Species (synonyms indented) Author, reference Type country 
? Rhagio samlandicus (as Leptis; Meunier, 1916: 277) Baltic Region 
Rhagio sardous Szil?dy, 1934: 247 Italy 
Rhagio scapulifer (as Leptis; Bigot, 1887: 113) Japan 
Rhagio schmidti Lindner, 1931: 205 Algeria 
Rhagio scolopaceus (as Musca; Linnaeus, 1758: 590) Sweden 
Rhagio scolopaceus 
hahnlei 
Lindner, 1923: 9 Germany 
Sylvicola solitarius Harris, 1780: 100 not given; Europe 
Sylvicola monotropus Harris, 1780: 101 not given; Europe 
Rhagio separatus Yang, Yang, and Nagatomi, 1997: 
233 
China 
Rhagio shaanxiensis Yang & Yang, 1997: 235 China 
Rhagio shennonganus Yang & Yang, 1991: 276 China 
Rhagio shimai Nagatomi & Nagatomi, 1990: 46 Malaysia 
Rhagio shirakii Szil?dy, 1934: 9 Taiwan 
Rhagio sikisimanus Nagatomi, 1972: 79 Japan 
Rhagio simushirus Soboleva, 1989: 403 Russia 
Rhagio sinensis Yang & Yang, 1993: 1 China 
Rhagio singularis Yang, Yang, and Nagatomi, 1997: 
240 
China 
Leptis sordidus (as Leptis; Loew, 1862: 74) Turkey 
Rhagio pilosus (as Leptis; Loew, 1865: 235) Turkey 
Rhagio stigmosus Yang, Yang, and Nagatomi, 1997: 
242 
China 
Rhagio strigosus (as Leptis; Meigen, 1804: 299) France 
Rhagio mellinus Becker, 1921: 48 Spain 
Rhagio subpilosus (as Leptis; Becker, 1892: 23) Switzerland 
Rhagio taorminae Becker 1921: 55 Italy 
Rhagio terminalis (as Leptis; Loew, 1861: 317) USA 
Rhagio tessella (as Leptis; Motschulsky, 1889: 
505) 
Russia 
Rhagio tipuliformis Fabricius, 1794: 273 Germany 
Rhagio tonsa (as Leptis; Loew, 1869: 29) Spain 
Rhagio triangulata (as Leptis; Brunetti, 1920: 127) India 
Rhagio tringarius (as Musca; Linnaeus, 1758: 590) Sweden 
Erax rufus Scopoli, 1763: 363 & 986 not given; Europe 
Rhagio vermileo Fabricius, 1775: 762 France 
Nemotelus scolopaceus De Geer, 1776: 162 not given; Europe 
Sylvicola solivagus Harris, 1780: 101 not given; Europe 
Musca vermileo Schrank, 1781: 441 Austria 
Rhagio tringarius 
vanellus 
Fabricius, 1794: 272 Denmark 
Leptis tringarius simplex Meigen, 1838: 61 not given; Europe 
Leptis tringarius 
punctatus 
Loew, 1840: 4 Poland 
 
 371
Species (synonyms indented) Author, reference Type country 
Leptis cinereus Zetterstedt, 1842: 221 Sweden 
Leptis ephippium Zetterstedt, 1842: 219 Sweden 
Leptis tringarius goebelii Strobl, 1893: 29 Austria 
Leptis tringarius 
nigriventris 
Loew, 1869: 33 Germany 
Rhagio tringarius 
tripustulatus 
Szil?dy, 1934: 246 not given; Europe 
Rhagio tristis (as Leptis; Schummel, 1837: 109). 
NOTE 13. 
Germany 
Rhagio tuberculatus Yang, Yang, and Nagatomi, 1997: 
244 
China 
Rhagio turcicus Lindner, 1930: 87 Turkey 
Rhagio venetianus Becker, 1921: 54 Italy 
Rhagio vermileonoides Frey, 1954: 12 Burma 
Rhagio vertebratus (as Leptis; Say, 1823: 38) USA 
Rhagio intermedius Walker, 1848: 212 Canada 
Leptis hirtus Loew, 1861: 318 USA 
Leptis scapularis Loew, 1861: 318 USA 
Rhagio vitripennis (as Leptis; Meigen, 1820: 91) not given; Europe 
Rhagio tringarius  Panzer, 1794: 20 Germany 
Leptis stigma Schummel, 1837: 108 Poland 
Leptis striola Meigen, 1838: 61 Austria 
? Rhagio wheeleri Melander, 1949: 29 USA (Miocene) 
Rhagio yasumatsui Nagatomi, 1972: 83 Japan 
Rhagio zhejiangensis Yang & Yang, 1989: 290 China 
 
Nomina nuda Author, reference 
Leptis albicornis Say, 1823: 38 
Leptis acutangulus Meunier, 1899: 177 
Leptis distans Hennig, 1967: 39 
Leptis flexus Meunier, 1899: 177 
Leptis recurvus Meunier, 1899: 177 
Leptis validus Meunier, 1899: 177 
  
. 
Genus SCHIZELLA Bezzi 
Figures 112-117. 
 
 372
SCHIZELLA Bezzi, 1917: 118.  Type species Schizella furcicornis Bezzi 1917, by 
original designation. 
DIAGNOSIS.  
The monophyly of the species of Schizella is supported by the autapomorphic form of 
the male first flagellomere, which is expanded conspicuously into a bifurcate process. 
 
Schizella species are delicate, small to mid-sized (3.7 to 6.3 mm) flies, with long, thin 
legs; thorax brown to orange-brown with blue-, purplish-, or golden-colored setae. 
Wings are hyaline, without markings; male holoptic, eyes separated in female; in 
female, arista produced ventrally; first flagellomere subcircular, laterally compressed; 
in male, first flagellomere enlarged and forked; arista short; mandibles absent; 
laterotergite setose; wing vein M
3
 present; tibial spur formula 0:2:1; hind tibia with 
short macrochaetae; tergite 9 without ventrolateral arms; female spermathecal ducts 
with accessory glands. Schizella is restricted to the Philippines and the males of this 
genus are unlikely to be confused with any other brachyceran genus, on account of 
their highly modified antenna. Females of Schizella are separated from most 
Chrysopilus species by having the arista produced ventrally from the first 
flagellomere. 
 
DESCRIPTION.  
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere of female antenna enlarged basally, bearing long stylus, which originates 
anteroventrally. In male, first flagellomere enlarged and forked; arista short. Eyes 
 
 373
dichoptic in both sexes, inconspicuously setulose; ocelli evenly spaced, of equal size. 
Labella with pseudotrachae, longer than palps. Theca short and stout. Palps one-
segmented. Mandibles absent. Thorax. Mesonotum without vittae. Dorsocentral 
bristles absent; all dorsal setae of equal length. Anepisternum setulose on dorsal 
margin only. Laterotergite setose. Postspiracular scale absent. Metallic- or scale-like 
thoracic setae, often with structural color present. Postspiracular sclerite bare. 
Proscutellum absent. Subscutellum not enlarged nor lengthened; inconspicuous. Wing 
hyaline, without markings; pseudostigma present or absent. Lower calypter reduced. 
Upper calypter well developed, but margin with curvature narrow, width more than 
twice length. Costa extends past wingtip (to at least R
5
). Humeral crossvein (h) well 
developed. Sc-r crossvein absent. Dorsal side of R
1
 setulose, ventral side bare. All 
other wing veins without setulae. Wing veins R
1
 and R
2+3
 close together at wing 
margin. R
2+3
 sinuous, apical third ultimately bends toward leading edge of wing 
margin (creating concave flex anteriorly), length of R
2+3
 about the same length as R
5
, 
or longer. Base of R
4
-R
5
 fork proximal or directly above distal end of cell dm. R
4
 at 
base strongly curved or angled,  often with short proximal offshoot at point of 
curvature near R
5
, nearly straight apically. R
4
 and R
5
 encompass wing tip, or R
5
 
anterior to wing tip. R
5
 clearly longer than R
4+5
 (r-m to R
4
 origin). R-m crossvein at 
proximal side of central one-third of discal cell (or more centrally). M
3
 wing vein 
present. Origin of CuA
1
 at bm cell. Wing cell m3 parallel-sided at margin. CuA
2
 
greater than 2/3 length of posterior vein of bm cell. Alula full, rounded, with broad 
curvature. Anal lobe well developed. Anal cell open. Halter knob approximately 1/2 
length of stem. Tibial spur formula 0:2:1. Hind coxal tubercle absent. Hind tibial 
 
 374
macrochaetae present, short. Postmetacoxal bridge absent. Abdomen. Abdominal 
segments evenly tapered. In female, last 3 abdominal segments telescoping; tergite 7 
much longer than wide; intersegmental membrane between segments 7 and 8 
especially long; sternite 8 length elongated; more than twice as long as wide. Male 
terminalia with epandrium simple, not containing hypandrium ventrally. Epandrial 
sclerite wider than long, strongly notched anteriorly. Subepandrial sclerite absent. 
Hypoproct tomentose, without setae. Cerci widely displaced from one another, 
separation distance greater than three quarters width of cercus; held at angle in 
relation to rest of abdomen; in posterior view cupped, forming circular outline 
medially. Hypandrium fused entirely to gonocoxites. Gonocoxite with sinuous dorsal 
ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short or long enough to 
reach anterior margin of hypandrium. Parameral sheath bulbous, without paired 
swellings ventrally. Lateral ejaculatory processes present, not part of sperm sac 
posteriorly. Ejaculatory apodeme moderately long, reaching to at least anterior 
margin of hypandrium. Ejaculatory apodeme rod-shaped (upside-down v-shaped in 
profile). Aedeagal tines absent. Endoaedeagal process absent. Female terminalia with 
three spermathecae, clubbed, moderately to well sclerotized. Spermathecal ducts 
longer than five times the length of sternite 9, but not so long as to be difficult to 
measure, not inflated at base of spermathecae. Spermathecal duct accessory glands 
present, arise at approximately the distal third of the spermathecal ducts. Ejection 
apparatus of spermathecal ducts lightly sclerotized, not thickened, without surface 
furrows. Common spermathecal duct thinner than individual ducts, about the same 
length as the longest diameter of genital chamber. Genital chamber elliptical, 
 
 375
elongate, occupying most of sternite 9 area. Accessory gland posterior to genital 
chamber inconspicuous, easily overlooked even after staining. Sternite 9 anterior end 
pointed, with broad extensions posteriorly that are held in horizontal plane. Tergite 10 
present. Sternite 10 split into two sclerites. Cercus two-segmented. First segment of 
cercus not elongate, without ventral process. Ventral lobes of first segment of cercus 
curve ventrally towards one another to form a ring, visible in the posterior 
perspective. Basal cerci adjacent dorsally. Second cercus segment not elongated, 
without apical sensory pits. Larva. The larvae are unknown. 
 
LITERATURE.  
Nagatomi (1982) gives a cursory treatment of the genus. Below is a key to all species. 
Key to Schizella species. 
1. Notopleural setae white, reflecting bluish light; clypeus gray, clearly longer 
than broad ...................................................................................................... 2 
 Notopleural setae black, reflecting golden light; clypeus brown, nearly 
spherical.......................................................................  woodleyi new species 
2(1). Notopleural setae purplish blue; abdomen brown throughout, body length 
greater than 4mm???.. .....................................................  pulchrina Frey 
 Notopleural setae light blue; abdomen brown anteriorly, dark brown 
posteriorly, body length less than 4mm..............................  furcicornis Bezzi 
 
 376
Schizella woodleyi new species 
Figs 46 - 50. 
 
Type material: The male holotype is labeled ?Mt. Apo, Mindanao Phil. Islds. C.F 
Clagg/Kidapayan Trail, 7-8,000ft, 20 Sep/HOLOTYPE ?  Schizella woodleyi P.H. 
Kerr/[USNM ENT 00025872].? The paratype is labeled ?Mt. Apo, Mindanao Phil. 
Islds., C.S. Clagg/Sibulan Riv. 7-8,000ft. 16 Sept/PARATYPE ?  Schizella woodleyi 
P.H. Kerr/[USNM ENT 00025871]?. The holotype is in good condition. The 
specimen is missing its entire left midleg, right hindleg, and the right wing is torn 
slightly at the anal lobe. Many of the scutal setae are missing, however distinctive 
setae remain on and around the notopleuron and posteriorly, adjacent to the 
scutoscutellar suture. The paratype is missing almost all thoracic setae, both mid and 
hind legs and the distal/posterior portion of the right wing. The abdomen beyond the 
sixth segment is macerated and preserved in glycerol, in a microvial attached to the 
specimen pin. The holotype and paratype are deposited in the MCZ. 
 
Species Diagnosis. 
Schizella woodleyi differs from S. pulchrina Frey and S. furcicornis Bezzi in having 
black setae that reflect gold color when illuminated, especially on the notopleuron 
and anterior to the scutoscutellar suture. It also differs in having a brown clypeus that 
is nearly spherical in shape. Additionally, it is larger than the two other species in the 
genus.  
 
 
 377
Species Description. 
Male: Total body length 5.5-6.3mm. General color light orange-brown, becoming 
dark brown posteriorly, legs light orange-brown becoming brown distally. Head (Fig. 
46) generally dark brown under gray tomentum; inner eye margins converging 
slightly toward apex; frons 0.20 width of head at anterior ocellus, with central crease 
that deepens ventrally between antennal sockets; fronto-facial region at antennal 
sockets slightly prominent; ocellar tubercle dark brown, bare; face with light brown 
tomentum; gena bare; vertex and occiput dark brown under gray tomentum; antennae 
prominent, extending beyond width of head laterally, first flagellomere bifurcate, 
ventral extension of first flagellomere bearing stylus, ratio of segments (scape: 
pedicel: [dorsal extension of first flagellomere: ventral extension of first 
flagellomere]: stylus) 7:3:[41:49]:5, scape with length/width ratio 1.55, scape bare, 
pedicel with  single longitudinal row of small dark brown setulae at middle of 
segment, flagellomere with dense vestiture of short silvery brown setulae, stylus 
vestiture similar but not as dense; clypeus light brown, nearly spherical, lightly 
covered with silver tomentum, otherwise bare; palp one-segmented, light orange-
brown becoming brown distally, covered with silver tomentum, and with silver setae 
on proximal areas and brown setae distally; proboscis light orange-brown becoming 
brown distally, labella with silver setae. Thorax color orange brown dorsally and light 
orange brown laterally, very lightly tomentose throughout; scutum with short black 
setae except posteriorly, near scutoscutellar suture, where setae are black, flattened in 
a scale-like manner, and become gold when reflecting light; these scale-like 
black/gold setae are also present on and just dorsal to the notopleuron; anepisternum 
 
 378
with black and yellow setae along its anterior and dorsal margins; katepisternum and 
meron bare; anatergite with yellow setae, densely arranged, primarily along anterior 
and dorsal margins; katatergite with longer, black and yellow setae; postalar wall 
bare; coxae and femora same color as pleura; coxae bearing weak, black and yellow 
setae, somewhat sparsely arranged; femora bearing short, weak yellow setae 
throughout, except along dorsal and especially the distal end where setae are almost 
exclusively brown and nearly entirely appressed against the sclerite surface; tibiae 
yellowish but covered in dark brown, densely arranged setulae which make them 
appear darker in color, tibial setae sparse; tarsi darkening to brown distally, with dark 
brown, densely arranged setulae, setae present ventrally; wing (Fig. 50) primarily 
clear with brownish hue, pterostigma dark brown; anal cell open; vein R
4
 arising 
nearly perpendicular to vein R
5
, sharply curved toward wing margin, usually bearing 
short proximal stem at R
4
 curvature; halter stem orange brown, knob orange dark 
brown. Abdomen slender, with tergites 1 and 2 light orange-brown or dirty yellow, 
tergites 3 and 4 completely dirty yellow or dark brown on anterior half and orange 
brown on posterior half; tergite 5 dark brown or dark brown on anterior half and 
orange brown or dirty yellow on posterior half; tergites 6-8 dark brown or orange 
brown or dirty yellow with brown along anterior and lateral margins; all sternites 
light orange-brown, darkening slightly towards terminalia; setae of tergites mostly 
black with a few yellow setae interspersed; setae of sternites brown and yellow, in 
about equal parts. Male terminalia (Figs. 47 - 49) brown, orange brown, or yellowish. 
Epandrial sclerite wider than long, with deep, rounded notch anteriorly; posterior 
margin weakly emarginate centrally; subepandrial sclerite absent; hypoproct present, 
 
 379
lacking setae (only tomentum present); cercus attached to hypoproct, displaced 
anteriorly from epandrial sclerite and laterally, from one another, by approximately 
the width of a single cercus; aedeagal guide forms bulbous sac, without paired 
swellings. 
Female: Unknown. 
Etymology: The specific epithet is given in honor of Norman E. Woodley who 
originally identified these specimens to genus at the Museum of Comparative 
Zoology in 1979 and recently called them to my attention. Without his assistance, 
these specimens and the new species they represent would remain unknown. 
List of included species 
Species (synonyms indented) Author, reference Type country 
Schizella furcicornis Bezzi 1917: 119 Philippines 
Schizella pulchrina Frey 1954: 25 Philippines 
Schizella woodleyi Kerr, present work. Philippines 
 
Genus SIERRAMYIA, NEW GENUS 
Figures 118-121. 
SIERRAMYIA Type species Sierramyia chiapasensis Kerr, new species, by present 
designation. NOTE 14. 
 
DIAGNOSIS.  
Characters that I regard as autapomorphic for the genus Sierramyia include the 
weakly developed or absent anal lobe, so that the alular incision is rounded and open 
broadly or absent; the upper calypter reduced; and the epimeron bare. The large size 
 
 380
of the labellum (which is longer than the width of the eye from a lateral perspective) 
and the swollen hypandrium, expanded anteriorly, may also be autapomorphic. The 
scarcity of examplars for this genus makes it difficult to assert more autapomorphies 
at the species group level. However, the Sierramyia species are very distinctive and 
their monophyly appears certain. 
 Species of Sierramyia are small to mid-sized flies (approximately 5.6 mm) that 
are slight of build and are similar to Rhagio in several aspects; laterotergite setose, 
tibial spur formula 0:2:2, antenna with round first flagellomere bearing terminal 
arista; arista bare; wings often infuscate or with markings; M
3
 present. Both sexes are 
dichoptic. All Sierramyia species are endemic to mountainous areas (at 5000 feet or 
greater) in Mexico. They may be distinguished immediately from Rhagio and all 
other genera by the unusually modified wing that is weakly developed along its 
posterior margin, including a very reduced upper calypter. Additionally, Sierramyia 
chiapasensis may be distinguished from all Rhagio species and the related genus 
Desmomyia by having a bare epimeron and by having setulae on the dorsal surface of 
wing vein R
5
 and on the ventral surface of wing veins R
2+3
 and R
4
.   Sierramyia 
species differ from Chrysopilus and related chrysopiline genera such as Schizella and 
Stylospania by having a bare arista, two hind tibial spurs, and by lacking thoracic 
setae with metallic sheen.  
 
DESCRIPTION.  
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere first flagellomere oval in cross section, lightly compressed laterally; 
 
 381
enlarged basally, bearing fused, bare arista-like extension. Eyes inconspicuously 
setulose, dichoptic. Labella enlarged, longer than palps. Palps one-segmented. 
Mandibles absent. Thorax. Mesonotum without vittae. Dorsocentral bristles absent, 
all dorsal setae of equal length. Anepisternum setulose on dorsal margin only. 
Laterotergite setose, on katatergite only. Postspiracular scale absent. Proscutellum 
shallowly present. Subscutellum not enlarged nor lengthened; inconspicuous. Wing 
membrane darkly infuscate, with hyaline markings; without pseudostigma. Lower 
calypter reduced. Upper calypter underdeveloped, with narrow curvature, width more 
than twice length. Costa approximately to wingtip. Humeral crossvein (h) well 
developed, Sc-r crossvein weakly developed, positioned distal to the humeral 
crossvein (h), by a little more than the length of h. Dorsal side of R
1
 setulose, ventral 
side of R
1
 bare. R
2+3
 sinuous, apical third of R
2+3
 ultimately bends toward wingtip 
(creating convex flex anteriorly); about the same length as R
5
; with setulaeon both 
sides of membrane. Base of R
4
-R
5
 fork proximal of distal end of cell dm. R
4
 at base 
relaxed, not strongly curved, nearly straight apically; dorsal side bare, ventral side 
with hair. R
4
 and R
5
 encompass wing tip. R
5
 clearly longer than R
4+5
 (r-m to R
4
 
origin); dorsal side setulose, ventral side bare. M
3
 wing vein present. Wing cell m3 
parallel-sided at margin. Origin of CuA
1
 at bm cell. Length of CuA
2
 v. posterior vein 
of bm cell about the same as 1/2 length of posterior vein of bm cell. Alula reduced. 
Anal lobe reduced. Anal cell open. Halter knob between 1/3?1/2 length of stem. 
Tibial spur formula 0:2:2. Hind coxal tubercle absent. Hind tibia without ventro-
apical swelling. In female, Abdomen. Terminal Abdomen. Abdominal segments 5?10 
evenly tapered from segments 1?4. Tergite 7 much longer than wide. Intersegmental 
 
 382
membrane between segments 7 and 8 especially long. Tergite 8 with pair of ducts 
arising from posterior margin of sclerite, terminating in clump of sclerotized tissue. 
Sternite 8 longer than wide. Male terminalia with epandrium simple, not containing 
hypandrium ventrally. Epandrial sclerite wider than long, modestly curved anteriorly. 
Subepandrial sclerite absent. Hypoproct triangular (rounded posteriorly), or 
pentagonal. Hypoproct flattened, distinct from sclerites above cerci; tomentose, 
without setae. Cerci partially displaced from one another, separation distance 
approximately half the width of single cercus. Hypandrium separated from the 
gonocoxites by a complete suture, expanded anteriorly. Gonocoxite with sinuous 
dorsal ridge, leading to gonocoxal apodeme. Gonocoxal apodemes long enough to 
reach anterior margin of hypandrium. Parameral sheath bulbous, without paired 
swellings ventrally. Lateral ejaculatory processes present, integrated into sperm sac 
membrane. Ejaculatory apodeme long, reaching beyond anterior margin of 
hypandrium; laterally compressed. Aedeagal tines absent. Endoaedeagal process 
present, short. Female terminalia with three spermathecae. Spermathecal ducts longer 
than five times the length of sternite 9, but not so long as to be difficult to measure. 
Spermathecal duct accessory glands not present. Ejection apparatus of spermathecal 
ducts lightly sclerotized, thickened, with ringed surface furrows. Spermathecal duct 
junction thickened. Common spermathecal duct thickened; short, shorter than longest 
diameter of genital chamber. Genital chamber teardrop shaped, moderately sized. 
Accessory gland posterior to genital chamber inconspicuous, easily overlooked even 
after staining. Sternite 9 anterior end broadly pointed; posterior end with broad lateral 
extensions, free, held in horizontal plane. Tergite 10 length approximately as long as 
 
 383
half measured width, or longer. Sternite 10 entire, pentagonal, pointed posteriorly; 
posterior half below first cercus segment. Cercus two-segmented. First segment of 
cercus not elongate, without ventral process. Ventral lobes of first segment of cercus 
curve ventrally towards one another to form a ring, visible in the posterior perspective 
(curving slightly ventrally; not obvious). Basal cerci adjacent dorsally. Second cercus 
segment not elongated. Cercus without apical sensory pits. Larva. The larvae are 
unknown. 
 
LITERATURE.  
Nagatomi (1982, 1984) illustrates wing, antenna, and male genitalia of two 
undescribed species belonging to this new genus. 
 
Sierramyia chiapasensis Kerr, new species 
Type material: The female holotype is labeled ?Mexico: Chiapas/ Lago Montebello/ 
16 07'N, 91 40'W/ 5000ft 14 July1969/ W.R.M. Mason /HOLOTYPE ?  Sierramyia 
chiapasensis P.H. Kerr/[USNM ENT 00022656].? The specimen is in good condition. 
The head, wing, and legs are all in perfect condition. Most thoracic setae are missing, 
however. The abdomen beyond the first segment is macerated and preserved in 
glycerol, along with the dissected genitalia, in a microvial attached to the specimen 
pin. The spermatheca, however, is missing and undescribed. The holotype is 
deposited in the CNC. 
 
Species Diagnosis.  
 
 384
Sierramyia chiapensis is distinguished from other species in the genus primarily by 
the coloration pattern on the wing, which is darkly infuscate brown, with two hyaline 
bands. The first band starts near the origin of the radial vein, passes through the distal 
portion of the basal cells, continues posteriorly and veers distally to occupy most of 
cell cua
1
. The second hyaline band spans the wing transversely, starting in cell r3 near 
the apex of vein R
2+3
, continuing posteriorly through cell r5 (slightly distal of R
4+5
 
fork), and in each of the wing cells posterior of this to cell m3 where the band 
terminates at the wing margin at the apex of CuA
1
.  This is the only described species 
of Sierramyia. However, another species of the genus apparently has been illustrated 
in Nagatomi (1982: Fig. 42), and this species appears to have uniformly infuscate 
wings throughout, without hyaline bands. 
 
Species Description 
Female: Total body length 5.6mm. General color yellow with abdomen darkening to 
light brown distally, legs yellow, except for fore tarsus, which is light brown. Head 
(Fig. 45) brown dorsally and along most of frons where it is microtomentose, yellow 
ventrally; inner eye margins converging toward apex; frons 0.07 width of head at 
anterior ocellus; ocellar tubercle brown, bare; parafacials very narrow, practically 
absent; gena bare; postocular setae short, dark brown; occiput pollinose yellow, with 
short, brown or yellow setae; antenna yellow, except for arista which is brown; scape 
and pedicel of equal size; scape with short yellow setae; pedical with short black 
setae; first flagellomere round, slightly laterally flattened, slightly smaller than scape 
and pedicel, bearing terminal arista; arista bare, approximately the same length as the 
 
 385
width of one eye, viewed from the dorsal perspective; clypeus yellow, nearly 
spherical, lightly microtomentose, without setae; palp one-segmented, yellow, with 
black setae; proboscis yellow, labella long (longer than palp and longer than the width 
of the eye from a lateral perspective) with short yellow setae ventrally. Thorax yellow 
throughout; scutum with short brown setae; proepimeron bare; anepisternum with a 
few brown seta near dorsal and posterior margins; katepisternum and meron bare; 
anatergite bare; katatergite with brown and yellow setae, few in number 
(approximately 10 to 12); postalar wall bare; coxae and femora same color as pleura; 
coxae bearing yellow setae of moderate length; femora bearing short, brown setae, 
mostly appressed near the sclerite surface; tibia yellow, bearing short setae; fore tibia 
with yellow and brown setae, mid- and hindtibia with brown setae only; all tarsi with 
short brown setae; fore tarsus light brown, mid- and hindtarsus yellow; wing darkly 
infuscate brown, with two hyaline bands; the first band starts near the origin of the 
radial vein, passes through the distal portion of the basal cells, continues posteriorly 
and veers distally to occupy most of cell cua
1
; the second hyaline band spans the wing 
transversely, starting in cell r3 near the apex of vein R
2+3
, continuing posteriorly 
through cell r5 (slightly distal of R
4+5
 fork), and in each of the wing cells posterior of 
this to cell m3 where the band terminates at the wing margin at the apex of CuA
1
; 
halter stem and knob yellow; stem long, with fine row of short black setae along 
trailing (dorsal) margin; short black setae also present, scattered on knob ventrally.   
Abdomen slender; tapering distally; yellowish brown, darkening to light brown 
distally; with brown and yellow setae. Sternite 8 longer than wide, at its widest point 
near posterior margin; sternite 9 broadly pointed posteriorly, lateral extensions free 
 
 386
anteriorly; genital chamber moderately sized, approximately the same length as the 
length of the common spermathecal duct; base of spermathecal ducts lightly 
sclerotized with conpicous transverse surface rings; spermathecal duct accessory 
glands absent; tergite 10 yellow, wider than long; sternite 10 yellow, pentagonal, 
pointed posteriorly; cercus yellow, two-segmented; first cercal segment with ventral 
process, which curves medially to form a ring when view from posterior perspective; 
second cercal segment without apical pits. 
 
Male: Unknown. 
 
Etymology. The specific epithet is in reference to the region in which the type 
specimen was collected. 
 
Genus SPANIA Meigen 
Figures 122-126. 
SPANIA Meigen, 1830: 335.  Type species nigra Meigen, by monotypy. 
 ARCHICERA Szil?dy, 1934: 264.  Type species Archicera avavorum Szil?dy, 
1934: 268, by monotypy. NOTE 15. 
  
DIAGNOSIS.  
I consider the reduced form of the female first segment of the cercus (approximately 
half the length or less of the fairly elongate second segment) and the form of female 
sternite 8, which lacks a medial invagination along its posterior margin, as 
 
 387
autapomorphic character states that define Spania. Since the closest relatives of 
Spania are mandibulate (Spaniopsis and Symphoromyia), the loss of mandibles may 
represent an additional autapomorphy.  
 
Spania is small (2.1 to 3.0 mm), dark brown to black in color, and slight in build. 
Wings are lightly infuscate, especially along costal vein; male holoptic, eyes 
separated in female; antenna with stylus arising ventrally or terminally from enlarged 
flagellomere base, laterally compressed; mandibles absent; laterotergite bare; wing 
vein M
3
 completely or incompletely present; tibial spur formula 0:2:0; hind tibia 
without macrochaetae; tergite 9 with ventrolateral arms, extending posteriorly, 
surrounding and fusing to sternite 9 laterally; female spermathecal ducts with 
accessory glands arising near base of spermathecae. Spania is most likely to be 
confused with Ptiolina or Bolbomyia. It may be distinguished most easily from 
Ptiolina species by lacking hind tibial spurs and its small size. Spania is 
approximately the same size as Bolbomyia, however, it has a stylate antenna and wing 
vein M
3
 at least incompletely present, and lacks fore and hind tibial spurs. Spania was 
once synonymized with Spaniopsis (Paramonov, 1962), however Spania may be 
distinguished from this genus by its small size, its delicate build, scape clearly smaller 
than the pedicel, and mandibles absent. Spania is restricted to the Holarctic Region, 
with a distribution that includes North America, Europe, and Japan.  
 
DESCRIPTION.  
 
 388
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere of antenna enlarged, laterally compressed, bearing ventro-apical stylus. 
Eyes in male ommatidia split into dorsal and ventral areas and smaller ventrally. Eyes 
inconspicuously setulose; in female, dichoptic; in male, holoptic. Parafacials in male 
not swollen. Labella with pseudotrachae, longer than palps. Theca short and stout, 
with medial suture. Palps one-segmented. Lacinia present, but very reduced. Lacinia 
apex not serrated. Mandibles absent. Cibarial pump short, as wide as long or wider. 
Cornu shorter than cibarial pump. Pharyngeal pump approximately same length as 
cibarial pump (including cornu). Mesonotum lacking vittae, black or brown, without 
dorsocentral bristles. Anepisternum bare. Postspiracular sclerite smooth, bare. 
Proscutellum narrowly present or absent. Subscutellum slightly swollen or absent. 
Laterotergite bare. Wing hyaline, without markings. Lower calypter reduced. Upper 
calypter well developed, with broad curvature, lobe-like, width twice length or less. 
Costa extends past wing tip. Humeral crossvein (h) well developed. Sc-r crossvein 
present, weakly developed, or absent; positioned distal to h, by the approximate 
length of h. Dorsal side of R
1
 setulose, ventral side bare. All other wing veins bare. 
Wing veins R
1
 and R
2+3
 separated at wing margin. R
2+3
 sinuous, apical third 
ultimately bends toward leading edge of wing margin, creating concave flex 
anteriorly. Length of R
2+3
 about the same length as R
5
. Base of R
4
-R
5
 fork proximal 
or directly above distal end of cell dm. R
4
 mostly straight apically. R
5
 ending at wing 
tip or anterior to wing tip, clearly longer than R
4+5
 (r-m to R
4
 origin). R-m crossvein 
at proximal one-third to near halfway of discal cell. M
3
 wing vein incompletely 
present (not reaching margin) or complete, reaching wing margin. Origin of CuA
1
 at 
 
 389
discal cell or at crossvein separating discal and basal medial cells; m-cu crossvein 
absent. Length of CuA
2
 v. posterior vein of bm cell greater than 1/2 length, less than 
2/3 length of posterior vein of bm cell. Alula with broad curvature that is slightly 
shifted distally. Anal lobe well developed. Anal cell open or closed. Halter knob 
approximately 1/2 length of stem. Tibial spur formula 0:2:0. Hind coxal tubercle 
absent. Hind tibial macrochaetae absent. Postmetacoxal bridge absent. Abdomen. 
Abdominal segments evenly tapered. In female, tergite 7 much wider than long; 
intersegmental membrane between segments 7 and 8 short, as throughout abdomen; 
sternite 8 sclerite entire, wider than long. Male terminalia with epandrium simple, not 
containing hypandrium ventrally. Epandrial sclerite wider than long, strongly notched 
anteriorly. Subepandrial sclerite absent. Hypoproct elliptical, flattened, tomentose, 
without setae. Cercus directly adjacent to epandrial sclerite; widely displaced from 
one another, separation distance greater than three quarters width of cercus; held 
vertical in relation to rest of abdomen; in posterior view cupped, forming circular 
outline medially. Hypandrium fused entirely to gonocoxites. Gonocoxite with sinuous 
dorsal ridge, leading to gonocoxal apodemes. Gonocoxal apodemes short, not 
reaching anterior margin of hypandrium. Parameral sheath slightly bulbous ventrally, 
produced into slightly developed paired lobes or smooth. Ejaculatory apodeme 
laterally compressed; long, reaching anterior margin of hypandrium. Lateral 
ejaculatory processes present, integrated into sperm sac membrane. Aedeagal tines 
absent. Endoaedeagal process absent. Female terminalia with tergite 9 entire, with 
narrow anteriorly-directed ventrolateral projections, enveloping sternite 9. 
Spermathecae three, spherical, lightly to moderately sclerotized. Spermathecal ducts 
 
 390
no more than three times the length of sternite 9, with swelling halfway between 
genital chamber and spermathecae. Spermathecal duct accessory glands present, arise 
at the base of each spermatheca. Spermathecal ducts sclerotized and thickened in a 
narrow ring near junction with common spermathecal duct, otherwise smooth, not 
enlarged, and unsclerotized. Common spermathecal duct thickened, short, shorter 
than longest diameter of genital chamber. Genital chamber teardrop shaped, small, 
occupying fraction of sternite 9 area. Accessory gland posterior to genital chamber 
present, inconspicuous, easily overlooked even after staining. Sternite 9 anterior end 
broadly paddle-shaped. Sternite 9 with broad posterolateral projections that are held 
at an angle. Tergite 10 present, entire, short (length less than half width). Sternite 10 
present, entire, roughly rectangular. Cercus two-segmented. Basal cerci not elongated, 
without ventral process, separated from each other dorsally by approximately the 
width of second cercal segment. Ventral lobes of first segment of cercusdo not curve 
ventrally towards one another to form a ring. Second cercus segment narrow, 
elongated (~2.5x longer than wide or more), without apical sensory pits. Larva. The 
larva is undescribed, however immature Spania nigra Meigen were found mining the 
thallus of Pellia neesiana (Bryophyta: Pelliaceae) (S?guy, 1927; Mik, 1896 in 
Nartshuk, 1995).  
LITERATURE.  
Nagatomi & Saigusa (1982) give a key to the Japanese fauna that includes all species. 
List of included species 
Species (synonyms indented) Author, reference Type country 
Spania kyushuensis Nagatomi & Saigusa, 1982: 226 Japan 
Spania naitoi Nagatomi & Saigusa, 1982: 229 Japan 
 
 391
Species (synonyms indented) Author, reference Type country 
Spania nigra Meigen, 1830: 335 not given; Europe 
Spania nigra americana Johnson, 1923: 70 USA 
 
Genus SPANIOPSIS White 
Figures 127-139. 
SPANIOPSIS White, 1914: 43.  Type species Spaniopsis tabaniformis White, 1914, 
by monotypy.  
 
DIAGNOSIS.  
Although Spaniopsis is a distinctive genus, I could only find a single feature of the 
internal mouthparts that I consider unambiguously autapomorphic. In Spaniopsis 
species, the cornu is fused apically to the pharyngeal pump. I have not seen this in 
any other tabanomorph. Nagatomi and Soroida (1985) illustrate Atherix ibis and 
Suragina caerulescens as having a fused cornu also, however I did not see this in any 
of the athericids I examined over the course of my study. In all other taxa that I have 
examined, the cornu extends beyond the pharyngeal pump, in line with the cibarium. 
 
Spaniopsis species are very stout bodied flies, small to moderately sized (3 to 6 mm), 
with generally gray or dark gray thorax, with the posterior margin of each abdominal 
tergite often lightened to light brown or faded yellow in color so that the abdomen 
appears banded. Wings are mostly hyaline and either only lightly infuscate in the 
costal vein area (more darkly in S. marginipennis), or infuscate near wing veins (as in 
S. mackerrasi); male holoptic, eyes separated in female; antenna with terminal stylus, 
 
 392
laterally compressed; mandibles present; laterotergite bare; wing vein M
3
 
incompletely present or absent; tibial spur formula 0:2:0; hind tibia without 
macrochaetae; tergite 9 with ventrolateral arms, extending posteriorly, surrounding 
and fusing to sternite 9 laterally; female spermathecal ducts with accessory glands. 
Spaniopsis is restricted to Australia and is more likely to be confused with local 
Tabanidae and Athericidae than with Rhagionidae, especially given the annoying 
bloodfeeding behavior of the females.  Spaniopsis may be distinguished from both 
Athericidae and Tabanidae by lacking a scale on the postspiracular sclerite and by the 
unsegmented, lanceolate form of the flagellum. Spaniopsis differs from Austroleptis 
by having a bulbous clypeus; a two-segmented palp; mandibles present; an 
unsegmented, lanceolate flagellum; and by lacking hind tibial spurs. The genus may 
be distinguished from Atherimorpha most easily by its robust body, the form of its 
antenna, wing vein M
3
 absent or incomplete, laterotergite bare, and hind tibia without 
spurs.  NOTE 16. 
 
DESCRIPTION.  
Head. Clypeus bulbous. Scape approximately the same size as pedicel. First 
flagellomere of antenna enlarged bearing stylus of single segment. Eyes in male 
ommatidia split into dorsal and ventral areas; smaller ventrally. Eyes inconspicuously 
setulose; in female, dichoptic; in male, holoptic. Parafacials in male not swollen. 
Labella longer than palps, with pseudotrachae. Theca short and stout; formed by two 
sclerites, slightly separated medially. Palps one-segmented. Stipes surrounded by 
membrane above theca, directed posteriorly. Lacinia longer than palps, with serrated 
 
 393
tip. Mandibles present. Cibarial pump short, as wide as long or wider. Cornu shorter 
than cibarial pump, apically fused to pharyngeal pump. Pharyngeal pump anteriorly 
broad, forming cup-like structure, longer in total length than length of cibarial pump. 
Thorax. Mesonotum vittate. Dorsocentral bristles absent; all dorsal setae of equal 
length. Anepisternum setulose throughout posterior half, except in S. mackerrasi 
Paramonov where anepisternum bare. Laterotergite bare. Postspiracular scale absent. 
Proscutellum present. Subscutellum slightly swollen or absent. Wing hyaline, with or 
without markings. Lower calypter reduced. Upper calypter well developed, with 
broad curvature, lobe-like, width twice length or less. Costa extends to wingtip. 
Humeral crossvein (h) well developed. Sc-r crossvein weakly developed, positioned 
distal to h, by the approximate length of h. Dorsal side of R
1
 setulose, ventral side 
bare. All other wing veins bare. Wing veins R
1
 and R
2+3
 separated at wing margin. 
Wing vein R
2+3
 sinuous, apical third ultimately bends toward leading edge of wing 
margin, creating concave flex anteriorly. Length of R
2+3
 clearly shorter than R
5
.  Base 
of R
4
-R
5
 fork proximal or directly above distal end of cell dm. R
4
 nearly straight 
apically. R
5
 ending at or near wing tip, clearly longer than R
4+5
 (r-m to R
4
 origin). R-
m crossvein at proximal one-third to near halfway of discal cell. M
3
 wing vein 
incompletely present (not reaching wing margin) or absent. Origin of CuA
1
 at discal 
cell; m-cu crossvein absent CuA
2
 approximately 2/3 length of posterior vein of bm 
cell). Alula with broad curvature that is slightly shifted distally. Anal lobe well 
developed. Anal cell closed. Halter knob between 1/2?2/3 length of stem. Tibial spur 
formula 0:2:0. Hind coxal tubercle absent. Hind tibial macrochaetae absent. 
Abdomen. Abdominal segments evenly tapered. In female, last 3 abdominal segments 
 
 394
telescoped; tergite 7 much wider than long; intersegmental membrane between 
segments 7 and 8 short, as throughout abdomen; sternite 8 as wide as long or wider 
than long. Male terminalia with epandrium simple, not containing hypandrium 
ventrally. Epandrial sclerite wider than long, strongly notched anteriorly. 
Subepandrial sclerite absent. Hypoproct triangular (rounded posteriorly), flattened, 
tomentose, without setae. Cercus directly adjacent to epandrial sclerite; widely 
displaced from one another, separation distance greater than three quarters width of 
cercus; held vertical in relation to rest of abdomen; in posterior view cupped, forming 
circular outline medially. Hypandrium fused entirely to gonocoxites. Gonocoxite with 
sinuous dorsal ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short, 
usually not long enough to reach anterior margin of hypandrium. Parameral sheath 
forming separate, distinct lobes ventrally. Ejaculatory apodeme short or long enough 
to reach anterior margin of hypandrium. Lateral ejaculatory processes present, not 
part of sperm sac posteriorly. Aedeagal tines absent. Endoaedeagal process absent. 
Female terminalia with tergite 9 with narrow anteriorly-directed ventrolateral 
projections, enveloping sternite 9. Spermathecae three, spherical, lightly to 
moderately sclerotized. Spermathecal ducts more than three times but less than five 
times the length of sternite 9, not inflated at base of spermathecae. Spermathecal duct 
accessory glands present, arise at approximately the distal third of the spermathecal 
ducts. Spermathecal ducts sclerotized and thickened in a narrow ring near junction 
with common spermathecal duct, otherwise slightly enlarged, lightly sclerotized, with 
small furrows on surface of ducts near base. Common spermathecal duct slightly 
thickened, about the same length as the longest diameter of genital chamber. Genital 
 
 395
chamber circular, small, occupying fraction of sternite 9 area. Accessory gland 
posterior to genital chamber inconspicuous, easily overlooked even after staining. 
Sternite 9 anterior end rounded; posterior end with narrow posterolateral extensions. 
Tergite 10 present, entire, short (length less than half width). Sternite 10 present, split 
into two sclerites. Cercus two-segmented. First segment of cercusnot elongated, 
without ventral process. Basal cerci separated from one another dorsally by 
approximately the width of the second cercal segment. Ventral lobes of first segment 
of cercus not curving ventrally towards one another to form a ring. Second cercus 
segment not elongated, without apical sensory pits. Larva. The larvae are unknown. 
 
LITERATURE.  
Paramonov (1962) gives a key to all Spaniopsis species (treated as Spania). 
List of included species 
Species (synonyms indented) Author, reference Type country 
Spaniopsis clelandi Ferguson, 1915: 240 Australia 
Spaniopsis longicornis Ferguson, 1915: 242 Australia 
Spaniopsis mackerrasi (as Spania; Paramonov, 1962: 140) Australia 
Spaniopsis marginipennis Ferguson, 1915: 239 Australia 
Spaniopsis rieki (as Spania; Paramonov, 1962: 145) Australia 
Spaniopsis tabaniformis White, 1914: 44 Australia 
Spaniopsis vexans Ferguson, 1915: 238 Australia 
 
Genus STYLOSPANIA Frey 
Figure 140. 
STYLOSPANIA Frey, 1954: 23.  Type species Stylospania lancifera Frey, by 
monotypy. 
 
 396
 
DIAGNOSIS.  
This genus is based on a single male specimen collected from Samar, Catbalogan, 
Philippines. Stylospania lancifera bears most features found in Chrysopilus, but may 
be distinguished from this genus by its stylate flagellum. It too has a reduced, bare 
proepimeron and its genitalia are indistinguishable from those of males of many 
Chrysopilus species.  The only known specimen of Stylospania lancifera is devoid of 
thoracic setae. The female is unknown. 
 
Stylospania lancifera is a delicate fly, fairly small in size (appx. 4.5 mm), with long, 
thin legs; orange-brown. Wings are hyaline, without markings; male dichoptic; 
flagellum subcircular with long, tapering stylus; mandibles absent; laterotergite 
setose; wing vein M
3
 present; tibial spur formula 0:2:1; hind tibia with short 
macrochaetae; tergite 9 without ventrolateral arms. Stylospania lancifera is known to 
occur only in the Philippines. Stylospania may be distinguished from all Chrysopilus 
by its antenna, which has a stylate flagellum.  
DESCRIPTION. 
Clypeus bulbous. Scape approximately the same size as pedicel. First flagellomere 
laterally compressed. First flagellomere of antenna enlarged bearing stylus of single 
segment, stylus narrow, nearly arista-like. Eyes inconspicuously setulose; male 
dichoptic; ommatidia evenly distributed, of equal size; not flattened dorsally. Head 
wider than thorax. Labella with pseudotrachae, longer than palps. Theca short. Palps 
one-segmented. Mandibles absent. Thorax. Mesonotum without vittae. Setae of 
 
 397
dorsum all of equal length. Specimen lacks metallic- or scale-like thoracic hairs, 
although these are likely present. Proepimeron reduced, bare. Anepisternum bare. 
Laterotergite setose, katatergite and anatergite indistinguishable, setae present 
throughout laterotergite. Thoracic surface between base of halter and postspiracular 
sclerite without setulae. Thoracic spiracle without flaps, not lined with setulae. 
Postspiracular scale absent. Postspiracular sclerite without setulae. Thoracic surface 
immediately posterior to postspiracular sclerite bare. Proscutellum absent. 
Subscutellum not bulbous, bare. Wing hyaline, without markings. Wing with 
pterostigma. Lower calypter reduced. Upper calypter well developed, full and 
rounded, with broad curvature, lobe-like, width twice length or less. Alula broad, 
rounded evenly. Anal lobe well developed. Sc-r crossvein present, weakly developed, 
positioned distal to the humeral crossvein (h), by the approximate length of h. Dorsal 
side of R
1
 setulose, ventral side bare. Wing veins R
1
 and R
2+3
 close together at wing 
margin (R
2+3
 clearly closer to R
1
 than to R4). Wing vein R
2+3
 directed toward wing 
margin, meeting margin abruptly. R
2+3
 bare on both dorsal and ventral surfaces of 
wing membrane. Base of R
4
-R
5
 fork proximal or directly above distal end of cell dm. 
R
4
 at base strongly curved or angled, straight or nearly straight apically. R
4
 and R
5
 
encompass wing tip. R
5
 aligned with R
4+5
. Wing vein M
3
 present, reaching wing 
margin. Tibial spur formula 0:2:1. Hind coxal tubercle present. Hind tibial 
macrochaetae present; small, easily overlooked, nearly flush with sclerite surface. 
Hind tibia without ventro-apical swelling. First hind metatarsus of male not swollen. 
Abdomen. Abdominal segments evenly tapered. Epandrial sclerite wider than long, 
modestly curved anteriorly. Epandrium simple, not containing hypandrium ventrally. 
 
 398
Subepandrial sclerite absent. Hypoproct rectangular; wider than long; tomentose, 
without setae. Cerci partially displaced from one another, separation distance 
approximately half the width of single cercus. Hypandrial sclerite fused entirely to 
gonocoxites. Gonocoxite smooth dorsally, without sinuous ridge leading to gonocoxal 
apodeme. Gonocoxal apodemes long enough to reach anterior margin of hypandrium. 
Parameral sheath developed into bulbous sac ventrally, without distinct lobes. Lateral 
ejaculatory processes present, part of sperm sac posteriorly, arising free of sperm sac 
membrane dorsally. Ejaculatory apodeme moderately long, reaching anterior margin 
of hypandrium, laterally compressed. Aedeagal tines absent. Endoaedeagal process 
absent. Gonostylus heavily setose. Larva unknown. 
 
LITERATURE.  
In addition to the original literature (Frey, 1954), Nagatomi (1982) provides a brief 
diagnosis of this genus.  
List of included species 
Species (synonyms indented) Author, reference Type country 
Stylospania lancifera Frey, 1954: 23 Philippines 
 
Genus SYMPHOROMYIA Frauenfeld 
Figures 141-154, 185, 186C-D, 187A, 194. 
SYMPHOROMYIA Frauenfeld, 1867: 496.  Type species Atherix melaena Meigen, 
by original designation. 
 SYMPHEROMYIA Schiner, 1868: 910 (lapsus). 
 
 399
 PARAPHEROMYIA Becker, 1921: 42. Type species Atherix crassicornis 
Panzer, by original designation. NOTE 17. 
 PARAPHOROMYIA Becker, 1922 (incorrect subsequent spelling, validated 
under Article 33.3.1). NOTE 18. 
 
DIAGNOSIS.  
The autapomorphic development of the enlarged scape of Symphoromyia provides 
support for the monophyly of the genus. Desmomyia also has an enlarged scape, 
however, the distant phylogenetic placements of this genera indicate that this 
character state has evolved independently. Another autapomorphy for the genus is the 
shape of the aristate flagellomere, which is produced ventrally and is often kidney-
shaped in profile. 
 
Symphoromyia species are stout bodied flies, moderately sized (4.7 to 9 mm), with 
black, gray or gold-gray thorax, and abdomen colored gray, black, mixed black and 
yellow, black terminating with yellow, or entirely yellow. Wings are hyaline or 
lightly infuscate; male holoptic, eyes separated in female; antenna aristate; first 
flagellomere kidney-shaped or subcircular, expanded ventrally; laterally compressed; 
mandibles present; laterotergite bare; wing vein M
3
 present; tibial spur formula 0:2:1; 
hind tibia without macrochaetae; tergite 9 with ventrolateral arms, extending 
posteriorly, surrounding and fusing to sternite 9 laterally; female spermathecal ducts 
with accessory glands. Symphoromyia is restricted to the Holarctic Region and 
reaches its greatest diversity in North America. Symphoromyia may be confused with 
 
 400
local Tabanidae and Athericidae since females of some species are known to be 
bloodfeeders and can be a nuisance.  Symphoromyia may be distinguished most easily 
from both Athericidae and Tabanidae by lacking a scale on the postspiracular sclerite 
and in addition to this, from the Athericidae by an elongated scape and wing vein R
2+3
 
meeting the margin at some distance away from R
1
; and from Tabanidae by having an 
aristate antenna and wing vein R
4
 not sinuate, and nearly parallel to R
5
. 
Symphoromyia is similar in form and color to several species of Ptiolina, but may be 
distinguished by the elongate scape (that is clearly larger than pedicel), the ventrally 
expanded first flagellomere, and setose laterotergite. Symphoromyia may be 
distinguished from Chrysopilus and Rhagio by the large size of the scape, the two-
segmented palp, and presence of mandibles. Additionally Symphoromyia may be 
separated from Rhagio by having a bare proepimeron and a single hind tibial spur; 
and from Chrysopilus by lacking thoracic setae that have a metallic sheen. 
 
DESCRIPTION.  
Head. Clypeus bulbous. Scape clearly larger than pedicel. First flagellomere of 
antenna laterally compressed, enlarged bearing terminal or anterodorsal arista. Eyes 
inconspicuously setulose; in male, often flattened dorsally, holoptic or dichoptic, 
ommatidia split into dorsal and ventral areas and smaller ventrally. Parafacials in 
male swollen slightly or not swollen. Labella with pseudotrachae, length variable. 
Theca short and stout, with medial suture. Palps two-segmented, distal segment 
longer than proximal segment. Stipes surrounded by membrane above theca, directed 
posteriorly. Lacinia present, longer than palps, with serrated tip. Mandibles present. 
 
 401
Cibarial pump long, clearly not as wide as long. Cornu nearly as long as or longer 
than cibarial pump, extending beyond pharyngeal pump. Pharyngeal pump anteriorly 
broad, forming cup-like structure, approximately same length as cibarial pump. 
Thorax. Thorax. Mesonotum with vittae. Dorsocentral bristles absent; all dorsal setae 
of equal length. Anepisternum setulose throughout posterior half. Laterotergite setose. 
Proscutellum present. Subscutellum inconspicuous. Wing hyaline, without markings 
or membrane lightly to moderately infuscate, brownish. Lower calypter reduced. 
Upper calypter well developed, with broad curvature, lobe-like, width twice length or 
less. Costa extends to wingtip (between R
4
 and R
5
). Humeral crossvein (h) well 
developed. Sc-r crossvein present, weakly developed, positioned distal to the humeral 
crossvein (h), by the approximate length of h. Wing veins R
1
 and R
2+3
 separated at 
wing margin. Dorsal side of R
1
 setulose, ventral side bare. All other wing veins 
without setulae. Wing vein R
2+3
 sinuous, apical third of R
2+3
 ultimately bends toward 
leading edge of wing margin creating concave flex anteriorly. Length of R
2+3
 shorter 
than R
5
. Base of R
4
-R
5
 fork proximal or directly above distal end of cell dm. R
4
 
nearly straight apically. R
5
 posterior or anterior to wing tip, clearly longer than R
4+5
 
(r-m to R
4
 origin). R-m crossvein at proximal one-third to near halfway of discal cell. 
M
3
 wing vein present. M-cu crossvein present. Origin of CuA
1
 at bm cell near discal 
cell. M
3
 cell at margin convergent. CuA
2
 length between ? and 2/3 the length of the 
posterior vein of bm cell. Alula with broad, evenly rounded curvature. Anal lobe well 
developed. Anal cell open. Halter knob approximately 1/2 length of stem. Tibial spur 
formula 0:2:1. Hind coxal tubercle absent. Hind tibial macrochaetae absent. 
Postmetacoxal bridge absent. Abdomen. Abdominal segments evenly tapered. In 
 
 402
female, last 3 abdominal segments telescoped; tergite 7 much wider than long; 
intersegmental membrane between segments 7 and 8 especially long; sternite 8 as 
wide as long or wider than long. Male terminalia with epandrium simple, not 
containing hypandrium ventrally. Epandrial sclerite wider than long, strongly notched 
anteriorly. Subepandrial sclerite present, divided medially, setose. Hypoproct present. 
Cercus base held underneath epandrial sclerite or directly adjacent to epandrial 
sclerite. Cerci partially displaced from one another, separation distance approximately 
half the width of single cercus. Cerci, in posterior view cupped, forming circular 
outline medially. Hypandrium fused entirely to gonocoxites. Gonocoxite with sinuous 
dorsal ridge, leading to gonocoxal apodeme. Gonocoxal apodemes short, usually not 
long enough to reach anterior margin of hypandrium. Parameral sheath forming 
separate, distinct lobes ventrally. Lateral ejaculatory processes present, not part of 
sperm sac posteriorly. Ejaculatory apodeme short or long. Ejaculatory apodeme 
laterally compressed. Aedeagal tines absent. Endoaedeagal process present. Female 
terminalia with tergite 9 entire, with narrow anteriorly-directed ventrolateral 
projections, enveloping sternite 9. Spermathecae three, spherical, moderately to well 
sclerotized. Spermathecal ducts more than three times but less than five times the 
length of sternite 9, inflated at base of spermathecae. Spermathecal duct accessory 
glands present, arise at approximately the distal third of the spermathecal ducts, 
thickened and lightly sclerotized, with furrows, near junction with common 
spermathecal duct. Spermathecal ducts furrows at base present. Common 
spermathecal duct thickened, long, clearly longer than longest diameter of genital 
chamber. Genital chamber teardrop shaped, moderately sized. Accessory gland 
 
 403
posterior to genital chamber inconspicuous, easily overlooked even after staining, 
common duct as long or shorter than sternite 9, with short paired extensions 
posteriorly. Sternite 9 anterior end rounded, posterior end with broad extensions 
posteriorly. Posterior end of sternite 9 joined together in horizontal plane centrally, 
held in vertical plane laterally. Tergite 10 entire, short, length less than half width. 
Sternite 10 split into two sclerites. Cercus two-segmented. First segment of cercus not 
elongate, without ventral process. Basal cerci adjacent dorsally. Ventral lobes of first 
segment of cercus sometimes curve ventrally towards one another to form a ring, 
usually are flat and compressed medially. Second cercus segment not elongated, with 
apical sensory pits. Larva. Body with 11 segments (not including head segment), 
smooth ventrally. Head capsule  not folded within second segment, composed of a 
single, undivided plate (dorsal plate), less than 4.5 times longer than greatest width (2 
width/7 length), not cone-shaped, with hole in dorsal shield around each antenna. 
Mandibular brush present, associated with simple fold of cuticle. Mandibular hook 
with external groove on adoral surface, serrate, otherwise smooth. Dorsal ridge of 
mandibular hook without anteriorly-directed microsetae. Labral teeth developed, 
sclerotized, in two rows, converging anteriorly. Maxilla sclerotized. Saw sclerite of 
mandibular base absent. Maxillary palp soft, segments three, poorly differentiated. 
Antenna last segment bifurcated, as nubs. Antenna apparently one-segmented. 
Unpaired salivary pump absent. Posterior tentorial expansion fused to each other, 
with thin extension produced anteriorly.  
LITERATURE.  
 
 404
Aldrich (1915) gives key to North American species. Narchuk (1969) gives key to 
eastern European species. Turner (1974) proposes species-groups within the genus. 
Sommerman (1962) provides natural history information for the larval stages.  
List of included species 
Species (synonyms indented) Author, reference Type country 
Symphoromyia algens Leonard, 1931: 1 USA 
Symphoromyia atripes Bigot, 1887: 111 USA 
Symphoromyia barbata Aldrich, 1915: 120 USA 
Symphoromyia cervivora Turner & Chillcott, 1973: 6 USA 
Symphoromyia cinerea Johnson, 1903: 25 USA 
Symphoromyia comata Bigot, 1887: 111 USA 
Atherix crassicornis (as Atherix; Panzer, 1806: 10) "Hartz;" Europe 
Atherix grisea Meigen, 1820: 109 not given; Europe 
Leptis griseola Fall?n, 1814: 7 Sweden 
Symphoromyia cruenta Coquillett, 1894: 55 USA 
Symphoromyia currani Leonard, 1931: 2 USA 
Symphoromyia fulvipes Bigot, 1887: 110 USA 
Symphoromyia hirta Johnson, 1897: 120 USA 
Symphoromyia flavipalpis Adams, 1904: 439 USA 
Symphoromyia immaculata (as Atherix; Meigen, 1804: 294) "Wiesen; " Europe 
Symphoromyia inconspicua Turner & Chillcott, 1973: 6 USA 
Symphoromyia incorrupta Yang, Yang, and Nagatomi, 1997: 
251 
China 
Symphoromyia inquisitor Aldrich, 1915: 127 USA 
Symphoromyia inurbana Aldrich, 1915: 127 USA 
Symphoromyia johnsoni Coquillett, 1894: 54 USA 
Symphoromyia kincaidi Aldrich, 1915: 129 USA 
Symphoromyia limata Coquillett, 1894: 54 USA 
Atherix melaena (as Atherix; Meigen, 1820: 109) Nepal 
Atherix maura Meigen, 1820: 109 not given 
Atherix pilosa Meigen, 1820: 109 not given 
Symphoromyia montana Aldrich, 1915: 133 USA 
Symphoromyia nana Turner & Chillcott, 1973: 15 USA 
Symphoromyia pachyceras Williston, 1886: 287 USA 
Symphoromyia pilosa Aldrich, 1915: 135 USA 
Symphoromyia plagens Williston, 1886: 287 USA 
Symphoromyia latipalpis Bigot, 1887: 108 USA 
Symphoromyia picticornis Bigot, 1887: 109 USA 
Symphoromyia pleuralis Curran, 1930: 40 USA 
Symphoromyia plumbea Aldrich, 1915: 138 USA 
Symphoromyia pullata Coquillett, 1894: 56 USA 
 
 405
Species (synonyms indented) Author, reference Type country 
Symphoromyia sackeni Aldrich, 1915: 139 USA 
Symphoromyia securifera Coquillett, 1904: 171 USA 
Symphoromyia sinensis Yang & Yang, 1997: 253 China 
Symphoromyia spitzeri Chv?la, 1983: 425 Uzbekistan 
Symphoromyia trivittata Bigot, 1887: 109 USA 
Symphoromyia fera Coquillett, 1894: 56 USA 
Symphoromyia trucis Coquillett, 1894: 55 USA 
Symphoromyia truncata Turner & Chillcott, 1973: 17 USA 
Symphoromyia varicornis (as Atherix; Loew, 1872: 58) USA 
Symphoromyia modesta Coquillett, 1894: 54 USA 
 
nomina dubia Author, reference 
Symphoromyia picea Walker 1848: 219 
 
Notes. 
1. In the Palearctic Catalogue (Majer, 1988), Arthroceras pollinosum Williston is 
given as the type species for the genus Arthroceras Williston, by original designation. 
Webb (1987) states that the type species is Arthroceras pollinosum Williston by 
?original description.? Both are incorrect. Williston (1886) created Arthroceras for 
two species, Arthroceras leptis (Osten Sacken) and Arthroceras pollinosum Williston 
and did not explicitly designate the type species for the genus. James (1965) correctly 
gives credit to Coquillett (1910) for the type species designation. 
 
2. Lindner (1923) designated the Palaearctic species Chrysopilus obscuripennis Loew 
as the type species for Bicalcar. However, Hennig (1955) located the type material of 
C. obscuripennis and found that there were actually two specimens. One was a typical 
Chrysopilus, labeled as the type, which evidently Lindner never saw, and the other 
was a specimen of Atherimorpha, from an unknown source. The latter specimen was 
 
 406
the basis of Lindner's description. Thus, Lindner cited Chrysopilus obscuripennis 
Loew as the type species of Bicalcar, but had described the genus on the basis of a 
misidentification. Sabrosky (1999) states that the type of Bicalcar is Chrysopila 
obscuripennis Loew = "Atherimorpha obscuripennis (Loew)". This new combination 
is in error, however, since the specimen that Lindner used for the type species of 
Bicalcar was not Chrysopilus; rather, it was an unidentified Atherimorpha species, 
misidentified as Chrysopilus obscuripennis.    
 
3. The original concept of Neorhagio Lindner has remained unresolved because the 
specimen that Lindner identified as Leptis setosa Philippi, the type species he 
designated for the genus, is lost. However, I concur with J. R. Malloch (1932: 206) 
who determined on the basis of his own knowledge of the Chilean fauna and on the 
basis of Philippi?s original descriptions, that all of Philippi?s species described in the 
genus Leptis belong to Atherimorpha White. I have a specimen on loan from the 
AMNH collection that bears on old determination label reading Atherimorpha setosa 
Philippi. ?Angol ? Chile, 12 Ocbre 1933, J Salazar / Atherimorpha setosa Phil. / 
USNMENT00025130.? I designate this specimen as the neotype of Leptis setosa 
Philippi, in the interests of a stable taxonomy.  
 
4. Chrysopilus nagatomii Evenhuis 1994 and Chrysopilus nagatomii Yang &Yang 
1991 are primary homonyms. By the principle of priority, C. nagatomii Yang &Yang 
remains valid and C.nagatomii Evenhuis is a junior homonym. The replacement name 
Chrysopilus meunieri Kerr, new name, is given here, named after the original worker.  
 
 407
 
5. I have seen the type of Chrysopilus arctica Frey and agree with Nagatomi (1982: 
56) and Nartshuk (1995: 18) that this species clearly belongs within Chrysopilus. 
Stuckenberg (1965) first discusses the confusion regarding C. arctica Frey, but did 
not have access to relevant material for making a determination at the time. Nartshuk 
(1995) eloquently describes the confusion regarding this species and justifies the 
correct placement, although she refers to Poppiusiella as Poppinsiella, in lapsus. 
 
6. Nagatomi (1982: 50) comments, in the same passage containing the description of 
Solomomyia, that ?this genus is certainly derived from Chrysopilus.? The recognition 
of this genus, therefore, by Nagatomi?s own admission, renders Chrysopilus a 
paraphyletic group. Solomomyia gressitti is distinguished from Chrysopilus species 
by having wing vein CuA
1
 arising from the discal cell. Vein R
4
 is also unusually long. 
Nagatomi mentions another character: ?the large area behind ocelli does not make an 
acute angle with front and is visible in a direct frontal aspect (i.e., when line from 
antenna to median ocellus is kept horizontal).? I interpret him to mean vertical when 
he says horizontal. If this is the case and my interpretation is correct, this is not an 
unusual character state for females of any genus within Tabanomorpha. While the 
features of the wing in Solomomyia appear unique, relatively minor differences in 
wing venation such as this hardly merit the recognition of species group, in my 
opinion. Intrageneric differences of wing venation (and aberrant wing venation) are 
not uncommon in this area of Diptera. The male genitalia illustrated by Nagatomi 
(1984: Figs. 118-120) are identical to many Chrysopilus species. Since this and all 
 
 408
other character states, aside from the aberrant CuA
1
 origin (and perhaps the long R
4
 
vein), are wholly consistent with the genus Chrysopilus, I do not believe that 
persistence of Solomomyia as a separate genus is justified. Paratypes of Solomomyia 
gressitti are not in the Bernice P. Bishop Museum, Honolulu, Hawaii; the Natural 
History Museum, London; nor at the National Museum of Natural History, 
Washington, D.C. as indicated in Nagatomi?s publication (1982). 
 
7. The species concept of Chrysopilus rufipes Macquart is uncertain, but placement in 
Chrysopilus is established (Hardy, 1920; Oldroyd in Paramonov, 1962) despite 
unplaced recognition for this species in Nagatomi & Evenhuis (1989).  Hardy (1920) 
wrote that this species was as a junior synonym of C. aequalis (Walker) but 
Paramonov (1962:135) preferred to treat these taxa as separate species, citing 
difficulty in interpreting original types. 
 
8. The holotype and paratype were collected in Cape Thompson, Alaska, 26-29 July 
1961, by B.S. Henning. There are two Cape Thompsons in Alaska. The one where 
Litoleptis alaskensis was collected is (most likely) near the Project Chariot 
bioenvironmental study (which was active in the late 1950s to early 1960s). It is 
located on the Chukchi Sea coast, 26 mi SE of Point Hope, Arctic Slope; 68.14?N, 
165.98?W. 
 
9. In the Palearctic Catalogue (Majer, 1988), Ptiolina obscura Fall?n is given as the 
type species for the genus Ptiolina, by original designation. However, Zetterstedt 
 
 409
(1842) placed two species in the new genus Ptiolina, without designating the type 
species. Frauenfeld (1867) designated the type species as Ptiolina obscura (Fall?n) 
(1867: 467, in key), which was originally described as Leptis obscura Fall?n.  
 
10. Yang, Yang & Nagatomi (1997: 256) described Spatulina sinensis from a single 
male specimen from Shaanxi, China and remark that if the new species is not a true 
Spatulina, it would belong to a new genus. The authors note that S. sinensis differs 
from Ptiolina in having mid-upper face deeply sunken and occiput above the neck 
strongly concave.  The head and abdomen are illustrated from the lateral view and the 
male genitalia are also illustrated, however, none of the putatively special features 
that they mention are visible. The specimen lacks antennal segments beyond the 
pedicel and no other potential autapomorphies are given in the text. The male 
genitalia are quite unlike those found in Ptiolina and it is unclear why the species is 
placed in Spatulina. Tergite 10 is split medially into two thin, lateral sclerites and the 
gonostyly are thick, with an obvious inward bend, exactly as it is in some species of 
Chrysopilus. The mid tarsus, hind femur, and all thoracic setae are also missing from 
the type specimen. All features illustrated and described are fully consistent with 
those found in species of Chrysopilus, including the deeply sunken face and concave 
occiput. For this reason, it is more appropriate to place this species in the genus 
Chrysopilus. Chrysopilus sinensis (Yang, Yang & Nagatomi), new combination. 
 
11. All characters of Rhagina are also seen in Rhagio species, except for the 
exceptional wing characters seen in some species. Yang et al. note that Rhagina male 
 
 410
lacks the subepandrial sclerite (tergite 10), whereas in Rhagio, it is present 
(1997:187), however, I find the male genitalia indistinguishable; both lack the 
subepandrial sclerite. The wing in Rhagina incurvatus de Meijere is distinctive, 
however, there appears to be a morphological grade in the group, especially as one 
examines the wing of Rhagina sinensis Yang and Nagatomi (1997:186) which has a 
sinuous R
2+3
 wing vein, but not distinctively so, and not far removed from venation 
found in some R. hirtus (Say) specimens and R. dichomaticus Chillcott. Another 
distinctive feature of Rhagina emphasized by Nagatomi (1982) and Yang et al. (1997) 
is a prominent ventro-apical ?hump? on the hind femur. This is a variable character in 
both Rhagio and Rhagina, however. Although most commonly absent in Rhagio, I 
have seen the ?hump? in an undescribed Rhagio species from Laos. Similarly, I have a 
Rhagina specimen which lacks such a hind tibial hump. Yang et al. indicate that the 
presence or absence of such a hump does not necessarily determine the genus 
(1997:115). 
 
12. No specimen was designated as the type of  Rhagio biroi Szil?dy (1934). 
 
13. The Catalogue of Palaearctic Diptera (Majer, 1982) errs in listing Leptis tristis 
Shummel, 1837:109 twice; as a junior synonym of Ptiolina obscura and as a valid 
species within the genus Rhagio. The species is recognized here as Rhagio tristis 
(Shummel), consistent with the Biosystematic Database of World Diptera 
(http://www.sel.barc.usda.gov/diptera/BIOSYS.HTM).  
 
 
 411
14. Nagatomi (1982) recognized that Neorhagio Lindner was likely a junior synonym 
of Atherimorpha White, but placed several undescribed specimens from Mexico in 
what he referred to as ?Neorhagio?. He gave no explanation why these specimens 
belong to Neorhagio Lindner. Judging from the photos, illustrations, and diagnosis 
given by Nagatomi, the undescribed species of ?Neorhagio? sensu Nagatomi actually 
belong to the new genus Sierramyia. The male genitalia are illustrated by Nagatomi 
(1984: 138). 
 
15. Szil?dy (1934: 264) distinguished Archicera from Ptiolina and Spania by the 
antennal flagellum, which he stated, had faintly visible divisions. This feature, as 
Nagatomi (1982: 54) has noted, is within the morphological variation present within 
Spania (see Nagatomi & Saigusa, 1982). On account of the small size of Archicera 
avarorum and the description of the flagellomere as being lanceolate, this species is 
certainly placed among the Spania and is likely to be a synonym of Spania nigra.  
The holotype of Archicera avarorum has been destroyed [Hungary National History 
Museum, Budapest]. Therefore a neotype, preferrably a specimen from either Austria 
or Croatia, should be designated for Archicera avarorum so that its appropriate 
taxonomic status may be documented formally. 
 
16. Paramonov (1962: 139) states in his diagnosis that Spaniopsis ?only has one spur 
on the hind tibia (often very weak).? All Spaniopsis species, however, have two 
midtibial spurs and the hind tibia lack spurs entirely. 
 
 
 412
17. Previous authors have referenced the genus name to Becker, 1921, page 59. 
However, the original article spans three issues with separate dates, hence the disjunct 
pagination (pp. 41-48, 15 August 1921; pp. 54-64, 15 November 1921; pp. 69-72, 15 
January 1922). Since the new genus name is published in the first installment of the 
article, even though the type species designation and full description is given on page 
59 in November, 1921, the name Parapheromyia was initially validly established in 
the key on page 42 in August, 1921 and this is the date and pagination which should 
be referenced. Becker (1921: 59) states that the type species for Parapheromyia is 
Symphoromyia crassicornix Panzer. This was a lapsus for Atherix crassicornis 
Panzer. 
 
18. Turner (1974: 859) explains that Becker used the name Paraphoromyia in 
subsequent publications in place of Parapheromyia and the subsequent spelling was 
followed by other workers. Furthermore, Symphoromyia is what he calls the 
?coordinate taxon? of Parapheromyia and therefore, should be similar in spelling 
(besides, he notes, Paraphoromyia is euphonious). Therefore, he concludes, the name 
Parapheromyia was introduced by Becker as a lapsus or perhaps by a copier?s or 
printer?s error. However, Parapheromyia is used consistently throughout the paper (in 
the key (p. 42), the generic description (p. 59), and twice in the discussion (p. 72)) 
and the name Paraphoromyia is never proposed. Therefore, there is no clear 
indication that Parapheromyia Becker was used in error in the original publication 
and the proposed emendation of Turner (1974: 859) does not satisfy article 32.5.1 of 
the code (ICZN, 1999). Furthermore, Becker makes no note of a spelling change in 
 
 413
the original paper, nor in corrigenda published by the author or the publisher and 
therefore, the proposed emendation also fails to satisfy article 33.2.1 of the code 
(ICZN, 1999). Since Becker?s original publication, most authors have used 
Paraphoromyia as an incorrect subsequent spelling (article 33.3 of ICZN, 1999). As 
perhaps a surprising new twist to the latest edition of the code, the name 
Paraphoromyia Becker is ultimately validated, therefore, by prevailing usage (ICZN, 
1999: article 33.3.1). 
 
 414
 Appendix A. Figures.  
 
 
Figure 14. Schematic of the male genitalia, showing primary structures.  
 
 
 
Figure 15. Dorsal outline of hypandrium for illustration of Character 118, origin of 
gonocoxal apodeme.  (A) Atherix pachypus is an example where the medial 
margins of the gonocoxal apodems are shorter than the lateral margins. (B) In 
Rhagio plumbeus and (C) Bolbomyia nana the lateral margins of the gonocoxal 
apodemes are shorter than the medial margins.  ga = gonocoxal apodeme. 
 
 415
 
Figure 16.  Aedeagus, lateral view. (A) Chrysopilus sp., with lateral ejaculatory 
apodemes [USNMENT00025242] (B) Suragina concinna, with aedeagal tines 
[USNMENT00025980] (C) Bolbomyia nana, with aedeagal tines. 
[USNMENT00024051] (D) Arthroceras pollinosum, with lateral ejaculatory 
apodemes and aedeagal tines [USNMENT00022737]. Scale bar = 0.1 mm. 
 
 416
 
Figure 17. Lateral view of Austroleptis multimaculata [USNMENT00025739]. Scale 
bar = 0.5 mm. 
 
 
Figure 18. SEM image of antenna of Austroleptis multimaculata. 
 
 417
 
 
Figure 19. Mouthparts of Austroleptis multimaculata [USNMENT00025905]. (A) 
Labellum, ventral view (B) Hypopharynx and associated structures, lateral view 
(C) Palps and associated structures. Scale bar = 0.1 mm. 
 
 418
 
Figure 20. Wing in species of Austroleptis. (A) A. multimaculata 
[USNMENT00025745] (B) Austroleptis sp. (South America) 
[USNMENT00022609]. Scale bar = 0.5 mm. 
 
 419
 
Figure 21. Male genitalia of Austroleptis multimaculata [USNMENT00025739]. (A) 
epandrium, dorsal view (B) hypandrium, dorsal view. Scale bar = 0.1 mm.  
 
 420
 
Figure 22. Internal structures of female genitalia of Austroleptis sp. 
[USNMENT00025761] (A) Sternite 9 and associated structures, dorsal view (B) 
Spermatheca. Scale bar = 0.1 mm.  
 
 421
 
Figure 23. Internal structures of female genitalia of Austroleptis multimaculata 
[USNMENT00024145] (A) Sternite 9 and associated structures, dorsal view (B) 
Detail of genital chamber and ejaculatory process of spermathecal ducts. Scale bar 
= 0.1 mm. 
 
 422
 
Figure 24. Female tergite 10 and cercus of Austroleptis sp. [USNMENT00025761]. 
Scale bar = 0.1 mm. 
 
 423
 
Figure 25. Habitus of Bolbomyia nana. (A) male, [USNMENT00025987] (B) female, 
[USNMENT00022909]. Scale bar = 0.5 mm. 
 
 424
 
Figure 26. Head and antennal form of Bolbomyia nana, male [USNMENT00022909]. 
Scale bar = 0.5 mm. 
 
 425
 
Figure 27. Mouthparts of Bolbomyia nana [USNMENT00025904]. (A) Labellum and 
palps (B) Hypopharynx and associated structures. Scale bar = 0.1 mm. 
 
 426
 
Figure 28. Male genitalia of Bolbomyia nana [USNMENT00024051]. (A) epandrium, 
dorsal view (B) hypandrium, dorsal view. Scale bar = 0.1 mm. 
 
 427
 
Figure 29. Internal structures of the female genitalia of Bolbomyia nana. (A) Sternite 
9 [USNMENT00022946] (B) Spermathecae [USNMENT00022920]. Scale bar = 
0.1 mm. 
 
 428
 
Figure 30. Female terminalia of Bolbomyia nana. (A) Dorsal view 
[USNMENT00022920] (B) Dorsal view of tergite 10, lateral view of left cercus 
[USNMENT00022946]. Scale bar = 0.1 mm. 
 
 
 429
 
Figure 31. Female sternite 8 of Bolbomyia nana [USNMENT00022920]. Scale bar = 
0.1 mm. 
 
 
 
 430
 
Figure 32. Antenna in species of Arthroceras. (A) A. fulvicorne, female, lateral view 
[USNMENT00022731] (B) A. gadi, male, oblique lateral view 
[USNMENT00022629]. Scale bar = 0.5 mm. 
 
 431
 
Figure 33. Antenna in species of Arthroceras. (A) A. pollinosum, female, dorsal view 
[USNMENT00025222] (B) A. leptis, male, oblique lateral view 
[USNMENT00022614]. Scale bar = 0.5 mm. 
 
 432
 
Figure 34. Mouthparts of Arthroceras pollinosum [USNMENT00022625]. (A) 
Labellum, ventral view (B) Hypopharynx, cibarial pump, and associated 
structures (C) Palps and lacinia, ventral view. Scale bar = 0.1 mm. 
 
 433
 
Figure 35. Wing in species of Arthroceras. (A) A. fulvicorne, female, dorsal view 
[USNMENT00022731] (B) A. gadi, female, ventral view [USNMENT00022628]. 
Scale bar = 0.5 mm. 
 
 
 434
 
Figure 36. Wings in species of Arthroceras, dorsal view. (A) A. pollinosum, female 
[USNMENT00025222]. (B) A. leptis, male [USNMENT00022614]. Scale bar = 
0.5 mm. 
 
 435
 
Figure 37. Epandria (male genitalia) in species of Arthroceras, dorsal view. (A) A. 
fulvicorne [USNMENT00022737] (B) A. leptis [USNMENT00022613]. Scale bar 
= 0.1 mm. 
 
 436
 
Figure 38. Dorsal view of hypandrium (male genitalia) in species of Arthroceras. (A) 
A. fulvicorne [USNMENT00022737] (B) A. leptis [USNMENT00022613]. Scale 
bar = 0.1 mm. 
 
 437
 
Figure 39. Aedeagus (male genitalia) in species of Arthroceras. Aedeagal tines (= at) 
are indicated. (A) A. fulvicorne, lateral view [USNMENT00022601] (B) A. 
pollinosum, dorsolateral view [USNMENT00022737]. Scale bar = 0.1 mm. 
 
 438
 
Figure 40. Female genitalia in Arthroceras fulvicorne USNM ENT 00025219. (A) 
Sternite 9, with associated spermathecal structures from the dorsal perspective. 
(B) Detail of spermathecal duct, with accessory gland. Scale bar = 0.1 mm. 
 
 439
 
Figure 41. Female terminalia structures in Arthroceras fulvicorne 
[USNMENT00022736]. (A) Cerci and associated structures, partially dissected 
showing sternite 9 from dorsal perpective (B) Sternite 8, ventral view. Scale bar = 
0.1 mm. 
 
 
 440
 
Figure 42. Arthroteles cinerea, female [USNMENT00023229]. Scale bar = 0.5 mm. 
 
 441
 
Figure 43.  Antenna in species of Arthroteles, lateral view. (A) A. cinerea, female 
[USNMENT00023229] (B) A. bombyliiformis, male [USNMENT00025009]. 
Scale bar = 0.5 mm.  
 
 442
 
 
Figure 44. SEM image of head of Arthroteles bombyliiformis. 
 
 443
 
Figure 45. Wing in species of Arthroteles. (A) A. cinerea, male, ventral view 
[USNMENT00023231] (B) A. bombyliiformis, male, dorsal view 
[USNMENT000233230]. Scale bar = 0.5 mm. 
 
 444
 
Figure 46. Male genitalia of Arthroteles bombyliiformis [USNMENT00024986]. (A) 
Epandrium, ventral view (B) Hypandrium, dorsal view. Scale bar = 0.1 mm. 
 
 445
 
Figure 47. Sternite 9 (female genitalia) and associated structures of Arthroteles 
bombyliiformis. 
  
 
 446
 
Figure 48. External structures of the female terminalia of Arthroteles bombyliiformis 
[USNMENT00025017]. (A) Tergite 10, dorsal view; and cerci, splayed flat, 
lateral view (B) Sternite 8, dorsal view. Scale bar = 0.1 mm. 
 
 447
 
Figure 49. Habitus of Atherimorpha sp. (South America), female 
[USNMENT00023370]. 
 
 
 
Figure 50. Antenna of Atherimorpha atrifemur. Scale bar = 0.5 mm. 
 
 448
 
Figure 51. Macerated heads showing mouthparts of Atherimorpha nemoralis. (A) 
Posterior view [USNMENT28402] (B) Lateral view [USNMENT00025903]. 
Scale bar = 0.1 mm. 
 
 
 449
 
 
Figure 52. Mouthparts of Atherimorpha nemoralis [USNMENT00025903]. (A) 
Labellum (B) Hypopharynx and associated structures (C) Palps and associated 
structures. Scale bar = 0.1 mm. 
 
 450
 
Figure 53. Epandrium in species of Atherimorpha. Dorsal view. (A) A. triangularis 
[USNMENT00028418] (B) Atherimorpha sp. (South America) 
[USNMENT00028520]. Scale bar = 0.1 mm. 
 
 451
 
Figure 54. Hypandrium in species of Atherimorpha. Dorsal view. (A) A. triangularis 
[USNMENT00028418] (B) Atherimorpha sp. (South America) 
[USNMENT00028520]. Scale bar = 0.1 mm. 
 
 452
 
Figure 55. Internal structure of the female genitalia in Atherimorpha nemoralis. (A) 
Dorsal view [USNMENT00025317] (B) Oblique lateral view 
[USNMENT00025109]. Scale bar = 0.1 mm. 
 
 453
 
Figure 56. External structures of the female terminalia of Atherimorpha nemoralis. 
(A) Dorsal view [USNMENT00025317] (B) Oblique lateral view 
[USNMENT00025109]. Scale bar = 0.1 mm. 
 
 454
 
Figure 57. Female sternite 8 of Atherimorpha nemoralis [USNMENT00025109]. 
Dorsal view. Scale bar = 0.1 mm. 
 
 455
 
Figure 58. Antennal form in species of Chrysopilus, lateral view. (A) C. ornatus, 
female [USNMENT00025947] (B) C. quadratus, female [USNMENT00025948]. 
Scale bar = 0.5 mm. 
 
 456
 
Figure 59. Wing of Chrysopilus species, dorsal view. (A) C. ornatus 
[USNMENT00025947] (B) C. quadratus [USNMENT00025948]. Scale bar = 0.5 
mm.  
 
 457
 
Figure 60. Epandrium in species of Chrysopilus, dorsal view. (A) C. quadratus 
[USNMENT00025951] (B) Chrysopilus sp. (New Caledonia) 
[USNMENT00025952] (C) Chrysopilus thoracicus [USNMENT00025242]. 
Scale bar = 0.1 mm. 
 
 458
 
Figure 61. Hypandrium in species of Chrysopilus, dorsal view. (A) C. quadratus 
[USNMENT00025951] (B) Chrysopilus rotundipennis [USNMENT00025980] 
(C) Chrysopilus sp. (New Caledonia) [USNMENT00025952] (D) Chrysopilus 
thoracicus [USNMENT00025242]. Scale bar = 0.1 mm.  
 
 
 
 459
 
Figure 62. Internal structures of the female genitalia in Chrysopilus thoracicus 
[USNMENT00025875], dorsal view. Scale bar = 0.1 mm. 
 
 460
 
Figure 63. Internal structures of the female genitalia in Chrysopilus sp. 
[USNMENT00025877], ventral view. (A) Detail of furrowed and sclerotized 
ejaculatory process of the spermathecal ducts (B) Spermatheca. Scale bar = 0.1 
mm. 
 
 461
 
Figure 64. Spermathecae and spermathecal duct accessory gland in Chrysopilus 
thoracicus [USNMENT00025875]. Scale bar = 0.1 mm. 
 
 462
 
Figure 65. Partially dissected female terminalia of Chrysopilus testaceipes 
[USNMENT00025876], dorsal view. Scale bar = 0.1 mm. 
 
 463
 
Figure 66. Male genitalia of Desmomyia thereviformis [USNMENT00025267]. (A) 
Epandrium, ventral view (B) Hypandrium, dorsal view. Scale bar = 0.1 mm.  
 
 464
 
Figure 67. Sternite 9 (female genitalia) and associated structures of Desmomyia 
thereviformis [USNMENT00025628]. Scale bar = 0.1 mm. 
 
 465
 
Figure 68. Spermathecae of Desmomyia thereviformis [USNMENT00025628]. Scale 
bar = 0.1 mm. 
 
 466
 
Figure 69. Litoleptis alaskensis, male. (A) Habitus, lateral view 
[USNMENT00024416, holotype] (B) head, anterior view [USNMENT00024417, 
paratype]. Scale bar = 0.5 mm. 
 
 467
 
Figure 70. Antennal form in species of Litoleptis. (A) L. alaskensis, male 
[USNMENT00024416, holotype] (B) L. alaskensis, male [USNMENT00024417, 
paratype] (C) L. chilensis [no USNM barcode, slide mounted holotype]. Scale bar 
= 0.1 mm. 
 
 468
 
Figure 71. Wing in species of Litoleptis. (A) L. alaskensis, male 
[USNMENT00024417, paratype] (B) L. chilensis, male [no USNM barcode, slide 
mounted holotype]. Scale bar = 0.5 mm. 
 
 469
 
Figure 72. Male genitalia of Litoleptis alaskensis, holotype [USNMENT00024416]. 
(A) Epandrium, oblique dorsal view (B) Hypandrium, dorsal view. Scale bar = 
0.1 mm. 
 
 470
 
Figure 73. Antennal form in species of Omphalophora. (A) O. fasciata, male 
[USNMENT00025460] (B) O. majuscula, female [USNMENT00025471]. Scale 
bar = 0.5 mm. 
 
 471
 
Figure 74. Antennal form in species of Omphalophora and Ptiolina. (A) O. 
nigripilosa, male [USNMENT00025204 HOLOTYPE] (B) P. nitida, male 
[USNMENT00022957]. Scale bar = 0.5 mm. 
 
 472
 
Figure 75. Wing in species of Omphalophora, ventral view. (A) O. fasciata 
[USNMENT00025460] (B) O. majuscula [USNMENT00025471]. Scale bar = 0.5 
mm. 
 
 473
 
Figure 76. Wing in species of Omphalophora and Ptiolina, ventral view. (A) O. 
nigripilosa [USNMENT00025205, allotype] (B) P. nitida 
[USNMENT00022957]. Scale bar = 0.5 mm. 
 
 474
 
Figure 77. Epandrium (male genitalia) in species of Omphalophora, dorsal view. (A) 
O. fasciata [USNMENT00025461] (B) O. lapponica [USNMENT00025921]. 
Scale bar = 0.1 mm. 
 
 475
 
Figure 78. Epandrium (male genitalia) of Omphalophora majuscula 
[USNMENT00025472]. Scale bar = 0.1 mm. 
 
 476
 
Figure 79. Hypandrium (male genitalia) in species of Omphalophora, dorsal view. 
(A) O. fasciata [USNMENT00025461] (B) O. lapponica [USNMENT00025921]. 
Scale bar = 0.1 mm. 
 
 477
 
Figure 80. Hypandrium (male genitalia) of Omphalophora majuscula 
[USNMENT00025472]. Scale bar = 0.1 mm. 
 
 
 478
 
Figure 81. Dissected female terminalia of Omphalophora majuscula, dorsal view 
[USNMENT00025474]. Scale bar = 0.1 mm. 
 
 479
 
Figure 82. Spermatheca and spermathecal duct accessory glands of Omphalophora 
majuscule [USNMENT00025474]. Scale bar = 0.1 mm. 
 
 
 480
 
Figure 83. Antennal form in species of Ptiolina. (A) P. edeta, male 
[USNMENT00022657] (B) P. edeta, female [USNMENT00023016]. Scale bar = 
0.5 mm. 
 
 
 481
 
Figure 84. Antennal form in species of Ptiolina. (A) P. obscura, male 
[USNMENT00023018] (B) P. zonata, female [USNMENT00022842]. Scale bar 
= 0.5 mm. 
 
 
 482
 
Figure 85. Labellum and theca in species of Ptiolina. (A) Ptiolina sp., male, posterior 
view [USNMENT00022845] (B) P. zonata, male, lateral view 
[USNMENT00023001]. Scale bar = 0.1 mm. 
 
 483
 
Figure 86. Hypopharynx, cibarial pump, and associated structures in species of 
Ptiolina, posterior view. (A) Ptiolina sp., male [USNMENT00022845] (B) P. 
zonata, male [USNMENT00023001]. Scale bar = 0.1 mm. 
 
 484
 
Figure 87. Palp and associated structures in Ptiolina. (A) Ptiolina sp., male 
[USNMENT00022845] (B) P. zonata, male [USNMENT00023001]. Scale bar = 
0.1 mm. 
 
 485
 
Figure 88. Wing in species of Ptiolina, ventral view. (A) P. obscura 
[USNMENT00023020] (B) P. zonata [USNMENT00022842]. Scale bar = 0.5 
mm. 
 
 486
 
Figure 89. Epandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
edeta [USNMENT00023011] (B) P. nigra [USNMENT00025932]. Scale bar = 
0.1 mm. 
 
 487
 
Figure 90. Epandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
nitida [USNMENT00025933] (B) P. obscura [USNMENT00025937]. Scale bar 
= 0.1 mm. 
 
 488
 
Figure 91. Epandrium (male genitalia) of Ptiolina zonata [USNMENT00022846].  
Dorsal view. Note: subepandrial sclerite flipped upside-down, facing wrong 
direction. Scale bar = 0.1 mm. 
 
 489
 
Figure 92. Hypandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
edeta [USNMENT00023011] (B) P. nigra [USNMENT00025932]. Scale bar = 
0.1 mm. 
 
 490
 
Figure 93. Hypandrium (male genitalia) in species of Ptiolina, dorsal view. (A) P. 
nitida [USNMENT00025933] (B) P. obscura [USNMENT00025937]. Scale bar 
= 0.1 mm. 
 
 491
 
Figure 94. Hypandrium (male genitalia) of Ptiolina zonata [USNMENT00022846]. 
Scale bar = 0.1 mm. 
 
 492
 
Figure 95. Dissected female terminalia in species of Ptiolina, dorsal view. (A) P. 
zonata [USNMENT00022841] (B) P. mallochi [USNMENT00022953]. Scale bar 
= 0.1 mm. 
 
 493
 
Figure 96. Spermatheca and spermathecal duct accessory glands in species of 
Ptiolina. (A) P. zonata [USNMENT00022841] (B) P. mallochi 
[USNMENT00022953]. Scale bar = 0.1 mm. 
 
 494
 
Figure 97. Male genitalia of Rhagina incurvatus [USNMENT00022728]. (A) 
Epandrium (B) Hypandrium. Scale bar = 1.0 mm. 
 
 495
 
Figure 98. Female terminalia of Rhagina incurvatus [USNM ENT 00025853]. (A) 
Dissected, showing spermathecae (B) Detail of cerci. Scale bar = 0.1 mm. 
 
 496
 
Figure 99. Internal structures of the female genitalia of Rhagina incurvatus [USNM 
ENT00025853]. (A) Dorsal view of sternite 9 (B) Spermatheca. Scale bar = 0.1 
mm. 
 
 
 497
 
Figure 100. Habitus of Rhagio mystaceus, a common species in the Maryland/ DC 
area. 
 
 498
 
Figure 101. Antennal form in species of Rhagio, lateral view. (A) R. punctipennis, 
male [USNMENT00025946] (B) R. scolopaceus, male [USNMENT00023193]. 
Scale bar = 0.5 mm. 
 
 499
 
Figure 102. SEM image of antenna of Rhagio mystaceus. 
 
 
 500
 
Figure 103. Mouthparts in Rhagio mystaceus. (A) Hypopharynx and associated 
structures, showing clypeus and scape and pedicel of antenna 
[USNMENT00025900] (B) Labellum [USNMENT00025908]. Scale = 0.1 mm. 
 
 501
 
Figure 104. Wing of species of Rhagio, dorsal view. (A) R. punctipennis, male 
[USNMENT00025946] (B) R. scolopaceus, male [USNMENT00023193]. Scale 
bar = 0.5 mm. 
 
 502
 
Figure 105. Epandrium in species of Rhagio, dorsal view. (A) R. mystaceus 
[USNMENT00025230] (B) R. punctipennis [USNMENT00025950]. Scale bar = 
0.1 mm. 
 
 503
 
Figure 106. Hypandrium in species of Rhagio, dorsal view. (A) R. plumbeus 
[USNMENT00025949] (B) R. punctipennis [USNMENT00025950]. Scale bar = 
0.1 mm. 
 
 504
 
Figure 107. Aedeagus (male genitalia) of Rhagio mystaceus [USNMENT00025230]. 
Scale bar = 0.1 mm. 
 
 505
 
Figure 108. Female genitalia of Rhagio palpalis [USNMENT00025879]. Dorsal 
view. (A) Partially dissected terminalia (B) Sternite 9. Scale bar = 0.1 mm. 
 
 506
 
Figure 109. Anterior end of sternite 9 and base of spermathecal ducts at genital 
chamber in species of Rhagio. (A) R. incisus [USNMENT00025873] (B) R. 
mystaceus [USNMENT00025244]. Scale bar = 0.1 mm.  
 
 507
 
Figure 110. Sternite 8 of Rhagio incisus [USNMENT00025873]. Scale bar = 0.1 mm. 
 
 508
 
Figure 111. Cercus in species of Rhagio. (A) R. hirtus [USNMENT00023150] (B) R. 
incisus [USNMENT00025873] (C) R. palpalis [USNMENT00025879], missing 
left second cercus segment. Scale bar = 0.1 mm. 
 
 509
 
 
Figure 112. Lateral view of Schizella furcicornis [USNMENT00025863]. Scale bar = 
0.5 mm. 
 
 510
 
Figure 113. Schizella woodleyi Kerr, new species. Head, anterior view. Scale bar = 
0.1 mm. 
 
Figure 114. Schizella woodleyi Kerr, new species. Wing. Scale bar = 1.0 mm.  
 
 511
 
Figure 115. Schizella woodleyi Kerr, new species. Epandrium, dorsal view. Scale bar 
= 0.1 mm. 
 
Figure 116. Schizella woodleyi Kerr, new species. Epandrium, posterior view. Scale 
bar = 0.1 mm. 
 
 512
 
Figure 117. Schizella woodleyi Kerr, new species. Hypandrium, dorsal view. Scale 
bar = 0.1 mm. 
 
 513
 
Figure 118. Lateral view of head, Sierramyia chiapasensis. 
 
Figure 119. Antenna of Sierramyia chiapasensis. Scale bar = 0.5 mm. 
 
 514
 
Figure 120. Female genitalia of Sierramyia chiapensis [USNMENT00022656]. (A) 
Terminal segments, lateral view (B) Sternite 9. Scale bar = 0.1 mm. 
 
 515
 
Figure 121. Female sternite 8 of Sierramyia chiapensis [USNMENT00022656], 
dorsal view. Scale bar = 0.1 mm. 
 
 
 516
 
Figure 122. Antenna of Spania nigra, male [USNMENT00025865]. (A) lateral view 
(B) anterodorsal view. Scale bar = 0.5 mm. 
 
 517
 
Figure 123. Wing in Spania nigra, male, dorsal view [USNMENT00025865]. Scale 
bar = 0.5 mm.  
 
 518
 
Figure 124. Male genitalia of Spania nigra [USNMENT00025868]. Dorsal view. (A) 
Epandrium (B) Hypandrium. Scale bar = 0.1 mm.  
 
  
 
 519
 
Figure 125. Female terminalia of Spania nigra [USNMENT00024389].  
 
 520
 
Figure 126. Spermatheca (imploded) and spermathecal duct accessory gland of 
Spania nigra [USNMENT00024389]. 
 
 
 521
 
Figure 127. Antennal form in species of Spaniopsis. Lateral view. (A) S. clelandi, 
female [USNMENT00025409] (B) S. longicornis, female 
[USNMENT00022643]. Scale bar = 0.5 mm. 
 
 522
 
Figure 128. Antennal form in species of Spaniopsis. Lateral view. (A) S. mackerrasi, 
male [USNMENT00025556] (B) S. marginipennis, female [USNMENT00025412 
? top specimen on pin]. Scale bar = 0.5 mm. 
 
 523
 
Figure 129. Antennal form in species of Spaniopsis. Lateral view. (A) S. rieki, female 
[USNMENT00022653] (B) S. tabaniformis, female [USNMENT00025568]. 
Scale bar = 0.5 mm. 
 
 524
 
Figure 130. Antenna of Spaniopsis vexans. Female, lateral view 
[USNMENT00025408]. Scale bar = 0.5 mm. 
 
 
 525
 
Figure 131. Mouthparts of Spaniopsis clelandi [USNMENT25907]. (A) Labellum, 
ventral view (B) Hypopharynx, cibarial pump, and associated structures, lateral 
view (C) Palp, mandible, and lacinia; lateral view. Scale bar = 0.1 mm.  
 
 526
 
Figure 132. Wing in species of Spaniopsis, ventral view. (A) S. clelandi 
[USNMENT00022649] (B) S. mackerrasi [USNMENT00025589]. Scale bar = 
0.5 mm. 
 
 527
 
Figure 133. Wing in species of Spaniopsis. (A) S. longicornis, ventral view 
[USNMENT00025293] (B) S. marginipennis, dorsal view 
[USNMENT00025252]. Scale bar = 0.5 mm. 
 
 528
 
Figure 134. Wing in species of Spaniopsis, ventral view. (A) S. rieki 
[USNMENT00022655] (B) S. tabaniformis [USNMENT00025568]. Scale bar = 
0.5 mm. 
 
 529
 
Figure 135. Wing in Spaniopsis vexans, dorsal view [USNMENT00025407]. Scale 
bar = 0.5 mm. 
 
 
 530
 
Figure 136. Epandrium in species of Spaniopsis, ventral view. (A) S. clelandi 
[USNMENT00025396] (B) S. marginipennis [USNMENT00025249]. Scale bar = 
0.1 mm. 
 
 531
 
Figure 137. Hypandrium in species of Spaniopsis, dorsal view. (A) S. clelandi 
[USNMENT00025396] (B) S. marginipennis [USNMENT00025249]. Scale bar = 
0.1 mm. 
 
 532
 
Figure 138. Spermathecae in species of Spaniopsis. (A) S. clelandi 
[USNMENT00025398] (B) S. marginipennis [USNMENT00025251]. Scale bar = 0.1 
mm. 
 
 533
 
Figure 139. Female terminalia in species of Spaniopsis, dorsal view. (A) S. clelandi 
[USNMENT00025398] (B) S. marginipennis [USNMENT00025251]. Scale bar = 
0.1 mm. 
 
 534
 
Figure 140. Stylospania lancifera [USNMENT00025234, holotype]. 
 
 535
mm.  
Figure 141.  Antennal form in Symphoromyia crassicornis. (A) S. crassicornis, 
female [USNMENT00023210] (B) S. crassicornis, male [USNMENT00023208]. 
Scale bar = 0.5 mm. 
 
 536
 
Figure 142. Antennal form in Symphoromyia cruenta (A) female 
[USNMENT00025941] (B) male [USNMENT00025942]. Scale bar = 1.0 mm. 
 
 537
 
Figure 143. Antennal form in Symphoromyia hirta, lateral view (A) male 
[USNMENT00028585] (B) female [USNMENT00028622]. Scale bar = 0.5 mm. 
 
 538
 
Figure 144. Mouthparts of Symphoromyia sp. [USNMENT00025909]. (A) Labellum, 
oblique ventral view (B) Hypopharynx, cibarial pump, and associated structures, 
lateral view (C) Palp and lacinia, lateral view. Scale bar = 0.1 mm.  
 
 539
 
Figure 145. Wing in species of Symphoromyia. (A) S. cruenta, female 
[USNMENT00025941] (B) S. crassicornis, female [USNMENT00023211]. Scale 
bar = 0.5 mm. 
 
 540
 
Figure 146. Wing in species of Symphoromyia. (A) S. hirta, female 
[USNMENT00028605] (B) S. flavipalpis, female [USNMENT00025944]. Scale 
bar = 0.5 mm. 
 
 541
 
Figure 147. Epandrium (male genitalia) in species of Symphoromyia. (A) S. 
crassicornis, dorsal view [USNMENT00028629] (B) S. hirta, ventral view 
[USNMENT00025791]. Scale bar = 0.1 mm. 
 
 542
 
Figure 148. Hypandrium (male genitalia) in species of Symphoromyia, dorsal view. 
(A) S. crassicornis [USNMENT00028629] (B) S. hirta [USNMENT00025791]. 
Scale bar = 0.1 mm. 
 
 543
 
Figure 149. Female genitalia in Symphoromyia hirta [USNMENT00028587]. Scale 
bar = 0.1 mm. 
 
 544
 
Figure 150. Female genitalia in Symphoromyia plagens [USNMENT00025792]. 
Scale bar = 0.1 mm. 
 
 545
 
Figure 151. Dissected terminalia in Symphoromyia species. (A) S. hirta 
[USNMENT00028587] (B) S. plagens [USNMENT00025792]. Scale bar = 0.1 
mm. 
 
 
 546
 
Figure 152. Spermathecae and spermathecal duct accessory glands of Symphoromyia 
plagens [USNMENT00025792]. Scale bar = 0.1 mm. 
 
 547
 
Figure 153. Spermathecae and spermathecal duct accessory glands of Symphoromyia 
hirta [USNMENT00028587]. Scale bar = 0.1 mm.  
 
 548
 
Figure 154. Sternite 8 in species of Symphoromyia, dorsal view. (A) S. hirta 
[USNMENT00028587] (B) S. plagens [USNMENT00025792]. Scale bar = 0.1 
mm. 
 
 549
 
Figure 155. Mouthparts of Glutops rossi [USNMENT00025333]. (A) Labellum, 
lateral view (B) Hypopharynx, posterior view (C) Palps. Scale bar = 0.1 mm. 
 
 550
 
Figure 156. Epandrium in species of Glutops. Dorsal view. (A) G. punctatus 
[USNMENT00025327] (B)  G. rossi [USNMENT00025231] (missing tergite 10). 
Scale bar = 0.1 mm. 
 
 551
 
Figure 157. Hypandrium in species of Glutops. Dorsal view. (A) G. punctatus 
[USNMENT00025327] (B) G. rossi [USNMENT00025231]. Scale bar = 0.1 mm. 
 
 552
 
Figure 158. Female genitalia of Glutops singularis [USNMENT00025338]. (A) 
Dissected terminalia (B) Spermathecae. Scale bar = 0.1 mm. 
 
 553
 
Figure 159. (A) Female sternites 7 and 8, showing long intersegmental region of 
Glutops singularis [USNMENT00025338] (B) Detail of sternite 8, lateral view 
(same specimen). Scale bar = 0.1 mm. 
 
 554
 
Figure 160. SEM image of antenna of Pelecorhynchus fusconiger. 
 
 555
 
Figure 161. Labellum in species of Pelecorhynchus. (A) P. fusconiger 
[USNMENT00025360] (B) P. personatus [USNMENT00025385]. Scale bar = 
0.1 mm. 
 
 556
 
Figure 162. Internal mouthparts in species of Pelecorhynchus, lateral view. (A) P. 
fusconiger [USNMENT00025360] (B) P. personatus [USNMENT00025385]. 
Scale bar = 0.1 mm. 
 
 557
 
Figure 163. Palps and associated structures in species of Pelecorhynchus. (A) P. 
fusconiger [USNMENT00025360] (B) P. personatus [USNMENT00025385]. 
Scale bar = 0.1 mm. 
 
 558
 
Figure 164. Hypandrium in species of Pelecorhynchus. (A) P. fusconiger, ventral 
view [USNMENT00025897] (B) P. personatus, dorsal view 
[USNMENT00025896]. Scale bar = 0.1 mm. 
 
 559
 
Figure 165. Epandrium in Pelecorhynchus personatus [USNMENT00025896]. 
Dorsal view. Scale bar = 0.1 mm.  
 
 560
 
Figure 166. Female terminalia in species of Pelecorhynchus elegans 
[USNMENT00025881], lateral view. (A) Partially dissected (B) Detail of cercus. 
Scale bar = 0.1 mm. 
 
 561
 
Figure 167. (A) Spermatheca of Pelecorhynchus fusconiger [USNMENT00025883] 
(B) Sternite 8, showing intersegmental membrane between sternites 7 and 8 in 
Pelecorhynchus personatus [USNMENT00025882]. Scale bar = 0.1 mm. 
 
 562
 
Figure 168. Tergite 8 of female Pelecorhynchus, showing ducts. (A) P. elegans 
[USNMENT00025880], anterior view (B) P. personatus [USNMENT00025882], 
ventral view. Scale bar = 0.1 mm. 
 
 563
 
Figure 169. Left lateral view of Pseudoerinna jonesi [USNMENT00025319]. Scale 
bar = 0.5 mm. 
 
 
 
 564
 
Figure 170. Internal structures of female genitalia of Pseudoerinna jonesi 
[USNMENT00025319]. (A) Sternite 9 and associated structures (B) Spermatheca. 
Scale bar = 0.1 mm. 
 
 565
 
Figure 171. External structures of female genitalia of Pseudoerinna jonesi 
[USNMENT00025319]. (A) Cerci (B) Sternite 8, with proceeding intersegmental 
membrane. Scale bar = 0.1 mm. 
 
 566
 
Figure 172. Lampromyia sp. [USNMENT00025955]. (A) Epandrium, ventral view 
(B) Hypandrium, dorsal view. Scale bar = 0.1 mm. 
 
 567
 
Figure 173. Male genitalia of Vermileo vermileo [USNMENT00025956]. (A) 
Epandrium, ventral view (B) Hypandrium, dorsal view. Scale bar = 0.1 mm. 
 
 568
 
Figure 174. Structures of the female terminalia of Vermileo vermileo 
[USNMENT00025793]. (A) Partially dissected terminalia, showing highly 
autapomorphic common spermathecal duct (B) Spermatheca. Scale bar = 0.1 mm. 
 
 569
 
Figure 175. Thoracic structure of Chrysopilus quadratus, lateral view. (A) Entire 
thorax (B) Detail showing reduced proepimeron (D) Detail of postspiracle and 
halter. 
 
 570
 
Figure 176. Thoracic structure of Rhagio mystaceus, lateral view. (A) Entire thorax 
(B) Detail showing anepisternum and setose proepimeron (D) Detail of 
postspiracle and halter. 
 
 571
 
 
Figure 177. Postspiracular scale of Pelecorhynchus fusconiger. 
 
 
 572
 
Figure 178. SEM of haltere of Austroleptis multimaculata.  
 
 573
 
Figure 179. SEM of halter of Bolbomyia nana. 
 
 574
 
 
Figure 180. SEM of haltere in Arthroceras leptis.  
 
 
 575
 
Figure 181. SEM of halter of Arthroteles bombyliiformis. 
 
 576
 
Figure 182. SEM images of haltere in Desmomyia thereviformis. 
 
 
 
 577
 
Figure 183. SEM images of halter in Ptiolina species. 
 
 
 578
 
Figure 184. SEM images of halter of  Ptiolina edeta. 
 
 579
 
Figure 185. Anterior view of larval head Symphoromyia sp. Abbreviations: ant = 
antenna; lb = labrum; lbt = labral teeth; lc = lacinia; mb = mandibular brush; mh = 
mandibular hook; mx = maxilla; mp = maxillary palpus. 
 
 
 
 580
 
Figure 186. SEM images of the larval head. (A) Chrysopilus sp. (B) Rhagio sp. (C, D) 
Symphoromyia sp. (E) Pelecorhynchus sp. (F) Glutops rossi. 
 
 
 581
 
Figure 187. SEM images of the larval hind segment, posterior view (except in B, 
where oblique ventral view). (A) Symphoromyia sp. (B) Rhagio sp. (C, D) 
Pelecorhynchus sp. 
 
 
 
 
 
Figure 188. Habitus of Ptiolina sp. larva, lateral view. 
 
 582
 
Figure 189. Larva of Ptiolina sp. Anterior view, bright field illumination. 
 
 
Figure 190. Lateral view of undissected Ptiolina sp. larva. Notice scalloping of first 
thoracic segment. 
 
 
 583
 
Figure 191. Lateral view of Ptiolina sp. larval head. Notice surface texture in absence 
of mandibular brush. Bright field illumination. 
 
 
Figure 192. Oblique dorsal view of dissected Ptiolina sp. larval head. Bright field 
illumination. 
 
 584
 
Figure 193. Adoral view of basal mandibular sclerite and mandibular hook of Rhagio 
sp., showing saw sclerite. Nomarski illumination. 
 
Figure 194. Mandibular hook of Symphoromyia sp., showing adoral groove. 
Nomarski illumination. 
 
 585
 
Figure 195. Larva of Dasyomma sp., lateral view, bright field illumination. Narrow, 
sclerotized palp evident.  
 
 
Figure 196. Larva of Dasyomma sp., lateral view, bright field illumination. Rod 
associated with mandibular brush and inner groove of mandibular hook evident. 
 
 586
 
Figure 197. Larva of Tabanus sp., with head capsule extended from (pulled out of) 
first thoracic segment. Lateral view. 
 
 
 
  
  
 
 587
References. 
Aldrich, J. M. 1905. A Catalogue of North American Diptera. Smithsonian 
Miscellaneous Collections 46:680. 
Amorim, D. S., and V. C. Silva. 2002. How far advanced was Diptera evolution in the 
Pangaea? Annales de la Societe Entomologique de France 38:177-200. 
Baker, R. H., and J. Gatesy. 2002. Is morphology still relevant? Pages 163-174 in 
Molecular systematics and evolution: Theory and practice (R. DeSalle, G. 
Giribet, and W. Wheeler, eds.). Birkh?user Verlag, Basel, Switzerland. 
Baum, B. 1992. Combining trees as a way of combining data sets for phylogenetic 
inference, and the desirability of combining gene trees. Taxon 41:3-10. 
Becker, T. 1900a. Beitr?ge zur Dipteren-Fauna Sibiriens. Acta Societatis Scientiarum 
Fennicae 26:3-66. 
Becker, T. 1900b. Bemerkungen zur Synonymie einiger Gattungen der nideren 
Brachycera (Diptera: Rhagionidae und Stratiomyidae). Acta Societatis 
Scientiarum Fennicae 26:3-66. 
Becker, T. 1900c. Die Leptiden Formen im Gebiete der Europaisch-Asitischen und 
Mittelmeer-Fauna. Entomologisches Nachrichten 26:97-116. 
Becker, T. 1921. Neue Dipteren Meiner Sammlung. Beitr?ge zur Systematischen 
Insektenkunde Berlin 2:41-48. 
Bezzi, M. 1903. Katalog der Pal?arktischen Dipteren, Budapest. 
Bezzi, M. 1917. Studies in Philippine Diptera. II. Philippine Journal of Science 
12:107-161. 
 
 588
Bezzi, M. 1926. South African Rhagionidae (Diptera) in the South African Museum. 
Annals of the South African Museum 23:297-324. 
Bigot, J. M. F. 1856. Essai d'une classification g?n?rale et synoptique de l'ordre des 
insectes Dipteres.  IIe. Annales de la Societe Entomologique de France ser. 3, 
4:51-91. 
Bigot, J. M. F. 1887. Dipt?res nouveaux ou peu connus. Leptidi J.B. Bulletin de la 
Societe Zoologique de France 12:97-118. 
Brauer, O. F. 1880. Die Zweifl?gler des Kaiserlichen Museums zu Wien.  I. 
Denkschriften der Mathematisch-Naturwissenschaftlichen classe der 
Kaiserlichen Akademie der Wissenschaften 42:1-112. 
Brauer, O. F. 1883. Die Zweifl?gler des Kaiserlichen Museums zu Wien.  III.  
Systematische studein auf Grundlage der Dipteren-Larven nebst einer 
Zusammenstellung von Beispielen aus der Literatur ?ber dieselben und 
Beschreibung neuer Formen. Denkschriften der Mathematisch-
Naturwissenschaftlichen classe der Kaiserlichen Akademie der 
Wissenschaften 47:1-100. 
Brindle, A. 1959. Notes on the larvae of the British Rhagionidae and Stratiomyidae 
with a key to the genera. Entomologist's Record 71:126-133. 
Brown, B. V. 2001. Flies, gnats, and mosquitos. Pages 815-826 in Encyclopedia of 
Biodiversity (S. A. Levin, ed.) Academic Press, San Diego. 
Brunetti, E. 1912. New Oriental Diptera, I. Records of the Indian Museum 7:445-513. 
 
 589
Bull, J. J., J. P. Huelsenbeck, C. W. Cunningham, D. L. Swofford, and P. J. Waddell. 
1993. Partitioning and Combining Data in Phylogenetic Analysis. Systematic 
Biology 42:384-397. 
Burgess, E. 1878. Two interesting American Diptera, Glutops singularis and Epibates 
osten-sackenii. Boston Society of Natural History, Proceedings 19:320-324. 
Carreno, R. A., and J. R. Barta. 1998. Small subunit ribosomal RNA genes of 
tabanids and hippoboscids (Diptera: Brachycera): evolutionary relationships 
and comparison with other Diptera. Journal of Medical Entomology 35:1002-
1006. 
Chillcott, J. G. 1961. The Genus Bolbomyia Loew (Diptera: Rhagionidae). Canadian 
Entomologist 93:632-636. 
Chillcott, J. G. 1963. A new genus of Rhagionidae (Diptera) with notes and 
descriptions of Bolbomyia Loew. Canadian Entomologist 95:1185-1190. 
Chillcott, J. G. 1965. A Revision of the Eastern Nearctic Species of Rhagio Fabricius 
(Diptera: Rhagionidae). Canadian Entomologist 97:785-795. 
Chvala, M. 1983. First record of blood-sucking in palaearctic Rhagionidae (Diptera), 
with description of a new Symphoromyia of the crassicornis-group from 
Central Asia. Acta Entomologica Bohemoslovaca 80:423-436. 
Cockerell, T. D. A. 1923. Symphoromyia hirta Johnson annoying in Colorado (Dipt.; 
Leptidae). Entomological News 34:21. 
Colless, D. H., and D. K. McAlpine. 1991. Diptera. Pages 717-? in Insects of 
Australia Melbourne University Press, Melbourne. 
 
 590
Coquillett, D. W. 1910. The type-species of the North American genera of Diptera. 
Proc. U.S. National Museum 37:499-647. 
Courtney, G. W., B. J. Sinclair, and R. Meier. 1997. Morphology and terminology of 
Diptera larvae. Pages 85-161 in Contributions to a manual of Palaearctic 
Diptera, vol. 2 (L. Papp, and B. Darvas, eds.). Science Herald, Budapest. 
de Queiroz, A., M. J. Donoghue, and J. Kim. 1995. Separate versus combined 
analysis of phylogenetic evidence. Annual review of Ecology and Systematics 
26:657-681. 
Douady, C. J., F. Delsuc, Y. Boucher, W. F. Doolittle, and E. J. P. Douzery. 2003. 
Comparison of Bayesian and maximum likelihood bootstrap measures of 
phylogenetic reliability. Molecular Biology and Evolution 20:248-254. 
Enderlein, G. 1922. Ein neues Tabanidensystem. Mitteilungen aus dem Zoologischen 
Museum in Berlin 10:333-351. 
Evenhuis, N. L. 1994. Cataloge of the Fossil Flies of the World (Insecta: Diptera). 
Backhuys Publishers, Leiden. 
Fabricius, J. C. 1775. Systema entomologiae, sistens insectorum classes, ordines, 
genera, species adiectis synonymis, locis, desriptionibus, observationibus., 
Flensburg and Leipzig. 
Fabricius, J. C. 1805. Systema antliatorum secundum ordines, genera, species., 
Brunswick. 
Felsenstein, J. 1985. Confidence limits on phylogenies: An approach using the 
bootstrap. Evolution 39:783-791. 
 
 591
Felsenstein, J. 2004. Inferring Phylogenies. Sinauer Associates, Inc., Sunderland, 
Massachusetts. 
Ferguson, E. W. 1915. Descriptions of new Australian blood-sucking flies to the 
family Leptidae. Journal of the Royal Society of New South Wales 49:233-
243. 
Foote, B. A. 1991. Rhagionidae (Tabanoidea). Pages 975 in Immature Insects (F. W. 
Stehr, ed.) Kendall / Hunt, Dubuque, Iowa. 
Frauenfeld, G. R. v. 1867. Zoologische Miscellen. Zool.-Bot. Gesell. Wien, Verhandl. 
17:425-502. 
Frey, R. 1954. Studien ?ber ostasiatische Dipteren.  III. Rhachiceridae, Rhagionidae, 
Hilarimorphidae. Natulae Entomologicae 34:1-25. 
Friedrich, M., and D. Tautz. 1997. Evolution and phylogeny of the diptera: A 
molecular phylogenetic analysis using 28S rDNA sequences. Systematic 
Biology 46:674-698. 
Frohne, W. C. 1953a. Another snipe fly pest in Alaska. Mosquito News 13:25. 
Frohne, W. C. 1953b. Habitat and males of Alaskan snipe fly pests, (Symphoromyia: 
Rhagionidae). Mosquito News 17:94-96. 
Frohne, W. C. 1959. Observations on Alaskan snipe-fly pests. Mosquito News 
19:172-178. 
Gatesy, J., C. Matthee, R. DeSalle, and C. Hayashi. 2002. Resolution of a 
supertree/supermatrix paradox. Systematic Biology 51:652-664. 
 
 592
Greene, C. T. 1926. Descriptions of larvae and pupae of two-winged flies belonging 
to the family Leptidae. Proceedings of the United States National Museum 
70:1-20. 
Griffiths, G. C. D. 1994. Relationships among the major subgroups of Brachycera 
(Diptera): a critical review. Canadian Entomologist 126:861-880. 
Grimaldi, D., and J. Cumming. 1999. Brachyceran Diptera in Cretaceous ambers and 
Mesozoic diversification of the Eremoneura. Bulletin of the American 
Museum of Natural History 239:124. 
Haliday, A. H. 1833. Catalogue of Diptera occurring about Hlywood in Downshire. 
Ent. Mag. (London) 1:147-180. 
Hardy, D. E., and J. U. McGuire. 1947. The nearctic Ptiolina (Rhagionidae-Diptera). 
Journal of the Kansas Entomological Society 20:1-15. 
Hardy, G. H. 1920a. Australian Rhyphidae and Leptidae (Diptera). Papers and 
Proceedings of the Royal Society Tasmania 1919:117-129. 
Hardy, G. H. 1920b. Notes on flies of the genus Pelecorhynchus (Tabanidae). 
Records of the Australian Museum 13. 
Hardy, G. H. 1955. The Phylogeny of Diptera 4.- Tabanoidea. Entomologist's 
Monthly Magazine 91:193-196. 
Heim, F., and C. Leprevost. 1892. Communication. Bull. Soc. Entomol. 61:155-157. 
Hennig, W. 1952. Die Larvenformen der Dipteren, Berlin. 
Hennig, W. 1955. Bemerkungen zur Synonymie einiger Gattungen der nideren. 
Beitrage zur Entomologie 5:426-428. 
 
 593
Hennig, W. 1967. Die sogenannten 'nideren Brachycera' im baltischen Bernstein 
(Diptera: Fam. Xylophagidae, Xylomyidae, Rhagionidae, Tabanidae). 
Stuttgarter Beitr?ge zur Naturkunde 174:1-51. 
Hennig, W. 1968. Kritische Bemerkungen ?ber den Bau der Fl?gelwurzel bei den 
Dipteren und die Frage nach der Monophylie der Nematocera. Stuttgarter 
Beitr?ge zur Naturkunde 193:1-23. 
Hennig, W. 1972. Eine neue Art der Rhagionidengattung Litoleptis aus Chile, mit 
Bemerkungen ?ber F?hlerbildung und Verwandtschaftsbexiebungen einiger 
Brachycerenfamilien (Diptera: Brachycera). Stuttgarter Beitr?ge zur 
Naturkunde 174:1-51. 
Hennig, W. 1973. Diptera (Zweiflugler). Pages 2-31 in Handbuch der Zoologie. 
Hillis, D. 1998. Taxonomic sampling, phylogenetic accuracy, and investigator bias. 
Systematic Biology 47:3-8. 
Hillis, D. M., J. P. Huelsenbeck, and C. W. Cunningham. 1994. Application and 
Accuracy of Molecular Phylogenies. Science 264:671-677. 
Hillis, D. M., D. D. Pollock, J. A. McGuire, and D. J. Zwickl. 2003. Is sparse taxon 
sampling a problem for phylogenetic inference? Systematic Biology 52:124-
126. 
Hillis, D. M., and J. J. Wiens. 2000. Molecules versus morphology in systematics: 
Conflicts, artifacts, and misconceptions. Pages 1-19 in Phylogenetic analysis 
of morphological data (J. J. Wiens, ed.) Smithsonian Institution Press, 
Washington, D.C. 
 
 594
Huelsenbeck, J. P. 1995. Performance of Phylogenetic Methods in Simulation. 
Systematic Biology 44:17-48. 
Huelsenbeck, J. P., and J. P. Bollback. 2001. Empirical and hierarchical Bayesian 
estimation of ancestral states. Systematic Biology 50:351-366. 
Huelsenbeck, J. P., and K. A. Crandall. 1997. Phylogeny estimation and hypothesis 
testing using maximum likelihood. Annual Review of Ecology and 
Systematics 28:437-466. 
Huelsenbeck, J. P., B. Larget, R. E. Miller, and F. Ronquist. 2002. Potential 
applications and pitfalls of Bayesian inference of phylogeny. Systematic 
Biology 51:673-688. 
Huelsenbeck, J. P., and B. Rannala. 1997. Phylogenetic methods come of age: 
Testing hypotheses in an evolutionary context. Science 276:227-232. 
Huelsenbeck, J. P., F. Ronquist, R. Nielsen, and J. P. Bollback. 2001. Evolution - 
Bayesian inference of phylogeny and its impact on evolutionary biology. 
Science 294:2310-2314. 
James, M. T. 1965. Family Rhagionidae (Leptidae)in A Catalog of the Diptera of 
America North of Mexico (A. Stone, C. W. Sabrosky, W. W. Wirth, R. H. 
Foote, and J. R. Coulson, eds.). Agricultural Research Service, United States 
Department of Agriculture, Washington, D. C. 
James, M. T. 1968. 29. Family Rhagionidae (Leptidae). Pages 29.1-29.12 in A 
Catalogue of the Diptera of the Americas South of the United States 
Departamento de Zoologia, Secretaria da Agricultura, Sao Paulo. 
 
 595
James, M. T. 1981. Xylophagidae. Pages 489-492 in Manual of Nearctic Diptera (J. 
F. McAlpine, ed.) Research Branch Agriculture Canada, Ottawa, Ontario. 
James, M. T., and W. J. Turner. 1981. Rhagionidae. Pages 483-488 in Manual of 
Nearctic Diptera (J. F. McAlpine, ed.) Research Branch Agriculture Canada, 
Ottawa, Ontario. 
Jenner, R. A. 2004. Accepting partnership by submission? Morphological 
phylogenetics in a molecular millennium. Systematic Biology 53:333-342. 
Kellogg, V. L. 1908. American Insects. Henry Holt and Company, New York. 
Kert?sz, K. 1908. Catalogus Dipterorum Hucusque Descriptorum, Leipzig, Budapest. 
Knab, F., and R. A. Cooley. 1912. Symphoromyia as a blood-sucker (Diptera, 
Leptidae). Proceedings of the Entomological Society of Washington 14:161-
162. 
Kovalev, V. G. 1981. The oldest representatives of the Diptera with short antennae 
from the Jurassic in Siberia. Paleontological Journal 15:84-100. 
Kovalev, V. G. 1982. Some Jurassic Diptera-rhagionids (Muscida, Rhagionidae). 
Paleontological Journal 16:87-99. 
Kovalev, V. G., and M. B. Mostovski. 1997. A new genus of snipe flies (Diptera: 
Rhagionidae) from the Mesozoic of Eastern Transbaikalia. Paleontological 
Journal 31:523-527. 
Krivosheina, K. P. 1991. Phylogeny of lower Brachycera (Dipter): A new view. Acta  
Entomologica Bohemoslovaca 88:81-92. 
Krivosheina, N. P. 1967. Comparative characteristics of the larva of Arthropeas 
sibirica Loew (Diptera, Xylophagidae). Zoologicheskii Zhurnal 46:954-956. 
 
 596
Krivosheina, N. P. 1971. The family Glutopidae, Fam. n. and its position in the 
system of Diptera Brachycera Orthorrhapha [In Russian]. 
Entomologischesdoe Obozrenie 50:681-694. 
Krzeminski, W. 1992. Triassic and Lower Jurassic stages of Dipteran Evolution. 
Mitteilungen der Schweizerischen Entomologischen Gesellschaft 65:39-59. 
Krzeminski, W. Year. Origin and the first stage of evolution of the Diptera 
Brachycera in IVth International Congress of Dipterology, Oxford 
University:113-114. 
Lameere, A. 1906. Notes pour la classification des Dipteres. Mem. Soc. Ent. Belg. 
12:105-140. 
Lane, R. S., and J. R. Anderson. 1982. Breeding sites of snipe flies (Rhagionidae) and 
other Diptera in woodland-grass soils. Journal Of Medical Entomology 
19:104-108. 
Latreille, P. A. 1802. Histoire naturelle, g?n?rale et particuli?re des Crustac?s et des 
Insectes.  Tome troisi?me. Pages 468 in Histoire naturelle par Buffon, 
nouvelle edition accompagn? de notes. (C. S. Sonnini, ed.) Dufart, Paris. 
Latreille, P. A. 1804. Histoire naturelle, g?n?rale et particuli?re des Crustac?s et des 
Insectes. Ouvrage faisant suite aux oeuvres de Leclerc de Buffon et partie du 
cours complet d'Histoire naturelle r?dig? p. C. S. Sonnini. Pages 415 pp. 
Dufart, Paris. 
Latreille, P. A. 1809. Genera Crustaceorum et Insectorum. Amand Koenig, Paris. 
 
 597
LeCointre, G., H. Philippe, H. L. V. Le, and H. Le Guyader. 1993. Species sampling 
has a major impact on phylogenetic inference. Molecular Phylogenetics and 
Evolution 2:205-224. 
Leonard, M. D. 1930. A revision of the Dipterous family Rhagionidae (Leptidae) in 
the United States and Canada. Memoirs of the American Entomological 
Society 7:181. 
Leonard, M. D. 1931. Some notes on my revision of the Rhagionidae (Diptera). 
Transactions of the American Entomological Society 57:321-323. 
Lewis, P. O. 2001. A likelihood approach to estimating phylogeny from discrete 
morphological character data. Systematic Biology 50:913-925. 
Lindner, E. 1923. Dipterologische Studien. I. Rhagionidae. Konowia 2:1-11. 
Lindner, E. 1924. Dipterologische Studien. II. Rhagionidae neotropicae. Konowia 
3:65-75. 
Lindner, E. 1925. Neue exotische Dipteren (Rhagionidae et Tabanidae). Konowia 
4:20-24. 
Linnaeus, C. 1758. Systema Naturae per regna tria naturae, secundum classes, 
ordines, genera, species, cum caracteribus, differentiis, synonymis, locis. 
Tomus I. Edito decima, reformata. Laurentii Salvii, Holmiae [=Stockholm]. 
Loew, H. 1850. Ueber den Bernstein und die Bernsteinfauna. Programm der 
K?niglichen Realschule zu Meseritz 1850:1-44. 
Loew, H. 1862. Diptera Americae septentrionalis indigena.  Centuria secunda. 
Berliner entomologische Zeitschrift 6:185-232. 
 
 598
Ludwig, P., U. Smola, and R. Melzer. 1996. Die Mundwerkzeuge des Wurml?wen 
Vermileo vermileo L. und ihre Funktion (Diptera, Vermileonidae). 
Nachrichtenblatt Der Bayerischen Entomologen 45:9-14. 
Mackerras, I. M., and M. E. Fuller. 1942. The genus Pelecorhynchus (Diptera, 
Tabanoidea). Proceedings Linnean Society of New South Wales 67:9-76. 
Macquart, J. 1826. Insectes Dipt?res du nord de la France.  Asiliques, Bombyliers, 
Xylotomes, Leptides, V?siculeux, Stratiomydes, Xylophagites, Tabaniens. L. 
Danel, Lille. 
Macquart, J. 1834. Histoire naturelle des Insectes.-Dipt?res, Tome premier. Roret, 
Paris. 
Macquart, J. 1840. Dipt?res exotiques nouveaux ou peu connus. M?moires de la 
Societe Royal de Sciences, de l'Agriculture et des Arts, Lille 1840:283-413. 
Macquart, J. 1850. Dipt?res exotiques nouveaux ou peu connus. 4e. suppl?ment. 
M?moires de la Societe Royal de Sciences, de l'Agriculture et des Arts, Lille 
1849:309-479. 
Macquart, J. 1855. Dipt?res exotiques nouveaux ou peu connus. 5e. suppl?ment. 
M?moires de la Societe Royal de Sciences, de l'Agriculture et des Arts, Lille 
1:25-156. 
Maddison, D., and W. Maddison. 2000. MacClade 4: Analysis of Phylogeny and 
Character Evolution. Sinauer Associates, Inc. 
Majer, J. 1988. Family Rhagionidae. Pages 14-29 in Catalogue of Palaearctic Diptera 
(A. Soos, and L. Papp, eds.). Elsevier Science Publishers & Akademiai Kaido, 
Hungary. 
 
 599
Malloch, J. R. 1917. A preliminary classification of Diptera, exclusive of pupipara, 
based upon larval and pupal characters, with keys to imagines in certain 
families.  Part I. Bulletin of the Illinois State Laboratory of Natural 
History:161-409. 
Malloch, J. R. 1931. Notes on Australian Diptera. XXVIII. Proc. Linn. Soc. N. S. 
Wales 51:273-276. 
Malloch, J. R. 1932a. Diptera of Patagonia and South Chile.  Part V, fascicle 3.- 
Rhagionidae (Leptidae), Therevidae, Scenopinidae, Mydaidae, Asilidae, 
Lonchopteridae. Pages 199-257 in Diptera of Patagonia and South Chile (B. 
M. o. N. History, ed.), London. 
Malloch, J. R. 1932b. Notes on exotic Diptera (1). Stylops 1:112-119. 
Meigen, J. W. 1820. Systematische Beschreibung der bekannten europ?ischen 
zweifl?geligen Insekten., 1 edition, Aachen. 
Mik, J. 1881. Ueber die systematische Stellung des Genus Hilarimorpha Schin. 
Verhandl. Zool-Bot. Gesellsch. Wien:327-329. 
Mik, J. 1896. Dipterologische Miscellen (2 Serie). Entomologische Zeitung 15:241-
278. 
Mostovski, M. B., and E. A. Jarzembowski. 2000. The First Brachycerous Flies 
(Diptera: Rhagionidae) from the Lower Jurassic of Gondwana. 
Paleontological Journal 34:367-369. 
Nagatomi, A. 1952. New Rhagio-species from Japan (1) (Diptera, Rhagionidae). 
Mushi 24:7-12. 
 
 600
Nagatomi, A. 1966. The Arthroceras of the world (Diptera: Rhagionidae). Pacific 
Insects 8:43-60. 
Nagatomi, A. 1977. Classification of Lower Brachycera (Diptera). Journal of Natural 
History 11:321-335. 
Nagatomi, A. 1982a. The genera of Rhagionidae (Diptera). Journal Of Natural 
History 16:31-70. 
Nagatomi, A. 1982b. The genus Pseudoerinna (=Bequaertomyia) (Diptera, 
Rhagionidae). Kontyu 50:97-99. 
Nagatomi, A. 1982c. Geographical distribution of the lower Brachycera (Diptera). 
Pacific Insects 24:139-150. 
Nagatomi, A. 1982d. Geographical-Distribution of the Lower Brachycera (Diptera). 
Pacific Insects 24:139-150. 
Nagatomi, A. 1982e. Ninth sternum and genital fork in female Nematocera and 
Orthorrhaphous Brachycera (Diptera). Kontyu 50:95-96. 
Nagatomi, A. 1984. Male genitalia of the lower Brachycera (Diptera). Beitrage zur 
Entomologie 34:99-157. 
Nagatomi, A. 1986. The Japanese Ptiolina (Diptera, Rhagionidae). Kontyu 54:309-
323. 
Nagatomi, A. 1991. History of some families of Diptera, chiefly those of the lower 
Brachycera (Insecta: Diptera). Bulletin of the Biogeographical Society of 
Japan 46:21-38. 
Nagatomi, A. 1992. Notes on the Phylogeny of Various Taxa of the Orthorrhaphous 
Brachycera (Insecta, Diptera). Zoological Science 9:843-857. 
 
 601
Nagatomi, A., and K. Iwata. 1976. Female terminalia of lower Brachycera - I 
(Diptera). Beitrage zur Entomologie 26:5-47. 
Nagatomi, A., and H. Nagatomi. 1987. The genus Austroleptis from south Chile and 
Patagonia (Diptera, Rhagionidae). Memoirs of the Kagoshima University 
Research Center for the South Pacific 8:139-156. 
Nagatomi, A., and H. Nagatomi. 1990. A revision of Atherimorpha White, 1915 from 
southern Africa (Diptera: Rhagionidae). Annals of the Natal Museum 31:33-
82. 
Nagatomi, A., and K. Soroida. 1985. The structure of the mouthparts of the 
orthorrhaphous Brachycera (Diptera) with special reference to blood-sucking. 
Beitrage zur entomologie 35:263-368. 
Nagatomi, A., and D. Yang. 1998. A review of extinct Mesozoic genera and families 
of  Brachycera (Insecta, Diptera, Orthorrhapha). Entomologist's Monthly 
Magazine 134:95-192. 
Narchuk, E. P. 1969. [Rhagionidae]. Pages 443-453 in Keys to the insects of the 
European part of the USSR (G. Y. Bei Bienko, ed.) Nauka, Leningrad. 
Narchuk, E. P. 1988. Family Rhagionidae (Leptidae). Pages 683-696 in Keys to the 
insects of the European part of the USSR (G. Y. Bei Bienko, ed.) Smithsonian 
Institution Libraries & National Research Foundation, Washington, DC. 
Nartshuk, E. P. 1995. Taxonomic and faunistic data on the Rhagionidae (Diptera, 
Brachycera) of the northern Palaearctic. Acta Zoologica Fennica:17-24. 
Oosterbroek, P. F. L. S., and G. Courtney. 1995. Phylogeny of nematocerous families 
of Diptera (Insecta). Zoological Journal of the Linnean Society 115:267-311. 
 
 602
Osten Sacken, C. R. 1858. Catalogue of the described Diptera of North America. 
Smithsonian Misc. Collect. 3:92. 
Osten Sacken, C. R. 1874a. A list of the Leptidae, Mydaidae and Daypogonina of 
North America. Bulletin of the Buffalo Society Natural Sciences 2:169-187. 
Osten Sacken, C. R. 1874b. A list of the species known in North America. Bulletin of 
the Buffalo Society Natural Sciences 2:171-174. 
Osten Sacken, C. R. 1878. Catalogue of the described Diptera of North America. 
Smithsonian Misc. Collect. 16:276. 
Osten Sacken, C. R. 1883. Synonymica concerning exotic dipterology.  No. II. 
Berliner entomologische Zeitschrift 27:295-298. 
Osten Sacken, C. R. 1890. Hilarimorpha Schin. is a Leptid. Berliner entomologische 
Zeitschrift 35:303-304. 
Ouchi. 1943. Diptera Sinica. Coenomyiidae 1. On a new genus belonging to the 
family Coenomyiidae from East China. Shanghai Sizenkagaku Kenkyusho Iho 
13:493-495. 
Palmer, C. M., and D. K. Yeates. 2000. Phylogenetic importance of immature stages: 
solving the riddle of Exeretonevra Macquart (Diptera: Xylophagidae). Annals 
of the Entomological Society of America 93:15-27. 
Paramonov, S. J. 1962. A review of the Australian Leptidae (Diptera). Australian 
Journal of Zoology 10:113-169. 
Pawlowski, J., R. Szadziewski, D. Kmieciak, J. Fahrni, and G. Bittar. 1996. 
Phylogeny of the infraorder Culicomorpha (Diptera: Nematocera) based on 
28S RNA gene sequences. Systematic Entomology 21:167-178. 
 
 603
Pechuman, L. L., and H. J. Teskey. 1981. [Chapter] 31: Tabanidae. Pages 463-478 in 
Manual of Nearctic Diptera (J. F. McAlpine, ed.) Research Branch, 
Agriculture Canada, Ottawa, Canada. 
Pelandakis, M., and M. Solignac. 1993. Molecular phylogeny of drosophila based on 
ribosomal rna sequences. Journal of Molecular Evolution 37:525?543. 
Peterson, A. 1916. The head-capsule and mouth-parts of Diptera. Illinois Biological 
Monographs 3:283. 
Poe, S. 1998. The effect of taxonomic sampling on accuracy of phylogeny 
reconstruction: test case of a known phylogeny. Molecular Biology and 
Evolution 15:1086?1090. 
Ragan, M. 1992. Phylogenetic inference based on matrix representation of trees. 
Molecular Phylogenetics and Evolution 1:53-58. 
Rannala, B., J. P. Huelsenbeck, Z. H. Yang, and R. Nielsen. 1998. Taxon sampling 
and the accuracy of large phylogenies. Systematic Biology 47:702-710. 
Regier, J. C., and J. W. Shultz. 1997. Molecular phylogeny of the major arthropod 
groups indicates polyphyly of crustaceans and a new hypothesis for the origin 
of hexapods. Molecular Biology and Evolution 14:902-913. 
Roberts, M. J. 1969. Structure of the mouthparts of the larvae of the flies Rhagio and 
Sargus in relation to feeding habits. Journal of Zoology, London 159:381-398. 
Rohdendorf, B. 1974. The Historical Development of Diptera. University of Alberta 
Press, Edmonton, Alberta. 
Rohdendorf, B. B. 1991. Order Diptera.  Two-winged insects. Pages 444-502 in 
Fundamentals of Paleontology Amerind Publishing Co., New Delhi. 
 
 604
Ronquist, F., and J. P. Huelsenbeck. 2003. MrBayes 3: Bayesian phylogenetic 
inference under mixed models. Bioinformatics 19:1572-1574. 
Rosen, B. R. 1990. From fossils to earth history: applied historical biogeography. 
Pages 437-481 in Analytical biogeography. (A. A. Myers, and G. P. S., eds.). 
Chapman and Hall, London. 
Rosenberg, M. S., and S. Kumar. 2001. Incomplete taxon sampling is not a problem 
for phylogenetic inference. Proceedings of the National Academy of Sciences 
98:10751-10756. 
Samouelle, G. 1819. The entomologist's useful compendium; or an introduction to the 
knowledge of British insects. Thomas Boys, London. 
Sanderson, M., A. Purvis, and C. Henze. 1998. Phylogenetic supertrees: Assembling 
the trees of life. Trends in Ecology & Evolution 13:105-109. 
Schiner, J. R. 1860. Vorlaufiger Commentar dipterologischen Theile der "Fauana 
austriaca" mit einer naheren Begrundung der in deselben aufgenommen neuen 
Dipteren-Gattungen. Wien entomologische Mschr. 4:47-55. 
Schiner, J. R. 1862. Fauna Austriaca. Druck und Verlag von Carl Gerold's Sohn, 
Wien. 
Schremmer, F. 1951. Morphologische und funktionelle Analyse der Mundteile und 
des Pharynx der Larve von Stratiomys chamaeleon L. ?sterreichische 
Zoologische Zeitschrift 3:326-397. 
Scotland, R. W., R. G. Olmstead, and J. R. Bennet. 2003. Phylogeny reconstruction: 
The role of morphology. Systematic Biology 52:539-548. 
 
 605
Shemanchuk, J. A., and J. Wintraub. 1961. Observations on the biting and swarming 
of snipe flies (Diptera: Symphoromyia) in the foothills of Southern Alberta. 
Mosquito News 21:238-243. 
Sinclair, B. J. 1992. A phylogenetic interpretation of the Brachycera (Diptera) based 
on the larval mandible and associated mouthpart structures. Systematic 
Entomology 17:233-252. 
Sinclair, B. J., J. M. Cumming, and D. M. Wood. 1993. Homology and phylogenetic 
implications of male genitalia in Diptera - lower Brachycera. Entomologica 
Scandinavica 24:407-432. 
Skevington, J. H., and P. T. Dange. 2002. Exploring the diversity of flies (Diptera). 
Pages 3-27 in Biodiversity. 
Sommerman, K. M. 1962. Alaskan snipe fly immatures and their habitat 
(Rhagionidae: Symphoromyia). Mosquito News 22:116-123. 
Steyskal, G. C. 1953. A suggested classification of the Lower Bracycerous Diptera. 
Annals of the Entomological Society of America 46:237-242. 
Stuckenberg, B. R. 1956. Notes on the genus Arthroteles Bezzi, with description of 
two new species (Diptera: Erinnidae). Annals of the Natal Museum 13:313-
335. 
Stuckenberg, B. R. 1966. A new genus and species of Rhagionidae from southern 
Brasil (Diptera). Proc. R. ent. Soc. Lond. (B) 35:57-60. 
Stuckenberg, B. R. 1973. The Athericidae, a new family in the lower Brachycera 
(Diptera). Annals of the Natal Museum 21:649-673. 
 
 606
Stuckenberg, B. R. 2001. Pruning the tree: a critical review of classifications of the 
Homeodactyla (Diptera, Brachycera), with new perspectives and an 
alternative classification. Studia dipterologica 8:3-41. 
Swofford, D. 2001. PAUP*: Phylogenetic analysis using parsimony (*and other 
methods), version 4.0b10. Sinauer Associates. 
Swofford, D., D. L. Sankoff, G. J. Olsen, P. J. Waddell, and D. M. Hillis. 1996. 
Phylogenetic Inference. Pages 407-514 in Molecular Systematics (D. M. 
Hilllis, C. Mortiz, and B. K. Mable, eds.). Sinauer Associates, Sunderland, 
Massachusetts. 
Szil?dy, Z. 1934. Die palaearktischen Rhagioniden. Annls hist.-nat. hung. 28:229-
270. 
Teskey, H. J. 1969. Larvae and pupae of some Eastern North American Tabanidae 
(Diptera). Memoirs of the Entomological Society of Canada 63:147. 
Teskey, H. J. 1970a. The immature stages and phyletic position of Glutops rossi 
(Diptera: Pelecorhynchidae). Canadian Entomologist 102:1130-1135. 
Teskey, H. J. 1970b. A review of the genus Glutops (Diptera: Pelecorhynchidae), 
with descriptions of four new species. Canadian Entomologist 102:1171-1179. 
Teskey, H. J. 1981a. Morphology and terminology- larvae [Chapter] 3. Pages 65-88 
in Manual of Nearctic Diptera Research Branch Agriculture Canada, Ottawa, 
Ontario. 
Teskey, H. J. 1981b. Pelecorhynchidae. Pages 459-461 in Manual of Nearctic Diptera 
Research Branch Agriculture Canada, Ottawa, Ontario. 
 
 607
Teskey, H. J. 1981c. Vermileonidae. Pages 529-532 in Manual of Nearctic Diptera 
Research Branch Agriculture Canada, Ottawa, Ontario. 
Thomas, A. 1974. Dipteres torrenticoles peu connus: I.- les Athericidae (larves et 
imagos) du sud de la France. Annls. Limnol. 10:55-84. 
Thomas, A. G. B. 1978a. Athericidae et Rhagionidae. Pages 477-478 in Limnofauna 
Europaea. Eine Zusammenstellung aller die europaischen Binnengewasser 
bewohnenden mehrzelligen Tierarten mit Angaben uber ihre Verbreitung und 
Okologie. ,  B (J. Illies, ed.) Gustav Fischer Verlag 
Swets & Zeitlinger, Stuttgart & New York  
Amsterdam. 
Thomas, A. G. B. 1978b. Dipteres torrenticoles peu connus: 5. Les Rhagionidae 
(genre Chrysopilus) du sud de la France (Brachycera, Orthorrhapha). Bulletin 
De La Societe D'Histoire Naturelle De Toulouse 114:305-331. 
Thomas, A. G. B. 1997. Rhagionidae and Athericidae, Snipe-flies. Pages 311-318 in 
Aquatic Insects of North Europe - A Taxonomic Handbook (A. N. Nilsson, 
ed.) Apollo Books, Stenstrup, Denmark. 
Thorley, J. L., and R. D. M. Page. 2000. RadCon: phylogenetic  T  ree comparison 
and consensus. Bioinformatics 16:486-487. 
Tsacas, L. 1962. Recherches sur la structure et le fontionnement de la tete et des 
pieces buccales larvaires de Rhagionidae (Dipteres). Memoires du Museum 
National d'Histoire Naturelle (Ser. A, Zoologie) 27:147-235. 
 
 608
Turner, W. J. 1979. A case of severe human allergic reaction to bites of 
Symphoromyia (Diptera: Rhagionidae). Journal of Medical Entomology 
15:138-139. 
Turner, W. J., and J. G. Chillcott. 1973. Four new species of the Symphoromyia 
pachyceras complex from California. Pan-Pacific Entomologist 49:5-20. 
Verrall. 1909. British Flies, London. 
Vockeroth, J. R. 2002. Introducing the ubiquitous Diptera. Biodiversity 3:3-5. 
Walker, F. 1851. Diptera. Pages 314 in Insecta Britannica Reeve and Benham, 
London. 
Walker, F. 1856. Dipterain Insecta Saundersiana: or Charcters of undescribed insects 
in the 
collection of William Wilson Saunders (W. W. Saunders, ed.) John Van Voorst, 
London. 
Walker, F. 1859. Catalogue of dipterous insects collected at Makessar in Celebes, by 
Mr. A. R. Wallace, with descriptions of new species. Proceedings of the 
Linnean Society of London, Zoology 4:90-144. 
Webb, D. W. 1977a. The Nearctic Athericidae (Insecta: Diptera). Journal of the 
Kansas Entomological Society 50:473-495. 
Webb, D. W. 1977b. Ptiolina Edeta (Walker) (Diptera: Rhagionidae), a Redescription 
and New Synonymy. Entomology News 88:153-157. 
Webb, D. W. 1978. A revision of the Nearctic genus Dialysis (Diptera: Rhagionidae). 
Journal of the Kansas Entomological Society 51:405-431. 
 
 609
Webb, D. W. 1983. The immature stages of Dialysis fasciventris (Loew) (Diptera: 
Coenomyiidae). Proceedings Entomological Society of Washington 85:691-
697. 
Westwood, H. O. 1840a. An Introduction to the Modern Classification of Insects, 
review. Isis (Oken's) 1840:584-588. 
Westwood, J. O. 1840b. Order XIII. Diptera Aristotle. (Antliata Fabricius.  
Halteriptera Clairv.). Pages 125-154 in An introduction to the modern 
classification of insects; founded on the natural habits and corresponding 
organisation of the different families.  Synopsis of the genera of British 
insects. (J. O. Westwood, ed.), London. 
Westwood, J. O. 1876. Notes dipterologicae. No. 3. -Descriptions of some new exotic 
species of the family Acroceridae. Transactions of the Entomological Society 
of London 1876:507-518. 
White, A. 1914. Diptera Brachycera of Tasmania. Papers and Proceedings of the 
Royal Society Tasmania 1914:35-75. 
Wiedemann, C. R. W. 1821. Diptera Exotica, Kiliae. 
Wiegmann, B. M., S. C. Tsaur, D. W. Webb, D. K. Yeates, and B. K. Cassel. 2000. 
Monophyly and relationships of the Tabanomorpha (Diptera : Brachycera) 
based on 28S ribosomal gene sequences. Annals of the Entomological Society 
of America 93:1031-1038. 
Wiegmann, B. M., D. K. Yeates, J. L. Thorne, and H. Kishino. 2003. Time Flies, a 
New Molecular Time-Scale for Brachyceran Fly Evolution Without a Clock. 
Systematic Biology 52:745-756. 
 
 610
Wiens, J. J. 1998. Combining data sets with different phylogenetic histories. 
Systematic Biology 47:568?581. 
Williston, S. W. 1885. Ueber einige Leptiden-Charaktere. Entologische Zeitschrift, 
Stuttgart 46:400. 
Williston, S. W. 1886. On two interesting new genera of Leptidae. Entomologica 
Americana 2:105-108. 
Williston, S. W. 1888. Hilarimorpha and Apiocera. Psyche 5:99-102. 
Williston, S. W. 1896. Manual of North American Diptera, Second Edition edition. 
James T. Hathaway, New Haven. 
Williston, S. W. 1908. Manual of North American Diptera, Third Edition edition. 
James T. Hathaway, New Haven. 
Wood, D. M., and A. Borkent. 1989. Phylogeny and classification of the Nematocera. 
Pages 1333-1370 in Manual of Nearctic Diptera (J. F. McAlpine, and D. M. 
Wood, eds.). Research Branch, Agriculture Canada, Ottawa. 
Woodley, N. E. 1989. Phylogeny and classification of the "orthorrhaphous" 
Brachycera. Pages 1371-1395 in Manual of Nearctic Diptera (J. F. McAlpine, 
ed.) Canadian Government Publishing Centre, Hull, Quebec. 
Yang, D., and A. Nagatomi. 1992. A study on the Chinese Rhagina (Dipt., 
Rhagionidae). Entomologist's Monthly Magazine 128:87-91. 
Yang, D., C.-k. Yang, and A. Nagatomi. 1997. The Rhagionidae of China (Diptera). 
South Pacific Study 17:113-262. 
Yeates, D. K. 1995. Groundplans and exemplars: paths to the tree of life. Cladistics 
11:343-357. 
 
 611
Yeates, D. K. 2002. Relationships of extant lower Brachycera (Diptera): a 
quantitative synthesis of morphological characters. Zoologica Scripta 31:105-
121. 
Yeates, D. K., and C. L. Lambkin. 1998. Cryptic species diversity and character 
congruence: review of the tribe Anthracini (Diptera : Bombyliidae) in 
Australia. Invertebrate Taxonomy 12:977-1078. 
Yeates, D. K., and B. M. Wiegmann. 1999. Congruence and Controversy: Toward a 
Higher-Level Phylogeny of Diptera. Annual Review of Entomology 44:397-
428. 
Zetterstedt, J. W. 1842. Diptera Scandinaviae.  Disposita et descripta, Lund. 
Zwickl, D. J., and D. M. Hillis. 2002. Increased taxon sampling greatly reduces 
phylogenetic error. Systematic Biology 51:588-598.