M
BRITISH MUSEUM
(NATURAL HiSi
26 JU
PRESENTED
GENERAL UC.-lARY
Bulletin of the
British Museum (Natural History)
The ichneumon-fly genus Banchus
(Hymenoptera) in the Old World
M. G. Fitton
Entomology series
Vol51 Nol 25 July 1985
The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.
Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers are
works of reference that will remain indispensable for years to come.
Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to:
Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW75BD,
England.
'
World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.)
Trustees of the British Museum (Natural History), 1985
The Entomology series is produced under the general editorship of the
Keeper of Entomology: Laurence A. Mound
Assistant Editor: W. Gerald Tremewan
ISBN 565 06011 2
ISSN 0524-6431 Entomology series
Vol 51 No 1 pp 1-60
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 25 July 1985
BRITISH MU:
The ichneumon-fly genus Banchus (Hymenoptera) in
the Old World -^1
M. G. Fitton
Department of Entomology, British Museum (Natural History), Cromwell Road, London
SW7 5BD
Contents
Synopsis 1
Introduction 1
Banchus Fabricius 3
Taxonomy 3
Biology and hosts 5
Systematic list of hosts 6
Nomenclatural summary 7
Format, material examined and depositories 8
Keys to species 11
Males 11
Females 12
Systematic section 14
Species excluded from Banchus 43
Acknowledgements 45
References 45
Index 59
Synopsis
This paper provides the first comprehensive account of the genus Banchus (Ichneumonidae, Banchinae) in
the Old World. Twenty-three species are recognized, described and keyed. Seven of the species are new.
Geographical distributions and data on biology and hosts are summarized. Known hosts are Lepidoptera
(mainly of the family Noctuidae) and Banchus species are known to play an important role in the control of
some pest species in Europe and North America. The taxonomic treatment includes 11 new synonymies,
one provisional synonymy, three new combinations, one revised combination, two revised statuses, three
lectotype designations and four nomina dubia.
Introduction
Banchus has had a chequered history. Taxonomists found it difficult to assign to one of the five
traditional subfamilies of the Ichneumonidae, sometimes placing it in the Ophioninae and
sometimes in the Pimplinae. As originally defined by Fabricius (1798; 1804) the genus included
members of the families Braconidae, Cephidae, Pompilidae, Proctotrupidae and Ibaliidae, as
well as the Ichneumonidae. This exotic assortment was reduced by Gravenhorst (1829) to a
grouping resembling the genus as understood at present. The main error in his treatment was the
transposition of the sexes. Developments in the following 150 years were mainly concerned with
description of new species, splitting the genus and faunistic studies. Much of this work was of
poor quality; no one made a comprehensive study and those describing and keying species failed
to appreciate all but the most superficial characters. A real advance in our understanding of the
taxonomy of Banchus came only with publication of the revision of the Nearctic species by
Townes & Townes in 1978.
Bull. Br. Mus. nat. Hist. (Ent.) 51 (1): 1-60 Issued 25 July 1985
M. G. FITTON
I
8
E
ICHNEUMON-FLY GENUS BANCHUS 3
BANCHUSFabridus
Banchus Fabricius, 1798: 209, 233. Type-species: Banchus pictus Fabricius, by subsequent designation
(Curtis, 1836: 588).
Corynephanus Wesmael, 1849: 631. Type-species: Banchus monileatus Gravenhorst, by monotypy.
Cidaphurus Foerster, 1869: 159. Type-species: Ichneumon volutatorius Linnaeus, by subsequent mono-
typy (Woldstedt, 1877: 439).
Nawaia Ashmead, 1906: 184. Type-species: Nawaia japoni ca Ashmead, by monotypy.
DIAGNOSIS. Moderately-built ichneumonids (Fig. 1), fore wing length 7-7-14-3 mm. Frons simple. Margin
of clypeus with a median notch. Genal carina joining hypostomal carina above base of mandible. Upper
tooth of mandible much wider than lower tooth and longer, its apex truncate and weakly subdivided.
Lower tooth pointed. Maxillary palp with segment 4 more or less widened and flattened distally, especially
in males. Labium not elongate. Antennal flagellum long, tapered to a slender apex. In males, flagellum
with poorly- to well-developed specialized setae (flag setae), 2 to 4 per segment, often arising from a
poorly- to well-marked, shallow trough on the dorsal surface. Epicnemial carina absent. Scutellum
produced posteriorly into a spine, although it is sometimes reduced to a very weak tubercle or even absent.
Propodeum short, its posterior transverse carina strong laterally, weak or obsolete medially. Pleural carina
present, at least in part. Tarsal claws pectinate. Laterotergites of segments 2 and 3 of gaster about 0-7 times
as deep as long. Gaster more or less compressed, relatively strongly so in most females. Ovipositor very
short, its sheath about 0-1 times as long as hind tibia.
REMARKS. The genus comprises 47 valid, described species. One is Holarctic in distribution, a
further 24 occur in the Nearctic (Townes & Townes, 1978) and another 22 are found in the
Palaearctic and the mountains of the Oriental region.
Banchus is the largest of a group of five genera united by the structure of the mandible. The
mandible is of an unusual, almost tridentate form unlikely to have arisen more than once within
the Banchinae. A mandible similar in structure is found only in the Diplazontinae (in which
subfamily it occurs in all species). The other genera in the Banchus-group are: Rhynchoban-
chus (with 3, possibly 4, Palaearctic species); Banchopsis (1 Palaearctic and 1 Oriental species);
Ceratogastra (1 Palaearctic and 1 Nearctic species) and; Philogalleria (7 Australian species and 1
species in Argentina which possibly deserves a separate genus). Within this group the most
important discriminant characters are the form of the claws, the presence or absence of the
epicnemial carina, the extent of the propodeal carinae, the form of the frons, the position of
cross- vein 2m-cu in the fore wing, and the relative sizes of the coxae.
The tribe Banchini, as currently recognized (Townes, 1970: 45; Townes & Townes, 1978:
378), includes two disparate elements- the Banchus-group of genera and Exetastes and its allies.
Cushman (1937), Perkins (in Beirne, 1941) and Beirne (1941) all supported separation of the
two groups on the basis of adult and larval characters. Cushman and Perkins associated the
Exetastes-gioup more closely with the tribe Lissonotini. Townes (1944) on the other hand linked
the two groups as the Banchini, separate from the Lissonotini, mainly on the basis of a rather
weak character in the hind wing venation. Townes has consistently maintained this position.
Viktorov (1967) supported Cushman's (1937) conclusion, citing several characters of Exetastes
which distinguish it from the Banchus-group. Aubert (1978) has gone further, placing Exetastes
in a tribe of its own, Exetastini. As with many ichneumonid subfamilies the generic and tribal
classification of the Banchinae, although workable, is in need of closer study. Townes' (1970)
treatment and key fail to take account of the structure of the mandible in Philogalleria (of which
Townes may have then seen only one specimen).
Taxonomy
Despite its long period of gestation this study, based on the examination of 3421 specimens of
Old World Banchus, cannot be regarded as definitive. It has raised as many taxonomic questions
as it has answered. The lack of specimens from some problem areas (taxonomic as well as
geographic), coupled with an inadequate knowledge of most aspects of the biology of the genus,
has effectively prevented discrimination of all of the species. A conservative approach was
adopted and some of the morphospecies recognized undoubtedly comprise complexes of two or
4 M. G. FITTON
more 'biological' species. With the material and resources available further splitting could not be
justified on scientific grounds.
Apart from obvious species-pairs or very small groups it is difficult to deduce anything of the
phylogenetic relationships of the species within the genus. As noted by Townes & Townes
(1978) in their revision of North American Banchus, the characters which distinguish species do
not correlate in ways which indicate clear-cut species-groups. Specialization by reduction (an
obvious example is in the length of the scutellar spine) is probably one of the complicating
factors. The best that can be done is to suggest trends of specialization for individual characters.
Many species show considerable sexual dimorphism (including colour patterns). The most
obvious morphological differences are in the shape of the gaster , the antennae and the maxillary
palps. The specializations in the female seem to be associated with oviposition and in the male
with mate-finding and courtship. In some species the overall colour pattern is the same in male
and female, in others it is completely different, for example, in B. falcatorius and B. volutatorius
the females are largely black with reddish legs while the males are conspicuously marked with
yellow. This female colour pattern, although still aposematic, probably renders her less
conspicuous while searching for hosts (which, in B. falcatorius, probably occurs very low in the
vegetation at soil level). As a general rule males have a yellowish face with a median black stripe
and females a black face, sometimes with yellowish orbital marks, and there are often also
differences in the colour of the antennae.
The characters which have been used in this study are discussed below. The formal
descriptions of species in the systematic section of the paper are restricted largely to the
characters used in the keys. When comparing specimens with figures allowance should be made
for some variation.
Head. The overall shape of the head is important in distinguishing species. The characters
most easily used are the relative width of the face and the size of the malar space. These are
measured as indicated in Figs 2 and 3. The shape of the head behind the eyes varies between
species but is much more difficult to use as a distinguishing character. The maxillary palps are
specialized in many species (Figs 11-32, 103-123). The relative lengths of the two terminal
segments are important and in males the way in which the fourth segment is widened and
flattened is characteristic. The antennae vary in length and in the shape of the terminal
segments, both characters being difficult to use in keys.
In the male the distal half of the antenna bears specialized setae ('flag setae') on its dorsal
surface (Figs 124-129). The development of these varies greatly between species. In the most
primitive cases the setae are not easily distinguished from others on the antenna. In the most
specialized forms the setae (two, three or four per segment) are erect, widened and flattened and
arise from a shallow polished trough. All segments do not bear these setae, there is a zone (which
varies in position between species) in which they are best developed and proximally and distally
to this the setae become progressively less modified. The descriptions given for each species
apply only to the zone bearing the most modified setae. The shape and number per segment are
subject to some variation, but together with the form of the male maxillary palps these setae give
the best 'key' characters for a species. Both are probably important in courtship.
Thorax and propodeum. The scutellum bears a spine on its posterior apex (Figs 33-58). The
primitive condition, it is supposed, is to have a well-developed spine. The length of the spine is
impossible to measure accurately because of the lack of reference points, and there is variation
within species. Nonetheless it is still useful taxonomically. The lower postero-lateral corners of
the mesothorax in some species are formed into weak tubercles which, although relatively
constant in their development, are difficult to use comparatively. In the descriptions the
'posterior part of the propodeum' refers to the posterior transverse carina and the area behind
and enclosed by it. The posterior transverse carina itself is often obsolete or absent medially.
The measured proportions of the hind femur (as shown in Fig. 5) are used to represent the
general proportions of the legs as a whole. The length of the fore wing (measured from the tip of
the tegula to the tip of the wing) is used as a measure of overall size.
Gaster. The form of the gaster in the female offers a number of very useful characters.
However, their practical value is much reduced by the very variable amounts of distortion found
ICHNEUMON-FLY GENUS BANCHUS 5
in dry specimens. Allowance must be made for this in comparing specimens with descriptions
and figures, particularly those of the apex of the gaster (Figs 82-102). The distortion most affects
the amount of compression of the posterior half of the gaster, the telescoping of segments and
the compression and profile of sternite 6 (the subgenital plate) which in some species has a
weaker, less sclerotised portion on its posterior margin. The form of tergite 1 is not subject to
distortion and it offers useful characters in males and females (Figs 59-81), although subject to
some variation (compare Figs 65 and 66 of male and female gudrunae). The proportions of
tergite 1 are measured as indicated in Fig. 4.
Sculpture and colour. The development of microsculpture and puncturation on the thorax,
coxae and gaster varies between species but variation within species and difficulties of descrip-
tion limit its value. Colour patterns are subject to variation but are easily observed. Care was
taken to ensure that such variation was taken into account in construction of those parts of the
keys relying on colour.
Biology and hosts
Townes & Townes (1978) state that Banchus is a genus of open shrubby country and that females
fly rather low and males higher and faster. However, there are few published observations on the
habits and habitats of individual species. Adults have been collected from flowers. The black and
yellow and/or reddish colour patterns may be aposematic (Townes & Townes note that
specimens give off a strong pungent odour when captured). B. falcatorius has been recorded as
the prey of an asilid fly.
Most species probably are univoltine and adults are on the wing for some period during late
spring or early summer. A few species (e.g. B. dilatatorius) occur in early spring. These probably
pass the winter as adults or pharate adults within the cocoon. In Europe three species (B.
zonatus, B. gudrunae and B. moppiti) have been taken in late autumn, winter and early spring.
Of these three, B. zonatus and B. gudrunae are Mediterranean in distribution whilst the few
known specimens of B. moppiti come from localities as scattered as Great Britain, Switzerland
and Spain. There appears to be a similar pattern in the species in the eastern Palaearctic and
mountains of the Oriental region, with some specimens of the more southern species occurring
in the period from October to May. However, some of these species might not be univoltine.
Courtship behaviour in B. hastator is described by van Veen (1982): when the male
encounters a female he orientates himself face to face and starts fanning with his wings, probably
creating an airstream along the female and himself. The gaster of the male is also raised and at
intervals these actions are interrupted by an attempt to mount the female. Using the tips of the
antennae the male gently strokes the legs and sides of the thorax of the female. If the female is
receptive copulation takes place.
The hosts of Banchus are Lepidoptera. All of the reliable observations relate to species of
Noctuidae, though species of seven other lepidopterous families are also recorded. Whether or
not any or all of the non-noctuid records should be rejected is difficult to assess. In the Nearctic
species B. flavescens it has been shown that females oviposited readily in the larvae of its usual
host Mamestra configurata Walker and also in Scotogramma trifolii (Rottemburg) and Tricho-
plusia ni (Hiibner) , but had to be induced to attack A utographa calif ornica (Speyer) , A . falcifera
(Kirby) and Heliothis ononis (Denis & Schiffermuller) by an immediately preceding exposure to
M. configurata larvae (Arthur & Ewen, 1975; Ewen & Arthur, 1976). All these hosts are
noctuids and in addition to the reluctance of the females to oviposit in some species the parasite
failed to develop in any except M. configurata, because of successful host defence reactions. Van
Veen (1982), working with B. hastator in the Netherlands, found 'that an antennal contact of a
few strokes with the cuticle of the host's skin was sufficient to reject a strange host by walking
away or to a susceptible instar of Pfanolis] flammea [the usual host] immediately by an acute
sting reaction'. He unfortunately does not identify the 'strange host'. All these results could be
used to support a contention that Banchus species are probably monophagous on species of
Noctuidae, but for one species, B. volutatorius , there are reliable rearings from at least three
noctuid species (Anarta myrtilli (Linnaeus), Lacanobia oleracea (Linnaeus) and Xestia xanth-
6 M. G. FITTON
ographa (Denis & Schiffermiiller)), so judgement must be reserved. The host records from two
Hymenoptera (Maneval, 1935) can probably be dismissed out of hand.
The larvae normally develop as solitary internal parasites and it is usually suggested that
oviposition is into an early instar host larva. In experiments, van Veen (1982) has demonstrated
a preference for early instar larvae in B. hastator, although females will attack all larval instars
except the last (the fifth) (van Veen, 1982; Bledowski & Krainska, 1926). The larval develop-
ment of B. hastator was studied in detail by Bledowski & Krainska (1926). The head sclerites of
the final larval instar of four species are illustrated by Short (1978). Short's work should be used
with caution: his figures of Banchus show why. The illustrations of B. femoralis and B. hastator
show differences which might be thought significant; however, both specimens are of the same
species (B. hastator, of which femoralis is a synonym) and one figure shows the view from outside
the head capsule while the other shows the view from within! Other figures are given by Beirne
(1941) and Barron (1976).
The full-grown Banchus larva kills and leaves the host larva after it has left its foodplant and
entered the soil to pupate (this has been reported in several species). Cocoons of Banchus are
elongate-ovoid, the silk dense and almost black. The adult emerges through a hole cut next to
one end.
Barron (1976) records a species of Banchus (from Maryland) parasitized by Euceros medialis
Cresson (Hymenoptera: Ichneumonidae).
Systematic list of hosts
The list below includes all the hosts of Banchus detailed elsewhere in this paper. Those details
can be accessed via the index.
LEPIDOPTERA
ZYGAENIDAE
Zygaena ephialtes (Linnaeus)
NYMPHALIDAE
Melitaea didyma (Esper)
THYATIRIDAE
Achlyaflavicornis (Linnaeus)
Habrosyne pyritoides (Hufnagel)
GEOMETRIDAE
Opisthograptis luteolata (Linnaeus)
SPHINGIDAE
Deilephila elpenor (Linnaeus)
Deilephila porcellus (Linnaeus)
Hyloicus pinastri (Linnaeus)
Smerinthus ocellata (Linnaeus)
NOTODONTIDAE
Phalera bucephala (Linnaeus)
LYMANTRIIDAE
Dasychira fascelina (Linnaeus)
Euproctis similis (Fuessly)
Leucoma salicis (Linnaeus)
NOCTUIDAE
Acronicta megacephala (Denis & Schiffermiiller)
Agrochola circellaris (Hufnagel)
Agrochola helvola (Linnaeus)
Agrotis exclamationis (Linnaeus)
Agrotis segetum (Denis & Schiffermiiller)
Anarta myrtilli (Linnaeus)
Aporophyla lutulenta (Denis & Schiffermiiller)
Aporophyla lutulenta luenebergensis (Freyer)
Atethmia ambusta (Denis & Schiffermiiller)
Bena prasinana (Linnaeus)
Blepharita adusta (Esper)
ICHNEUMON-FLY GENUS BANCHUS
Ceramica pisi (Linnaeus)
Euxoa nigricans (Linnaeus)
Hadena compta (Denis & Schiffermuller)
Hadena rivularis (Fabricius)
Heliothis viriplaca (Hufnagel)
Lacanobia contigua (Denis & Schiffermuller)
Lacanobia oleracea (Linnaeus)
Lacanobia suasa (Denis & Schiffermuller)
Lycophotia porphyrea (Denis & Schiffermuller)
Mamestra brassicae (Linnaeus)
Panolis flammea (Denis & Schiffermuller)
Phlogophora meticulosa (Linnaeus)
Xestia xanthographa (Denis & Schiffermuller)
HYMENOPTERA
DIPRIONIDAE
Diprion pini (Linnaeus)
ICHNEUMONIDAE
Ophion luteus (Linnaeus)
Nomenclatural summary
agathae sp. n.
altaiensis Meyer, 1927 (nomen dubium)
cerinus Chandra & Gupta, 1977
cre/e/densisUlbricht, 1916
croaticusHensch, 1928
dilatatorius (Thunberg, 1822) stat. rev.
acuminator (Fabricius, 1787) (homonym) syn. n.
compressus (Fabricius, 1787) (homonym)
sibiricus Meyer, 1927 syn. n.
fa/catorius (Fabricius, 1775)
variegator (Fabricius, 1775)
intersectus (Geoffrey, 1785)
aries (Christ, 1791)
labiatus (Schrank, 1802)
histrio (Schrank, 1802) (homonym)
tricolor (Schrank, 1802)
falcat or Fabricius, 1804
sachalinensis (Matsumura, 1911)
luteofasciatus Ulbricht, 1911
nobilitatorMorley, 1915
sanguinator Meyer, 1922
lavrovi Meyer, 1927
nigromarginatus Constantineanu & Pisica, 1960
propitius Kuslitzky, 1979 (provisional synonymy)
flavomaculat us (Cameron, 1904)
gudrunae sp. n.
hasta tor (Fabricius, 1793)
pungitor (Thunberg, 1822)
reticulator (Thunberg, 1822) syn. n.
femoralis Thomson, 1897
kolosovi Meyer, 1925 syn. n.
insulanus Roman, 1937
japonicus(Ashmead, 1906)
mauricettae sp. n.
moppitisp. n.
noxMorley, 1913
pa/pa/JsRuthe, 1859 stat. rev.
spinosus Cresson, 1865 syn. n.
8 M. G. FITTON
formidabilis Provancher, 1874 syn. n.
groenlandicus Aurivillius, 1890 syn. n.
alticola (Ashmead, 1901) syn. n.
picf us Fabricius, 1798
cultratus (Gmelin, 1790) (homonym)
mutillatus (Christ, 1791) (homonym)
zagoriensis Hensch, 1928
bipunctatus Hensch, 1928
russiator Aubert, 1981 syn. n.
poppitisp. n.
punkettaisp. n.
sanjozanus Uchida, 1929
tholussp. n.
tumidus Chandra & Gupta, 1977
furcaf or Aubert, 1981
volutatorius (Linnaeus, 1758)
venator (Linnaeus, 1758)
umbellatarum (Schrank, 1786) syn. n.
certator (Thunberg, 1822) (homonym)
monileatus Gravenhorst, 1829 syn. n.
farrani Curtis, 1836
calcaratus Szepligeti, 1910
alticola Schmiedeknecht, 1910 (homonym)
obscurus Meyer, 1926
zonafusRudow, 1883
algericus Schmiedeknecht, 1910 syn. n.
Species excluded from Banchus
Andricus villosulus (Gravenhorst) (nomen dubium) comb. n.
Banchus rofeusfusRudow, 1883 (nomen dubium)
Cephas pygmeus (Linnaeus, 1767)
spinipes (Panzer, 1800)
vim/atar (Fabricius, 1804)
Earinuse/af or (Fabricius, 1804) comb. n.
Exetastes fornicator (Fabricius, 1781)
Exetastes tomentosus (Gravenhorst, 1829)
Ibalia leucospoides (Hochenwarth, 1785)
cultellator (Fabricius, 1793)
Ichneumon vigilatorius Panzer, 1804 (nomen dubium) comb. rev.
Lissonota histrio (Fabricius, 1798)
Megarhyssa quadrator (Schellenberg, 1802) comb. n.
Phytodietus armillatus (Morley, 1913)
Pompilus (Episyron) annulatus (Fabricius, 1793)
Proctotrupes gravidator (Linnaeus, 1758)
Theronia atalantae (Poda, 1761)
varius (Fabricius, 1793)
Format, material examined and depositories
In the systematic section the species are arranged in alphabetical order. For each (except
altaiensis) information is under the following headings.
Synonymy. When types have not been available for study (because of loss or destruction or
because they could not be obtained on loan) the inclusion of a nominal species in a synonymy is
based on consideration of the original description and/or it is substantiated by reference to a
published treatment or is considered more fully under 'Nomenclature'.
Nomenclature. This includes discussion of synonymies, type-restrictions, etc. It takes up a
considerable amount of space (and it took up a lot of time) because the nomenclatural problems
ICHNEUMON-FLY GENUS BANCHUS 9
associated with work on European species are related to human historical and sociological
factors rather than biological ones.
Description. The descriptions for females and males are separate. The characters included are
mainly those used in the keys, the ones which apply equally to both sexes are given for the female
only (except where the female is not known). The characters, including details of measure-
ments, are discussed in the 'Taxonomy' section above.
Remarks. These include brief notes on recognition and relationships of the species and
discussion of its taxonomy.
Biology and hosts. This is a brief summary based on data associated with material examined
and information gleaned from the literature. Names of hosts have been up-dated to conform
with current usage.
Distribution. As given this is based entirely on data taken from specimens examined, as are
the distribution maps. Map 1 shows the distribution of the genus (that is, the combined
distributions of all the species) in the Old World for comparison with the distributions of
individual species (Maps 2-11). The records in the literature are not reliable and only a few
references are made to them (Aubert (1978) should be consulted for more details of these).
Material examined. Except for new species and those known from only a few specimens, the
data under 'material examined' are restricted to totals of specimens, the names of countries (and
Map 1 Distribution of Banchus in the Old World.
10 M. G. FITTON
larger islands) from which they come and the depositories in which they are to be found. All
specimens on which this revision is based bear my determination labels. Some detail of collecting
localities is presented in the distribution maps for each species. However, a feature of much
European material is inadequate locality data on labels and there are no points on the maps for
perhaps a third of all the specimens examined. Dates of collection are also often lacking; for
instance, in an attempt to relate morphological to seasonal variation in Banchus pictus only 90 of
the 184 specimens to hand could be utilized because the rest had no date of collection.
The names of depositories are abbreviated as in the list below.
AC Collection of J. Aubert, Paris, France
ANS Academy of Natural Sciences, Philadelphia, U.S.A.
BC Collection of R. Bauer, Grossschwarzenlohe, B.R.D.
BMNH British Museum (Natural History)
BRI Biosystematics Research Institute, Ottawa, Canada
CM Castle Museum, Norwich, England
EIHU Entomological Institute, Hokkaido University, Sapporo, Japan
FSA Faculte des Sciences Agronomique de 1'Etat, Gembloux, Belgium
GC Collection of V . Gupta , University of Florida , Gainsville , U . S . A .
HC Collection of R. Hinz, Einbeck, B.R.D.
IBMPP Ail-Union Institute of Biological Methods of Plant Protection, Kishinev, U.S.S.R.
IEAU Istituto di Entomologia Agraria dell'Universita, Sassari, Italy
IEE Instituto Espanol de Entomologia, Madrid, Spain
IEUB Istituto di Entomologia, Universita degli studi di Bologna, Bologna, Italy
IP Institut fur Pflanzenschutzforschung, Eberswalde, D.D.R.
IRSNB Institut Royal des Sciences Naturelles de Belgique, Brussels, Belgium
ITZ Institut voor Taxonomische Zoologie, Zoologisch Museum, Amsterdam, Netherlands
IZPAN Instytut Zoologiczny, Polska Akademia Nauk, Warsaw, Poland
JC Collection of R. Jussila, Paattinen, Finland
JKC Collection of J . Kolarov , Sadovo , Bulgaria
JPM Jena Phyletisches Museum , Jena , D . D . R .
KC Collection of K. Kusigemati, Kagoshima, Japan
KHC Collection of K. Horstmann, Wiirzburg, B.R.D.
LELW Laboratorium voor Entomologie van de Landbouwhogeschool, Wageningen, Netherlands
LSL Linnean Society, London, England
MCSN Museo Civico di Storia Naturale, Genoa, Italy
MHN Museum d'Histoire Naturelle, Geneva, Switzerland
MIZS Museo ed Istituto di Zoologia Sistematica, Turin, Italy
MLSU Zoological Museum, Moscow Lomonosov State University, Moscow, U.S.S.R.
MNHN Museum National d'Histoire Naturelle, Paris, France
MNHU Museum fur Naturkunde der Humboldt-Universitat , Berlin , D.D.R.
MUM Manchester University Museum, Manchester, England
NC Collection of A. Nakanishi, Fukuoka, Japan
NM Naturhistorisches Museum, Vienna, Austria
NMB Naturhistorisches Museum, Basle, Switzerland
NMV National Museum of Victoria, Melbourne, Australia
NR Naturhistoriska Riksmuseet, Stockholm, Sweden
PC Collection of C. Pisica, lasi, Rumania
RNH Rijksmuseum van Natuurlijke Historic, Leiden, Netherlands
RSM Royal Scottish Museum, Edinburgh, Scotland
SC Collection of H. Schnee, Markkleeberg, D.D.R.
SMT Staatliches Museum fur Tierkunde , Dresden , D . D . R .
TAU Department of Zoology, Tel- Aviv University, Tel- Aviv, Israel
TC Collection of H. & M. Townes, American Entomological Institute, Ann Arbor, U.S.A.
TM Termeszettudomanyi Muzeum, Budapest, Hungary
UL Universite Laval, Quebec, Canada
UM University Museum, Oxford, England
USNM U.S. National Museum of Natural History, Washington, U.S.A.
UU Department of Entomology, University of Uppsala, Uppsala, Sweden
UZI Universitetets Zoologiska Institutionen, Lund, Sweden
ICHNEUMON-FLY GENUS BANCHUS 11
UZM Universitetets Zoologisk Museum, Copenhagen, Denmark
VRC Collection of G. van Rossem, Ede, Netherlands
ZC Collection of K. Zwart, Wageningen, Netherlands
ZI Zoological Institute, Wroclaw, Poland
ZIL Zoological Institute, Leningrad, U.S.S.R.
ZIM Zoological Institute and Museum, Sofia, Bulgaria
ZMU Zoological Museum of the University, Helsinki, Finland
ZPZ Zavod Za Poljoprivednu Zoologiju, Zagreb, Yugoslavia
ZSBS Zoologische Sammlung des Bayerischen Staates, Munich, B.R.D.
Keys to species
The keys to females and males are separate. Sexing specimens of Banchus is not always
straightforward. The short, broad ovipositor sheaths are sometimes mistaken for the male
genitalia. The most reliable way to sex specimens is to examine the last visible tergites and
sternites of the gaster: in females they are relatively long (Figs 6-10) and in males they are short
(Fig. 1).
Males (the males of cerinus and insulanus are unknown)
1 Maxillary palp (Figs 104, 114) with segment 5 less than 0-2 as long as segment 4; segment 4 long
and cylindrical , with only its extreme apex sharply expanded and flattened 2
- Maxillary palp (Figs 103, 105-113, 115-123) with segment 5 more than 0-4 as long as segment 4;
segment 4 gradually and evenly widened and flattened from base to apex or more sharply
expanded from near base 3
2 Hind femur less than 6-2 times as long as deep; entirely reddish yellow in colour. Gaster usually
entirely black, the segments sometimes brownish or yellowish on their posterior margins,
very rarely as conspicuous yellow bands on tergites 1, 2 and 3. Scutellar spine usually long
(more than 0-8 as long as scutellum) (Fig. 47) palpalis Ruthe (p. 30)
- Hind femur more than 6-5 times as long as deep; black, yellow proximally and distally and
sometimes dorsally. Gaster with each segment black anteriorly and yellow posteriorly.
Scutellar spine short (about 0-3 as long as scutellum) (Fig. 34) crefeldensis Ulbricht (p. 15)
3 Antennal flag setae well developed (Figs 126-129): upright, flattened, 2 or 3 (rarely 4) per
segment, arising from a more or less well-defined and polished trough 4
- Antennal flag setae not or only poorly differentiated (as in Fig. 124, or at most as in Fig. 125 but
shorter): at 60 degrees or less to the segment surface, not flattened, 2 per segment, never
arising from a trough 17
4 Antennal flag setae 3 (occasionally 4) per segment (Fig. 129)
- Antennal flag setae 2 per segment (Figs 126-128) (very rarely a few segments with 3 setae)
5 Scutellar spine long (about 0-8 as long as scutellum) (Fig. 44) mauricettae sp. n. (p. 28)
- Scutellar spine short (about 0-3 as long as scutellum) to absent (Figs 33, 37, 38) 6
6 Hind femur less than 5-3 times as long as deep, usually entirely reddish yellow in colour but can
be darkened medially. Malar space more than 0-6 times basal width of mandible
falcatorius (Fabricius) (p. 19)
- Hind femur about 6 1 times as long as deep , yellow in colour with the median part black . Malar
space about 0-5 times basal width of mandible agathae sp. n. (p. 14)
7 Tergite 1 of gaster at least 2-1 times as long as broad
Tergite 1 of gaster at most 2-0 times as long as broad 11
8 Hind femur at least 7-0 times as long as deep. Tergite 3 of gaster with a crease separating
laterotergite along its whole length gudrunaesp. n. (p. 23)
Hind femur less than 6-5 times as long as deep. Tergite 3 of gaster with a crease separating
laterotergite on only about its anterior 0-5
9 Meoscutum shining , with only weak microsculpture between punctures . Scutellar spine at most
about 0-5 as long as scutellum (Fig. 39) ffavomaculatus (Cameron) (p. 23)
- Mesoscutum not shining, with well-developed microsculpture between punctures. Scutellar
spine at least about 0-7 as long as scutellum (Figs 46, 52)
10 Hind femur about 6-0 times as long as deep. Antennal flagellum black with underside of
segment 1 and distal section reddish yellow. Segments of gaster each black anteriorly and
yellow posteriorly , with j unction of the two areas reddish nox Morley (p . 30)
12 M. G. FITTON
Hind femur at most 5-6 times as long as deep. Antennal flagellum reddish, darkened dorsally
(especially first few segments and distal third). Segments of gaster each reddish with some
blackish anteriorly, especially on tergites 1 and 2 sanjozanus Uchida (p. 37)
11 Maxillary palp (Figs 110, 116, 122) with segment 5 about 0-5 as long as segment 4 12
Maxillary palp (Figs 105, 113, 121 , 123) with segment 5 more than 0-7 as long as segment 4 14
12 Antennal flagellum black ventrally. Hind femur more than 6-2 times as long as deep. Gaster
with tergite 1 at most 1-65 times as long as broad; tergite 3 with a crease separating the
laterotergite on at least its anterior 0-8 japonicus (Ashmead) (p. 26)
Antennal flagellum yellowish ventrally. Hind femur at most 6-1 times as long as deep. Gaster
with tergite 1 at least 1-60 times as long as broad; tergite 3 with a crease separating the
laterotergite on its anterior 0-5 or less 13
13 Hind femur reddish yellow, usually entirely, rarely darkened ventrally; at most 5-4 times as
long as deep. Hind tibia and segment 1 of hind tarsus reddish yellow, the tibia and sometimes
the tarsal segment distally blackish volutatorius (Linnaeus) (p. 40)
Hind femur yellow, black-marked ventrally and internally; at least 5-6 times as long as deep.
Hind tibia and segment 1 of hind tarsus blackish, the tibia partly reddish yellow ventrally
poppiti sp. n. (p. 35)
14 Malar space at most 0-6 times basal width of mandible. Antennal flagellum with its proximal 0-6
orange turcator Aubert (p. 39)
- Malar space more than 0-6 times basal width of mandible. Antennal flagellum with at least its
dorsal surface entirely brownish or blackish 15
15 Antennal flagellum yellowish ventrally (and posterior margins of gastral tergites 1, 2 and 3
yellow) or, if antennal flagellum is black ventrally, the posterior margins of gastral tergites 1 ,
2 and 3 yellowish brown or reddish brown. Hind coxa entirely black
dilatatorius (Thunberg) (p. 17)
- Antennal flagellum blackish ventrally. Posterior margins of gastral tergites 1, 2 and 3 yellow.
Hind coxa black, usually with at least a dorsal yellow spot 16
16 Tergite 1 of gaster at most 1-6 times as long as broad. Hind femur at most 5-5 times as long as
deep. Fore wing length at most 10-0 mm moppitisp. n. (p. 28)
- Tergite 1 of gaster more than 1 6 times as long as broad . Hind femur more than 5 3 times as long
as deep. Fore wing length more than 10-5 mm zonatus Rudow (p. 41)
17 Tergite 1 of gaster with a strong median swelling in front of level of spiracles, the anterior face
of swelling at almost 90 degrees to its dorsal surface (Figs 77, 78). Scutellar spine long (at
least 0-7 as long as scutellum) (Figs 53, 54) 18
- Tergite 1 of gaster at most weakly swollen at level of spiracles (Figs 64, 73, 75). Scutellar spine
moderately long (about 0-6 as long as scutellum) to very small (Figs 41 , 49, 51) 19
18 Segment 4 of maxillary palp relatively weakly expanded (Fig. 119); uniformly reddish yellow in
colour tholussp. n. (p. 37)
Segment 4 of maxillary palp considerably expanded from the base (Fig. 120); bi-coloured
tumidus Chandra & Gupta (p. 38)
19 Maxillary palp (Fig. 117) with segment 5 about 0-6 as long as segment 4; segment 4 considerably
widened and flattened. Hind femur about 6-0 times as long as deep punkettai sp. n. (p. 36)
Maxillary palp (Figs 109, 115) with segment 5 about 0-8 as long as segment 4; segment 4 slightly
or moderately widened and flattened. Hind femur less than 5-7 times as long as deep 20
20 Posterior part of propodeum entirely black. Hind coxa almost always entirely black. Segments
of gaster each black, usually with the posterior margin brownish yellow. Segment 4 of
maxillary palp relatively slightly widened (Fig. 109). Antennal flag setae poorly differenti-
ated and very small (Fig. 124) hastator (Fabricius) (p. 24)
Posterior part of propodeum almost always with yellow marks. Hind coxa black, often with a
yellow dorsal patch. Segments of gaster each black anteriorly, broadly yellow posteriorly.
Segment 4 of maxillary palp relatively more widened (Fig. 115). Antennal flag setae at about
60 degrees to the segment surface, small (as in Fig. 125 but setae shorter) pictus Fabricius (p. 33)
Females (the female of sanjozanus is not included in the key)
1 Maxillary palp (Figs 13, 24) with segment 5 at most about 0-5 as long as segment 4; segment 4
relatively very long and slender 2
Maxillary palp (Figs 11, 12, 14-23, 25-32) with segment 5 at least 0-7 as long as segment 4;
segment 4 variable in proportions, but much less slender 3
ICHNEUMON-FLY GENUS BANCHUS 13
2 Hind femur less than 6-0 times as long as deep; reddish yellow in colour, sometimes dark-
ened ventrally. Gaster usually entirely black, posterior edges of some segments may be
brownish. Scutellar spine long (usually more than 0-8 as long as scutellum) (Fig. 47)
I ml pa Us Ruthe (p. 30)
- Hind femur more than 6-3 times as long as deep; black and yellow (proximally and distally and
sometimes dorsally). Gaster with each segment black anteriorly and yellow posteriorly.
Scutellar spine short (about 0-3 as long as scutellum) (Fig. 34) crefeldensis Ulbricht (p. 15)
3 Posterior part of propodeum entirely black 4
- Posterior part of propodeum with at least some yellow or reddish marks 9
4 Malar space at least 0-8 times basal width of mandible. Width of lower face at least 1-1 times
vertical length of eye. Hind femur mainly black dilatator ius (Thunberg) (p. 17)
- Malar space at most 0-7 times basal width of mandible. Width of lower face at most 1-0 times
vertical length of eye. Hind femur entirely reddish yellow or with more or less extensive
black areas 5
5 Tergite 7 of gaster (Fig. 85) elongate, subacute, its upper surface rounded posteriorly
falcatorius (Fabricius) (p. 19)
- Tergite 7 of gaster (Figs 88, 89, 95, 101) of normal length, subtruncate, its upper surface not
markedly rounded posteriorly 6
6 Hind femur more than 6-0 times as long as deep. Tergite 3 of gaster with a crease separating
laterotergite along its whole length insulanus Roman (p. 26)
- Hind femur at most 5-5 times as long as deep. Tergite 3 of gaster with a crease separating
laterotergite on less than its anterior 0-5 7
7 Tergite 1 of gaster at least 1-8 times as long as broad. Hind femur entirely reddish yellow
volutatorius (Linnaeus) (p. 40)
- Tergite 1 of gaster at most 1-7 times as long as broad. Hind femur reddish yellow with more or
less extensive black areas 8
8 Tergites of gaster each black anteriorly, broadly yellow posteriorly, the boundary between the
two areas sharply defined pictus Fabricius (p. 33)
- Tergites of gaster each black, grading posteriorly, to a greater or lesser extent, into a brownish
and sometimes yellowish marginal area hastator (Fabricius) (p. 24)
9 Tergite 7 of gaster (Figs 84, 92, 102) relatively elongate, subacute, its upper surface rounded
posteriorly AND width of lower face at least 1-0 times vertical length of eye 10
- Tergite 7 of gaster (Figs 6, 82, 86, 87, 90, 91, 93, 95-100) of normal length, subtruncate, its
upper surface not markedly rounded posteriorly AND/OR width of lower face at most 0-9 times
vertical length of eye 12
10 Antennal flagellum yellowish orange in colour, with the base of segment 1 and the distal
segmentsdark dilatatorius (Thunberg) (p. 17)
- Antennal flagellum black, brownish distally 11
11 Tergite 7 of gaster (Fig. 102) more elongate. Hind femur at least 5-3 times as long as deep. Fore
wing length at least 11-0 mm zonatus Rudow (p. 41)
- Tergite 7 of gaster (Fig. 92) less elongate. Hind femur at most 5-2 times as long as deep. Fore
wing length at most 10-8 mm moppiti sp. n. (p. 28)
12 Tergite 1 of gaster with a strong median swelling in front of level of spiracles, the anterior face
of the swelling at almost 90 degrees to its dorsal surface (Figs 77, 78)
tholussp. n. (p. 37) and tumidus (Chandra & Gupta) (p. 38)
(It has proved impossible to separate reliably the females of these two species. See notes
under tumidus.}
- Tergite 1 of gaster at most weakly swollen in front of level of the spiracles (Figs 6 , 59 , 63 , 65 , 66 ,
68,69,71,73,74,75,79) 13
13 Antennal flagellum entirely black. Hind femur largely black (yellow or reddish proximally and
distally) or yellow with a black area medially (which extends onto its dorsal surface) 14
- Antennal flagellum reddish, orange, or yellow, at least on its proximal 0-5, or dark dorsally
(black or brownish) and paler ventrally (yellow or reddish). If antennal flagellum is dark
(blackish or brownish dorsally and dark reddish ventrally) then hind femur reddish with a
black ventral mark, which does not reach to dorsal surface 16
14 Tergite 3 of gaster with a crease separating laterotergite along its whole length. Hind femur
more than 6-6 times as long as deep. Hind tibia yellow, with its distal 0-2 blackish
gudrunae sp. n. (p. 23)
- Tergite 3 of gaster with a crease separating laterotergite on only its anterior 0-5. Hind femur at
14 M. G. FITTON
most 6-5 times as long as deep. Hind tibia yellow with its distal 0-2 blackish or black with a
median yellowish band 15
15 Hind tibia yellow with its distal 0-2 black. Hind femur about 5-4 times as long as deep
agathae sp. n. (p. 14)
Hind tibia black with a median yellowish band. Hind femur more than 5-5 times as long as deep
japonicus (Ashmead) (p. 26)
16 Hind femur entirely yellow or entirely reddish 17
Hind femur black and yellow or black and reddish , at least reddish with a ventral black mark .... 18
17 Hind femur more than 6-0 times as long as deep. Tergite 1 of gaster more than 2-5 times as long
as broad cerinus Chandra & Gupta (p. 15)
Hind femur less than 5-5 times as long as deep. Tergite 1 of gaster less than 2-5 times as long as
broad mauricettae sp. n. (p. 28)
18 Tergites of gaster each black anteriorly, broadly to narrowly yellow posteriorly, the boundary
between the two areas sharply defined 19
- Tergites of gaster (at least 1, 2 and 3) each black anteriorly, reddish or yellow posteriorly, if
yellow with an intermediate reddish band grading into the black and into the yellow 21
19 Hind tibia and tarsus mainly blackish, tibia and segment 1 of tarsus slightly lighter proximally.
Tergites of gaster each only narrowly yellow posteriorly poppiti sp. n. (p. 35)
- Hind tibia, on at least its proximal 0-5, and tarsus yellow. Tergites of gaster each broadly to
relatively narrowly yellow posteriorly 20
20 Gaster (about as in Figs 7,8) relatively less compressed posteriorly ; tergite 7 and sternite 6 (Fig .
100) subtruncate. Antennal flagellum entirely yellowish orange (at least its proximal 0-7),
sometimes with a very small dark mark dorsally on segment 1 turcator Aubert (p . 39)
- Gaster relatively more compressed posteriorly; tergite 7 and sternite 6 (Fig. 95) more acute.
Antennal flagellum yellowish orange, sometimes with proximal section dark dorsally and
almost always with at least proximal 0-5 of segment 1 blackish dorsally . . . pictus Fabricius (p. 33)
21 Tergites of gaster each black anteriorly (at least 1, 2 and 3), reddish posteriorly. Tergite 1 of
gaster at least 2-25 times as long as broad 22
Tergites of gaster each black anteriorly, yellow posteriorly and with an intermediate reddish
band. Tergite 1 of gaster at most 2 -25 times as long as broad punkettai sp. n. (p. 36)
22 Posterior part of propodeum entirely reddish. Gaster with tergites 1 to 5 each black anteriorly
ffavomaculatus (Cameron) (p. 23)
- Posterior part of propodeum black, with an irregular reddish yellow band along posterior
transverse carina. Gaster with only tergites 1 , 2 and 3 black anteriorly no* Morley (p. 30)
Systematic section
Hunch us agathae sp. n.
(Figs 11, 33, 59, 82, 103)
[Banchus pictus Fabricius; Bischoff, 1930: 225. Misidentification (and wrong sex).]
DESCRIPTION. Female. Width of lower face 0-90 times vertical length of eye. Malar space 0-60 times basal
width of mandible. Maxillary palp as in Fig. 11. Antenna with apical segments as long as broad. Scutellar
spine (Fig. 33) about 0-3 as long as scutellum. Mesopleuron and mesoscutum shining, only very weakly
sculptured, with strong punctures, on the mesopleuron separated by much less than their diameter. Hind
femur 5-40 times as long as deep. Fore wing length 10-1 mm. Tergite 1 of gaster 2-00 times as long as broad,
its dorsal profile as in Fig. 59. Gaster compressed from the posterior of segment 3, reaching just beyond the
tips of the fore wings (when folded back). Tergite 3 with a crease along its anterior 0-5 separating
laterotergite. Tergite 7 and sternite 6 as in Fig. 82.
Colour: black and yellow. Face yellow with a median black stripe. Antenna black with scape and pedicel
yellow ventrally. Maxillary palp yellowish with dark marks, including distal 0-7 of segment 4 and all of
segment 5. Posterior part of propodeum yellow, narrowly black posteriorly. Hind coxa black with a yellow
dorsal patch and distal margin narrowly yellow. Hind femur black, yellow proximally and distally.
Segments of gaster each black anteriorly, very narrowly yellow posteriorly.
Male. Width of lower face 0-90 times vertical length of eye. Malar space 0-45 times basal width of
mandible. Maxillary palp as in Fig. 103; segment 4 flattened and considerably widened; segment 5 about
0-7 as long as 4. Antennal flag setae (similar to Fig. 129, but the setae only about half as long) upright, short,
flattened, 3 per segment, arising from a polished trough. Hind femur 6-10 times as long as deep. Fore wing
ICHNEUMON-FLY GENUS BANCHUS 15
length 8-5 mm. Tergite 1 of gaster 1-80 times as long as broad. Tergite 3 with a crease along its anterior 0-5
separating the laterotergite.
Colour: black and yellow. Antenna black with scape, pedicel and proximal and distal parts of flagellum
yellow ventrally. Maxillary palp with segments 1 yellowish, 2 and 3 reddish, 4 and 5 blackish. Posterior part
of propodeum yellow, narrowly black posteriorly. Hind coxa black with a yellow dorsal spot and narrow
postero-ventral margin. Hind femur yellow, black medially. Segments of gaster each black anteriorly,
yellow posteriorly.
REMARKS. The male should be readily recognized by the form of the antennal flag setae. The female is
keyed to a large extent on colour characters but the form of the posterior part of the gaster is also
important.
BIOLOGY AND HOSTS. Unknown. The specimens were taken in July and August.
DISTRIBUTION (Map 2, p. 16). Known from only two specimens collected in the Pamir.
MATERIAL EXAMINED
Holotype cf , U.S.S.R.: Pamir, Umss-Tugai, 25.vii.1928 (Rickmers) (ZSBS).
Paratype $, U.S.S.R.: Pamir, Maz, 3580 m, 15.viii.1928 (Reinig) (MNHU).
Banchus altaiensis Meyer nomen dubium
Banchus altaiensis Meyer, 19276: 295. Type(s) cf , U.S.S.R. (destroyed).
This species was described from material, probably only one specimen, collected in the Altai mountains.
The description does not fit adequately any of the species from Central Asia known to me and I hesitate to
identify it with any species not proven to occur there. Its identification will have to wait until more extensive
collections have been made in the area.
Banchus cerinus Chandra & Gupta
(Figs 6, 12)
Banchus cerinus Chandra & Gupta, 1977: 185. Holotype $, INDIA (GC) [examined].
DESCRIPTION. Female. Width of lower face 0-85 times vertical length of eye. Maxillary palp as in Fig. 12.
Antenna with apical segments longer than broad. Scutellar spine of moderate length. Mesopleuron and
mesoscutum weakly coriaceous, with dense, fairly small punctures. Hind femur 6-25 times as long as deep.
Fore wing length 14-3 mm. Tergite 1 of gaster 2-75 times as long as broad, its dorsal profile as in Fig. 6.
Gaster subcylindrical, only weakly compressed posteriorly, not reaching to tips of fore wings (when folded
back). Tergite 3 with a crease along its anterior 0-6 separating laterotergite. Tergite 7 and sternite 6 as in
Fig. 6.
Colour: mainly yellow, with black marks. Face yellow with a black mark between and just below
antennal sockets. Antenna yellow with distal half of flagellum and an externo-lateral mark on pedicel
blackish. Maxillary palp yellowish. Posterior part of propodeum yellow. Hind coxa yellow, blackish
distally. Hind femur yellow. Segments of gaster yellow, with basal third of tergites 1,2 and 3 and of sternite
2 black.
Male. Unknown.
REMARKS. A pale-coloured species with relatively long appendages. Its closest relatives seem to be tholus
and tumidus.
DISTRIBUTION (Map 4, p. 20), BIOLOGY AND HOSTS. The species is known only from the holotype, which was
collected amongst mixed bushes in coniferous forest in the NW. Himalayas (Gupta, 1975: (appendix) 50).
MATERIAL EXAMINED (1 $)
India: 1 $ (holotype), Himachal Pradesh, Ahla, 2286 m, 18.vii.1971 (Gulati) (GC).
Banchus crefeldensis Ulbricht
(Figs 4, 13,34,60,83,104, 125)
Banchus crefeldensis Ulbricht, 1916: 12. Holotype cf, WEST GERMANY (A. Ulbricht coll., Krefeld)
[examined].
Banchus croaticus Hensch, 1928: 99. Lectotype $, YUGOSLAVIA (ZPZ), designated by Horstmann, 1982a:
82 [examined].
16
M. G. FITTON
Map 2 Distribution of Banchus agathae, B. japonicus, B. mauricettae, B. nox, B. sanjozanus, B. tholus
and B. tumidus.
DESCRIPTION. Female. Width of lower face 0-90-1-00 times vertical length of eye. Malar space 0-60-0-75
times basal width of mandible. Maxillary palp as in Fig. 13. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 34) about 0-3 as long as scutellum. Mesopleuron and mesoscutum strongly
coriaceous, with moderately strong punctures, on mesopleuron separated by about their diameter. Hind
femur 6-35-7-50 times as long as deep. Fore wing length 9-9-10-7 mm. Tergite 1 of gaster 1-80-2-00 times as
long as broad, its dorsal profile as in Fig. 60 (male). Gaster compressed from segment 4, reaching about to
tips of fore wings (when folded back). Tergite 3 with a crease along its entire length separating
laterotergite. Tergite 7 and sternite 6 as in Fig. 83.
ICHNEUMON-FLY GENUS BANCHUS 17
Colour: black and yellow. Face black with broad yellow orbital stripes and a vertical yellow mark below
each antennal socket. Antenna entirely black except for a yellow patch on underside of scape, and rarely
also on pedicel. Maxillary palp blackish with part or all of segment 2 and proximal 0-7 of segments 3 and 4
reddish yellow. Posterior part of propodeum yellow anteriorly, black posteriorly. Hind coxa black with a
yellow dorsal patch. Hind femur black, yellow proximally and distally and sometimes dorsally. Segments of
gaster each black anteriorly and broadly yellow posteriorly.
Male. Width of lower face 0-90-1-00 times vertical length of eye. Malar space 0-55-0-70 times basal
width of mandible. Maxillary palp as in Fig. 104; segment 4 narrow and cylindrical, flattened and widened
only at its extreme distal apex; segment 5 less than 0-2 as long as 4. Antennal flag setae (Fig. 125) at about 50
degrees, not flattened, 2 or 3 per segment, not arising from a trough. Hind femur 6-60-7-65 times as long as
deep. Fore wing length 9-1-10-8 mm. Tergite 1 of gaster 1-80-2-10 times as long as broad (Fig. 4). Tergite 3
with a crease along its entire length separating laterotergite.
Colour: black and yellow. Antenna black, with scape and pedicel yellow ventrally. Maxillary palp
blackish with segments 2 wholly and 3 and 4 proximally reddish yellow. Posterior part of propodeum black
with a variable amount of yellow anteriorly, ranging from a very broad band along posterior transverse
carina to a small median spot to absent entirely. Hind coxa black, usually with a yellow dorsal patch. Hind
femur black, yellow proximally and distally and sometimes dorsally. Segments of gaster each black
anteriorly, yellow posteriorly (on at least tergites 2 and 3 relatively broadly so).
REMARKS. Like the related B. palpalis, this is a very distinctive species; it is readily recognized by the
structure of the maxillary palps and its coloration.
BIOLOGY AND HOSTS. Dates of collection range from mid-February (in Spain and Portugal) to mid-May (in
Scotland and Ireland). Stelfox (1936: 63) records males flying in numbers round ivy (Hedera helix) and
gorse (Ulex europaeus) in Ireland. Specimen labels indicate capture of males on Betula and around Buxus
and females from pine.
I have seen 11 reared specimens, 3 without a satisfactory host identification, the remaining 8 (including
both sexes) fromAporophyla lutulenta (Denis & Schiffermuller) or A. lutulenta 'subspecies' lueneburgensis
(Freyer) (Noctuidae). The named hosts were collected at Bussum, Netherlands; Aviemore, Great Britain;
and Bremen and near Hamburg, West Germany.
DISTRIBUTION (Map 3, p. 18). Widely distributed in Europe but not found in Scandinavia or the U.S.S.R.
In Great Britain it is restricted to the Scottish Highlands.
MATERIAL EXAMINED (40 $ , 79 cf )
Austria, Belgium, France, Great Britain, Hungary, Ireland, Italy, Netherlands, Portugal, Spain,
Switzerland, Tunisia, Turkey, West Germany, Yugoslavia (BMNH, FSA, IEE, IRSNB, LELW, MHN,
MNHN, NM, NMB, NR, RNH, UM, USNM, ZC, ZSBS).
Banchus dilatatorius (Thunberg) stat. rev.
(Figs 2, 14, 35, 36, 61, 84, 105)
Ichneumon acuminator Fabricius, 1787: 268. Type(s) cf , EAST GERMANY (lost). [Junior primary homonym
of Ichneumon acuminator Miiller, 1776: 157.] Syn. n.
Ichneumon compressus Fabricius, 1787: 381. Holotype $, SWEDEN (UZM) [examined]. [Junior primary
homonym of Ichneumon compressus Sulzer, 1776: 190.]
Ichneumon dilatatorius Thunberg, 1822: 279; 1824: 360. Holotype $, GERMANY (UU) [not examined].
Banchus sibiricus Meyer, 19276: 294. Syntypes $, cf , U.S.S.R. (destroyed). Syn. n.
NOMENCLATURE. In most of the literature this species is referred to by the invalid name compressus.
Although Ichneumon acuminator has been placed previously in Banchus it has never been identified.
The description fits certain males of dilatatorius if one allows that Fabricius failed to observe the pale colour
of the underside of the antennae.
The description of Banchus sibiricus fits well specimens from Kamchatka (which include those identified
by Roman (1931: 29) as lavrovi var.) which are considered conspecific with dilatatorius but are discussed
under 'Remarks' below. The identity of the female from Rumania determined as sibiricus by Constan-
tineanu & Pisica (1959: 190) is not known.
DESCRIPTION. Female. Width of lower face 1-15-1-25 times vertical length of eye. Malar space 0-80-1-00
times basal width of mandible (Fig. 2). Maxillary palp as in Fig. 14. Antenna with apical segments about as
broad as long. Scutellar spine (Figs 35, 36) very small or virtually absent. Mesopleuron and mesoscutum
coriaceous, with strong punctures, on mesopleuron separated by less than diameter. Hind femur 4-35-4-85
M. G. FITTON
o *
crefeldensis
Map 3 Distribution of Banchus crefeldensis.
times as long as deep. Fore wing length 8-7-11-5 mm. Tergite 1 of gaster 1-30-1-40 times as long as broad,
its dorsal profile as in Fig. 61. Gaster strongly compressed from posterior of tergite 3, reaching beyond tips
of fore wings (when folded back). Tergite 3 with a crease along its anterior 0-3 separating laterotergite.
Tergite 7 and sternite 6 as in Fig. 84.
Colour: black and yellow (the yellow often rather creamy). Face black with yellow orbital marks.
Antenna with scape and pedicel black, scape, and usually pedicel, yellowish or reddish ventrally.
Flagellum yellowish orange, with base of segment 1 and distal segments dark. Maxillary palp blackish,
entirely or with segments 2 and 3 brownish. Posterior part of propodeum black, entirely or, more usually,
with a yellow band along posterior transverse carina, the band usually not extensive and narrowly absent
medially. Hind coxa black. Hind femur black, narrowly yellowish proximally and usually widely yellowish
distally. Segments of gaster each black anteriorly, yellow posteriorly, the yellow bands usually less than 0-3
of length of segments.
Male. Width of lower face 1-10-1-25 times vertical length of eye. Malar space 0-85-0-95 times basal
width of mandible. Maxillary palp as in Fig. 105; segment 4 only slightly flattened and widened distally;
segment 5 about 0-9 as long as 4. Antennal flag setae (similar to Fig. 126) upright, flattened, relatively
narrow, 2 per segment, arising from a polished trough. Hind femur 4-35-5-00 times as long as deep. Fore
wing length 8-6-10-7 mm. Tergite 1 of gaster 1-25-1-55 times as long as broad. Tergite 3 with a crease along
its anterior 0-3 separating laterotergite.
Colour: black and yellow. Antenna blackish or brownish dorsally, yellow or yellowish ventrally (except
for apex of flagellum which is dark). Maxillary palp blackish or brownish. Posterior part of propodeum
black, entirely or with a yellow band (sometimes broken medially) along posterior transverse carina. Hind
coxa black. Hind femur yellow with a black mark on interne-lateral and ventral surface, and sometimes
also dorsally near base. Segments of gaster each black anteriorly, yellow posteriorly, the yellow bands
usually less than 0-3 of length of segment.
ICHNEUMON-FLY GENUS BANCHUS 19
REMARKS. This species is distinctive and easily identified. The very wide face is a conspicuous feature.
However, the poor characters used in earlier keys led to confusion with other species in several of the
collections examined.
The details of colour given in the description above do not apply fully to all of the material examined.
One female from Iran (AC) has the yellow coloration much more extensive than in any other specimens
seen. All the specimens from Kamchatka (1 $ (MLSU) and 4 $ , 2 0" (NR)) have yellow absent but with
reddish in place of the yellow on the legs and on the face of the male, and with the margins of some or all of
tergites 1, 2 and 3 of the gaster narrowly to broadly reddish brown. Other than in these colour characters
the Iranian and Kamchatka specimens do not differ much from 'typical' dilatatorius . It is impossible to
decide whether or not they represent extremes of colour variation or distinct species. Unfortunately the
character which might enable a decision to be reached, the male flag setae, cannot be investigated. Of the
two male specimens one lacks the head completely and the other lacks both antennae. The Kamchatka
specimens match the description of B. sibiricus which was described from Irkutsk (see 'Nomenclature'
above). The material from Mongolia identified by Momoi (1973: 242-243) as lavrovi might also be this dark
form of dilatatorius. I could not examine these Mongolian specimens because Momoi has so far failed to
return them to Budapest (Zombori, pers. comm.). The six specimens in Stockholm (NR) had been
identified by Roman (1931: 29) as lavrovi var.
BIOLOGY AND HOSTS. This is an early spring species with most records relating to the period from late March
to late April. The earliest dates of collection (7 and 13 March) come from sand dune areas in the
Netherlands. The label on one female records that it was found on catkins of Salix caprea. I have seen no
reared specimens of this species, although the following hosts are recorded in the literature: Blepharita
adusta (Esper) (Noctuidae) (Bajari, 1960: 260); Euxoa nigricans (Linnaeus) (Noctuidae) (Schmiedek-
necht, 1910: 1926; Meyer, 1934: 228); Phlogophora meticulosa (Linnaeus) (Noctuidae) (Gyorfi, 1944:
106); Euproctis similis (Fuessly) (Lymantriidae) (de Gaulle, 1907: 19); and Leucoma salicis (Linnaeus)
(Lymantriidae) (Leonardi, 1928: 83).
DISTRIBUTION (Map 4, p. 20). Western and central Europe, mainly north of 50 degrees N. (but with very
few records from the area bordering the Atlantic (France, the British Isles and Norway) and only a few
from Denmark and Sweden); the European Alps; and with some scattered records in the western
U.S.S.R. , north-west Iran and the Soviet Far East (see 'Remarks' above).
MATERIAL EXAMINED (86 9 , 96 cf)
Austria, Belgium, Czechoslovakia, Denmark, East Germany, Finland, France, Great Britain, Hungary,
Iran, Italy, Netherlands, Sweden, Switzerland, U.S.S.R., West Germany (AC, BMNH, BRI, HC, IP,
IRSNB, ITZ, LELW, MNHN, MHN, MIZS, NMB, NR, RNH, SMT, USNM, UZI, UZM, ZC, ZIL,
ZMU, ZSBS).
Banchus falcatorius (Fabricius)
(Figs 5, 9, 10, 15, 37, 38, 62, 85, 106, 129)
Ichneumon falcatorius Fabricius, 1775: 332. Holotype d", DENMARK (UZM) [examined].
Ichneumon variegator Fabricius, 1775: 339. LECTOTYPE d", SWEDEN (UZM), here designated
[examined] .
Ichneumon intersectus Geoffroy, 1785: 414. Type(s) cf , FRANCE (lost) (Horstmann, 19826: 243).
Ichneumon aries Christ, 1791: 339. Type(s) $, no type-locality (lost).
Ichneumon labiatus Schrank, 1802: 264. Syntypes $ [not cf as stated by Schrank], WEST GERMANY (lost).
Ichneumon histrio Schrank, 1802: 265. Syntypes cf , WEST GERMANY and FRANCE (lost) (Horstmann,
19826: 243). [Junior primary homonym of Ichneumon histrio Christ, 1791: 356.]
Ichneumon tricolor Schrank, 1802: 286. Syntypes cf , WEST GERMANY (lost).
Banchus falcator Fabricius, 1804: 128. [Unjustified emendation of Ichneumon falcatorius Fabricius, 1775.]
CorynephanessachalinensisMatsumura, 1911: 92. Lectotype d" [not $ as stated by Matsumura] , U.S.S.R.
(EIHU), designated by Townes, Momoi & Townes, 1965: 236 [examined].
Banchus falcatorius var. luteofasciatus Ulbricht, 1911: 151. Type(s) , HUNGARY (?lost).
Banchus nobilitator Morley, 1915: 138. Holotype $, U.S.S.R. (BMNH) [examined].
Banchus falcatorius var. sanguinator Meyer, 1922: 139. Holotype d", U.S.S.R. (destroyed).
Banchus lavrovi Meyer, 19276: 294. Syntypes $, cf, U.S.S.R. (destroyed).
Banchus falcatorius var. nigromarginatus Constantineanu & Pisica, 1960: 710. Syntypes 3 $, RUMANIA
(?Constantineanu coll.) [not examined].
Banchus propitius Kuslitzky, 1979: 351. Holotype d", MONGOLIA (ZIL) [not examined]. [Provisional
synonymy.]
M. G. FITTON
Map 4 Distribution of Banchus cerinus, B. dilatatorius and B. turcator.
NOMENCLATURE. Ichneumon labiatus was listed by Dalla Torre (1901: 63) in the synonymy of compressus
(= dilatatorius). However, as noted by Gravenhorst (1829: 390) the description applies to B. falcatorius.
Schrank misdetermined the sex of his species.
Ichneumon tricolor has been synonymised usually with compressus (= dilatatorius) (e.g. Aubert, 1978:
153). The description and the figure to which Schrank refers (Schaeffer, 1768: pi. 116, fig. 5), differ from
that species in the coloration of the gaster. However, they fit the male of B. falcatorius and there can be
little doubt about the identity of the two species. This synonymy was first proposed by Gravenhorst (1829:
396).
The synonymy of sanguinator might be questioned. I have seen no male of B. falcatorius as red as
described by Meyer; the most reasonable explanation of the colour is that the specimen had suffered from
overlong exposure to cyanide in a killing jar.
The description of lavrovi fits specimens from the northern part of the eastern Palaearctic and here
included in falcatorius . One such specimen determined as lavrovi by Meyer himself has been examined.
The oldest name applying to this form of falcatorius is nobilitator Morley.
In terms of falcatorius as recognized in this revision (see 'Remarks' below) I have thought it best
provisionally to synonymize the species, propitius, recently described by Kuslitzky (1979).
DESCRIPTION. Female. Width of lower face 0-95-1-00 times vertical length of eye. Malar space 0-60-0-70
times basal width of mandible. Maxillary palp as in Fig. 15. Antenna with apical segments just longer than
broad. Scutellar spine (Fig. 37) minute or absent (but see 'Remarks'). Mesopleuron and mesoscutum
ICHNEUMON-FLY GENUS BANCHUS 21
coriaceous, with moderately strong punctures, on mesopleuron separated by less than their diameter. Hind
femur 4-25^4-70 times as long as deep. Fore wing length 10-1-11-2 mm. Tergite 1 of gaster 1-45-1-70 times
as long as broad, its dorsal profile as in Fig. 62. Gaster (Figs 9, 10) strongly compressed, reaching beyond
tips of fore wings (when folded back). Tergite 3 with a crease along its anterior 0-3 separating laterotergite.
Tergite 7 and sternite 6 as in Fig. 85.
Colour: mainly black, with legs largely reddish and usually some reddish and very rarely yellow on the
gaster. Face black, often with very small reddish or yellowish orbital spots. Antenna dark brown or
blackish dorsally, brown or reddish ventrally. Maxillary palp reddish with segments 4 apically and 5 wholly
dark. Posterior part of propodeum entirely black. Hind coxa black, very rarely with a postero-ventral
reddish spot. Hind femur reddish yellow, entirely or, in a few specimens, blackish medially. Segments of
gaster black, rarely entirely so, usually with variable reddish marks on segments 1,2 and 3 and more rarely
on other segments. Very rarely reddish areas grade to yellow posteriorly.
Male. Width of lower face 0-95-1-10 times vertical length of eye. Malar space 0-65-0-75 times basal
width of mandible. Maxillary palp as in Fig. 106; segment 4 considerably widened and flattened; segment 5
about 0-5 as long as 4. Antennal flag setae (Fig. 129) upright, long, flattened and considerably widened, 3
(sometimes 4) per segment, arising from a polished trough. Hind femur (Fig. 5) 4-45-5-25 times as long as
deep. Fore wing length 9-2-11-7 mm. Tergite 1 of gaster 1-40-1-70 times as long as broad. Tergite 3 with a
crease along its anterior 0-3 separating the laterotergite.
Colour: black, yellow and reddish. Antenna black dorsally (and ventrally at distal apex of flagellum),
yellow ventrally. Maxillary palp with segment 1 yellow, segments 2 and 3 wholly and 4 proximally reddish
yellow, segment 4 distally and 5 wholly brown or blackish. Posterior part of propodeum yellow anteriorly,
black posteriorly, the yellow sometimes reduced to a pair of latero-median spots or entirely absent. Hind
coxa black, usually with a yellowish or reddish patch postero-ventrally and sometimes also dorsally. Hind
femur reddish yellow, almost always entirely but in a few specimens blackish medially. Gaster with tergite 1
black anteriorly, reddish and then yellow posteriorly; tergites 2 and 3 each reddish anteriorly, yellow
posteriorly (with some blackish laterally); remaining tergites black, 4, 5 and 6 each with a postero-median
yellow spot.
REMARKS. B. falcatorius has several specialized features and it is the species in which sexual dimorphism is
most marked. There is also some geographical variation in characters; for instance, specimens from the
eastern part of the range have the hind femur darkened medially and a few specimens from Turkey have a
well-developed spine on the scutellum. More than one species might be included in the taxon here
recognized, but in the limited material from the eastern Palaearctic available for study no clear-cut
segregates could be recognized.
Although the sex association in this species is not in doubt, interesting confirmation comes from the
discovery of four gynandromorph specimens in the extensive material examined. These were the only
gynandromorphs found (or at least recognized) during the present study. Details of the specimens are as
follows.
Italy: Piemonte, Susa, 24. vi. 1872 (Gribodo) (MCSN). Head male (unfortunately the antennal flagella
are missing), thorax and abdomen female. Italy: Cadore, Valle del Boite (IEUB). Head female, thorax
and propodeum apparently mainly female but with yellow marks on the left of the mesoscutum and
scutellum (male characters), gaster apparently male (including the genitalia) but with some bilateral
asymmetry in colour anteriorly. ?locality [illegible] (Giraud coll.) (MNHN). Head (including antennae
and palps) male on the left and female on the right, prothorax and anterior of mesothorax mainly female
but partly male on left, posterior of mesothorax and remainder of body female. Interestingly this specimen
had been labelled (anonymously) as a hybrid between falcatorius and monileatus [= palpalis}. Denmark:
Ordrup, 19.viii. 1877 (Drewsen) (UZM). Head (including appendages) female, remainder male except that
the dark colour of the left fore and mid coxae suggests some female influence.
BIOLOGY AND HOSTS. This species is on the wing from early June to mid August (with a few records for late
May and late August). The adults have often been collected from umbellifer flowers. One female is
labelled as being the prey of Dasypogon teutonics (Diptera; Asilidae).
Despite the abundance of specimens in collections there are very few reared examples, and not all of
those have an attached host name or host remains. The most reliable host data relate to Agrotis segetum
(Denis & Schiffermiiller) (Noctuidae) and this is certainly an important host species. It is (or was) a major
pest of root crops (particularly various beets and carrots) and its association with B. falcatorius was the
subject of research in the U.S.S.R. in the 1920s and 30s (Kosobutzkii, 1928; Meyer, 1927a, 1928; Pospelov,
1924; Samoilova, 1936). The long, knife-like gaster of the female is undoubtedly adapted to enable it to
reach the host larvae, which feed at soil level. Other hosts recorded on specimen labels are: Deilephila
porcellus (Linnaeus) and D. elpenor (Linnaeus) (Sphingidae), Dasychira fascelina (Linnaeus) (Lyman-
22
M. G. FITTON
triidae), Acronicta megacephala (Denis & Schiffermiiller) (Noctuidae) and Melitaea didyma (Esper)
(Nymphalidae). The last named host record seems unlikely to be accurate, but it is difficult to judge the
others (all the specimens are without locality or date !). Other hosts recorded in the literature are: Agrotis
exclamationis (Linnaeus) (Noctuidae) (Meyer, 1927a: 81) and, erroneously, two Hymenoptera (Diprion
pini (Linnaeus) (Diprionidae) and Ophion luteus (Linnaeus) (Ichneumonidae) (Maneval, 1935: 74)).
DISTRIBUTION (Map 5, p. 22). Most of the Palaearctic, but few specimens seen from large areas of the
U.S.S.R. and further east than the Caucasus none from south of 50 degrees N. In the more southern parts
of its range in the western Palaearctic it is apparently restricted to higher altitudes.
MATERIAL EXAMINED (652 $ , 806 cf )
Austria, Belgium, Bulgaria, Corsica, Czechoslovakia, Denmark, East Germany, Finland, France, Great
Britain, Greece, Italy, Netherlands, Norway, Poland, Rumania, Sakhalin, Spain, Sweden, Switzerland,
Turkey, U.S.S.R., West Germany, Yugoslavia (AC, BC, BMNH, BRI, CM, EIHU, FSA, HC, IEE,
IEUB, IP, IRSNB, ITZ, IZPAN, JC, JKC, KHC, LELW, MCSN, MHN, MNHN, MNHU, MUM, NMB,
NR, PC, RNH, SC, SMT, UM, USNM, UZI, UZM, VRC, ZIL, ZIM, ZMU, ZC, ZSBS).
falcatorius
T flavomaculatus
Map 5 Distribution of Banchus falcatorius and B. flavomaculatus.
ICHNEUMON-FLY GENUS BANCHUS 23
Banchus ffavomaculatus (Cameron)
(Figs 16, 39, 63, 86, 107)
Cidaphurusflavomaculatus Cameron, 1904: 346. Lectotype $ [not cf as stated by Cameron and by Morley,
1913], INDIA (BMNH), fixed by Morley, 1913: 255.
NOMENCLATURE. The lectotype and one paralectotype female (BMNH) are the only known syntype
specimens. Both bear labels 'Cidaphurus flavomaculatus Cam. Simla' in Cameron's handwriting. The
lectotype has, in addition, Nurse's typewritten data label 'Simla 5.97.'. It is unfortunate that Cameron's
original description applies better to the paralectotype because it is not conspecific with the lectotype and is
B. punkettai.
DESCRIPTION. Female. Width of lower face 0-75-0-80 times vertical length of eye. Malar space 0-50-0-55
times basal width of mandible. Maxillary palp as in Fig. 16. Antenna with apical segments about as broad as
long. Scutellar spine (Fig. 39) varying from very short to about 0-5 as long as scutellum. Mesopleuron and
mesoscutum shining, weakly coriaceous, with moderate punctures, on mesopleuron separated by about or
a little more than their diameter. Hind femur 5-60-5-65 times as long as deep. Fore wing length 8-8-9-2
mm. Tergite 1 of gaster 2-25-2-35 times as long as broad, its dorsal profile as in Fig. 63. Gaster strongly
compressed from posterior of segment 3, reaching to about tips of fore wings (when folded back). Tergite 3
with a crease along its anterior 0-4 separating laterotergite. Tergite 7 and sternite 6 as in Fig. 86.
Colour: black and reddish, with some yellow marks. Face reddish, yellowish laterally and with a black
median stripe. Antenna blackish dorsally, scape and pedicel yellowish ventrally, flagellum brown
ventrally. Maxillary palp reddish yellow, segments 4 distally and 5 wholly dark. Posterior part of
propodeum entirely reddish. Hind coxa black with large dorsal and postero-ventral reddish patches. Hind
femur reddish with a black stripe ventrally. Segments of gaster black anteriorly, very broadly reddish
posteriorly.
Male. Width of lower face 0-80 times vertical length of eye. Malar space 0-60 times basal width of
mandible. Maxillary palp as in Fig. 107; segment 4 considerably widened and flattened; segment 5 about
0-5 as long as 4. Antennal flag setae (similar to Fig. 126, but setae slightly longer) upright, long, flattened, 2
per segment, arising from a polished trough. Hind femur 5-25 times as long as deep. Fore wing length 8-3
mm. Tergite 1 of gaster 2-30 times as long as broad. Tergite 3 with a crease along its anterior 0-5 separating
laterotergite.
Colour: black and yellow, with some reddish, especially on legs. Antenna black dorsally, scape and
pedicel yellow ventrally, underside of flagellum reddish yellow proximally and distally, black medially.
Maxillary palp with segments 1 brownish, 2 reddish yellow, 3 brownish and 4 and 5 blackish. Posterior part
of propodeum entirely yellow. Hind coxa black, with a yellow dorsal spot contiguous interno-laterally with
a postero-ventral spot. Hind femur reddish yellow with a black mark ventrally. Segments of gaster each
black anteriorly, yellow posteriorly, the yellow tending to reddish, especially on sternites.
REMARKS. See 'Remarks' under nox.
BIOLOGY AND HOSTS. Virtually unknown. Adults have been collected in May and October. The specimens
collected at Simla by Gupta and Joseph were from coniferous and deodar forest (Gupta, 1971: (appendix)
2, 3, 54).
DISTRIBUTION (Map 5, p. 22). Along the southern flank of the Himalayas at about 2000 m.
MATERIAL EXAMINED (11 $, 2 cT).
India: 1 $, Himachal Pradesh, Simla, v.1897 (Nurse) (BMNH) (lectotype); 4 $, 1 cf, Himachal
Pradesh, Simla, 2133 m, 1 & lO.x.1962, 3.X.1966 (Gupta, Joseph) (GC); 1 $, Uttar Pradesh, Khurpatal,
ll.x.1978 (Gupta) (GC); 1 $, 1 cT, Uttar Pradesh, Mussoorie, 2100 m, l.x.1962, 21.x. 1972 (Gupta,
Khana) (GC); 3 $ , Uttar Pradesh, Nainital, 1938 m, 2-10 & 12.X.1978 (Gupta) (GC). Nepal: 1 $ , 3-2 km
SE. of Sikha, 2100-2400 m, 23.V.1954 (Quinlan) (BMNH).
Banchus gudrunae sp. n.
(Figs 17, 40, 65, 66, 87, 108)
DESCRIPTION. Female. Width of lower face 0-75-0-85 times vertical length of eye. Malar space 0-45-0-55
times basal width of mandible. Maxillary palp as in Fig. 17. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 40) about 0-7 as long as scutellum. Mesopleuron and mesoscutum coriaceous,
with moderately strong punctures, on mesopleuron separated by about their diameter. Hind femur
6-75-7-20 times as long as deep. Fore wing length 9-7-10-7 mm. Tergite 1 of gaster 2-00-2-50 times as long
24 M. G. FITTON
as broad, its dorsal profile as in Figs 65 (male), 66 (female). Gaster compressed from tergite 4, reaching
about to tips of fore wings (when folded back). Tergite 3 with a crease along its whole length separating
laterotergite. Tergite 7 and sternite 6 as in Fig. 87.
Colour: black and yellow. Face yellow with a black median stripe. Antenna black (sometimes slightly
brownish distally), scape, and sometimes pedicel, with a yellow patch beneath. Maxillary palp reddish
yellow with segment 1 and distal parts of 4 and 5 blackish. Posterior part of propodeum yellow, narrowly
black posteriorly. Hind coxa black with a yellow dorsal patch. Hind femur yellowish with proximal half
largely black. Segments of gaster each black anteriorly, yellow posteriorly, the yellow bands relatively
wide.
Male. Width of lower face 0-80-0-90 times vertical length of eye. Malar space 0-45-0-65 times basal
width of mandible. Maxillary palp as in Fig. 108; segment 4 flattened and widened distally; segment 5 about
0-5 as long as 4. Antennal flag setae (similar to Fig. 127, but with setae slightly shorter and very much
broader) upright, relatively short, flattened and very broad, 2 (occasionally 3) per segment, arising from a
polished trough. Hind femur 7-00-7-45 times as long as deep. Fore wing length 8-8-10-0 mm. Tergite 1 of
gaster 2-10-2-40 times as long as broad. Tergite 3 with a crease along its whole length separating
laterotergite.
Colour: black and yellow. Antenna black dorsally, scape and pedicel yellow ventrally, flagellum
brownish ventrally. Maxillary palp reddish yellow, segment 1 dorsally narrowly black, segment 4 distally
and 5 wholly blackish. Posterior part of propodeum yellow, very narrowly black posteriorly. Hind coxa
black with a large yellow dorsal patch. Hind femur yellow, blackish on its proximal 0-3, especially ventrally
and laterally. Segments of gaster each black anteriorly, very broadly yellow posteriorly.
REMARKS. This species has a number of characters which make relating it to any others in the Palaearctic
difficult. The form of the male palp suggests a relationship withpalpalis and crefeldensis but the flag setae
are of a very much more specialized kind and the female palp is not specialized in the same way.
BIOLOGY AND HOSTS. Unknown. Dates of collection of adults range from the middle of November to the
middle of January.
DISTRIBUTION (Map 6, p. 27). Known only from the island of Cyprus.
MATERIAL EXAMINED
Holotype cf , Cyprus: Polemedia Hills, 14.xii.1948 (Mavromoustakis) (BMNH).
Paratypes 10 $ , 9 cf . Cyprus: 6 $ , 3 cf , Polemedia Hills, 14 & 20.xii.1948 (Mavromoustakis) (BMNH);
3 $, Palodkia, 5 & 12. i. 1949 (Mavromoustakis} (BMNH); 1 $, Limassol, 4.U940 (Mavromoustakis)
(USNM); 4 cf, Zakaki Marshes, 29.xi.1946 (Mavromoustakis) (TC); 1 cf, near Limassol, 18.xi.1946
(Mavromoustakis) (TC); 1 cf , Paphos district, near Panayia, 900 m, 29.xi.1946 (Mavromoustakis) (TC).
Banchus hastator (Fahricius )
(Figs 18, 41, 64, 88, 109, 124)
Ichneumon hastator Fabricius, 1793: 167. Holotype $, EUROPE (UZM) [examined].
Ichneumon pungitor Thunberg, 1822: 265; 1824: 320. [Replacement name for Ichneumon hastator
Fabricius.]
Ichneumon reticulatorThunberg, 1822: 265; 1824: 321. Type(s) cf , SWEDEN (lost). Syn. n.
Banchus femoralis Thomson, 1897: 2411. Lectotype 9, SWEDEN (UZI), designated by Townes, Momoi &
Townes, 1965: 237 [examined].
Banchus kolosovi Meyer, 1925: 10. Syntypes3 <J>, U.S.S.R. (destroyed). Syn. n.
NOMENCLATURE. This species until recently has been referred to by the junior synonym femoralis. It is dealt
with in a relatively large number of non-taxonomic papers so the name change is particularly unfortunate.
A further source of confusion is the incorrect use of the name hastator by Townes & Townes (1978) (see
nomenclatural notes under palpalis).
Ichneumon pungitor is a junior objective synonym of hastator. It was proposed apparently because of the
secondary homonymy in Ichneumon, in Thunberg's work, with Foenus hastator Fabricius, 1804 [which,
incidentally, was miss-spelled as vastator by Thunberg (1822: 262; 1824: 315)]. The status of pungitor as a
replacement name has not been recognized previously and it had been placed incorrectly in the synonymy
of B. volutatorius (Linnaeus) (e.g. Aubert, 1978: 158).
Ichneumon reticulator has not been identified since its description. Roman (1912: 277) reported that
there were no specimens in Thunberg's collection and suggested that the species was a Banchus. The
description fits some males of hastator, with which it is here synonymised.
After its description, except for an entry in the Zoological Record, B. kolosovi was not referred to (even
ICHNEUMON-FLY GENUS BANCHUS 25
by Meyer) until Aubert (1978: 166) noted its existence. Aubert miss-spelled the name as kozlovi. The
description fits particularly dark specimens ofhastator, with which it is here synonymised. Notwithstanding
Meyer's comments on kolosovi andfemoralis, it seems that he realised his mistake because in his treatment
of the genus (1934: 229) he mentions, under femoralis, material from Sverdlovsk (the type-locality of
kolosovi) collected by Kolosov. This is presumably the type-material of kolosovi. No material collected by
Kolosov is noted under other Banchus species by Meyer.
DESCRIPTION. Female. Width of lower face 0-85-0-90 times vertical length of eye. Malar space 0-50-0-65
times basal width of mandible. Maxillary palp as in Fig. 18. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 41) about 0-2 as long as scutellum. Mesopleuron and mesoscutum coriaceous,
usually fairly weakly so, with moderate punctures, on mesopleuron separated by about their diameter.
Hind femur 4-50-5-15 times as long as deep. Fore wing length 8-8-10-9 mm. Tergite 1 of gaster 1-50-1-70
times as long as broad, its dorsal profile as in Fig. 64. Gaster weakly compressed, reaching about to tips of
fore wings (when folded back). Tergite 3 with a crease along its anterior 0-3 separating laterotergite.
Tergite 7 and sternite 6 as in Fig. 88.
Colour: mainly black, with yellow and reddish, particularly on legs. Face black with yellow orbital
marks, which are usually small but are sometimes large and with a pair of very small reddish marks in centre
of face. Antenna with scape and pedicel black, scape, and usually pedicel, yellow ventrally. Flagellum
brown dorsally, reddish ventrally, darker at extreme base and apex. Maxillary palp reddish with segment 1
black and segments 2 proximally, 4 distally and 5 wholly darker. Posterior part of propodeum entirely
black. Hind coxa black. Hind femur reddish or yellowish proximally and distally, medially varying from
reddish dorsally and blackish ventrally to all blackish. Segments of gaster black, sometimes almost entirely
except for brownish posterior margin of sternites 1 and 6 and tergite 7, but usually with posterior margins of
further segments narrowly yellowish and with the yellow grading into a wider brownish band.
Male. Width of lower face 0-90-1-00 times vertical length of eye. Malar space 0-50-0-65 times basal
width of mandible. Maxillary palp as in Fig. 109; segment 4 only slightly widened and flattened; segment 5
almost as long as 4. Antennal flag setae (Fig. 124) relatively poorly differentiated, small, not flattened, 2
per segment, not arising from a trough. Hind femur 4-60-5-50 times as long as deep. Fore wing length
7-2-10-7 mm. Tergite 1 of gaster 1-45-1-80 times as long as broad. Tergite 3 with a crease along its anterior
0-3 separating laterotergite.
Colour: mainly black, with some reddish and yellow, particularly on legs. Antenna black dorsally, scape
and pedicel yellow ventrally, flagellum reddish ventrally. Maxillary palp with segment 1 blackish or
blackish and yellow, 2 and 3 reddish, 4 and 5 blackish or brownish. Posterior part of propodeum entirely
black. Hind coxa black (in a single specimen with a small yellowish dorsal spot and the postero- ventral
margin reddish). Hind femur reddish or yellowish proximally and distally, medially varying from reddish
dorsally and blackish ventrally to all blackish. Gaster sometimes almost entirely black except for brownish
posterior margins of some segments, but usually with posterior margins of segments narrowly yellowish
with adjacent brownish bands.
REMARKS. This is an isolated species, with few obvious specializations. The primitive condition of the
antennal flag setae in the male is particularly notable.
BIOLOGY AND HOSTS. The species is an important parasite of Panolis flammea (Denis & Schiffermiiller)
(Noctuidae), which is sometimes a serious pest in pine forests. B. hastator is univoltine and adults occur in
May, June and early July. Reared specimens have emergence dates in March and April also. Some details
of the biology and ecology of the species have been investigated as a result of the association with an
economically important host. Most of the work has been connected with its effect as a natural control agent
(Friederichs, 1936; Habermehl, 1922, 1924; Pfeffer, 1933; Scheidter, 1934; Schwerdtfeger, 1952; Smits van
Burgst, 1927) but has included studies of embryonic and larval development and morphology (Beirne,
1941; Biedowski & Krainska, 1926; Shevyrev, 1913), courtship, host identification and host-stage
preferences, egg-laying and the reproductive system (van Veen, 1982).
There are many specimens reared from P. flammea in collections. Apart from that species, the following
species are recorded as hosts on specimen labels: Blepharita adusta (Esper) (Noctuidae) (1 $, 2 cf,
Germany (Smits van Burgst coll.) LELW); Deilephila porcellus (Linnaeus) (Sphingidae) (1 cf , no locality
(Adkin) BMNH); Hyloicus pinastri (Linnaeus) (Sphingidae) (1 $, Poland (Mazur) IZPAN) and ?Achlya
flavicornis (Linnaeus) (Thyatiridae) (1 $, Great Britain (Lyle) BMNH). While one might cast doubt on
these records it is difficult to investigate them critically. Kolubajiv (1934: 114, 116) records Lymantria
dispar (Linnaeus) (Lymantriidae) as a host.
DISTRIBUTION (Map 6, p. 27). Widely distributed in northern Europe as far east as Leningrad, perhaps
extending further east to the Urals (if my synonymy of kolosovi is correct (see 'Nomenclature' above)). It
also occurs in the European Alps and I have seen single individuals collected in Corsica and Yugoslavia.
26 M. G. FITTON
MATERIAL EXAMINED (113 $, 151 cf)
Belgium, Corsica, Denmark, East Germany, Finland, France, Great Britain, Ireland, Netherlands,
Poland, Sweden, Switzerland, U.S.S.R., West Germany, Yugoslavia (AC, BC, BMNH, BRI, HC, IP,
IRSNB, ITZ, IZPAN, LELW, KC, MNHN, MUM, NMB, NR, RNH, RSM, SC, SMT, UM, USNM,
UZI, UZM, VRC, ZC, ZIL, ZMU, ZSBS).
Banchus insulanus Roman
(Figs 19, 42, 67, 89)
Banchus insulanus Roman, 1937: 18. Holotype $, MADEIRA (NR) [examined].
DESCRIPTION. Female. Width of lower face 0-85 times vertical length of eye. Malar space 0-55 times basal
width of mandible. Maxillary palp as in Fig. 19. Antenna with apical segments longer than broad. Scutellar
spine (Fig. 42) indistinct. Mesopleuron and mesoscutum coriaceous, with moderate punctures, on
mesopleuron separated by more than their diameter. Hind femur 6-55 times as long as deep. Fore wing
length 9-3-9-5 mm. Tergite 1 of gaster 1-65 times as long as broad, its dorsal profile as in Fig. 67. Gaster
only weakly compressed, not reaching to tips of fore wings (when folded back). Tergite 3 with a crease
along its entire length separating laterotergite. Tergite 7 and sternite 6 as in Fig. 89.
Colour: mainly black, with yellow and reddish marks. Face black with irregular, yellow orbital marks.
Antenna black. Maxillary palp blackish. Posterior part of propodeum black. Hind coxa black with a small
yellow dorsal patch. Hind femur yellowish, blackish ventrally and posteriorly on its proximal 0-5. Segments
of gaster blackish, in one specimen with only posterior 0-5 of tergite 3 and posterior 0-3 of tergite 7 reddish;
in the other specimen with posterior 0-7 of tergite 2, almost all of tergite 3, posterior 0-5 of tergite 4, and
tergites 5 and 6 laterally, reddish.
Male. Unknown (but see 'Remarks').
REMARKS. The specimen described as the male by Hellen (1949: 13) proved, on examination, to be a
female. The material collected in 1959 and supposedly including both sexes (Hellen, 1961: 37) cannot be
found in Hellen's collection (A. Albrecht, pers. comm.).
The relationships of the species are not at all obvious from the characters of the female.
BIOLOGY AND HOSTS. Unknown. Adults have been collected in May and July-August.
DISTRIBUTION (Map 6, p. 27). Known only from the island of Madeira.
MATERIAL EXAMINED (2 9)
Madeira: 1 $, 25 Fontes, Rabacal, 17.vrM.viii.1935 (Lunblad) (NR) (holotype); 1 $, Ribeiro Frio,
3.v.l938(Fre>0(ZMU).
Banchus japonicus (Ashmead)
(Figs 20, 43, 68, 90, 110)
Nawoia japonica Ashmead, 1906: 185. LECTOTYPE 9, JAPAN (USNM), here designated [examined].
NOMENCLATURE. Ashmead described this species from two specimens, also stating 'Type. - Cat. No. 7259,
U.S.N.M.', thus giving the impression that one was the holotype. However, both specimens are labelled
Type No. 7259 U.S.N.M.' and must be considered as syntypes. Other labels include 'Nawaia japonica
Ashm' on one specimen and 'Nawaia japonica Ash. Paratype' on the other, the latter on a similar label but
in a different hand and obviously more modern than the former. The specimen labelled as paratype also
bears a label '47' and is in much better condition than the other. I have labelled and hereby designate it as
lectotype.
DESCRIPTION. Female. Width of lower face 0-80-0-90 times vertical length of eye. Malar space 0-35-0-45
times basal width of mandible. Maxillary palp as in Fig. 20. Antenna with apical segments about as long as
broad. Scutellar spine (Fig. 43) about 0-3 as long as scutellum. Mesopleuron and mesoscutum shining and
only weakly coriaceous, with moderate punctures, on mesopleuron separated by about their diameter.
Hind femur 5-55-6-50 times as long as deep. Fore wing length 9-2-12-1 mm. Tergite 1 of gaster 1-45-1-85
times as long as broad, its dorsal profile as in Fig. 68. Gaster only weakly compressed, reaching to tips of
fore wings (when folded back). Tergite 3 with a crease along its anterior 0-5 separating laterotergite.
Tergite 7 and sternite 6 as in Fig. 90.
Colour: black and yellow. Face yellow with a broad median black stripe and sometimes a black mark
below each antennal socket. Antenna black, with a small yellow mark on scape. Maxillary palp blackish,
with a longitudinal yellow stripe ventrally on segments 1 and 2. Posterior part of propodeum yellow,
ICHNEUMON-FLY GENUS BANCHUS
27
Map 6
^ gudruni
hastator
V insulanus
Map 6 Distribution of Banchus gudrunae, B. hastator and B. insulanus.
entirely or with a black area postero-medially. Hind coxa black with a yellow dorsal patch. Hind femur
blackish, yellow proximally and distally. Segments of gaster black with posterior margins broadly yellow
and with a pair of yellow spots on tergite 2 and usually also on 1 and 3.
Male. Width of lower face 0-90 times vertical length of eye. Malar space 0-35-0-45 times basal width of
mandible. Maxillary palp as in Fig. 110; segment 4 flattened and considerably widened; segment 5 about
0-5 as long as 4. Antennal flag setae (similar in appearance to Fig. 126, but with setae longer) upright, long,
flattened, 2 per segment, arising from trough which is not polished. Hind femur 6-25-6-75 times as long as
deep. Fore wing length 9-3-10-3 mm. Tergite 1 of gaster 1-50-1-65 times as long as broad. Tergite 3 with a
crease along almost its entire length separating laterotergite.
Colour: black and yellow. Antenna black with scape, pedicel and proximal end of flagellar segment 1
yellow ventrally. Maxillary palp with segment 1 yellow, 2 and 3 brownish yellow to blackish, 4 and 5 black.
Posterior part of propodeum yellow, entirely or with small black marks posteriorly. Hind coxa black and
yellow. Hind femur yellow with a black mark ventrally and interno-laterally, usually extending also to
dorsal surface. Segments of gaster black with posterior margins broadly yellow and with a pair of yellow
spots on tergite 2 and usually also on 1 and 3.
REMARKS. Although japonicus has been confused withpoppiti the two species are easily separated.
BIOLOGY AND HOSTS. Unknown. Adults have been collected in May and June.
DISTRIBUTION (Map 2, p. 16). Japan. Recorded also from Korea (Kim, 1955: 493), from which country the
only species of Banchus examined is B. palpalis.
MATERIAL EXAMINED (6 $ , 6 c?)
Japan: 1 $, Gifu, Fujishiro (Nawa) (USNM) (lectotype); 1 $, Gifu, Gifu-yama (Nawa) (USNM)
(paralectotype); 1 $, Hokkaido (Uchidd) (TC); 1 $, 3 cf, Hokkaido, Mt Soranuma, 26.vi.1965 and
28 M. G. FITTON
15.vi.1968 (Kusigemati) (KC); 1 cf, Hokkaido, Toya-ko, 14.vi.1967 (Miyazaki) (KC); 1 $, Honshu,
Tokyo, Komaba, 9.V.1916 (Hirayama) (EIHU); 1 cf , Honshu, Hyogo Pref., Ryuzoji, Sasayama, 5. v. 1965
(Nakanishi) (NC); 1 $, Honshu, Tottori Pref., Mt Daisen, 12.vi.1963 (Nakanishi) (NC); 1 cf , Honshu,
Yatsugatake, 2.vi.l967 (Kocha) (KC).
Banchus mauricettae sp. n.
(Figs 21, 44, 69, 91, 111)
DESCRIPTION. Female. Width of lower face 0-90-0-95 times vertical length of eye. Malar space 0-55 times
basal width of mandible. Maxillary palp as in Fig. 21. Antenna with apical segments longer than broad.
Scutellar spine (Fig. 44) about as long as scutellum. Mesopleuron and mesoscutum coriaceous, with
moderate to strong punctures, on mesopleuron separated by a little more than their diameter. Hind femur
5-20-5-30 times as long as deep. Fore wing length 10-5-10-6 mm. Tergite 1 of gaster 1-90-2-35 times as long
as broad, its dorsal profile as in Fig. 69 (male). Gaster moderately compressed, not quite reaching tips of
fore wings (when folded back). Tergite 3 with a crease along its anterior 0-6 separating laterotergite.
Tergite 7 and sternite 6 as in Fig. 91.
Colour: yellowish red with black and yellow marks. Face yellowish red with orbits yellow and sometimes
with a blackish median stripe. Antenna entirely reddish yellow except for extreme apex of flagellum and
scape and pedicel dorsally, which are blackish. Maxillary palp reddish yellow with segments 4 distally and 5
wholly dark. Posterior part of propodeum reddish, entirely or with black marks posteriorly. Hind coxa
varying from reddish with black patches to almost entirely black. Hind femur entirely reddish. Segments of
gaster entirely reddish or with some black anteriorly.
Male. Width of lower face 0-85-0-90 times vertical length of eye. Malar space 0-60-0-65 times basal
width of mandible. Maxillary palp as in Fig. Ill; segment 4 flattened and widened; segment 5 about 0-8 as
long as 4. Antennal flag setae (similar in appearance to Fig. 129, but with setae only two-thirds as long)
upright, fairly short and broad, 3 (occasionally 4) per segment, arising from a polished trough. Hind femur
5 95-6 20 times as long as deep . Fore wing length 11-2-12-3 mm . Tergite 1 of gaster 2 20-2 35 times as long
as broad. Tergite 3 with a crease along its anterior 0-6 separating laterotergite.
Colour: yellow, reddish and black. Antenna blackish dorsally, scape and pedicel yellow ventrally,
flagellum reddish yellow ventrally. Maxillary palp reddish yellow with segments 4 distally and 5 wholly
blackish. Posterior part of propodeum entirely yellow. Hind coxa largely reddish yellow, blackish or
brownish antero-ventrally and dorso-posteriorly. Hind femur reddish, sometimes with a blackish mark
ventrally. Segments of gaster each black, reddish and yellow (from anterior to posterior), the area of black
reduced or absent on posterior segments.
REMARKS. This is a distinctive species; its relationships are difficult to discern.
BIOLOGY AND HOSTS. Unknown. Dates of collection are in June and July.
DISTRIBUTION (Map 2, p. 16). The Szechuen province of China. Altitudes given on data labels range from
300-4500 m.
MATERIAL EXAMINED
Holotype cf , China: Szechuen, Yao-Gi, 1200-2400 m, 3.vii.l929 (Graham) (USNM).
Paratypes 4 $, 10 cf. China: 1 9, 2 cf, Szechuen, Yao-Gi, 1200-2400 m, 3 & 16.vii.1929 (Graham)
(USNM); 2 cf, Szechuen, near Mupin, 600-2400 m, 28.vi & 22.vii.1929 (Graham) (USNM); 1 cf,
Szechuen, Ningyuenfu, 1800-3240 m, 24-26. vii. 1928 (Graham) (USNM); 1 cf , Szechuen, Mt Omei, 840
m, vi.1937 (Graham) (USNM); 1 cf , Szechuen, Suifu, 300-450 m, l-21.vi.1928 (Graham) (USNM); 2 cf ,
Szechuen, 14-5 km SW. of Tatsienlu, 2550-3900 m, 25-27.vi.1923 (Graham) (USNM); 1 cf , Szechuen, U
Long Kong, near Tatsienlu, 3000-4500 m, 25-30.vi.1923 (Graham) (USNM); 2 9, Szechuen, Yachow,
v-vi.1928 (Graham) (USNM); 1 $, Szechuen, Yachow to Mupin, 600-1500 m, 23-27.vi.1929 (Graham)
(USNM).
Banchus moppitisp. n.
(Figs 22, 45, 70, 92, 113, 127)
DESCRIPTION. Female. Width of lower face 1-10-1-20 times vertical length of eye. Malar space 0-85-1-00
times basal width of mandible. Maxillary palp as in Fig. 22. Antenna with apical segments about as broad as
long. Scutellar spine (Fig. 45) about 0-6 as long as scutellum, usually downcurved at the tip. Mesopleuron
and mesoscutum coriaceous, usually strongly so, with strong punctures, on mesopleuron separated by
much less than their diameter. Hind femur 4-75-5-20 times as long as deep. Fore wing length 8-2-10-8 mm.
Tergite 1 of gaster 1-60-1-80 times as long as broad, its dorsal profile as in Fig. 70. Gaster compressed from
ICHNEUMON-FLY GENUS BANCHUS 29
segment 4, reaching just beyond tips of fore wings (when folded back). Tergite 3 with a crease along its
anterior 0-3 separating laterotergite. Tergite 7 and sternite 6 as in Fig. 92.
Colour: black and yellow. Face black with yellow orbital stripes, usually broad but varying considerably
in width. Antenna black, usually with a small yellow patch on underside of scape. Maxillary palp blackish.
Posterior part of propodeum black with a broad yellow stripe along posterior transverse carina. Hind coxa
black, usually with a small yellow dorsal patch. Hind femur black, yellow proximally and distally and
usually dorsally. Segments of gaster each black anteriorly, broadly yellow posteriorly.
Male. Width of lower face 1-10-1 -20 times vertical length of eye. Malar space 0-75-0-95 times basal
width of mandible. Maxillary palp as in Fig. 113; segment 4 flattened and widened; segment 5 not quite as
long as 4. Antennal flag setae (Fig. 127) upright, relatively short, flattened and widened, 2 per segment,
arising from a polished trough. Hind femur 4-80-5-50 times as long as deep. Fore wing length 8-5-10-9 mm.
Tergite 1 of gaster 1-45-1-60 times as long as broad. Tergite 3 with a crease along its anterior 0-3 separating
laterotergite.
Colour: black and yellow. Antenna black with scape and pedicel yellow ventrally. Maxillary palp
blackish. Posterior part of propodeum yellow anteriorly, black posteriorly. Hind coxa black, usually with a
yellow dorsal spot and sometimes also a postero-lateral one. Hind femur yellow with an extensive black
mark, extending ventrally and laterally. Segments of gaster each black anteriorly, yellow posteriorly.
REMARKS. This species seems closely related to zonatus. Its generally much smaller size and relatively
shorter appendages (and gaster in females) readily differentiate it.
BIOLOGY AND HOSTS. Unknown. The specimens were collected early in the year (in February, March and
April) with the exception of a female taken in October.
DISTRIBUTION (Map 7, p. 29). The few known specimens come from widely scattered localities in western
Europe.
Map?
% moppiti
Q moppiti. located only to country
Map 7 Distribution of Banchus moppiti.
30 M. G. FITTON
MATERIAL EXAMINED
Holotype $, Spain: Madrid, Cercedilla, 22.X.1978 (Noyes) (BMNH).
Paratypes 6 $, 4 cf . France: 1 cf , Vincennes, 20.ii.1885 (de Gaulle coll.) (MNHN); 1 cf , Boulogne,
22.iii (Giraud coll.) (MNHN). Great Britain: 1 $ (Desvignes coll.) (BMNH). Spain: 4 $, 2 cf , Teruel,
15.iii.82 (Hiendlmayr coll.) [note - only 1 $ of these 6 specimens has the locality and date but all are
identically mounted and set] (ZSBS). Switzerland: 1 $ , 7.iv.l861 (Sichel coll.) (MNHN).
Banchus nox Morley
(Figs 23, 46, 71, 93, 112)
Banchus nox Morley, 1913: 255. Holotype cf , INDIA (BMNH) [examined].
NOMENCLATURE. The species misidentified as nox by Chandra & Gupta (1977: 182) is B. punkettai.
DESCRIPTION, female. Width of lower face 0-80 times vertical length of eye. Malar space 0-55 times basal
width of mandible. Maxillary palp as in Fig. 23. Antenna with apical segments longer than broad. Scutellar
spine (Fig. 46) about 0-7 as long as scutellum. Mesopleuron and mesoscutum coriaceous, with moderate
punctures, on mesopleuron separated by a little more than their diameter. Hind femur 6-40 times as long as
deep. Fore wing length 11-2 mm. Tergite 1 of gaster 2-70 times as long as broad, its dorsal profile as in Fig.
71 (male). Gaster strongly compressed from posterior of segment 3, not reaching to tips of fore wings
(when folded back). Tergite 3 with a crease along its anterior 0-4 separating laterotergite. Tergite 7 and
sternite 6 as in Fig. 93.
Colour: black and reddish, with some yellow marks. Face black with very broad yellow orbital marks
which also extend beneath antennal sockets. Antenna dark brown, blackish dorsally, scape and pedicel
yellowish ventralty. Maxillary palp reddish yellow, with segment 1 dorsally and segment 4 distally blackish.
Posterior part of propodeum blackish with an irregular reddish yellow band along posterior transverse
carina. Hind coxa black with a large reddish dorsal patch. Hind femur reddish with a black stripe ventrally.
Segments of gaster reddish, tergites 1, 2 and 3 black anteriorly.
Male. Width of lower face 0-85 times vertical length of eye. Malar space 0-50 times basal width of
mandible. Maxillary palp as in Fig. 112; segment 4 flattened and considerably widened; segment 5 about
0-6 as long as 4. Antennal flag setae (similar to Fig. 126, but setae slightly longer) upright, long, flattened, 2
per segment, arising from a polished trough. Hind femur 6-05 times as long as deep. Fore wing length 9-7
mm. Tergite 1 of gaster 2-15 times as long as broad. Tergite 3 with a crease along its anterior 0-5 separating
laterotergite.
Colour: black and yellow, with some reddish, especially on legs. Antenna black dorsally, scape and
pedicel yellow ventrally, underside of flagellum with segment 1 and distal part reddish yellow, remainder
black. Maxillary palp with segment 1 yellowish, 2 and 3 wholly and 4 proximally reddish yellow, 4 distally
and 5 wholly brownish. Posterior part of propodeum yellow, narrowly black postero-medially. Hind coxa
black, with a yellow dorsal spot contiguous interno-laterally with a postero- ventral spot. Hind femur
reddish with a black mark ventrally. Segments of gaster each black anteriorly, yellow posteriorly, with
junction of the two areas reddish.
REMARKS. Closely related toflavomaculatus. The aggregate differences between these taxa warrant their
separation, but this will need to be reassessed when more material becomes available for study.
BIOLOGY AND HOSTS. The only two known specimens were collected in October and November, the male on
flowers of Spiraea (Morley, 1913: 255).
DISTRIBUTION (Map 2, p. 16). Assam, between 1800 and 2400 m.
MATERIAL EXAMINED (1 $, 1 cf)
India: 1 cf, Assam, Shillong, Khasi Hills, 1800 m, x.1903 (Turner) (BMNH) (holotype); 1 $, Assam,
Mishmi Hills, Delai Valley, Cha Che, 2200-2400 m, 21.xi.1936 (Steele) (BMNH).
Banchus palpalis Ruthe
(Figs 1,24, 47, 72, 94, 114)
[Banchus monileatus Gravenhorst, 1829: 393; in part. Misidentification.]
Banchus palpalis Ruthe, 1859: 377. Syntypes 2 cf , ICELAND (lost).
Banchus spinosus Cresson, 1865: 274. Holotype 9 [not cf as stated by Cresson], U.S.A. (ANS) [not
examined]. Syn. n.
Banchus formidabilis Provancher, 1874: 61. Holotype $, CANADA (UL) [not examined]. Syn. n.
ICHNEUMON-FLY GENUS BANCHUS 31
Banchus (Corynephanus) groenlandicus Aurivillius, 1890: 30. Lectotype cf , GREENLAND (NR), designated
by Townes, 1961: 104 [examined]. Syn. n.
Cidaphurus alticola Ashmead, 1901 : 148. Holotype $ [not cf as stated by Ashmead], U.S. A. (USNM) [not
examined]. Syn. n.
NOMENCLATURE. For almost 150 years the name monileatus was applied consistently to this species.
However, two species were mixed in Gravenhorst's original material and a series of unfortunate events
culminated in Townes & Townes (1978: 532) designating as lectotype of monileatus a female which was
volutatorius , rather than the species as it had been identified previously. In addition, as a result of
misinformation about my examination of the holotype of hastator, Townes & Townes incorrectly
synonymized monileatus with that species. Female volutatorius and palpalis superficially are strikingly
similar, so much so that Aubert (1978: 156), without realising that he was examining volutatorius,
commented that Townes' lectotype differed from volutatorius only in certain details! When he originally
examined Gravenhorst's syntypes (in 1964) Townes probably checked the identity of the species and then
made the mistake of selecting the specimen in best condition for designation as lectotype without
re-checking it. Under the International Code the lectotype designation has priority over the previous
restriction of the taxonomic species.
Thus monileatus becomes a junior synonym of volutatorius while the species previously known as
monileatus must take the next available name -palpalis. The name hastator correctly applies to the species
previously known by the junior synonym femoralis.
The identity of palpalis was established by Roman (1928: 24; 1930: 285) and Fitton (1978a: 76).
Banchus spinosus, B. formidabilis and Cidaphurus alticola are included in the synonymy of palpalis on
the basis of their treatment in Townes & Townes (1978).
DESCRIPTION. Female. Width of lower face 0-80-0-90 times vertical length of eye. Malar space 0-65-0-75
times basal width of mandible. Maxillary palp as in Fig. 24. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 47) long, more than 0-8 as long as scutellum (in a very few specimens only 0-3 as
long as scutellum). Mesopleuron and mesoscutum coriaceous, with fine to moderate punctures, on
mesopleuron separated by a little more than their diameter. Hind femur 5-00-5-95 times as long as deep.
Fore wing length 9-8-11-5 mm. Tergite 1 of gaster 1-65-2-15 times as long as broad, its dorsal profile as in
Fig. 72 (male). Gaster compressed from posterior of segment 3, not quite reaching to tips of fore wings
(when folded back). Tergite 3 with a crease along its anterior 0-4 separating laterotergite. Tergite 7 and
sternite 6 as in Fig. 94.
Colour: almost entirely black except for mainly reddish legs and a few small yellowish marks. Face black
with yellow orbital marks, usually very small and sometimes absent. Antenna black, entirely or with
yellowish marks ventrally on scape and pedicel. Maxillary palp yellowish or brownish with segments 1 and 5
wholly and 3 and 4 distally blackish. Posterior part of propodeum entirely black. Hind coxa black. Hind
femur reddish yellow, usually entirely, rarely darkened ventrally. Segments of gaster black, sternites and
posterior edges of tergites 6 and 7 sometimes brownish.
Male (Fig. 1). Width of lower face 0-90-0-95 times vertical length of eye. Malar space 0-60-0-70 times
basal width of mandible. Maxillary palp as in Fig. 114; segment 4 narrow and cylindrical, flattened and
widened only at its extreme distal apex; segment 5 less than 0-2 as long as 4. Antennal flag setae (similar to
Fig. 125, but setae slightly longer and showing signs of flattening) at about 50 degrees, very weakly
flattened, 3 per segment, not arising from a trough. Hind femur 5-35-6- 10 times as long as deep. Fore wing
length 9-8-11-4 mm. Tergite 1 of gaster 1-90-2-25 times as long as broad. Tergite 3 with a crease along its
anterior 0-5 separating laterotergite.
Colour: black, with some yellow and with legs mainly reddish yellow. Antenna black, scape and pedicel
yellow ventrally, proximal and distal parts of flagellum often brownish yellow ventrally. Maxillary palp
with segments 1 and 2 wholly and 3 and 4 proximally reddish yellow, 3 and 4 distally and 5 wholly blackish.
Posterior part of propodeum usually entirely black, sometimes with a transverse yellow mark (or marks)
immediately behind posterior transverse carina. Hind coxa sometimes entirely black, usually with an
externo-lateral reddish yellow mark and sometimes also dorsal and/or ventral yellowish spots. Hind femur
entirely reddish yellow. Segments of gaster black, sometimes brownish or yellowish on their posterior
margins, very rarely with conspicuous yellow bands on tergites 1, 2 and 3.
REMARKS. Easily identified from the structure of the maxillary palps. Nonetheless, it is confused with
female volutatorius by incompetent or overconfident identifiers. Its presumed sister-species is crefeldensis.
The males with well-developed yellow marking and conspicuous yellow bands on the gaster come
mainly, but not exclusively, from the southern parts of the range. In North America, where the
geographical variation in colour may be more consistent and also applies to females, this form has been
recognized as a separate 'subspecies' (Townes & Townes, 1978: 533).
32
M. G. FITTON
BIOLOGY AND HOSTS. Adults normally occur from mid-June to mid-August, with one record from
Switzerland as early as 22 May and one from Yugoslavia as late as September. In Europe the usual host
seems to be Blepharita adusta (Esper) (Noctuidae), from which I have seen 39 reared specimens. Of these,
34, possibly 35, come from a single mass rearing (with the only data: Germany (Smits van Burgs t coll.)
(LELW)). In Great Britain Blepharita adusta is more common in the north (Bretherton, Goater &
Lorimer, 1983) and the records ofpalpalis show a similar pattern.
Hosts recorded in the literature are: Panolis flammea (Denis & Schiffermuller) (Noctuidae) (Smits van
Burgst, 1927: 239), Lacanobia oleracea (Linnaeus) (Noctuidae) (Meyer, 1934: 231) and Deilephila
porcellus (Linnaeus) (Sphingidae) (Leonardi, 1928: 87). The record from L. oleracea, at least, can
probably be dismissed as having resulted from the misidentification of female volutatorius.
DISTRIBUTION (Map 8, p. 32). The only Holarctic species of Banchus. Very widely distributed; in the more
southern parts of its range it occurs at higher altitudes (for example, altitudes of 1800 to 2100 m are
recorded for the specimens from Austria, Yugoslavia, Bulgaria, Greece and Turkey).
Map 8
tt.
palpal is
Map 8 Distribution of Banchus palpalis (the North American records are taken from Townes & Townes ,
1978).
ICHNEUMON-FLY GENUS BANCHUS 33
MATERIAL EXAMINED (103 $ , 88 cf )
Belgium, Bulgaria, Canada, Denmark, East Germany, Finland, France, Great Britain, Greece, Iceland,
Ireland, Italy, Japan, Korea, Netherlands, Poland, Sweden, Switzerland, Turkey, West Germany, U.S.A.,
U.S.S.R., Yugoslavia (BC, BMNH, BRI, EIHU, FSA, HC, IEUB, IRSNB, IZPAN, JC, JKC, KC
LELW, MHN, MNHN, MNHU, MUM, NC, NR, RNH, USNM, UZM, VRC, ZC, ZI, ZMU, ZSBS).
Banchus pictus Fab rid us
(Figs 25, 49, 50, 73, 95, 115)
Ichneumon cultratus Gmelin, 1790: 2708. Type(s) [?sex], EUROPE (destroyed). [Junior (by first reviser
choice of Gravenhorst, 1829: 382, 1006) primary homonym of Ichneumon cultratus Gmelin, 1790: 2699.]
Ichneumon mutillatus Christ, 1791: 358. Type(s) cf , no type-locality (lost). [Junior primary homonym of
Ichneumon mutillatus Gmelin, 1790: 2716.]
Banchus pictus Fabricius, 1798: 234. Lectotype $, GERMANY (UZM), designated by Townes, Momoi &
Townes, 1965: 238 [examined].
Banchus zagoriensis Hensch, 1928: 100. ?Syntypes 2 <J>, 1 cf , YUGOSLAVIA (ZSBS) [examined].
Banchus bipunctatus Hensch, 1928: 101. Holotype cf , YUGOSLAVIA (ZPZ) [examined].
Banchus russiator Aubert, 1978: 157. [Unavailable name published conditionally (Article 15 of the Code).]
Banchus russiator Aubert, 1981: 18. Holotype $, U.S.S.R. (AC) [examined]. Syn. n.
NOMENCLATURE. The synonymy of/, cultratus with pictus was queried by Aubert (1978: 157) but he gave no
reasons for so doing and I can find none.
The description of Ichneumon mutillatus Christ fits well the male of pictus and the synonymy, proposed
by Gravenhorst (1829: 383) but queried by Aubert (1978: 157), is accepted. The Christ name is unavailable
because of its homonymy with /. mutillatus Gmelin, which is a replacement name for /. mutillarius
Fabricius, 1787: 271 (a junior homonym of/, multillarius Fabricius, 1775: 342).
I did not succeed in obtaining on loan material from the Hensch collection. However, Dr K. Horstmann
was able to visit Zagreb in 1980 and in the course of his work on the species of Ichneumonidae described by
Hensch he made the types of the Banchus species available to me. Unfortunately the specimen tentatively
selected for designation as lectotype of B. zagoriensis had the date of collection 28 August 1928 and could
not therefore have been a syntype as the description was published on 1 September 1928 (Horstmann,
1982a: 82). Three specimens labelled as cotypes (= syntypes) of B. zagoriensis are present in the Bauer
collection (ZSBS). Because they bear no dates of collection and because the cotype labels were almost
certainly added by Bauer and not Hensch (Horstmann, pers. comm.) they are only tentatively regarded as
syntypes. Selection and designation of a lectotype for B. zagoriensis is best deferred until the other material
in the Hensch collection can be examined. The identity of the species is not in doubt.
DESCRIPTION. Female. Width of lower face 0-80-0-95 times vertical length of eye. Malar space 0-40-0-50
times basal width of mandible. Maxillary palp as in Fig. 25. Antenna with apical segments slightly longer
than broad. Scutellar spine (Figs 49, 50) ranging from very small to about 0-6 as long as scutellum.
Mesopleuron and mesoscutum coriaceous, with strong punctures, on mesopleuron separated by less than
their diameter. Hind femur 5-00-5-50 times as long as deep. Fore wing length 8-1-10-6 mm. Tergite 1 of
gaster 1-45-1-65 times as long as broad, its dorsal profile as in Fig. 73. Gaster compressed, reaching just
beyond tips of fore wings (when folded back). Tergite 3 with a crease along its anterior 0-2 to 0-3 separating
later otergite. Tergite 7 and sternite 6 as in Fig. 95.
Colour: black and yellow. Face yellow, with a black median stripe (rarely rather wide). Antenna with
scape and pedicel black dorsally, yellow ventrally. Flagellum orange, sometimes slightly darker dorsally,
almost always with segment 1 proximally blackish and distal segments dark. Maxillary palp reddish yellow,
with segment 1 black and sometimes segments 2 proximally and 4 distally darkened. Posterior part of
propodeum yellow anteriorly, black posteriorly; the yellow area varying considerably in extent, from
covering almost all of the area to being divided by the black area medially, reduced to a pair of spots, or
(very rarely) entirely absent. Hind coxa black, entirely or with a yellow dorsal patch. Hind femur medially
black (sometimes brownish red, especially dorsally), yellow proximally and distally. Segments of gaster
each black anteriorly, broadly yellow posteriorly.
Male. Width of lower face 0-85-1-00 times vertical length of eye. Malar space 0-45-0-55 times basal
width of mandible. Maxillary palp as in Fig. 115; segment 4 moderately widened and flattened; segment 5
about 0-9 as long as 4. Antennal flag setae (similar to Fig. 125, but setae shorter) at about 60 degrees, short,
not flattened, 2 per segment, not arising from a trough. Hind femur 5-05-5-65 times as long as deep. Fore
wing length 8-1-10-3 mm. Tergite 1 of gaster 1-45-1-70 times as long as broad. Tergite 3 with a crease along
its anterior 0-3 separating laterotergite.
34 M. G. FITTON
Colour: black and yellow. Antenna with scape and pedicel black dorsally, yellow ventrally. Flagellum
reddish orange, dorsally darker (blackish at base and apex). Maxillary palp blackish, sometimes entirely,
but usually with segments 2 and 3 wholly and 4 and 5 proximally yellowish brown. Posterior part of
propodeum black, entirely or with a pair of antero-lateral yellow patches (rarely the yellow more
extensive, very rarely posterior part of propodeum entirely yellow). Hind coxa black, often with a dorsal
yellow patch. Hind femur medially black (sometimes brownish red, especially dorsally), yellow proximally
and distally. Segments of gaster each black anteriorly, broadly yellow posteriorly.
REMARKS. Specimens of turcator have been confused with pictus. The two species can be distinguished
using the characters given in the key.
The morphospecies/7/ctaj, as here defined, has variations in time of occurrence of adults, length of the
scutellar spine, and sex ratios which need further investigation and explanation.
There are two distinct periods of occurrence of adults. When this was noted 184 specimens were to hand.
Of these 90 had a date of collection, 48 were captured between 10 April and 3 June and 42 between 1 July
and 2 October (most of the latter batch after 10 August). The length of the scutellar spine varies from very
small to about 0-6 as long as the scutellum. Although there is continuous variation in the length of the spine
most of the specimens with a short spine were collected in the first period and most of those with a longer
spine in the second. The sex ratio of the 48 specimens collected in the first period was 1 18 $ : 1 cf and of the
42 in the second period 4-25 $ : Icf .
Three possible biological explanations of these observations are that (1) pictus is bivoltine, with the
generations varying slightly in morphology and more distinctly in sex ratio; (2) the early and late summer
groups represent separate species; or (3) that one species has a partial second generation and a second
species (possibly parthenogenetic) occurs in late summer.
If two species were involved diligent study should have revealed variation in some morphological
characters, colour, geographical distribution, or host or other biological data correlated with date of
capture or length of scutellar spine. None was found. Assignment of the 94 specimens without a date of
capture to two subjective classes - with short or long scutellar spine - gave groups which varied in sex ratio
in the same way as the early and late summer groups (short spine, 58 specimens, 1-55 $ : 1 cf ; long spine, 36
specimens, 5 $: 1 cf). The only other relevant evidence comes from two reared specimens: two larvae of
the noctuid Agrochola helvola (Linnaeus) collected near Sheffield, Great Britain (Ford) (UM) on 13 June
1969 proved to be parasitized. The parasite larvae emerged in August 1969 and spun cocoons. The resulting
adults (1 $ Id") did not emerge until early May 1970.
None of this evidence gives unambiguous support to any of the three suggested explanations of the
observations. Even the one year life-cycle of the specimens reared in Sheffield could be explained by
postulating univoltinism in the northern part of the range of the species and bivoltinism in the south
(although no evidence of this was found). However, the balance is perhaps in favour of the involvement of
more than one species, if only because there is no undisputed evidence of bivoltinism in any other species of
Banchus. The poor quality and quantity of the available data and the possible complexity of the situation
mean that the problem remains unresolved - a fruitful field for future study.
BIOLOGY AND HOSTS. For information on phenology see 'Remarks' above. I have seen only three reared
specimens of pictus, from Agrochola helvola (Linnaeus) (Noctuidae) (all from Great Britain (Lyle and
Ford) BMNH and UM). There are several hosts noted in the literature, but as black-and-yellow Banchus
are frequently misidentified as pictus they should be treated with caution. The recorded hosts are:
Agrochola circellaris (Hufnagel) (Noctuidae) (Habermehl, 1922: 269), Agrotis segetum (Denis & Schiffer-
muller) (Noctuidae) (Bajari, 1960: 261), Atethmia ambusta (Denis & Schifferrmiller) (Noctuidae)
(Schmiedeknecht, 1910: 1928), Lycophotia porphyrea (Denis & Schiffermiiller) (Noctuidae) (de Gaulle,
1907: 119), Hadena rivularis (Fabricius) (Noctuidae) (Wagner, 1929: 11), Phalera bucephala (Linnaeus)
(Notodontidae) (Leonardi, 1928: 83) and Smerinthus ocellata (Linnaeus) (Sphingidae) (Meyer, 1934: 228).
DISTRIBUTION (Map 9, p. 35). Widely distributed in the western Palaearctic as far north as 56N in the west,
and extending south-east as far as the Pamirs.
MATERIAL EXAMINED (162 $ , 75 cf ).
Austria, Belgium, Czechoslovakia, Denmark, East Germany, France, Great Britain, Greece, Hungary,
Italy, Morocco, Netherlands, Poland, Portugal, Rumania, Sardinia, Spain, Sweden, Switzerland, Turkey,
U.S.S.R., West Germany, Yugoslavia (AC, BMNH, BRI, CM, FSA, HC, IBMPP, IEAU, IEE, IP,
IRSNB, KHC, LELW, MCSN, MHN, MIZS, MLSU, MNHN, MUM, NM, NMB, NR, PC, RNH, RSM,
TM, UM, USNM, UZI, UZM, VRC, ZC, ZIL, ZMU, ZSBS).
ICHNEUMON-FLY GENUS BANCHUS
Map 9 Distribution of Banchus pictus, B. poppiti and B. punkettai.
Banchuspoppitisp. n.
(Figs 26, 48, 74, 96, 116)
DESCRIPTION. Female. Width of lower face 0-90 times vertical length of eye. Malar space 0-55-0-65 times
basal width of mandible. Maxillary palp as in Fig. 26. Antenna with apical segments longer than broad.
Scutellar spine (Fig. 48) about 0-4 as long as scutellum. Mesopleuron and mesoscutum coriaceous, with
moderate punctures, on mesopleuron separated by about their diameter. Hind femur 5-25-5-65 times as
long as deep. Fore wing length 10-3-11-1 mm. Tergite 1 of gaster 1-65 times as long as broad, its dorsal
profile as in Fig. 74 (male). Gaster very weakly compressed, not reaching to tips of fore wings (when folded
back). Tergite 3 with a crease along its anterior 0-4 separating laterotergite. Tergite 7 and sternite 6 as in
Fig. 96.
Colour: black and yellow, the margins of some yellow areas slightly reddish. Face black with yellow
orbital stripes. Antenna with scape and pedicel black dorsally, yellow ventrally. Flagellum orange
ventrally, dark dorsally. Maxillary palp yellowish with segments 1 dorsally, 4 distally, and 5 wholly dark.
Posterior part of propodeum yellow, entirely or with posterior margin black. Hind coxa black with a large
yellow dorsal patch. Hind femur black, narrowly reddish yellow proximally and distally. Segments of
gaster black, narrowly yellow posteriorly.
Male. Width of lower face 0-95 times vertical length of eye. Malar space 0-50-0-65 times basal width of
mandible. Maxillary palp as in Fig. 116; segment 4 flattened and widened; segment 5 about 0-5 as long as 4.
36 M. G. FITTON
Antennal flag setae (similar to Fig. 127, but with setae narrower) upright, flattened, 2 per segment, arising
from a polished trough. Hind femur 5-60-6-10 times as long as deep. Fore wing length 11-1-11-2 mm.
Tergite 1 of gaster 1-65-1 -90 times as long as broad. Tergite 3 with a crease along its anterior 0-4 separating
laterotergite.
Colour: black and yellow, the margins of some yellow areas slightly reddish. Antenna black dorsally,
yellow ventrally. Maxillary palp yellow with segments 4 distally and 5 wholly blackish. Posterior part of
propodeum yellow, entirely or narrowly black postero-medially. Hind coxa black with very large dorsal
and ventral yellow patches. Hind femur yellow, black-marked ventrally and internally, tending to reddish
distally. Segments of gaster each black anteriorly, black posteriorly.
REMARKS. This species has not previously been differentiated fromjaponicus.
BIOLOGY AND HOSTS. Unknown. Adults have been collected in May and June.
DISTRIBUTION (Map 9, p. 35). Japan.
MATERIAL EXAMINED
Holotype cf , Japan: Hakodate, 12.vi.1926 (Malaise) (NR).
Paratypes 2 $, 1 cf. Japan: 1 $, 1 cf, Hakodate, 12.vi.1926 (Malaise) (NR); 1 , Mt Mino, 6.V.1929
(Teranishi) (TC).
Banchus punkettai sp. n.
(Figs 27, 51, 75, 97, 117)
[Banchus nox Morley; Chandra & Gupta, 1977: 182. Misidentification.]
DESCRIPTION. Female. Width of lower face 0-75-0-80 times vertical length of eye. Malar space 0-45-0-60
times basal width of mandible. Maxillary palp as in Fig. 27. Antenna with apical segments just broader than
long. Scutellar spine (Fig. 51) about 0-5 as long as scutellum. Mesopleuron and mesoscutum shining, only
very weakly sculptured, with strong punctures, on mesopleuron separated by less than their diameter.
Hind femur 5-60-6-10 times as long as deep. Fore wing length 8-6-9-7 mm. Tergite 1 of gaster 1-80-2-25
times as long as broad, its dorsal profile as in Fig. 75. Gaster strongly compressed from posterior of segment
3, reaching to tips of fore wings (when folded back). Tergite 3 with a crease along its anterior 0-3 separating
laterotergite. Tergite 7 and sternite 6 as in Fig. 97.
Colour: black and yellow, with some reddish, especially on legs. Face yellow with a black median stripe,
the stripe sometimes rather wide but with reddish yellow patches remaining below antennal sockets.
Antenna brown, often rather darker dorsally and distally, with scape and pedicel blackish dorsally and
yellowish ventrally. Maxillary palp reddish yellow with distal parts of segments 4 and 5 dark. Posterior part
of propodeum yellowish with a black patch medio-posteriorly, the yellow area sometimes reduced to a
band along posterior transverse carina. Hind coxa black with a yellow dorsal patch and a red postero-
ventral patch. Hind femur reddish with a blackish ventral stripe. Segments of gaster each black anteriorly,
yellow posteriorly, with junction between two colours reddish. Tergite 1 also with posterior margin black
and yellow area reduced to two lateral and/or one median spot(s).
Male. Width of lower face 0-90 times vertical length of eye. Malar space 0-65 times basal width of
mandible. Maxillary palp as in Fig. 117; segment 4 flattened and considerably widened; segment 5 about
0-6 as long as 4. Antennal flag setae (similar to Fig. 124) relatively poorly differentiated, at about 50
degrees, very short, not flattened, 2 per segment, without a polished trough. Hind femur 6-05 times as long
as deep. Fore wing length 8-6 mm. Tergite 1 of gaster 1-90 times as long as broad. Tergite 3 with a crease
along its anterior 0-5 separating laterotergite.
Colour: black and yellow, with some reddish, especially on legs. Antenna black dorsally, scape, pedicel
and proximal flagellar segments yellow ventrally, remainder of flagellum brownish yellow ventrally.
Maxillary palp with segment 1 yellow, 2 and 3 reddish yellow, 4 and 5 blackish. Posterior part of
propodeum yellow, black postero-medially. Hind coxa black, with a dorsal yellow spot which is contiguous
interno-laterally with a postero-ventral spot. Hind femur reddish, black ventrally and with a separate
dorsal blackish area. Segments of gaster each black anteriorly, yellow posteriorly, the junction between the
two colours very narrowly reddish.
REMARKS. The colour pattern of this species is very similar to two other species, flavomaculatus and nox,
found in the same area. The males are easily identified using the flag setae but the females are more difficult
to separate.
ICHNEUMON-FLY GENUS BANCHUS 37
BIOLOGY AND HOSTS. Virtually unknown. Adults have been collected in April and May. The specimens
collected by Kamath and Gupta came from mixed vegetation in coniferous forest (Gupta, 1975: (appendix)
20,22).
DISTRIBUTION (Map 9, p. 35). Along the southern flank of the Himalayas and south-east into Burma,
between about 2000 and 3000 m.
MATERIAL EXAMINED
Holotype cf , Nepal: 2756'N, 85WE, 3030 m, 23-29.V.1967 (Can. Nepal Exped.) (BRI).
Paratypes 13 $ . Burma: 5 $ , NE. , Kambaiti, 2000 m, 4, 6 & 7.iv. 1934 (Malaise) (NR); 1 $ , Mt Victoria,
2800 m, 27.iv.1938 (Heinrich) (TC). India: 1 $, Himachal Pradesh, Dalhousie, 2132 m, 29.iv.1971 (Ram)
(GC); 1 $, Himachal Pradesh, Kalatop, 2438 m, S.v.1971 (Kamath) (GC); 1 $, Himachal Pradesh, Simla
(Nurse) (BMNH) (paralectotype of Cidaphurus flavomaculatus Cameron). Nepal: 2 $, 2756'N, 8500'E,
3030 m, 23-29.V.1967 (Can. Nepal Exped.) (BRI); 2 $, 2800'N, 8500'E, 21-23.V.1967 (Can. Nepal
Exped.)(BRI).
Banchus sanjozanus Uchida
(Figs 52, 76, 118)
Banchus volutatorius var. sanjozanus Uchida, 1929: 184. Holotype cf , JAPAN (EIHU) [examined].
DESCRIPTION. Female. See 'Remarks'.
Male. Width of lower face 0-86 times vertical length of eye. Malar space 0-50-0-60 times basal width of
mandible. Maxillary palp as in Fig. 118; segment 4 flattened and considerably widened; segment 5 about
0-6 as long as 4. Antennal flag setae (similar to Fig. 128) upright, long, flattened, 2 per segment, arising
from a polished trough. Scutellar spine (Fig. 52) about as long as scutellum. Mesopleuron and mesoscutum
coriaceous, with moderately fine punctures, on mesopleuron separated by a little more than their
diameter. Hind femur 5-15-5-60 times as long as deep. Fore wing length 10-2-10-4 mm. Tergite 1 of gaster
2-10-2-22 times as long as broad, its dorsal profile as in Fig. 76. Tergite 3 with a crease along its anterior 0-4
separating laterotergite.
Colour: black and reddish, with some yellow. Face entirely reddish yellow or yellow with a narrow
median black stripe. Antenna with scape and pedicel blackish dorsally, yellow or reddish yellow ventrally.
Flagellum reddish, darkened dorsally, especially first few segments and distal 0-3. Maxillary palp with
segments 1 , 2 and 3 reddish yellow , 4 and 5 blackish . Posterior part of propodeum reddish , entirely or with
posterior margin narrowly black. Hind coxa blackish with a large (but not sharply defined) reddish dorsal
patch and at least some reddish ventrally. Hind femur reddish with a blackish ventral mark. Segments of
gaster each reddish, with some blackish anteriorly, especially on tergites 1 and 2.
REMARKS. Closely related to volutatorius. I have seen no females which I can associate with the males,
although females were recorded by Uchida (1931: 52).
BIOLOGY AND HOSTS. Unknown. Adults have been collected in August.
DISTRIBUTION (Map 2, p. 16). Japan.
MATERIAL EXAMINED (2 cf)
Japan: 1 cf , Sanjodake, Yamato, 9.viii.l913 (Isshiki) (EIHU) (holotype); 1 cf , Hokkaido, Mt Yubari,
ll.viii.1966 (Kusigemati) (KC).
Banchus tholussp. n.
(Figs 28, 53, 77, 98, 119)
DESCRIPTION. Female. Width of lower face 0-70-0-80 times vertical length of eye. Malar space 0-65-0-85
times basal width of mandible. Maxillary palp as in Fig. 28. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 53) about as long as scutellum. Mesopleuron shining, very weakly sculptured,
with moderate punctures, separated by more than their diameter. Hind femur 6-60-7-05 times as long as
deep. Fore wing length 10-8-12-2 mm. Tergite 1 of gaster 1-90-2-20 times as long as broad, its dorsal profile
as in Fig. 77. Gaster subcylindrical, only weakly compressed apically, not reaching to tips of fore wings
(when folded back). Tergite 3 with a crease along its whole length separating laterotergite. Tergite 7 and
sternite 6 as in Fig. 98.
Colour: variable, from brown and yellowish cream to largely reddish orange (see 'Remarks' below).
Face yellowish to reddish orange with orbits yellow. Antenna blackish, with scape, pedicel, and sometimes
base of flagellum, yellowish or reddish ventrally. Maxillary palp entirely reddish yellow. Posterior part of
38 M. G. FITTON
propodeum pale yellow cream with a postero-median brown area or entirely reddish. Hind coxa yellowish
cream and brown or entirely reddish. Hind femur entirely reddish. Segments of gaster reddish with
segment 1 pale cream or yellowish anteriorly, and sometimes with segments 1 to 4 each with a dark
brownish transverse median band.
Male. Width of lower face 0-80 times vertical length of eye. Malar space 0-70 times basal width of
mandible. Maxillary palp as in Fig. 119; segment 4 moderately widened and flattened; segment 5 about as
long as 4. Antennal flag setae not differentiated (antennae in poor condition and proper observation
difficult). Hind femur 7-15 times as long as deep. Fore wing length 11-5 mm. Tergite 1 of gaster 1-60 times
as long as broad. Tergite 3 with a crease along its whole length separating laterotergite.
Colour: brown and yellowish cream, with some reddish. Antenna brownish dorsally, brownish yellow
ventrally. Maxillary palp entirely reddish yellow. Posterior part of propodeum reddish yellow, narrowly
dark brownish postero medially. Hind coxa yellowish cream and brown. Hind femur entirely reddish.
Segments of gaster reddish, with segment 1 pale creamy yellow anteriorly and 1 and 2 more or less brownish
medially.
REMARKS. The females show a considerable range of variation in colour, the specimens from Burma
resembling those of tumidus, while those from the Philippines are almost entirely reddish orange. The
specimens from Sumatra and Java have an intermediate coloration. Unfortunately only a single male (from
Burma) is known. The females might represent more than one species. See also 'Remarks' under B.
tumidus.
BIOLOGY AND HOSTS. Virtually unknown. Adults have been collected in December, March, April, May and
June. The specimen from Mount Data was collected in oak forest.
DISTRIBUTION (Map 2, p. 16). On mountains in Burma, Sumatra, Java and the Philippines.
MATERIAL EXAMINED
Holotype cf , Burma: NE., Kambaiti, 1800 m, ll.vi.1934 (Malaise) (NR).
Paratypes 10 <J>. Burma: 1 $, Mt Victoria, 1400 m, iii.1938 (Heinrich) (GC) (paratype of Banchus
tumidus Chandra & Gupta); 2 $,Maymyo, 800m, xii. 1937 (HeinricK) (TC) (paratypes of Banchus tumidus
Chandra & Gupta). Sumatra: 1 $, Sungei Kumbang, Korinchi, 1370 m, iv.1914 (Robinson & Klass)
(BMNH). Java: 1 $, Gedeh, Tjibodas, 1700 m, xii.1935 (Lieftinck) (TC). Philippines: 4 $, Mindoro,
Hong, Mt Halcon, 1370 m, 9, 10 & ll.v.1954 (Townes) (TC); 1 $, Mt Data, 2380 m, 31.xii.1952 (Townes)
(TC).
Banchus tumidus Chandra & Gupta
(Figs 29, 54, 78, 99, 120)
Banchus tumidus Chandra & Gupta, 1977: 183. Holotype $, INDIA (GC) [examined].
DESCRIPTION. Female. Width of lower face 0-75-0-80 times vertical length of eye. Malar space 0-65-0-75
times basal width of mandible. Maxillary palp as in Fig. 29. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 54) about as long as scutellum. Mesopleuron shining, only very weakly
sculptured, with moderate punctures, separated by more than their diameter. Hind femur 6-80-7-05 times
as long as deep. Fore wing length 10-5-11-9 mm. Tergite 1 of gaster 1-85-1-95 times as long as broad, its
dorsal profile as in Fig. 78. Gaster subcylindrical, only weakly compressed posteriorly, not reaching to tips
of fore wings (when folded back). Tergite 3 with a crease along its whole length separating laterotergite.
Tergite 7 and sternite 6 as in Fig. 99.
Colour: yellowish cream, reddish and brown. Face yellowish. Antenna dark brown, with scape, pedicel
and base of flagellum yellowish ventrally. Maxillary palp entirely reddish yellow. Posterior part of
propodeum yellowish cream with a postero-median brown area. Hind coxa yellowish cream and brown.
Hind femur entirely reddish. Segments of gaster reddish yellow, with segment 1 pale anteriorly and
segments 1 to 4 each with a brownish transverse median band.
Male. Width of lower face 0-80-0-85 times vertical length of eye. Malar space 0-60-0-70 times basal
width of mandible. Maxillary palp as in Fig. 120; segment 4 very considerably widened and flattened;
segment 5 about 0-76 as long as 4. Antennal flag setae not differentiated. Hind femur 6-70-6-90 times as
long as deep. Fore wing length 10-8 mm. Tergite 1 of gaster 1-70-1-75 times as long as broad. Tergite 3 with
a crease along its whole length separating laterotergite.
Colour: creamy yellow and brown. Antenna dark brown dorsally, yellow ventrally. Maxillary palp
reddish yellow with segment 4 partly blackish (Fig. 120). Posterior part of propodeum reddish yellow,
brownish postero-medially. Hind coxa creamy yellow and brown. Hind femur entirely reddish. Segments
of gaster reddish yellow, with segment 1 pale cream anteriorly and 1 and 2 brownish medially.
ICHNEUMON-FLY GENUS BANCHUS 39
REMARKS. This species is very closely related to tholus. The males are easily distinguished on the form of the
maxillary palp, but not otherwise. The females are impossible to separate on morphological characters;
those placed here as tumidus have segment 4 of the maxillary palp very weakly bicoloured (it is strongly
bicoloured in males) . The females of tholus show a much wider range of variation in colour and the division
between the two species was made largely on the basis of geography. See also 'Remarks' under tholus.
BIOLOGY AND HOSTS. Virtually unknown. Adults have been collected in April, May, June, September and
November.
DISTRIBUTION (Map 2, p. 16). On the southern flank of the western Himalayas between 600 and 2000 m.
MATERIAL EXAMINED (9 9 , 2 cf )
India: 1 $ , Uttar Pradesh, Kumaon Himalaya, Jeolikote, 1219 m, 12.ix.1965 (Tikar) (GC) (holotype); 1
$ , Himachal Pradesh, Khajjiar, 1920 m, 24. vi. 1965 (Joseph) (GC); 1 $ , Himachal Pradesh, Khajjiar, 1828
m, 30.iv. 1971 (Ram) (GC); 2 $ , Himachal Pradesh, Manali, 1828 m, 17 & 20. v. 1970 (Ram & Gulati) (GC);
1 $, Uttar Pradesh, 1949 (Bianchi} (TC); 2 $, Uttar Pradesh, Dehra Dun, 600 m, 27.xi.1965 (Gupta}
(GC); 1 cT, Uttar Pradesh, Kumaon Hills, Bhowali, 1700 m, 5-8.vi.1968 (Gupta) (GC); 1 cf, Uttar
Pradesh, Dehra Dun, 7.iv.l967 (Tikar) (GC); 1 $, Uttar Pradesh, Gharwal Himalaya, Phata, 1524 m,
l2.v.l961(Kamath)(GC).
Banchus turcator Aubert
(Figs. 30, 55, 79, 100, 121)
Banchus turcator Aubert, 1978: 157. [Unavailable name published conditionally (Article 15 of the Code).]
Banchus turcator Aubert, 1981: 18. Holotype cf , TURKEY (AC) [examined].
DESCRIPTION. Female. Width of lower face 0-90-1-00 times vertical length of eye. Malar space 0-50-0-55
times basal width of mandible. Maxillary palp as in Fig. 30. Antenna with apical segments longer than
broad. Scutellar spine (Fig. 55) usually very small, sometimes absent or reasonably well developed.
Mesopleuron and mesoscutum strongly coriaceous, with moderate to strong punctures, on mesopleuron
separated by slightly less than their diameter. Hind femur 4-70-5-60 times as long as deep. Fore wing length
8-0-10-7 mm. Tergite 1 of gaster 1-45-1-90 times as long as broad, its dorsal profile as in Fig. 79. Gaster
relatively weakly compressed, reaching about to tips of fore wings (when folded back). Tergite 3 with a
crease along its anterior 0-3 separating laterotergite. Tergite 7 and sternite 6 as in Fig. 100.
Colour: mainly black and yellow, the yellow areas on the appendages tending to reddish. Face varying
from yellow with a black median stripe to black with yellow orbital stripes. Antenna with scape and pedicel
black dorsally, yellow ventrally. Flagellum entirely orange except that it is dark at extreme apex. Maxillary
palp reddish yellow, with segments 1 and 5 wholly, 2 proximally and 4 distally dark. Posterior part of
propodeum black with a yellow band (sometimes interrupted medially) along posterior transverse carina.
Hind coxa black, entirely or with a yellow dorsal patch. Hind femur black, yellow or reddish yellow
proximally and distally, the yellowish area sometimes extending along most of dorsal surface. Segments of
gaster each black anteriorly, yellow (sometimes broadly so) posteriorly.
Male. Width of lower face 0-90-1-05 times vertical length of eye. Malar space 0-45-0-55 times basal
width of mandible. Maxillary palp as in Fig. 121 ; segment 4 flattened and considerably widened; segment 5
about as long as 4. Antennal flag setae (similar to Fig. 127, but setae more widened toward their apices)
upright, flattened and widened, 2 per segment, arising from a polished trough. Hind femur 4-60-5-30 times
as long as deep. Fore wing length 8-7-10-0 mm. Tergite 1 of gaster 1-55-1-65 times as long as broad. Tergite
3 with a crease along its anterior 0-3 separating laterotergite.
Colour: black and yellow, with some reddish. Antenna with scape and pedicel black dorsally, yellow
ventrally. Flagellum orange, darkened distally. Maxillary palp with segment 1 yellowish (sometimes partly
darkened), 2 and 3 wholly and 4 proximally reddish yellow, 4 distally and 5 wholly blackish or brownish.
Posterior part of propodeum black with a yellow band along posterior transverse carina, the band
sometimes broken medially, reduced to two lateral spots or even absent. Hind coxa black, usually with a
ventral and sometimes also a dorsal yellow spot. Hind femur yellowish red with a large medial black area.
Segments of gaster each black anteriorly, yellow posteriorly.
REMARKS. This species has been confused with pictus in collections. It is, however, quite unrelated as
shown, for instance, by the rather different flag setae.
BIOLOGY AND HOSTS. Unknown. Adults have been collected in April, May and June.
DISTRIBUTION (Map 4, p. 20). Turkey and in the mountains of Tadzhikistan.
40 M. G. FITTON
MATERIAL EXAMINED (5 9 , 9 cf )
Turkey: 1 of, Guriin, 12-15.vi.1976 (Heinrich) (AC) (holotype); 1 $, Alem-Dag, 600 m, 26-30. vi.,
(Demelt) (FSA); 1 cf , Antalya, Termessus, 9.V.1968 (Halliri) (NR); 1 $ , Erzurum, 20 km tspir to Ikizdere
rd, 1700 m, 2.vi.l962 (Guichard & Harvey) (BMNH); 1 cf , Gumus,ane, near Maden, 1800 m, 29.V.1962
(Guichard & Harvey) (BMNH); 1 $, Isik-Dag, 1200 m, vi.1966 (Real) (FSA); 1 $, Istanbul, 2Q.iv.(de
Gaullecoll.) (MNHN); 1 cf , Kizilcahamam, 1000m, 26-28.V.1964 (FSA); 1 $, 1 cf , Nevehir, Urgiip, 4 &
6.vi.l978 (Schwarz) (AC); 1 cf, Zara Taiger (UZM). U.S.S.R.: 2 cf , Gissarskiy Khrebet, K-K chanch-ov
on Aliche, 25.iv.1960 (Malyavin) (MLSU); 1 cf , Khorog region, Shugnansk Khrebet, 2600 m, 5.vi.l956
(Zhelokhovtsev) (MLSU).
Banchus volutatorius (Linnaeus)
(Figs 3, 7, 8, 31, 56, 80, 101, 122, 128)
Ichneumon volutatorius Linnaeus, 1758: 562. Lectotype cf , EUROPE (LSL), fixed by Roman, 1932: 14
[examined] .
Ichneumon venator Linnaeus, 1758: 564. Type(s) $, EUROPE (lost).
Ichneumon umbellatarum Schrank, 1786: 261. Type(s) $ [not cf as stated by Schrank], WEST GERMANY
(lost). Syn. n.
Ichneumon certatorThunberg, 1822: 266; 1824: 322. Holotype $, SWEDEN (UU) [not examined].
Banchus monileatus Gravenhorst, 1829: 393. Lectotype $, POLAND (ZI), designated by Townes &
Townes, 1978: 532 [examined]. Syn. n.
Banchus farrani Curtis, 1836: 588. Lectotype cf , IRELAND (NMV), designated by Fitton, 1976: 322
[examined].
Banchus calcaratus Szepligeti, 1910: 186. Holotype $, HUNGARY (TM) [examined].
Banchus volutatorius var. alticola Schmiedeknecht, 1910: 1931 . Syntypes 9 , cf , EAST GERMANY (7MNHU)
[not examined]. [Junior secondary homonym of Cidaphurus alticola Ashmead, 1901: 148.]
Banchus obscurus Meyer, 1926: 263. Type(s) $ [not cf as stated by Meyer], U.S.S.R. (destroyed).
NOMENCLATURE. The identity of venator was established by Fitton (19786: 375).
Gravenhorst (1829: 389) synonymised /. umbellatarum with B. falcatorius but subsequently it has been
included in the synonymy of B. compressus (= dilatatorius) (e.g. Aubert, 1978: 152 [with the date given
incorrectly as 1802]) without, however, being used as the valid name for that species, despite its seniority.
The description does not fit dilatatorius; it could apply to some females of falcatorius or volutatorius. I
consider that it best fits certain females of volutatorius, with which it is here synonymised.
Although the types of var. alticola were not located a male determined by Schmiedeknecht, and from the
type-locality, was examined (ZSBS).
Meyer undoubtedly mistook the sex of the type-material (probably a single specimen) of obscurus
because the face coloration cannot apply to any male Banchus. The description fits perfectly the female of
volutatorius and the host recorded by Meyer (Lacanobia oleracea) is one of those known for this species.
DESCRIPTION. Female. Width of lower face 0-85-0-90 times vertical length of eye (Fig. 3). Malar space
0-55-0-60 times basal width of mandible. Maxillary palp as in Fig. 31. Antenna with apical segments longer
than broad. Scutellar spine (Fig. 56) distinct, often almost 0-5 as long as scutellum. Mesopleuron and
mesoscutum coriaceous, with moderate punctures, on mesopleuron separated by about their diameter.
Hind femur 4-85-5-35 times as long as deep. Fore wing length 8-1-9-8 mm. Tergite 1 of gaster 1-80-2-05
times as long as broad, its dorsal profile as in Fig. 80. Gaster (Figs 7, 8) moderately compressed, not
reaching to tips of fore wings (when folded back). Tergite 3 with a crease along its anterior 0-4 separating
laterotergite. Tergite 7 and sternite 6 as in Fig. 101.
Colour: predominantly black, with legs largely reddish. Face black, occasionally with brownish orbital
marks. Antenna black, with scape, pedicel and proximal flagellar segments reddish brown ventrally.
Maxillary palp black, with segments 2 and 3 wholly and 4 proximally reddish or brownish. Posterior part of
propodeum black, often entirely, sometimes with reddish areas (of varying extent) on segments 1, 2 and 3
(in extreme cases with posterior margins of these segments narrowly yellowish).
Male. Width of lower face 0-90-1-00 times vertical length of eye. Malar space 0-45-0-60 times basal
width of mandible. Maxillary palp as in Fig. 122; segment 4 considerably widened and flattened; segment 5
about 0-5 as long as 4. Antennal flag setae (Fig. 128) upright, very long, flattened and widened, 2 per
segment, arising from a polished trough. Hind femur 4-85-5-40 times as long as deep. Fore wing length
8-0-9-9 mm. Tergite 1 of gaster 1-65-2-00 times as long as broad. Tergite 3 with a crease along its anterior
0-4 separating laterotergite.
Colour: black and yellow, with some reddish. Antenna black with flagellum yellowish ventrally except at
ICHNEUMON-FLY GENUS BANCHUS 41
its distal apex. Maxillary palp with segment 1 yellow, 2 and 3 wholly and 4 proximally reddish yellow, 4
distally and 5 wholly blackish. Posterior part of propodeum black or with a yellow mark or marks along
(usually just anterior to) posterior transverse carina. Hind coxa black, entirely or with small yellow marks
externo-laterally and/or postero-ventrally. Hind femur reddish yellow, rarely with a blackish ventral mark.
Segments of gaster each black anteriorly, yellow posteriorly, tergites 1, 2 and 3 very broadly yellow and
with anterior area often partly (sometimes largely) reddish.
REMARKS. The male is similar, superficially, to falcatorius but can be separated readily by the number of
flag setae and the lack (usually) in falcatorius of a distinct spine on the scutellum. The female has a gaster
which is quite unlike that of falcatorius and would not be confused with it. The female is, however,
frequently confused with palpalis, although the maxillary palp is different in form and volutatorius
generally has a much shorter scutellar spine.
BIOLOGY AND HOSTS. Adults have been collected mainly in the period from mid- June to early August, but
there are a few records as early as mid- April and as late as September. However, there is no evidence of two
generations in the distribution of the records and information from rearings also indicates that volutatorius
is univoltine. I have seen 58 reared specimens, many with detailed and reliable host data. The hosts (in
decreasing order of number of rearings and reared specimens) are: Anarta myrtilli (Linnaeus) (Noctuidae),
12 $, 11 cf from 12 rearings, Great Britain, Germany, the Netherlands and Switzerland; Lacanobia
oleracea (Linnaeus) (Noctuidae), 14 $ , 1 cf from 5 rearings, Great Britain; Mamestra brassicae (Linnaeus)
(Noctuidae), 1 $, 1 cf from 2 rearings, Great Britain; Hadena compta (Denis & Schiffermuller)
(Noctuidae), 3 $, 1 cf from 1 rearing, Denmark; Ceramica pisi (Linnaeus) (Noctuidae), 1 $, 1 cf from 1
rearing, Great Britain; Heliothis viriplaca (Hufnagel) (Noctuidae), 2 cf from 1 rearing, Germany; Xestia
xanthographa (Denis & Schiffermuller) (Noctuidae), 1 cf from 1 rearing, Great Britain; Habrosyne
pyritoides (Hufnagel) (Thyatiridae), 1 $ from 1 rearing, Germany; Opisthograptis luteolata (Linnaeus)
(Geometridae), 1 $ from 1 rearing, Germany; and Zygaena ephialtes (Linnaeus) (Zygaenidae), 1 cf from
1 rearing, Germany.
The literature additionally records these noctuid hosts: Agrotis segetum (Denis & Schiffermuller)
(Meyer, 1927a: 81), Lycophotia porphyrea (Denis & Schiffermuller) (Leonardi, 1928: 83; Meyer, 1934:
231); Lacanobia suasa (Denis & Schiffermuller) and L. contigua (Denis & Schiffermuller) (Ljungdhal,
1918: 82; Meyer, 1934: 231); and Bena prasinana (Linnaeus) (Hedwig, 1939: 22). Zorin & Zorina (1929)
give some biological information on the association with Lacanobia oleracea.
DISTRIBUTION (Map 10, p. 42). Widespread in northern and central Europe, extending into Turkey, and in
the U.S.S.R. occurring as far south as Alma-Ata and as far east as the Chitinskaya Oblast.
MATERIAL EXAMINED (486 $, 301 cf)
Austria, Belgium, Denmark, East Germany, Finland, France, Great Britain, Hungary, Ireland, Nether-
lands, Norway, Poland, Rumania, Sweden, Switzerland, Turkey, U.S.S.R., West Germany (AC, BC,
BMNH, BRI, CM, FSA, HC, IBMPP, IEUB, IP, ITZ, IZPAN, JC, KHC, LELW, MHN, MLSU,
MNHN, MUM, NMB, NR, PC, RNH, RSM, TM, UM, USNM, UZI, UZM, VRC, ZC, ZI, ZIL, ZIM,
ZMU, ZSBS).
Banchus zonatus Rudow
(Figs 32, 57, 58, 81, 102, 123, 126)
Banchus zonatus Rudow, 1883o: 57. Type(s) $, EUROPE 'Sudeuropa' (?JPM) [not examined].
Banchus algericus Schmiedeknecht, 1910: 1927. Holotype $, ALGERIA (MNHU) [examined]. Syn. n.
NOMENCLATURE. Apart from some exaggeration of size, the description of zonatus fits perfectly, and only,
the female of this species. The synonymy is further confirmed by the type-locality 'Vaterland Sudeuropa'
(the species was not described from Germany as stated by Aubert, 1978: 166). It has not been possible to
obtain on loan material from the Rudow collection. Rudow's methods and the poor condition of his
collection were criticized during his lifetime. The collection, still neglected, was seen recently by
Horstmann (pers. comm.), who made notes on its contents, including the fact that three specimens stand as
B. zonatus. However, nothing is known of the status or identity of these specimens.
DESCRIPTION. Female. Width of lower face 1-00-1 -15 times vertical length of eye. Malar space 0-75-0-95
times basal width of mandible. Maxillary palp as in Fig. 32. Antenna with apical segments broader than
long. Scutellar spine (Figs 57, 58) about 0-5 as long as scutellum or almost absent (see 'Remarks')-
Mesopleuron and mesoscutum coriaceous with moderately strong punctures, on mesopleuron separated
by less than their diameter. Hind femur 5-30-6-15 times as long as deep. Fore wing length 11-1-13-5 mm.
M. G. FITTON
Map 10 Distribution of Banchus volutatorius.
Tergite 1 of gaster 1-35-1-65 times as long as broad, its dorsal profile as in Fig. 81. Gaster strongly
compressed from posterior part of segment 3, reaching beyond tips of fore wings (when folded back).
Tergite 3 with a crease along its anterior 0-4 separating laterotergite. Tergite 7 and sternite 6 as in Fig. 102.
Colour: black and yellow. Face yellow with a black median stripe. Antenna black (sometimes brownish
distally), scape with a yellow patch beneath. Maxillary palp dark brown or black. Posterior part of
propodeum black with a yellow stripe along posterior transverse carina. Hind coxa black, usually with a
yellow dorsal patch. Hind femur black, yellow at extreme base, distally, and often dorsally. Segments of
gaster each black anteriorly, yellow posteriorly.
Male. Width of lower face 1-00-1-20 times vertical length of eye. Malar space 0-60-0-90 times basal
width of mandible. Maxillary palp as in Fig. 123; segment 4 flattened and slightly widened; segment 5 about
0-8 as long as 4. Antennal flag setae (Fig. 126) upright, moderately long, flattened, 2 per segment, arising
from a polished trough. Hind femur 5-35-6-65 times as long as deep. Fore wing length 10-5-14-0 mm.
Tergite 1 of gaster 1-65-1-90 times as long as broad. Tergite 3 with a crease along its anterior 0-4 separating
laterotergite.
Colour: back and yellow. Antenna black, scape and pedicel (and sometimes flagellum segment 1
proximally) yellow ventrally, flagellum sometimes slightly brownish ventrally. Maxillary palp blackish or
brownish. Posterior part of propodeum yellow anteriorly, black posteriorly, the yellow varying from a
stripe along posterior transverse carina to covering almost entire area. Hind coxa black with a dorsal and
often a postero- ventral yellow patch. Hind femur yellow, blackish laterally and ventrally towards base.
ICHNEUMON-FLY GENUS BANCHUS
Map 1 1 Distribution of Banchus zonatus.
Segments of gaster each black anteriorly, yellow posteriorly, the yellow less variable in extent than in
females, occupying about 0-5 of each tergite.
REMARKS. This is the largest species of the genus and the appearance of the females in particular is very
striking. B. zonatus seems to be most closely related to dilatatorius and moppiti.
The specimens from the eastern Mediterranean (Cyprus and Israel) consistently have the scutellar spine
very small or virtually absent and it is possible that they represent a separate species. However, this is
difficult to investigate without additional material. There are no specimens from peninsular Italy, Greece
or Turkey in collections.
BIOLOGY AND HOSTS. Unknown. Dates of collection range from early September to mid-December. There
is a single male with a date in March, unfortunately without an intelligible locality but with the intriguing
data'Salixcaprea'.
DISTRIBUTION (Map 11, p. 43). A 'Mediterranean' species. Meyer (19276: 291) records the species from
the Caucasus but I have seen no specimens to substantiate this north-eastward extension of its distribution.
MATERIAL EXAMINED (28 $ , 21 C?)
Algeria, Cyprus, France, Israel, Italy, Morocco, Spain, Tunisia (AC, BMNH, FSA, IEE, MHN,
MNHN, MNHU, NR, TAU, TC, UZM).
Species excluded from Banchus
The following Old World species were described incorrectly in Banchus.
Banchus armitfafiisMorley, 1913: 254. Holotype cf, NICOBAR ISLANDS (BMNH) [examined].
Identity. Phytodietus armillatus (Morley). In Kaur & Jonathan's (1979) treatment of Oriental Phyto-
44 M. G. FITTON
dietus the holotype runs to P. alasuffuscus Kaur & Jonathan. However, it differs in some details,
particularly in the degree of constriction of tergite 1 of the gaster.
Banchus elator Fabricius, 1804: 128. LECTOTYPE cf, AUSTRIA (UZM), here designated (selected by
G. E. J. Nixon) [examined].
Identity. Earinus elator (Fabricius) comb. n. (Braconidae) (see Nixon, 1986, in press).
Banchus histrio Fabricius, 1798: 234. Lectotype cf, EAST GERMANY (UZM), designated by Horstmann,
19826: 243 [not examined].
Identity. Lissonota histrio (Fabricius) (Horstmann, 19826: 243).
Itaiic/iifsqiiadratorSchellenberg, 1802: 21. Type(s) $, SWITZERLAND (lost).
Identity. Megarhyssa quadrator (Schellenberg) comb. n. Schellenberg's description and figures are
obviously of a rhyssine (Pimplinae). The species concerned corresponds to that named as M. citraria
(Olivier) in the BMNH collection. The two recent catalogues (Oehlke, 1967; Aubert, 1969) covering
European Megarhyssa differ in so many details that it is not clear what name is the valid one for this species.
Whatever it is, quadrator will be a junior synonym. Gravenhorst's invitation (1829: 959) to compare
quadrator with M. superba (Schrank) has been overlooked by all who have dealt subsequently with the
European Rhyssini.
Banchus robustusRudow, 18836: 246. Type(s) $, EAST GERMANY (?JPM) [not examined].
Identity. Unknown, it remains a nomen dubium. It is excluded from Banchus because of the length of the
ovipositor: 'Legestachel fast halb so lang als Hinterlieb'.
Banchus spinipes Panzer, 1800: 17. Type(s) $, GERMANY (lost).
Identity. Junior synonym of Cephus pygmeus (Linnaeus) (Cephidae) (Muche, 1981: 283).
Banchus tomentosusGravenhorst, 1829: 376, Holotype $, EAST GERMANY (ZI) [examined].
Identity. Exetostes tomentosus (Gravenhorst) (Townes, Momoi & Townes, 1965: 235).
Banchus viUosulus Gravenhorst, 1807: 267. Syntypes [?number] [?$], no type-locality (lost).
Identity. The description of this species fits a cynipoid rather than an ichneumonid. Gravenhorst himself
compared it to Ibalia and did not refer to it in any of his subsequent works on ichneumonids. It is unlikely
that any type-material will be found. The general nature of the description makes application of the name
to a particular species difficult, but it is desirable to formally transfer the name to the Cynipoidea so that
workers on that group can consider its identity. The species is here placed as a nomen dubium in Andricus
(comb, n.) (Cynipidae), to the agamic females of which the description seems best to apply.
Banchus viridator Fabricius, 1804: 127. Syntypes $, AUSTRIA (UZM) [one of four putative syntypes
examined].
Fabricius cited Banchus spinipes Panzer as a synonym of his new species viridator. I can see no reason for
him not to have used Panzer's name for the species. Whatever the reasons, it complicates consideration of
the name. It could be treated as an independent species, as published in synonymy, or as a replacement
name (in which case the type-specimens would be those of B. spinipes rather than those cited above).
Identity. Junior synonym of Cephus pygmeus (Linnaeus) (Cephidae) (Muche, 1981: 283).
The following species have been placed incorrectly in Banchus at some time.
Ichneumon annulatus Fabricius, 1793 is a pompilid, Pompilus (Episyron) annulatus (Fabliau's) (Schulz,
1912: 73).
Ichneumon cultellator Fabricius, 1793 is a junior synonym of Ibalia leucospoides (Hochenwarth) (Iba-
liidae) (Kerrich, 1973: 73).
Ichneumon fornicator Fabricius, 1781 is a species oiExetastes (Townes, Momoi & Townes, 1965: 229).
Ichneumon gravidator Linnaeus, 1758 is a species of Proctotrupes (Proctotrupidae) (Fitton, 19786: 378;
Townes & Townes, 1981: 179).
Ichneumon varius Fabricius, 1793 is a junior synonym of Theronia atalantae Poda (Aubert, 1978: 167).
Ichneumon vigilatorius Panzer, 1804 was synonymised with Banchus falcatorius by Gravenhorst (1829:
390). This was later rejected in favour of synonymy with compressus (Aubert, 1978: 153) but was reinstated
by Horstmann (19826: 238). However, Schaeffer's illustration (1767: pi. 61, fig. 6), to which Panzer's
description and name refer, clearly shows an ichneumonid with segment 1 of the gaster strongly petiolate
and which probably belongs in the Ichneumoninae. Although I cannot find an ichneumonine with all
ICHNEUMON-FLY GENUS BANCHUS 45
details of the colour pattern correct, I think it best to place the species as a nomen dubium in Ichneumon
(comb. rev.).
Acknowledgements
I am grateful to the following for the loan of material and provision of information relating to collections in
their care: A. Albrecht, A. A. Allen, A. Athanasov, J. F. Aubert, C. Baroni Urbani, J. R. Barron, R.
Bauer, C. Besuchet, A. Brindle, R. W. Carlson, R. Danielsson, G. Delrio, P. Dessart, E. Diller, M. Dorn,
R. Eck, W. A. Ely, P. P. d'Entreves, M. Fischer, D. Gerling, V. K. Gupta, R. Hinz, K. Horstmann, A. G.
Irwin, R. Jussila, D. R. Kasparyan, S. Kelner-Pillault, J. Kolarov, E. Konigsmann, K. Kusigemati, W.
Kuslitzky, E. Mellini, E. Mingo, A. Nakanishi, A. Neboiss, J. Oehlke, P. Oosterbroek, C. O'Toole, J.
Papp, P. I. Persson, B. Petersen, C. Pisica, R. Poggi, W. J. Pulawski, G. van Rossem, J. Sawoniewicz, H.
Schnee, M. R. Shaw, M. Sorg, M. Suwa, C. Thirion, H. K. Townes, M. C. Townes, L. Zimina, L.
Zombori, and K. W. R. Zwart. I am indebted to my colleagues in the BMNH for discussion of various
points; help with translations; and technical assistance.
References
Arthur, A. P. & Ewen, A. B. 1975. Cuticular encystment: a unique and effective defense reaction by
cabbage looper larvae against parasitism by Banchus flavescens (Hymenoptera: Ichneumonidae).
Annals of the Entomological Society of America 68: 1091-1094.
Ashmead, W. H. 1901. Some insects of the Hudsonian zone in New Mexico. II. Hymenoptera. Psyche,
Cambridge 9: 147-148.
1906. Descriptions of new Hymenoptera from Japan. Proceedings of the United States National
Museum 30: 169-201.
Aubert, J. F. 1969. Les Ichneumonides ouest-palearctiques et leurs holes. 1 (Pimplinae, Xoridinae,
Acaenitinae). 302pp. Paris.
1978. Les Ichneumonides ouest-palearctiques et leurs holes. 2 (Banchinae, et suppl. aux Pimplinae).
318 pp. Paris.
1981. Syllogismes, illogismes et innovations chez les Ichneumonides. Bulletin de la Societe Entomolo-
gique de Mulhouse 1981: 17-22.
Aurivillius, C. 1890. Gronlands Insektfauna. I. Lepidoptera, Hymenoptera. Kungliga Svenska Vetenskap-
sakademiens Handlingar 15 (IV) (1): 1-33.
Bajari. E. 1960. Furk6szdarazs-Alkatuak I. Ichneumonoidea I. Fauna Hungariae 54, 266 pp.
Barron, J. R. 1976. Systematics of Nearctic Euceros (Hymenoptera: Ichneumonidae: Eucerotinae). Le
Naturaliste Canadien 103: 285-375.
Beirne, B. P. 1941 . A consideration of the cephalic structures and spiracles of the final instar larvae of the
Ichneumonidae (Hym.). Transactions of the Society for British Entomology 7: 123-190.
Bischoff, H. 1930. Entomologische Ergebnisse der Deutsch-Russischen Alai-Pamir-Expedition 1928 (I). 3.
Hymenoptera I (Sphecidae, Vespidae, Scoliidae, Mutillidae, Chrysididae, Ichneumonidae, Evaniidae).
Mitteilungen aus dem Zoologischen Museum in Berlin 16: 215-225.
Bledowski, R. & Krairiska, M. K. 1926. Die Entwicklung von Banchus femoralis Thorns. (Hymenoptera,
Ichneumonidae). Bibliotheca Universitatis Liberae Polonae (A) 16: 3-50.
Bretherton, R. F., Goater, B. & Lorimer, R. I. 1983. Noctuidae [part]. In Heath, J. & Emmet, A. M.
(Eds), The moths and butterflies of Great Britain and Ireland 10: 36-413.
Cameron, P. 1904. Descriptions of new genera and species of Ichneumonidae from India (Hym.).
(Continued). Zeitschrift fur systematische Hymenopterologie und Dipterologie 4: 337-347.
Chandra, G. & Gupta, V. K. 1977. Ichneumonologia Orientalis. Part VII. The tribes Lissonotini
and Banchini (Hymenoptera: Ichneumonidae: Banchinae). Oriental Insects Monographs 7, 290 pp.
Delhi.
Christ, J. L. 1791. Naturgeschichte, Klassification und Nomenclatur der Insekten vom Bienen, Wespen und
Ameisengeschlecht. 535 pp. Frankfurt am Main.
Constantineanu, M. I. & Pisica, C. 1959. Specii de Pimpline (Pimplinae Cresson) noi pentru fauna
Republicii Populare Romine. In Omagiu lui Traian Sdvulescu cuprilejul tmplinirii a 70 de ani: 187-192.
1960. Ichneumonide din Valea Bistrijrei dintre Poiana Teiului sj Vaduri (subfamiliile: Pimplinae,
Lissonotinae, Xoridinae, Acoenitinae, Colyriinae i Banchinae). Analele Stiintifice de Universitdti "Al.
I. Cuza" din lasl. (New series) (II. Stiinte naturale) 6: 701-712.
Cresson, E. T. 1865. Catalogue of Hymenoptera in the collection of the Entomological Society of
46 M. G. FITTON
Philadelphia, from Colorado Territory. Proceedings of the Entomological Society of Philadelphia 4:
242-313.
Curtis, J. 1836. British Entomology 13: 578-625. London.
Cushman, R. A. 1937. Revision of the North American species of ichneumon-flies of the genus Exetastes
Gravenhorst. Proceedings of the United States National Museum 84: 243-312.
Dalla Torre, C. G. de. 1901. Catalogus Hymenopterorum 3 (1): 1-544.
Ewen, A. B. & Arthur, A. P. 1976. Cuticular encystment in three noctuid species (Lepidoptera): induction
by acid gland secretion from an ichneumonid parasite (Banchusflavescens). Annals of the Entomological
Society of America 69: 1087-1090.
Fabricius, J. C. 1775. Systema Entomologiae. 832 pp. Flensburgi et Lipsiae.
1781. Species Insectorum 1, 552 pp. Hamburgii et Kilonii.
1787. Mantissa Insectorum 1, 348 pp. Hafniae.
1793. Entomologia Systematica 2, 519 pp. Hafniae.
1798. Supplementum Entomologiae Systematicae. 572 pp. Hafniae.
1804. Systema Piezatorum. 439 pp. Brunsvigae.
Fitton, M. G. 1976. The western Palaearctic Ichneumonidae (Hymenoptera) of British authors. Bulletin of
the British Museum (Natural History) (Entomology) 32: 301-373.
19780. The Ichneumonidae (Hymenoptera) described by J. F. Ruthe. Zeitschrift der Arbeitsgemein-
schaft Osterreichischer Entomologen 30: 75-79.
19786. The species of 'Ichneumon' (Hymenoptera) described by Linnaeus. Biological Journal of the
Linnean Society 10: 361-383.
Foerster, A. 1869. Synopsis der Familien und Gattungen der Ichneumonen. Verhandlungen des Naturhis-
torischen Vereines der Preussischen Rheinlande und Westphalens 25: 135-221.
Friederichs, K. 1936. Zur Phanologie einiger Parasiten der Forleule (Panolisflammea). Entomologische
Zeitschrift 50: 89-91.
Geoffrey, E. L. 1785. In Fourcroy, A. F. de (ed.), Entomologia Parisiensis 2: 233-544. Parish's.
Gmelin, J. F. 1790. In Linnaeus, C., Systema Naturae 13th edn, 1 (5): 2225-3020. Lipsiae.
Gravenhorst, J. L. C. 1807. Vergleichende Uebersicht des Linneischen und einiger neueren zoologischen
Systeme. 476 pp. Gottingen.
1829. Ichneumonologia Europaea 3, 1097 pp. Vratislaviae.
Gupta, V. K. 1971. Ichneumon hunting in India. 109 + 80 pp. Delhi.
1975. Ichneumonological explorations in India. 62 + 64 pp. Delhi.
Gyorfi, J. 1944. Beitrage zur Kenntnis der Ichneumoniden Ungarns. III. Fragmenta Faunistica Hungarica
7: 103-107.
Habermehl, H. 1922. Beitrage zur Kenntnis der palaearktischen Ichneumonidenfauna. Konowia 1:
266-282.
1924. Beitrag zur Kenntnis der primaren und sekundaren Schmarotzerwespen der Kieferneule
(Panolis flammea Schiff. = P. griseovariegata Goeze) (Hym.). Deutsche Entomologische Zeitschrift
1924: 183-184.
Hedwig, K. 1939. Verzeichnis der bisher in Schlesien aufgefundenen Hymenopteren. Zeitschrift filr
Entomologie 18 (3): 12-28.
Hellen, W. 1949. Zur Kenntnis der Ichneumonidenfauna der Atlantischen Inseln. Commentationes
BiologicaeS(lT):l-23.
1961. Ichneumonidenfunde aus Madeira. Notulae Entomologicae 41: 35-38.
Hensch, A. 1928. Beitrag zur Kenntnis der jugoslawischen Ichneumonidenfauna. Konowia 7: 99-112.
Horstmann, K. 1982a. Typenrevision einiger von A. Hensch beschriebener Ichneumoniden-Arten
(Hymenoptera). Zeitschrift der Arbeitsgemeinschaft Osterreichischer Entomologen 33: 81-88.
19826. Revision der von Panzer beschriebenen Ichneumoniden-Arten (Hymenoptera). Spixiana 5:
231-246.
Kaur, R. & Jonathan, J. K. 1979. Ichneumonologia Orientalis. Part VIII. The tribe Phytodietini from
India (Hymenoptera: Ichneumonidae). Oriental Insects Monographs 9: 276 pp. Delhi.
Kerrich, G. J. 1973. On the taxonomy of some forms of Ibalia Latreille (Hymenoptera: Cynipoidea)
associated with conifers. Zoological] ournal of the Linnean Society 53: 65-79.
Kim, C. W. 1955. A study on the ichneumon-flies in Korea. [In Chinese and German.] Commemoration
Theses, Fiftieth Anniversary , Korea University: 423-497.
Kolubajiv, S. 1934. Die Ergebnisse der Ziichtung von parasitischen Insekarten aus ihren Wirten in der
staatlichen Versuchsanstalt in Prag in den J. J. 1929-1933. Casopis Ceskoslovenske Spolecnosti
Entomologicke3l: 59-68, 113-120, 155-163.
Kosobutzkii, M. I. 1928. [The noctuid attacking winter crops (Euxoa segetum Schiff.) in the Votsk
ICHNEUMON-FLY GENUS BANCHUS 47
Autonomous Region (Biology, Ecology and Measures of Control), 1926-1928. (Attempt at a mono-
graphic investigation).] 192 pp. [In Russian.] [Original not seen. Information taken from Review of
Applied Entomology (A) 21: 217.]
Kuslitzky, W. 1979. Banchini (Hymenoptera, Ichneumonidae) of the Mongolian People's Republic. [In
Russian.] Insects of Mongolia 6: 343-353.
Leonard!, G. 1928. Elenco delle species di insetti dannosi e loro parassiti ricordati in Italia fino all' anno 1911.
Partell. 592pp. Portici.
Linnaeus, C. 1758. Systema Naturae, 10th edn, 1, 824 pp. Holmiae.
Ljungdhal, D. 1918. Lepidoperologiske anteckningar. Entomologisk Tidskrift 39: 82-91.
Maneval, H. 1935. Observations sur des Hymemopteres de la faune franchise et description d'une espece
nouvelle. Revue Frangaise d' Entomologie 2: 65-76.
Matsumura, S. 1911. Erster Beitrag zur Insekten-Fauna von Sachalin. Journal of the College of Agricul-
ture, Tohoku Imperial University, Sapporo 4: 1-144.
Meyer, N. F. 1922. [Towards the ichneumonid fauna of the Astrakhan province.] [In Russian and
German.] Izvestiya Otdela Prikladnoi Entomologii 2: 125-140.
1925. [A new ichneumonid species from the Urals.] (Hymenoptera, Ichneumonidae). [In Russian and
German.] Izvestiya Entomologicheskogo i fitopatlogicheskogo byuro Ural'skogo obshchestva lyubitelei
estestvoznaniya 4: 10.
1926. Einige neue Ichneumoniden und Cynipiden. Russkoe Entomologicheskoe Obozrenie 20:
260-264.
1927a. Schlufwespen, die in Russland in den Jahren 1881-1926 aus Schadlingen gezogen sind. [In
Russian.] Izvestiya Otdela Prikladnoi Entomologii 3: 75-91.
1927ft. Zur Kenntnis der Tribus Banchini (Familie Ichneumonidae) und einiger neuer Schlupfwespen
aus Russland. Konowia 6: 291-311.
1928. Zur Biologic der Parasiten von Feltia segetum Schiff . [In Russian.] Izvestiya Otdela Prikladnoi
Entomologii 3: 201-218.
1934. Tables systematiques des Hymenopteres parasites (fam. Ichneumonidae) de 1'URSS et des pays
limitrophes. 3. [In Russian.] Opredeliteli po Faune SSSR, isdawaemyje soologitscheskim Institutom
Akademii Nauk 15: 271 pp.
Momoi, S. 1973. Ergebnisse der zoologischen Forschungen von Dr. Z. Kaszab in der Mongolei. 332.
Einige mongolischen Arten der Tribus Banchini (Hymenoptera: Ichneumonidae). Folia Entomologica
Hungarica (Series Nova) 26 (Suppl.): 241-250.
Morley, C. 1913. Ichneumonidae: I. Ichneumones Deltoidei. Fauna of British India. Hymenoptera 3:
1-531.
1915 . A revision of the Ichneumonidae based on the collection in the British Museum (Natural History) .
4, 167 pp. London.
Muche, H. 1981. Die Cephidae der Erde (Hym., Cephidae). Deutsche Entomologische Zeitschrift (N.F.)
28: 239-295.
Miiller, O. F. 1776. Zoologiae Danicae prodromus. 274pp. Havniae.
Nixon, G. E. J. 1986. A revision of the European Agathidinae (Hymenoptera: Braconidae). Bulletin of the
British Museum (Natural History) (Entomology) [in press] .
Oehlke, J. 1967. Westpalaarktische Ichneumonidae I: Ephialtinae. Hymenopterorum Catalogus (nova
editio) 2, 49 pp. 's-Gravenhage.
Panzer, G. W. F. 1800. Faunae Insectorum Germanicae 73, 24 pp. Nurnberg.
1804. D. Jacobi Christiani Schaefferi iconum insectorum circa Ratisbonam indigenorum enumeratio
systematica. 260pp. Erlangae.
Pfeffer, A. 1933. [Invasion de Panolis flammea en Slovaquie occidentale. Biologic. Moyens repressifs, etc]
Recueil des Travawc de L'Institut Recherche Agronomique de Tschecoslovaquie 116 (2): 3-45. [Original
not seen. Information taken from Review of Applied Entomology (A) 22: 382-385.]
Pospelov, V. P. 1924. [Annual report of the department of applied entomology.] [In Russian.] [Annals of
the state institute of experimental agronomy] 2: 243-252. [Original not seen. Information taken from
Review of Applied Entomology (A) 14: 604.]
Provancher, L. 1874. Les ichneumonides de Quebec avec description de plusieurs especes nouvelles. [part]
Le Naturaliste Canadien 6: 55-63.
Roman, A. 1912. Die Ichneumonidentypen C. P. Thunbergs. Zoologiska Bidragfrdn Uppsala 1: 229-293.
1928. Liste der aus Island bekannten Ichneumoniden. In Lindroth, C. H., Zur Land-Everte-
bratenfauna Islands. I. Goteborgs Kungl. Vetenskaps och Vitterhets Samhalles Handlingar (B) 1 (6):
15-24.
1930. Ichneumonidae collected by Major R. W. G. Hingston on the Oxford University Expedition to
48 M. G. FITTON
Greenland, 1928. Annals and Magazine of Natural History (10) 5: 281-288.
1931. Entomologische Ergebnisse der scwedischen Kamtchatka-Expedition 1920-1922. 33.
Ichneumonidae, Subfamilien Pimplinae und Tryphoninae. Arkivfor Zoologi 23 A (6): 1-32.
1932. The Linnean types of Ichneumon flies. Entomologisk Tidskrift 53: 1-16.
1937. Die Arthropoden fauna von Madeira nach den Ergebnissender Reise von Prof. Dr. O.
Lundblad Juli-August 1935. II. Hymenoptera: Ichneumonidae. Arkiv for Zoologi 30A (1): 1-26.
Rudow, F. 1883o. Einige neue Hymenoptera. Entomologische Nachrichten 9: 57-64.
18836. Neue Ichneumoniden. Entomologische Nachrichten 9: 232-247.
Ruthe, J. F. 1859. Verzeichnis der von Dr. Staudinger im Jahre 1856 auf Island gesammelten Hymenop-
teren. Stettiner Entomologische Zeitung 20: 305-322, 362-379.
Samoilova, Z. 1. 1936 [Influence of parasitism on the quantity of food consumed by infected or non-infected
larvae.] [In Russian.] [Summary of scientific research work of the Institute of Plant Protection, 1935]:
316-317. [Original not seen. Information taken from Review of Applied Entomology (A) 25: 153.]
Schaeffer, J. C. 1767. Icones insectorum circa Ratisbonam indigenorum 1 (2): LI-C. Regensburg.
1768. Icones insectorum circa Ratisbonam indigenorum 2 (1): CI-CL. Regensburg.
Scheidter, F. 1934. Forstentomologische Beitrage. 33. Sekundarparasiten aus Banchus femoralis Thorns.
Zeitschriftfur Pflanzenkrankheiten 44: 507-508.
Schellenberg, J. R. 1802. Entomologische Beytrage. 1, 24pp. Winterthur.
Schmiedeknecht, 0. 1910. Opuscula Ichneumonologica 4 (fasc. 25): 1921-2000. Blankenburg.
Schrank, F. P. 1786. Baiersche Reise. 276 pp. Munchen.
1802. Fauna Boica 2 (2), 412 pp. Ingolstadt.
Schulz, W. A. 1912. Aelteste und alte Hymenopteren skandinavischer Autoren. Berliner Entomologische
Zeitschrift 57: 52-102.
Schwerdtfeger, F. 1952. Untersuchungen iiber den 'Eisernen Bestand' von Kiefernspanner (Bupalus
piniarius L.), Forleule (Panolis flammea Schiff.) und Kiefernschwarmer (Hyloicus pinastri L.). Zeit-
schriftfur angewandte Entomologie 34: 216-283.
Shevyrev, I. Y. 1913. Parasiti i swerchparasiti is mira nasekomych. [In Russian.] Entomologicheskil
Vestnik 1:117-222.
Short, J. R. T. 1978. The final larval instars of the Ichneumonidae. Memoirs of the American Entomo-
logical Institute 25, 508 pp.
Smits van Burgst, C. A. L. 1927. Lijst van de namen der in Midden- en West-Europa waargenomen
parasieten en hyperparasieten van de Gestreepte Dennenrups (Panolis griseovariegata Goeze). Ento-
mologische Berichten 7: 237-240.
Stelfox, A. W. 1936. The Irish ichneumon-flies of the genus Banchus. Irish Naturalist's Journal 6:
63-64.
Sulzer, J. H. 1776. Abgekiirzte Geschichte der Insekten nach dem Linnaeischen System. 1: 274 pp.
Winterthur.
Szepligeti, G. 1910. Ubersicht einiger Gattungen der Banchoinae. Archivum Zoologicum 1: 183-186.
Thomson, C. G. 1897. Opuscula entomologica 22: 2407-2452. Lund.
Thunberg, C. P. 1822. Ichneumonidea, Insecta Hymenoptera, illustrata. Pars I. Memoires de I'Academie
Imperiale des Sciences de St. Petersbourg 8: 249-281.
1824. Ichneumonidea, Insecta Hymenoptera, illustrata. Pars II. Memoires de I'Academie Imperiale
des Sciences de St. Petersbourg 9: 285-368.
Townes, H. K. 1944. A catalogue and reclassification of the Nearctic Ichneumonidae. Memoirs of the
American Entomological Society 11 (1): 1-477.
Townes, H. 1970. The genera of Ichneumonidae. Part 3. Memoirs of the American Entomological Institute
13: 307 pp.
Townes, H., Momoi, S. & Townes, M. 1965. A catalogue and reclassification of the eastern Palearctic
Ichneumonidae. Memoirs of the American Entomological Institutes, 661 pp.
Townes, H. & Townes, M. 1978. Ichneumon-flies of America north of Mexico: 7. Subfamily Banchinae,
tribes Lissonotini and Banchini. Memoirs of the American Entomological Institute 26: 614 pp.
1981. A revision of the Serphidae (Hymenoptera). Memoirs of the American Entomological Institute
32, 541 pp.
Townes, H., Townes, M. & Gupta, V. K. 1961. A catalogue and reclassification of the Indo- Australian
Ichneumonidae. Memoirs of the American Entomological Institute 1, 522 pp.
Uchida, T. 1929. Drei neue Gattungen, neunzehn neue Arten und fuenf neue Varietaten der Ichneumo-
niden aus Japan, Korea und Formosa (Hym). Insecta Matsumurana 3: 168-187.
1931. Beitraege zur Kenntnis der Tribus Banchini aus Japan. Insecta Matsumurana 6: 45-54.
Ulbricht, A. 1911. Ichneumonidenstudien. Archiv fur Naturgeschichte 77 (1(2)): 144-152.
ICHNEUMON-FLY GENUS BANCHUS 49
1916. Niederrheinische Ichneumoniden. 3. Nachtrag. Abhandlungen des Vereins fur Naturwissen-
schaftliche Erforschung des Niederrheins 2: 1-21.
Veen, J. C. van 1982. Notes on the biology of Banchus femoralis Thomson (Hym., Ichneumonidae) an
endoparasitoid oiPanolisflammea (D. & S.) (Lep. , Noctuidae). Zeitschriftfur angewandte Entomologie
94:300-311.
Viktorov, G. A. 1967. On the systematical position of ichneumonids of the genera Odinophora Foerster
and Exetastes Gravenhorst and their new species (Hymenoptera, Ichneumonidae). [In Russian.]
Entomologicheskoe Obozrenie 46: 864-868.
Wagner, A. C. W. 1929. Schlupfwespen und ihre Wirte. Verhandlungen des Vereins fur Naturwissenschaft-
liche Unterhaltung zu Hamburg 20: 1-17.
Wesmael, C. 1849. Sur les ichneumonides de Belgique, appartenant aux genres Metopius, Banchus et
Coleocentrus. Bulletin de I' Academic Royale des Sciences, des Lettres et des Beaux-Arts de Belgique 16:
621-634.
Woldstedt, F. W. 1877. Beitrag zur Kenntnis der um St. Petersburg vorkommenden Ichneumoniden.
Bulletin de I' Academic Imperiale des Sciences de St. Petersbourg 23: 432-460.
Zorin, P. V. & Zorina, L. M. 1929. K biologii ogordnoj ssowki, Polia oleracea L. [In Russian.] Zashchita
RasteniiS: 475-486. [Abstract in Review of Applied Entomology (A) 17: 588-589.]
50
M. G. FITTON
10
Figs 2-10 2-5, measurements used in descriptions, of (2) malar space, dilatatorius, female; (3) face,
volutatorius, female; (4) gaster tergite 1, crefeldensis , male; (5) hind femur, falcatorius, male. 6-10,
female gaster of (6) cerinus, lateral; (7) volutatorius, lateral; (8) volutatorius, dorsal; (9) falcatorius,
lateral; (10) falcatorius , dorsal. Scale lines represent 1 mm.
ICHNEUMON-FLY GENUS BANCHUS
51
13
15
16
\l
17
18
20
22
23
25
26
27
28 29
30
31
Figs 11-40 11-32, segments 4 and 5 of female maxillary palp, right, of (11) agathae; (12) cerinus; (13)
crefeldensis; (14) dilatator ius; (15) falcatorius; (16) flavomaculatus; (17) gudrunae; (18) hastator; (19)
insulanus; (20) japonicus; (21) mauricettae; (22) moppiti; (23) nox; (24) palpalis; (25) p/cft; (26)
poppiti; (27) punkettai; (28) tfio/ws; (29) tumidus; (30) turcator; (31) volutatorius; (32) zonatus. 33-40,
outline of scutellum, left lateral, of (33) agathae; (34) crefeldensis; (35, 36) dilatatorius; (37, 38)
falcatorius; (39) flavomaculatus; (40) gudrunae. Scale lines represent 1 mm.
52
M. G. FITTON
57
58
Figs 41-58 Outline of scutellum, left lateral, of (41) hastator; (42) insulanus; (43) japonicus; (44)
mauricettae; (45) moppiti; (46) nox; (47) palpalis; (48) poppiti; (49, 50) pictus; (51) punkettai; (52)
sanjozanus; (53) tholus; (54) tumidus; (55) turcator; (56) volutatorius; (57, 58) zonatus. Scale line
represents 1 mm.
ICHNEUMON-FLY GENUS BANCHUS
53
Figs 59-73 Gaster tergite 1 , left lateral, of (59) agathae $ ; (60) crefeldensis cf ; (61) dilatatorius $ ; (62)
falcatorius $ ; (63) flavomaculatus $ ; (64) hastator $ ; (65) gudrunae cf ; (66) gudrunae $ ; (67) insulanus
$; (68) japonicus $; (69) mauricettae cf; (70) moppiti <j>; (71) no* cf; (72) palpalis cf; (13) pictus $.
Scale line represents 1 mm.
54
M. G. FITTON
74
75
82
Figs 74-84 74-81, gaster tergite 1, left lateral, of (74) poppiti d"; (75) punkettai 9 ; (76) sanjozanus d";
(77) tholus $ ; (78) tumidus <j> ; (79) turcator $ ; (80) volutatorius $ ; (81) zonatus $ . 82-84, apex of gaster
of female, (a) dorsal, (b) left lateral, of (82) agathae; (83) crefeldensis; (84) dilatatorius. Scale line
represents 1 mm.
ICHNEUMON-FLY GENUS BANCHUS
55
Figs 85-93 Apex of gaster of female, (a) dorsal, (b) left lateral, of (85) falcatorius; (86)flavomaculatus;
(87) gudrunae; (88) hastator; (89) insulanus; (90) japonicus; (91) mauricettae; (92) moppiti; (93) nox.
Scale line represents 1 mm.
56
M. G. FITTON
97
Figs 94-102 Apex of gaster of female, (a) dorsal, (b) left lateral, of (94)palpalis; (95)pictus; (96) poppiti;
(97) punkettai; (98) tholus; (99) tumidus; (100) turcator; (101) volutatorius; (102) zonatus. Scale line
represents 1 mm.
ICHNEUMON-FLY GENUS BANCHUS
57
'103 \ 104
105
111
118
119
123
Figs 103-123 Segments 3, 4 and 5 of male maxillary palp, right, of (103) agathae; (104) crefeldensis; (105)
dilatatorius; (106) falcatorius; (107) flavomaculatus; (108) gudrunae; (109) hastator; (110) japonicus;
(111) mauricettae; (112) nox; (113) moppiti; (114) palpalis; (115)pictus; (116) poppiti; (117) punkettai;
(118) sanjozanus; (119) //ZO/MJ; (120) tumidus; (121) turcator; (122) volutatorius; (123) zonatus. Scale
line represents 1 mm.
58
M. G. FITTON
Figs 124-129 Segments of male antennal flagellum, showing flag setae, right antenna, interno-lateral, of
(124) hastator; (125) crefeldensis; (126) zonatus; (127) moppiti; (128) volutatorius; (129) falcatorius .
ICHNEUMON-FLY GENUS BANCHUS
Index
59
The index includes host names. Principal entries for the valid Banchus species are in bold.
Achlya 6, 25
Acronicta 6, 22
acuminator 7, 17
adusta 6, 19,25,32
agathae7, 11, 14
Agrochola 6, 34
Agrotis6,21,22,34,41
alasuffuscus 44
algericus 8, 41
altaiensis 7, 15
alticola (Ashmead) 8,31
alticola (Schmiedeknecht) 8, 40
ambusta 6, 34
AnartaS, 6,41
Andricus 8, 44
annulatus 8, 44
Aporophyla 6, 17
aries 7, 19
armillatus 8, 43
atalantae 8, 44
Atethmia 6, 34
Autographa 5
Banchopsis 3
Banchus 3
Bena 6, 41
bipunctatus 8, 33
Blepharita 6, 19, 25, 32
brassicae 7, 41
bucephala 6, 34
calcaratus 8, 40
californica 5
Cephus 8, 44
Ceramica 7, 41
Ceratogastra 3
cerinus 7, 14, 15
certator 8, 40
Cidaphurus3,23,31
circellaris 6, 34
citraria 44
compressus 7, 17, 44
compta 7, 41
configurata 5
contigua 7, 41
Corynephanes 19
Corynephanus 3,31
crefeldensis 7, 11, 13, 15, 24, 31
croaticus 7, 15
cultellator 8, 44
cultratus 8, 33
Dasychira 6, 21
Deilephila6,21,25,32
didyma 6, 22
dilatatorius 5, 7, 12, 13, 17, 40, 43
Diprion 7, 22
Diprionidae 7
dispar 25
Earinus 8, 44
elator8,44
elpenor 6, 21
ephialtes 6, 41
Episyron 8, 44
Euceros 6
Euproctis 6, 19
Euxoa 7, 19
exclamationis 6, 22
Exetastes 3, 8, 44
falcator 7, 19
falcatorius4, 5, 7, 11, 13, 19, 41, 44
falcifera 5
farrani 8, 40
fascelina 6, 21
femoralis 6, 7, 24
flam me a 5, 7, 25, 32
flavescens 5
flavicornis 6, 25
flavomaculatus 7, 11, 14, 23, 30, 36
formidabilis 8, 30
fornicator 8, 44
Geometridae 6
groenlandicus 8, 31
gravidator 8, 44
gudrunae 5, 7, 11,13,23
Habrosyne 6, 41
Hadena7,34,41
hastator5,6,7, 12, 13,24,31
Heliothis5,7,41
helvola 6, 34
histrio (Fabricius), 8, 44
histrio(Schrank),7, 19
Hyloicus 6, 25
Ibalia 8, 44
Ichneumon 8, 17, 19, 24, 33, 40, 45
insulanus7, 13,26
intersectus 7, 19
japonicus 7, 12, 14, 26, 36
kolosovi 7, 24
kozlovi 25
labiatus 7, 19
Lacanobia 5, 7, 32, 41
lavrovi 7, 17, 19
Leucoma 6, 19
leucospoides 8, 44
Lissonota 8, 44
luenebergensis 6, 17
luteofasciatus 7, 19
luteolata 6, 41
luteus 7, 22
lutulenta 6, 17
Lycophotia 7, 34, 41
Lymantria 25
Lymantriidae 6
Mamestra5,7, 41
mauricettae 7, 11, 14, 28
medialis 6
megacephala 6, 22
Megarhyssa 8, 44
Melitaea 6, 22
meticulosa 7, 19
monileatus 8, 30, 40
moppiti 5, 7, 12, 13, 28, 43
mutillarius 33
mutillatus 8, 33
myrtilli 5, 6, 41
Nawaia3,26
ni5
nigricans 7, 19
nigromarginatus 7, 19
nobilitator 7, 19
Noctuidae 6
Notodontidae 6
nox 7, 11,14,23,30,36
Nymphalidae 6
obscurus 8, 40
ocellata 6, 34
oleracea5,7,32, 41
ononis 5
Ophion 7, 22
Opisthograptis 6, 41
palpalis 7, 11, 13, 17, 24, 27, 30
Panolis5,7,25,32
Phalera 6, 34
Philogalleria 3
Phlogophora 7, 19
Phytodietus 8, 43
pictus 8, 12, 13, 14, 33, 39
pinastri 6, 25
pini 7, 22
pisi 7, 41
Pompilus 8, 44
poppiti 8, 12, 14, 27, 35
porcellus6,21,25,32
porphyrea 7, 34, 41
prasinana 6, 41
Proctotrupes 8, 44
propitius 7, 19
pungitor7,24
punkettai 8, 12, 14, 23, 36
pygmeus 8, 44
pyritoides 6, 41
quadrator 8, 44
reticulator7,24
Rhynchobanchus 3
rivularis 7, 34
robustus 8, 44
russiator 8, 33
sachalinensis 7, 19
salicis 6, 19
60 M. G. FITTON
sanguinator 7, 19 tholus 8, 12, 13, 15, 37 venator 8, 40
sanjozanus 8, 12, 37 Thyatiridae 6 vigilatorius 8, 44
Scotogramma 5 tomentosus 8, 44 villosulus 8, 44
segetum 6, 21 , 34, 41 Trichoplusia 5 viridator 8, 44
sibiricus 7, 17 tricolor 7, 19 viriplaca 7, 41
similis 6, 19 trifolii 5 volutatorius 4, 5, 8, 12, 13, 31, 37, 40
Smerinthus 6, 34 tumidus 8, 12, 13, 15, 38
Sphingidae 6 turcator 8, 12, 14, 34, 39 xanthographa 5 , 7, 41
spinipesS, 44 Xestia5,7,41
spinosus 7, 30 umbellatarum 8, 40
suasa 7, 41 zagoriensis 8, 33
superba 44 variegator 7, 19 zonatus 5, 8, 12, 13, 29, 41
varius 8, 44 Zygaena 6, 41
Theronia 8, 44 vastator 24 Zygaenidae 6
British Museum (Natural History)
An introduction to the Ichneumonidae of Australia
/. D. Gauld
In the important field of biological and integrated control of pests the parasitic Hymenoptera
are of particular significance, and this work considers one of the largest families of Parasitica,
the Ichneumonidae. The group has received little attention in Australia - though it has already
been utilized successfully in curtailing the ravages caused by accidentally introduced pests. For
selective control programmes to be effective, however, a sound knowledge of the biology of
both the pest and its parasites is essential - and a sound taxonomic base is vital for the
development of such knowledge.
Ironically, considering the group's economic importance, the parasitic Hymenoptera is
amongst the least studied of any group of living organisms, and taxonomic difficulties have
presented major problems to many entomologists working with the Parasitica. An
Introduction to the Ichneumonidae of Australia will go a long way towards rectifying this
situation, being a taxonomic treatment, by genus, of the Australian ichneumonids, a
comprehensive illustrated identification guide, and a summary of all available information on
the group. It will also serve as an introduction to the biology and distribution of Australian
ichneumonids, and provide a check-list of the described species and an index to their known
hosts. It provides an important revision of ichneumonid nomenclature in order to bring the
group into line with the generally accepted principles of zoological nomenclature.
1984, 413pp, 3 maps, 580 figs. Paperback. 565 00896 X 40.00
Titles to be published in Volume 51
The ichneumon-fly genus Banchus Banchus (Hymenoptera) in the Old World
By M. G. Fitton
The phylogeny, classification and evolution of parasitic wasps of the subfamily Ophioninae
(Ichneumonidae) .
By I. D.Gauld
A cladistic analysis and classification of trichodectid mammal lice (Phthiraptera: Ischnocera).
By C. H. C. Lyal
The British and some other European Eriococcidae (Homoptera: Coccoidea).
ByD. J.Williams
Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk
Printed in Great Britain by Henry Ling Ltd, Dorchester
Bulletin of the
British Museum (Natural I
BRITISH MUbtUM
.(NATURAL HISTORY)
30AU61985
PRESENTED
GENERAL LIBRARY
istory)
The phylogeny, classification and
evolution of parasitic wasps of the
subfamily Ophioninae
(Ichneumonidae)
lanD. Gauld
Entomology series
VolSl No 2 29 August 1985
The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.
Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers are
works of reference that will remain indispensable for years to come.
Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to:
Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW75BD,
England.
World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.)
Trustees of the British Museum (Natural History), 1985
The Entomology series is produced under the general editorship of the
Keeper of Entomology: Laurence A. Mound
Assistant Editor: W. Gerald Tremewan
ISBN 565 06012
ISSN 0524-6431
British Museum (Natural History)
Cromwell Road
London SW7 5BD
Entomology series
Vol51No2pp61-185
Issued 29 August 1985
The phytogeny, classification and evolution of
parasitic wasps of the subfamily Ophioninae
(Ichneumonidae)
Ian D. Gauld
Department of Entomology, British Museum (Natural History), Cromwell Road, London
SW7 5BD
Contents
Synopsis 62
Introduction 62
The research potential of Ophioninae 63
The aim of the present work 63
Terminology 63
Material examined 63
Discussion of methodology 63
Choice of parsimony method utilized 64
The advantages and disadvantages of the parsimony method 65
Compatibility methods utilized 66
Value of LEQU . B AS for overcoming some problems of compatibility analysis 67
The operational procedure adopted in this study 70
An apologia for subjectivity in cladistic analysis 70
The holophyly of the subfamily 70
The systematic position of the subfamily 71
Polarity determination and character coding 73
Characters used in the study 74
The characters and their polarity 74
Preliminary remarks on the relative values of the characters 79
The phylogenetic analysis 80
Preliminary analysis of data 80
Analysis using Le Quesne test program, LEQU. B AS 80
Analysis using FOURS program 85
Phylogenetic analyses of groups of ophionine taxa 87
The Euryophion group 87
The Thyreodon group 88
The Dictyonotus/Ophionopsis group 89
The Thyreodon + Euryophion + Dictyonotus complex 90
The Stauropoctonus group 95
The Enicospilus + Stauropoctonus complex 95
The major groups and their relationship with the unassigned taxa 107
The inter-relationship of the two major generic complexes 107
The position of the unassigned taxa 107
Discussion of suggested phylogeny 116
A classification of the Ophioninae 118
Zoogeographic discussion 119
The subfamily Ophioninae 122
The Ophion genus-group 122
The Sicophion genus-group 128
The Eremotylus genus-group 130
The Thyreodon genus-group 133
BRITISH MUSEl
{NATURAL HISTORY
30AUGI9&
PRESENTED
GENERAL LIBRA
Bull. Br. Mus. not. Hist. (Ent.) 51 (2): 61-185 Issued 29 August 1985
62 I. D. GAULD
The Enicospilus genus-group 141
The Orientospilus subgroup 142
The Ophiogastrella subgroup 144
The Stauropoctonus subgroup 145
The Leptophion subgroup 149
The Enicospilus subgroup 154
Acknowledgements 162
References 163
Appendices 168
Index to generic names 185
Synopsis
An attempt is made to reconstruct the phylogeny of the genera of the ichneumonid subfamily Ophioninae
using both parsimony and compatibility methods of analysis. The problems of phylogenetic analysis of
highly homoplastic data are discussed and the strengths and weaknesses of the various methods evaluated.
It was concluded that there is no wholly adequate objective method of analysing highly homoplastic data,
but it is suggested that analysis may be undertaken using subjective evaluation of characters supporting
conflicting patterns. Within the Ophioninae five major evolutionary lineages are recognized, the Ophion,
Sicophion, Eremotylus, Thyreodon and Enicospilus genus-groups. The largest, the Enicospilus genus-
group, is subdivided into five subgroups, the Orientospilus, Ophiogastrella, Stauropoctonus, Leptophion
and Enicospilus subgroups. A scenario for the possible evolution of the subfamily is suggested. Thirty-two
genera are recognized, one of which, Janzophion, is described as new. Rictophion is treated as a synonym
of Euryophion, Ophionopsis is treated as a synonym of Dictyonotus, and Aulophion treated as a synonym
of Stauropoctonus. The majority of genera are shown to be holophyletic, but Ophion and Enicospilus are
thought to be paraphyletic, though for purposes of classificatory convenience they are retained as valid
genera. Diagnostic descriptions of the genera are given and the inter-relationships of their component
species are discussed.
Introduction
Ophionines are mostly quite large, slender, orange-brown ichneumonids that can frequently be
observed flying around lights at night in virtually any part of the world. In temperate regions
there are rather few species, but in the tropics there are very large numbers of taxa, and
ophionines form a conspicuous component of the ichneumonid fauna. Almost all members of
the subfamily are nocturnal or crepuscular, though in drier areas a few are diurnally active.
Males of many species fly at dusk, but most females are not active until an hour or two after dark.
Ophionines are solitary protelean endoparasites of holometabolous insect larvae. The hosts
of the vast majority of species are not known, but what records there are usually refer to species
parasitizing lepidopterous larvae, most often exposed, solitary larvae, and in particular species
of the families Noctuidae, Geometridae and Lymantriidae. Larvae of a wide range of other
families are also attacked, but there are extremely few records of ophionines parasitizing
Microlepidoptera or Rhopalocera. One Nearctic species is exceptional in that it is known to
parasitize coleopterous larvae (Townes, 1971).
Oviposition is usually into the host larva. The parasitoid egg is apparently free in the host's
haemocoel where it hatches to produce a caudate first instar larva (Moutia & Curtois, 1952).
Species attacking mature larvae undergo rapid development, but species (e.g. Enicospilus
americanus) that oviposit in very young larvae have a protracted first larval instar (Price, 1975).
Several species are apparently host specific (Janzen, pers. comm.), whilst others (e.g.
Thyreodon atriventris) parasitize a variety of taxonomically related hosts. Some temperate
species seem to attack a number of different hosts in a similar niche, whilst a few species are
bivoltine, with different generations attacking different hosts (Brock, 1982). A few species seem
to attack a wide range of hosts (Gauld & Mitchell, 1981). The parasitoid larva completes
development just prior to host-pupation, often after the host has constructed a cocoon. The
ichneumonid larva spins a characteristic fibrous, ovoid cocoon which is generally dark brown
with a pale equatorial band. Species may remain as mature larvae or even adults in this cocoon
for the greater part of the year in seasonal habitats.
PHYLOGENY OF THE OPHIONINAE 63
The research potential of Ophioninae
The ease with which ophionines may be collected (using m.v. light-traps) makes them particu-
larly suitable for zoogeographic and ecological study. Large samples may be collected in terrain
where sweep netting and Malaise trapping yield poor results, or, as in the case of rain forest
canopy, where collecting can only be achieved by cumbersome, expensive and (for fast-flying
insects) unproven techniques. Unlike Lepidoptera, which are similarly easy to collect, most
tropical ophionines are readily separable without recourse having to be made to time-consuming
genitalic preparation. Illustrated keys are available to facilitate identification of most Old World
tropical species (Gauld, 1977; Gauld & Mitchell, 1978; 1981).
The aim of the present work
If the Ophioninae is to be used as a serious vehicle for zoogeographic study then it is necessary to
have an understanding of the phylogenetic inter-relationships of the genera, as many authors
contend that biogeographic speculations are valid only when related to the evolutionary history
of a group (Mackerras, 1962; Nelson & Platnick, 1981). As little has been published on the
phylogeny of Ophioninae, a cladistic study of the group is necessary before much of the
distributional data available can be interpreted in an evolutionary manner. The aim of the
present study is to investigate the phylogeny of the Ophioninae using a variety of modern
cladistic techniques, and to relate the results to what is known about the distribution and biology
of the group in general.
Terminology
The morphological terminology in this work follows that proposed by Richards (1956) and
interpreted for the Ophioninae by Gauld & Mitchell (1978; 1981). Specialist terms, relevant to
only some ophionines, are defined in these works. Family-group names used conform with the
directives of the International Code of Zoological Nomenclature and with various Opinions of
the International Commission. Some recent authors (e.g. Townes, 1969; 1971) have chosen to
disregard certain of these opinions and have not followed the Code when forming family-group
names (see Fitton & Gauld, 1976).
Material examined
The majority of specimens examined are contained in the collections of the British Museum
(Natural History) (BMNH), but valuable additional material was furnished by the Australian
National Insect Collection (ANIC), Canberra, the Bernice P. Bishop Museum (BPBM),
Hawaii, the Canadian National Collection (CNC), Ottawa, the Gupta Collection (GC), the
Musee Royal de 1'Afrique Centrale (MRAC), Tervuren, the Museum National d'Histoire
Naturelle (MNHN), Paris, the Taiwan Agricultural Research Institute (TARI), Taichung, the
Townes Collection (TC), Ann Arbor, the United States National Collection (USNM),
Washington and the Zoological Institute (ZI), Leningrad. The types of virtually all Old World
species have been examined and exhaustive lists of material examined are contained in recent
revisionary studies (Gauld, 1977; Gauld & Mitchell, 1978; 1981; Gauld & Carter, 1983).
Discussion of methodology
The present work is an attempt to elucidate the phylogeny of the Ophioninae and therefore only
phylogenetic methods of data analysis have been used. These methods are based on ideas
initially expounded by Hennig (1966). Although Hennig made a major contribution to system-
atic philosophy, he greatly underestimated the difficulty that would occur in phylogenetic
analyses due to morphologically undetectable evolutionary parallelism and character-state
reversal (i.e. homoplasy). Such events result in there being incompatible character sets in
taxonomic data. An estimation of the extent of incompatibility can be made using the simple test
64 I. D. GAULD
outlined by Le Quesne (1969) (see also Gauld & Mound, 1982; Underwood & Gauld, in prep.),
and such incompatibilities are far from uncommon in most real data.
In the past decade two main approaches have been adopted by cladists in an attempt to resolve
conflicting character sets - parsimony methods, which seek to minimize the number of
character-state transformations, and compatibility methods, which seek to find a series of
nesting groups supported by the largest number of compatible characters (Felsenstein, 1982).
For simple data sets where there is a low incidence of homoplasy, these methods yield very
similar results (e.g. Gauld, 1983), but as the frequency of homoplasy increases so the methods
usually give increasingly different results.
Parsimony methods have gained wide acceptance in the literature, with the shortest rooted
Wagner tree often being uncritically presented as the best phylogenetic hypothesis. Simul-
taneously, many authors have criticized compatibility analysis either for producing numerous
apparent best solutions (Kluge, 1976), or as being less useful at producing congruent phy-
logenies from different developmental stages (e.g. Mickevich, 1978; Rohlf & Sokal, 1980) (but
see also Rohlf etal., 1983). However, Felsenstein (1981; 1982) stated that both methods can be
justified as maximum likelihood methods, but under somewhat different circumstances. If
homoplasy is expected to be scattered at random throughout all characters then a parsimony
method is favoured, but if homoplasy is expected to be concentrated in certain characters then
compatibility is supported. Felsenstein (op. cit.) continues by noting that both methods require
both homoplasy to be rare and characters to have a low rate of change. In the case of
Ophioninae, homoplasy demonstrably is not rare, suggesting results obtained by either method
should be viewed circumspectly. The additional assumption necessary for parsimony methods to
approach maximum likelihood, that is that homoplasy is randomly scattered across all charac-
ters, contradicts the consensus of opinion of most practising entomological taxonomists. Both
parsimony and compatibility methods were used, with caution, in this study.
Choice of parsimony method utilized
Of the parsimony methods available at the start of this study, the program selected was PHYLIP
(package for inferring phytogenies) written by Dr J. Felsenstein. This program offers four
routines for dealing with non-polymorphic, discrete state data: (a) the Camin-Sokal parsimony
method; (b) the Dollo parsimony method; (c) the Wagner parsimony method and (d) a mixed
method allowing a, b or c to be specified for each character. The Camin-Sokal and Dollo
methods were not used in the present study. Neither gives as short a tree as that obtained by the
Wagner method as both place additional constraints on the 'tree-growing' method (Felsenstein,
1982), the former by not allowing reversal, the latter by not permitting forward parallelism.
In the insects being studied there is virtually no evidence at all for either of the additional a
priori assumptions these methods necessitate. Indeed there is some biological evidence to
suggest that these extra assumptions are unwarranted for ichneumonids. Dollo parsimony may
justifiably be invoked for treating complex structures (such as the vertebrate eye), but virtually
all characters used in the present study involve small changes in simple structures. Very often the
derived state involves reduction of a structure, and observation of other ichneumonid taxa
suggests that certain of these apomorphies (such as reduction of extent of occipital carina) have
been developed in parallel in numerous different evolutionary lineages. Even when the
apomorphic state is the development of a novel structure (such as an alar sclerite), the evidence
strongly suggests (as the feature occurs in a few otherwise specialized species of several different
genera) parallel development in closely related lineages.
Reversal is more difficult to demonstrate, but clearly it does occur. For example, the alar
sclerites apparently have been lost by some Enicospilus species on oceanic islands (Gauld &
Carter, 1983) and the posterior transverse carina of the propodeum, which is absent in most
Enicospilus species, has been redeveloped in some members of the otherwise highly specialized
E. signativentris species-complex (Gauld & Mitchell, 1981).
Because of the objections to the Camin-Sokal and Dollo methods the option employed in this
study was that which necessitates no extra a priori assumptions, the Wagner method. Clado-
PHYLOGENY OF THE OPHIONINAE 65
grams were rooted using a hypothetical all zero ancestor, as use of any outgroup taxon is liable to
introduce more incidences of homoplasy.
The advantages and disadvantages of the parsimony method
The major advantage of the parsimony method seemed to be that, because of its 'averaging
procedure' (see below), generally all species of a particular genus were associated. As ophionine
genera are polythetic, compatibility methods invariably exclude certain taxa (see Davies &
Boratyriski, 1979).
Despite its widespread acceptance there are serious flaws in parsimony methodology when
applied to complex data sets. (There are also philosophical objections (see Pratt, 1972;
Felsenstein, 1981; Friday, 1982; Panchen, 1982), but these are beyond the discussion of the
present work.) One, and perhaps the most obvious methodological flaw, is that because the
method considers each character as an identical piece of binary information, and is attempting
only to minimize transformation, a large set of coincidental 'bad' characters will be favoured at
the expense of even a very slightly smaller set of 'good' characters. This is best illustrated in the
case of some Hawaiian genera of ophionines (taxa 924-6 in the following study) and Ophio-
gastrella (taxa 910-11). In both cases, the sets of characters responsible for positioning these taxa
in the Wagner analysis are composed of characters that are likely to be highly homoplastic in
other taxa. Although a traditional taxonomist intuitively recognizes these as poor characters
(and taxonomist's intuitive judgement may be very good, see Davies, 1981) and has no
confidence in them as indicators of phylogenetic affinity, it was not found to be possible to
translate this subjective bias into an objective taxonomic weighting scheme. Without an
objective weighting scheme the averaging procedure adopted by the Wagner method has the
effect of reducing the number of transformation steps of many 'weak' characters at the expense
of 'moderately good' ones (Strauch, 1984).
Although some hybrid parsimony/compatibility trees produced in this study were longer than
the shortest Wagner tree (mainly because certain characters appear to have been 'written-off,
i.e. allowed to have a large number of transformations) it is noteworthy that these cladograms
were supported by more characters with minimal or nearly minimal homoplasy. This can best be
illustrated by reference to two alternative cladograms produced during this study, the most
parsimonious 235-step one and the favoured 239-step one (Gauld, unpubl. PhD). The numbers
of characters with various numbers of transformation steps are shown in Fig. 1.
A second disadvantage of the parsimony method is that a multiplicity of almost equally
parsimonious cladograms may be generated from the same data set by the same procedure
(Strauch, 1984), but altered merely by re-ordering the input order of the taxa (e.g. Figs 21-24).
In this study it soon became apparent, when rearrangements are practised, that certain taxa are
prone to 'hopping', i.e. moving from one lineage to another in different reconstructions. Taxon
906 was found to be such an OTU - its final position in a cladogram depended solely upon the
position in which it was entered into the analysis. This was presumed to operate in the following
manner: taxon 906 is almost equally associated (see Appendix 3) with three separate lineages,
Simophion, the Thyreodon complex and the Enicospilus complex. This equality of association is
so pronounced that once the taxon has linked to one or other of the alternatives (the first
encountered) it does not share enough derived features with any other taxon to disassociate
itself. Thus if the first three taxa entered are 905 (Simophion), 906 and 940 (a more distantly
related taxon), 906 remains associated with 905, whilst if the first three taxa entered are 935 (one
of the Thyreodon complex), 906 and 940, 906 remains associated with the Thyreodon complex.
Some programs attempt to circumvent this problem by computing an 'advancement index' for
ordering the taxon input, but I fail to see the intellectual justification for this.
A third disadvantage of the parsimony method is that it is not possible to predict, from any
given data set, the actual minimum tree length (Felsenstein, 1982; Day, 1983). This means that,
for large data sets, in practice one is guessing that the minimum length tree obtained is the
shortest possible. This study shows that for highly homoplastic data, a number of equally short,
quite different cladograms may be obtained. It appears to be merely an act of faith that a slightly
66
I. D. GAULD
Number of
characters
with N
transform-
ation
steps
OQ _
25-
20-
15-
10-
5-
239 step cladogram
235 step cladogram
/ A \
;\
I \
is
. \
i \
\
\\
\ '
\
\
\
+
+
4-
12345678 9
number of transformation steps, N
Fig. 1 A comparison of the incidences of transformation required to fit all characters to the most
parsimonious 242-step cladogram and a favoured 249-step cladogram (From Gauld, unpubl. PhD
thesis). It is noteworthy that the latter, although 7 steps longer, is supported by more minimally
homoplastic characters.
shorter and yet totally different arrangement does not exist, a rather disconcerting observation if
one accepts the principle of parsimony as paramount in phylogenetic reconstruction.
Compatibility methods utilized
Felsenstein's package, PHYLIP, provides an option for producing compatible character sets.
This option, CLIQUE, was used to find the largest cliques for various sets of taxa considered in
this study. A second program (written by Dr G. Underwood) was developed by Underwood and
Gauld during the course of this study. This program was developed from the work of Le Quesne
(1969; 1972) and involves calculating the probability of incompatibility between two characters
on a null hypothesis of random distribution of states of both. This value was then used to
calculate for each character a coefficient of character-state randomness (herein called the O/E
value). Typically the output of this program, LEQU.BAS, consists of a list of characters with
their observed incompatibilities, calculated expected number of incompatibilities, and O/E
PHYLOGENY OF THE OPHIONINAE 67
values tabulated; the characters are ranked by O/E value as a postscript (e.g. Tables 2, 6). A
facility has been incorporated so that multistate characters which have been coded in binary
fashion and numbered as decimal increments (e.g. 1-1, 1-2), and cannot logically fail, are not
compared.
Included also in LEQU.BAS is a labelling feature. This is based on an idea given by Guise et
al. (1982) for counting frequencies of each of the four possible character-state combinations
(0,0; 0,1; 1,0; 1,1) for each character pair. If a pair of characters fail the test (because all four
combinations are found in the data) on account of a single occurrence of one combination, the
taxon having this combination is labelled. The results of this operation are printed in the form of
a table (e.g. Appendix 4). The figure in brackets adjacent to the taxon is the total number of
labels the taxon has received. The taxa most frequently labelled will be the ones with the most
discordant character sets. A particularly high label score for a species for any one character
strongly suggests homoplasy for that character with respect to other taxa in the set. Some care is
needed in the interpretation of results as characters with very unequal character-state distri-
bution (either only two or two 1 scores) often have high label values. These are easily
recognized as equally high scores are given for the two taxa with the minority states. Even these
values can sometimes be seen to be unusually high (i.e. suggestive of homoplasy) if the character
label values are compared with the label values of similarly uninformative characters.
Value of LEQU.BAS for overcoming some problems of compatibility analysis
Two major disadvantages with compatibility analysis are: (1) a maximum character set is often
far too small to allow full resolution of data (Felsenstein, 1982) and (2) the data set may yield a
number of more or less equally large cliques (Kluge, 1976). LEQU.BAS is of some value in
overcoming both these problems.
As the largest compatible character set is generally very small, it is useful to have a way of
ordering characters. Homoplastic characters that would be eliminated from any compatible set
are not all equally bad; some are far more discordant than others. For example, allowance of just
one incidence of parallelism or reversal may be enough to make certain characters compatible
with a cladogram, whereas others will necessitate the invocation of multiple incidences of
homoplasy to achieve congruence. It can be seen from the hypothetical data set 1 (Tables 1-3)
that LEQU.BAS offers a way of grading characters from slightly to extremely homoplastic.
Eight taxa, A-H, exhibit 18 binary characters (Table 1). Characters 1-12 support the nested set
(((AB)C)((DE)(F(GH)))). Characters 13-18 are differentially homoplastic, 13 and 14 requiring
one extra transformation to fit the specified cladogram, 15 and 16 requiring two extra
transformations and 17 and 18 needing three extra. Analysing these data with LEQU.BAS gave
the results presented in Table 2. It can be seen that the six homoplastic characters are ranked in
order of the extra number of transformations required to make them fit the specified cladogram.
Several taxa are highly labelled for certain characters, suggesting homoplasy. For example,
character 15, which is present in the apomorphic state in B and C, but is plesiomorphic in A, is
highly labelled for A, suggesting a reversal may have occurred.
Table 1 Character state matrix for hypothetical data set 1. A-H represent taxa, 1-18 their independent
characters. = plesiomorphic, 1 = apomorphic.
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18
A 111100000000110110
B 111100000000111101
C 001100000000101001
D 000011110000010010
E 000011110000010001
F 000011001100001111
G 000011001111101010
H 000011001111101101
68
I. D. GAULD
Table 2 Result of LeQuesne test on hypothetical data set 1 . The number in italics is the character number ,
the second column is the number of observed failures, the third column the number of expected failures
and the final column in each case the ratio of observed over expected failures. Characters with a low O/E
value are considered to be good characters, and are ranked in the final line higher than those with high
O/E scores.
Character number: failures observed expected O/E ratio
1
3
8.68
0.35
.2
3
8.68
35
.3
5
12.21
0.41
4
5
12.21
0.41
1
5
12.21
41
5
12.21
0.41
7
2
8.68
0.23
a
2
8.68
23
2
4
12.21
0.33
12
4
12.21
0.33
jj_
3
8.68
35
I?.
3
8.68
0.35
J_3
7
12.21
0.57
i*
8
13.09
61
J_5
10
12.21
0.82
Ii
13
13.09
0.99
\j_
16
13.09 1
22
IS
16
12.21
1.31
Grand totals- f
n lures
observed
expected
O/E
ratio
57
100.63
0.57
Ranking ratios
7 8 9 10 1 2
11 12 3 4 5 6 13 14 15 16 17 18
Although there is a clear relationship between the number of extra transformations required
to make a character 'fit' a favoured cladogram and that character's O/E value, the O/E value also
varies with information value of the character, i.e. the number of 1 scores in relation to scores.
For example, in data set 1, character 14 may be modified so it has the following distribution of 1
states in taxa A-H, 11100000, 11001000 or 10001001. These require one, two and three
transformations, respectively, and have O/E values of 041, 0-90 and 1-24. However, if the
number of transformations necessary is kept constant (at say two) and the number of derived
states varied, the following relationship may be observed: 10001000 (O/E = 1-41), 11001000
(O/E = 0-90), 11011000 (O/E = 0-61), 11100010 (O/E = 0-53), 11100011 (O/E = 0-50),
11100111 (O/E = 0-23). Similar variations of character 16 (requiring three transformations)
produced values with the range 0-74-1-24 (the largest being for 10010001), whilst a range of
1-16-1-35 was found for a character requiring four transformations. Thus a homoplastic
character with a high proportion of derived states supporting various subgroups in the definitive
cladogram may have a lower O/E value than a less homoplastic, but less informative character.
Kluge (1976) pointed out that compatibility analysis may produce a large number of equally
large cliques. From his data set of 139 binary characters he obtained six almost equally large
cliques of 80 or 81 characters. Kluge remarked that not only is there no reason to prefer one to
any other of the sets, but also pointed out that the most primitive species in each of the cliques
did not correspond to the most primitive as assessed on the basis of the best documented
characters. To consider Kluge's objections necessitates re-examining his data. Of his six largest
cliques, it is apparent that no less than 76 characters are common to all sets. The six dendrograms
supported by these cliques are essentially similar, differing mostly in the relative order of the
most primitive taxa (im-ti in Kluge's fig. 7a) so a great deal of information is common to all, and
Table 3 Labels matrix derived from hypothetical data set 1 . The column to the right of the taxon letter is
the total number of times that taxon is solely responsible for a character failing the LeQuesne test. For
further details see text.
A (21) 333333
B ( 6} 3 3
C ( 8) - -
D ( 5) - -
E ( 7) - -
F ( 5) - -
G (12) - -
H ( 8) - -
9 10 11 12 13 14 15 16 17 18
- - - - 1 -
2
222-------4
c
4
<
. A
-
_ _ -22----1-
- -
2
.' J
-
_ _ _ 2 2 - - - - 1
2
4
3
-
- - - - - 1 1 - - 5 1
1 1
-
-
-
-----2233-1
1 4
2
6
.
-----1133--
2
4
2
PHYLOGENY OF THE OPHIONINAE 69
this agrees with Kluge's preferred phylogenetic arrangement. Most of the differences between
cliques (and between the dendrograms supported by these and the preferred arrangement)
involve relatively uninformative characters. It is mathematically demonstrable that the prob-
ability of chance compatibility of a particular character with any other set of characters is
inversely proportional to the information value of the character. Therefore, one criterion of
which clique to select is to favour that with the highest information value.
However, in practice one frequently finds that the subsets which are not common to all the
similar-sized cliques comprise equally poorly informative characters. In such cases the average
O/E value for a clique is a useful indicator of the degree to which the clique is consistent with
slightly homoplastic (and hence excluded) characters which are not normally considered in a
compatibility analysis. For example, consider the hypothetical data set 2 (Tables 4, 5). There are
two equally large, equally informative cliques:
A (10) [1, 2, 3, 4, 5, 6, 7, 8, 9, 10] Average O/E value = 0-308
B (10) [1, 2, 3, 4, 5, 6, 7, 8, 9, 14] Average O/E value = 0-323
A DB = (1,2, 3, 4, 5, 6, 7, 8, 9}
Table 4 Character state matrix for hypothetical data set 2. Conventions follow table 1.
1
2
3
4
5
6
7
8
9
10
11
12
13
14
J
1
1
1
1
1
1
1
1
1
K
1
1
1
1
1
1
1
1
L
1
1
1
1
1
1
M
1
1
1
1
1
N
1
1
1
1
1
1
1
P
1
1
1
1
1
1
Q
1
1
1
1
1
1
Table 5 Results of LeQuesne test on hypothetical data set 2. Conventions follow table 2.
Character number: failures observed expected O/E ratio
li
3
9,
,86
0,
,30
2 :
3
9,
,86
0,
.30
4 :
3
9,
,86
0,
,30
_5 :
3
10
,63
0,
,28
I :
3
9,
,86
0,
,30
1 ;
3
9,
,86
0,
,30
12 :
1
6,
,89
0,
,15
ii :
11
9.
,86
1,
,12
13 :
12
10,
,63
1,
13
14 :
4
6,
,89
0,
,58
_3 : 3 9.86 0.3
6^ : 3 10.63 0.28
2 : 2 6.89 0.29
12 : 12 10.63 1.13
Ranking ratios
10 6 5 9 1 2 3 4 7 8 14 11 12 13
These sets differ in one character, possession of either 10 or 14. Character 10 supports J + K
whereas 14 supports J + M. No other character that is compatible offers support for one or other
of these alternatives but some homoplastic characters do support J + K (plus other taxa also)
reinforcing the J -I- K grouping suggested by character 10. No homoplastic character reinforces
the J + M suggestion. This reinforcement can be detected by differences in the values of the O/E
ratio, as character 10 scores 0-15 whereas character 14 scores 0-58. This difference in O/E value
is responsible for the difference in average O/E values of the two cliques A and B, and it thus
seems reasonable to prefer the clique with the lowest average O/E value. This method offers a
simple objective procedure for selecting one of a number of large cliques as 'most favourable' , as
it will be the one reinforced by most secondary characters.
In the present study compatibility analysis was found to give, up to a point, results that were
similar to conventional taxonomic assessment. Frequently, characters considered important in
generic classification by traditional taxonomists were found to be the intersect of the largest
cliques.
70 I. D. GAULD
The operational procedure adopted in this study
This study was commenced by tentative recognition of groups of taxa from the shared derived
character matrix (Appendix 3). The rigidity of these groups was tested using Underwood's
FOURS program (Underwood, 1982) and then further analyses using the compatibility and
parsimony methods outlined above. Resulting cladograms were compared and differences
subjectively evaluated by appraisal of characters supporting different options. The various
groups were gradually related to each other until the majority of taxa had been placed. The
remaining unplaced taxa all exhibited confusing affinities. To facilitate their placement a
number of hypothetical taxonomic units were proposed to represent the various genera already
placed in the analysis. The data set comprising hitherto unplaced taxa and HTUs was analysed
using parsimony and compatibility techniques, and the resulting cladograms subjectively
evaluated.
An apologia for subjectivity in cladistic analysis
In recent years there has been a great deal of intellectual activity devoted to making taxonomy
more 'scientifically respectable' by attempting to remove subjectivity. Phenetic methods were
the first to be claimed as objective taxonomic procedures (Sokal & Sneath, 1963), but these
methods were essentially a movement away from evolutionary taxonomy (Nelson & Platnick,
1981) and have largely been eclipsed by the development of numerical cladistics. Despite the
seeming objectivity of such numerical methods, they are really only objective ways of analysing
largely subjective data, as character selection is a highly subjective procedure (Pratt, 1972).
Furthermore, the claim by many cladists that their favoured numerical methods (parsimony
analyses) are scientific (in the Popperian sense) is flawed by the assumption that parsimony per
se is a criterion for formulation of a scientific hypothesis (i.e. the least falsified hypothesis is
accepted - Gaffney, 1979), as is eloquently shown by Panchen (1982). A phylogenetic
hypothesis derived from a cladistic analysis rests on the supposition (or hypothesis) that a
particular derived character-state is a genuine synapomorphy which thus indicates commonality
of descent of two or more taxa. In a highly homoplastic data set, for every character that is
accepted as a genuine synapomorphy, others are rejected as showing false patterns. Farris
(1969) defends parsimony in such cases, on the grounds that hierarchic correlations are more
likely to occur among cladistically reliable characters than among cladistically unreliable
characters (i.e. random variables). This is true only if cladistically unreliable characters vary
randomly and is not a corollary of the generally accepted statement that random variables are
cladistically unreliable characters. Clearly it is untrue; most taxonomists can cite examples of
suites of characters being size related, habitat related, associated with a particular biological
strategy and so forth. How then is a taxonomist supposed to choose between competing
character sets if not by 'biggest is best' criteria? I think this can only be done at present,
subjectively. Competing character sets are examined in the light of a taxonomist's knowledge of
the variability of characters over a much wider group of organisms. Highly variable features
(such as the number of ribs or vertebrae in some birds - Strauch, 1984 - or the form of
mammalian teeth - Butler, 1982) are rejected as unlikely to be indicative of phylogenetic
affinity. Other sets clearly comprise adaptive features associated with certain habitats and these
may likewise be rejected. It is noteworthy that very many cladistic analyses seem to have been
undertaken on data sets that have initially been subjectively edited (with no explanation) by
discarding characters which are presumably thought to have no phylogenetic value. In this work
I have tried to explain my reasons for rejecting certain character sets in preference to others.
The holophyly of the subfamily
A prerequisite to meaningful cladistic analysis is the establishment of the holophyly of the group
being studied. The holophyly of the Ophioninae is supported by the following apomorphies.
(a) Possession of a spurious vein extending from the vannal notch to the tornus of the fore wing.
This feature, first recognized by Perkins (1959), is an autapomorphy of the subfamily.
PHYLOGENY OF THE OPHIONINAE 71
(b) Possession of numerous setae on the labial sclerite of the final instar larva (Short, 1978).
This feature is also an autapomorphy of the subfamily.
(c) Presence of a single radio-medial cross vein (? 3r-ra) far distal to 2m-cu. This unusual
condition is elsewhere only found in isolated genera of the Campopleginae, Tersilochinae and
Anomaloninae (Townes, 1970; 1971; Gauld, 1976). In none of these taxa is the cross vein as far
distal to 2m-cu as is the case in ophionines.
(d) Possession of a sinuous pectinal comb on the tarsal claws. Many ichneumonids have a few
scattered pectinal teeth on the tarsal claws but very few have a strongly developed sinuous comb
like that found in virtually all ophionines.
(e) Total loss of glymmae. These structures are present in at least some members of virtually all
other subfamilies. No trace has ever been observed in ophionines.
(f) Enlargement of ocelli and related nocturnal specializations (Gauld & Huddleston, 1976).
The pale colour, enlarged ocelli and eyes, long antennae and slender legs are features found in
many nocturnal ichneumonoids, but this suite of characters is apparently an apomorphic feature
of the Ophioninae.
The systematic position of the subfamily
The subfamily Ophioninae belongs to the Ichneumonidae, a holophyletic group of apocrite
Hymenoptera (Konigsmann, 1978). Virtually nothing has been published about the phylo-
genetic relationships of ichneumonid taxa and, if a reasonable assessment is to be made of
character polarity in Ophioninae, it is necessary to attempt to place the group in relation to some
other subfamilies.
From study of Cretaceous fossil ichneumonoids (Townes, 19730 & b; Rasnitsyn, 1983),
comparison with putative ancestral groups (such as siricids and cephids) and examination of
primitive extant Apocrita (e.g. Megalyridae, Stephanidae, some Braconidae), it can be deduced
that primitive ichneumonpid features are likely to include: possession of an un-notched
(possibly long) ovipositor; possession of a gaster that is broadly attached to the propodeum and
has a free first sternite; possession of an areolated propodeum; possession of notaular grooves;
possession of simple or basally lobate claws. Study of the biological evolution of parasitoids (e.g.
Shaw, 1983) suggests that the ectoparasitic habit is primitive with respect to endoparasitism, and
as both types occur in the Ichneumonidae (Clausen, 1940; Gauld, 19846), one expects that the
most primitive taxa will be found amongst ectoparasitoids. Some authors (e.g. Cushman, 1926;
Telenga, 1969; Achterberg, 1976; 1984) suggest that primitive ichneumonoids are likely to have
been parasites of xylophagous coleopterous larvae, as are archaic extant parasitoids such as
orussids (Quinlan & Gauld, 1981; Middlekauf, 1983). These data suggest that the most primitive
extant ichneumonids are to be found amongst the Pimplinae, Labeninae and perhaps also
Tryphoninae. This hypothesis is supported by studies of larval morphology (Short, 1978) and is
concomitant with the informal higher classificatory scheme outlined by Townes (1969).
Compared with the Pimplinae and Labeninae, the Ophioninae appears to be a rather
specialized subfamily. It seems to belong to a large, holophyletic group of taxa that includes the
subfamilies Ophioninae, Campopleginae, Cremastinae, Tersilochinae, Banchinae and Cteno-
pelmatinae. The holophyly of this group is suggested by the following apomorphies:
(a) possession of a dorsal subapical notch on the ovipositor;
(b) possession of similar female reproductive tract (Pampel, 1913);
(c) having endoparasitic larvae which lack a labral sclerite and possess a Y-shaped prelabial
sclerite (see also Fig. 2).
Two of the included taxa, the Ctenopelmatinae and Banchinae, usually possess a plesio-
morphic first gastral segment which is broadly attached to the propodeum, has a more or less
centrally positioned pair of spiracles and a free sternite. In the Ophioninae, Campopleginae,
Cremastinae and Tersilochinae this segment is specialized. It is more slender, has the tergite and
sternite intimately fused and is lengthened anteriorly so that the spiracles are nearer to the
posterior end. Furthermore, all these taxa possess apomorphic, laterally compressed gasters,
suggesting they constitute a holophyletic clade.
m
m
14
L5
Fig. 2 Cladogram showing putative phylogenetic inter-relationship of the ophionoid group of ichneu-
monid subfamilies. The apomorphic features supporting this arrangement are: 1 , possession of a fringed
clypeus ; 2 , fusion of Rs with M in central part of fore wing ; 3 , development of sclerotized bridge between
tibial spurs; 4, possession of striae on tergite 2 of gaster; 5, loss of distal abscissae of veins in hind wing; 6,
fusion of areae superomedia and petiolaris; 7, development of spurious vein in fore wing; 8, loss of 2r-m
in fore wing ; 9 , enlargement of pterostigma ; 10 , reduction of length of Rs in hind wing ; 1 1 , development
of sinuous apex to ovipositor; 12, possession of expanded larval labial sclerite; 13, development of
tubular petiole with reduced glymmae (note this is developed in parallel in some tersilochines); 14,
gaster laterally compressed; 15, petiolar spiracle near hind end of tergite; 16, tergite and sternite 1
intimately associated; 17, possession of a dorsal subapical notch on ovipositor; 18, possession of a
similarly modified female reproductive tract (see Pampel, 1913); 19, endoparasitic larva with a Y-shaped
prelabial sclerite.
PHYLOGENY OF THE OPHIONINAE 73
Within this clade two apparent groups are recognizable. One comprises the Tersilochinae +
Cremastinae and is supported by two apomorphies, possession of an enlarged pterostigma and
strongly shortened vein Rs in the hind wing. In addition both these taxa share a number of
developmental trends (underlying synapomorphies of Saether, 1979), most notably, the de-
velopment of a sinuous ovipositor apex and the presence of dorsally convergent eyes in males.
The second group, Ophioninae + Campopleginae, is supported by a larval specialization, the
possession of an expanded labial sclerite. The petioles of these two taxa are similarly modified,
although this development appears to have been paralleled in some tersilochines (primitive
tersilochines have a more 'cremastine-like' petiole). Most ophionines and campoplegines
possess an apomorphic short, straight ovipositor; even when it is long it is very similar in the two
groups, being robust and up-curved.
For the purposes of this study the Ophioninae is considered to be the sister-group of the
Campopleginae, and the Campopleginae + Ophioninae is treated as the sister-group of the
Cremastinae + Tersilochinae (Fig. 2).
Polarity determination and character coding
The polarity of the majority of characters has largely been determined by the method of
out-group comparison (Watrous & Wheeler, 1981) using as out-groups the taxa mentioned
above. Unfortunately this method does not work for all characters, usually because both states
occur in both the group under study and the out-groups. In these cases the polarity assignment is
based on unsatisfactory criteria such as common equals primitive, but this is stated in the
discussion.
In many cases in the Ichneumonidae, character-states are progressive steps in the loss of
ancestral features such as the occipital carina, propodeal carinae or segments of the palp. Such
characters can be arranged in simple transformation series, from plesiomorphic to the most
derived state. For example, -for maxillary palps - 5-segments to 4-segments to 3-segments. Such
transformation series may simply be scored in binary form as two characters -
8-1 8-2
maxillary palp 5-segmented
maxillary palp 4-segmented 1
maxillary palp 3-segmented 1 1
where = plesiomorphic and 1 = apomorphic. Hence 8-1 represents reduction from 5 to 4
segments and 8-2 further reduction to 3. Such characters are obviously not independent for it is
impossible to have a 0,1 coding as, if extreme reduction is observed, it is assumed that an
intermediate stage has been passed through.
In other cases a particular structure may have been modified in one of several ways. For
example, in the Ichneumonidae generally the mandible is bidentate equally, so in the Ophio-
ninae teeth of the same length must be considered a plesiomorphic feature. Although the lower
tooth is fairly constant in size, the upper may either be lengthened or shortened. Such a bifurcate
series can be scored as two binary characters -
4-1 4-2
mandible equally bidentate
upper tooth enlarged 1
upper tooth reduced 1
Thus character 4-1 represents enlargement of the upper tooth, whilst 4-2 represents reduction.
Clearly a 1,1 coding cannot exist for such character pairs. All transformation series in the
characters dealt with below have been treated in similar fashion.
From the very large number of characters exhibited by the Ophioninae a number have been
selected which show the greatest range of variation between the different genera. Included are
all characters previously considered to be diagnostic of genera, and a number of additional
features that have been found to be useful in delineating species-groups (e.g. in Gauld &
74 I. D. GAULD
Mitchell, 1978). Many other characters, features in which a single species differs from others in a
species-complex, were excluded from this analysis. These characters, such as the shape of the
alar sclerites, form of the microsculpture of the alitrunk, density of hairs on the wing surface,
general colour pattern, relative length of the mid tibial spurs etc. , show a considerable range of
variation within the species of any single genus and, in most cases, this range of variation is
repeated in many genera. Such characters were considered to be unlikely to make significant
contribution to resolving the phylogeny of the genera.
Characters used in the study
In the following section considerable space has been allocated to explaining the rationale
involved in character scoring and polarity determination. Although this practice is not usual in
all cladistic studies (characters are frequently relegated to an appendix) it is considered to be
important in the present work. It is upon these scores that the results of any numerical analysis
depend; consequently, this section is seminal to the entire work.
The characters and their polarity
1-1, 1-2 Occipital carina. A complete occipital carina is plesiomorphic (0,0). It may be centrally
interrupted (1,0) or absent (1,1).
2 Mandibular axis. The plesiomorphic condition is for the axis of articulation of the mandible to be at 90 to
the longest axis of the head. In a few ophionines the head is narrowed and the mandibular axis tilted
almost into the vertical plane (1).
3 Mandibular swelling. A flat outer mandibular surface is plesiomorphic for ichneumonids; the swelling
present near the mandibular base of some ophionines is apparently a derived feature (1).
4-1, 4-2 Mandibular teeth. The relative lengths of the teeth vary considerably in ichneumonids, but equally
bidentate appears to be the plesiomorphic state (0). In ophionines the alternative specializations are -
upper tooth elongate (1,0) and upper tooth reduced (0,1).
5-1,5-2 Torsion of the mandibles. The plesiomorphic condition for ichneumonids is to have the teeth of the
mandible aligned in the same plane as the mandibular axis (Figs 38, 39). Many ophionines have the
mandible twisted from 5-50 so that the lower tooth is directed forwards (Fig. 40). In a few species the
mandible is exceptional in being twisted more than 70 so that, when closed, the lower tooth occludes
the internal, upper one. This torsion is here considered a serial development, from untwisted (0,0)
through slightly twisted (1,0) to exceptionally twisted (1,1). The selective advantage of this torsion is
not clearly understood, but it is suggested that it is important in facilitating egress from certain types of
cocoon. It is noteworthy that not only do all species of the huge and successful genus Enicospilus have
such mandibles, but the totally unrelated tryphonine genus Netelia (which is also nocturnal and has a
similar host range) also has twisted mandibles.
6-1, 6-2 Shape of the clypeal margin in anterior aspect. The shape of the clypeus is often characteristic of a
group of ichneumonids. For example, in the pimplines it is usually bilobate, in the anomalonines
convex with a central tooth, whilst in the ichneumonines it is usually flat and truncated. The weakly
convex condition found in many ophionines, campoplegines and cremas tines is considered plesio-
morphic. In the Ophioninae the clypeus may be modified one of two ways - either being strongly
concave (1,0), or centrally produced and pointed (0,1).
7-1, 7-2 Clypeal profile. It is difficult to assign polarity to this character, but the widespread condition in the
subfamily and in the closely related out-groups was assumed to be plesiomorphic. This is the
possession of a virtually flat clypeus (0,0) (Fig. 39). The clypeus may be modified in one of two ways -
either flared outwards (1 ,0) (Fig. 44) or with a groove present parallel to the margin, so that the actual
margin is sharp (0,1) (Fig. 38).
8-1, 8-2 Number of maxillary palp segments. The plesiomorphic condition for the Hymenoptera is
6-segmented palps, but in the Ichneumonidae the number is reduced to 5. Fusion may occur between
the distal two or three segments leading to a reduction in apparent segment number to 4 or 3. This
character is scored as a transformation series, 5-segmented being 0,0, 4 being 1,0 and 3 being 1,1.
9 Shape of central segments of maxillary palp. In most primitive ichneumonids, virtually all members of the
out-groups and many ophionines, the maxillary palp segments are slender and elongate. This is
considered to be plesiomorphic. In a few ophionines the central segments are specialized, being
globose.
10 Maxillae. The maxillae of most ichneumonids resemble those of the more primitive mandibulate
PHYLOGENY OF THE OPHIONINAE 75
hexapods (Richards, 1956). A few ophionines have the maxillae specialized, unusually elongated so
that the galea projects below the mandibles. This apomorphic condition is found in species inhabiting
dry areas and is presumed to be an adaptation to feeding from certain flowers.
11 Labium. Like the maxillae, the labium is remarkably unspecialized in most ichneumonids. A few
eremic ophionines possess greatly elongated glossae, an apomorphy that presumably has a similar
function to character 10.
12 Ocelli. The majority of ophionines exhibit a set of features, the so-called ophionoid facies (Gauld &
Huddleston, 1976), common to nocturnal Hymenoptera. These features are here regarded as
apomorphies of the entire Ophioninae. They include possession of greatly enlarged ocelli. The
majority of species in the out-groups, being diurnally active, lack this specialization. Consequently the
possession of enlarged ocelli must, within the Ophioninae, be regarded as a plesiomorphy. In a few
areas where competition from other ichneumonids is low, e.g., remote islands, deserts, tops of high
mountains, several ophionines have adopted a diurnal habit. This is presumed to be a secondary
feature as several species retain some nocturnal features. However, these diurnal species have small
ocelli, and in these cases small ocelli are considered apomorphic. Polarity determination of this
character can be questioned as it is based on an a posteriori inference.
13 Frontal grooves. Most ichneumonids, including almost all members of the out-groups, have the lateral
part of the frons flat. In some ophionines a groove is present on either side, parallel to the inner orbit.
Possession of this groove is regarded as a specialization.
14-1, 14-2 Flagellum length. In most ichneumonids the flagellum is setaceous and consists of a large
number of more or less identical segments (flagellomeres). In the majority of ophionines there are
between 45 and 65 such segments, and the flagellum is about as long as the fore wing. This is
considered, because of its common occurrence within the group, to be the plesiomorphic condition.
The flagellum appears to be modified in one of two ways. Either the segments may be very short and
transverse (scored 1 ,0) or the number may be greatly increased (to between 75 and 95) so the flagellum
is very much longer than the fore wing (0,1). A short flagellum is found in many species inhabiting hot,
dry areas, and is perhaps an attempt at reducing an evaporative surface. A very long flagellum is found
in species inhabiting humid areas, particularly rain forest canopies. Similarly shortened or elongated
antennae can be observed in other groups, e.g. the Anomaloninae (Gauld, 1976).
15-1, 15-2 Spiracular sclerite. In most ichneumonids, including the majority of species in the closely related
out-groups, the spiracular sclerite is exposed and is clearly visible near the upper hind corner of the
pronotum (Fig. 42). This is considered to be plesiomorphic. A large number of ophionines have the
upper corner of the pronotum somewhat broadened and notched to partially occlude the spiracular
sclerite (1,0), whilst in a few species this flap completely covers the sclerite (1,1) (Fig. 43).
16-1, 16-2 Notauli. The notauli are grooves in the mesoscutum that appear to extend backwards from the
anterior margin, and at their most extreme, reach the scuto-scutellar groove. The presence of these
grooves seems to be a plesiomorphic feature (0,0), as they are well developed in many sawflies and are
visible in many fossil apocritans. However, notauli are not present in all members of the out-groups so
some doubt remains about the correct polarity of this character. The majority of ophionines only have
vestigial notauli impressed at the extreme anterior margin of the mesoscutum (1,0), whilst in several
they are absent entirely (1,1).
17 Pronotal crest. The plesiomorphic condition for ichneumonids is possession of a more or less flat
pronotum dorsally. In several ophionines a crest is present to protect the neck region, possibly against
attacks by asilids which habitually kill Hymenoptera by piercing the cervical region. The presence of
this crest is regarded as an apomorphy.
18-1, 18-2 Mesopleural furrow. The plesiomorphic condition of the ichneumonid mesopleuron is for it to
be relatively flat with a small pit (the episternal scrobe) near to the middle of the mesopleural suture
(Townes, 1969). Amongst ophionines two specializations have apparently arisen - the presence of a
diagonal groove from the pit to just below the subalar prominence (1 ,0) or possession of a groove from
the pit to the upper end of the epicnemial carina (0,1).
19 Epicnemial carina. It is a plesiomorphic feature for ichneumonids to have this carina complete,
extending laterally from the medioventral line to near the subalar prominence. In a number of taxa the
lateral portion of this carina is lost (1).
20-1, 20-2 Scutellar carinae. These carinae are of sporadic occurrence throughout the Ichneumonidae but
in many of the more primitive groups such as Pimplinae they are not developed. In the majority of
species in the out-groups these carinae are absent, suggesting that their absence in ophionines may be
a plesiomorphic feature. When present these carinae may be short, reaching to or not quite to the
centre (1,0), or they may be virtually complete, reaching 0-8 or more of the length of the scutellum
(1,1).
76 I. D. GAULD
27-7, 27-2 Metanotal protuberances. The hind rim of the metanotum of most ophionines, as well as most
campoplegines and cremastines, is unspecialized. In a few ophionines a small lateral tooth is
discernible (1,0), whilst in some taxa this tooth is apparently enlarged to form a protuberance that
extends back almost to the propodeal spiracle (1,1).
22-7, 22-2 Propodeal anterior area. The anterior part of the propodeum (Fig. 50), immediately behind the
metanotum, is, in ichneumonids, characteristically depressed to form a transverse groove which is
often somewhat broadened medially behind the postscutellum. This presumably plesiomorphic state
is found widely in Ophioninae and in virtually all members of the out-groups. Two modifications from
this pattern have been observed. In some species the groove is broadened (lengthened) and striate,
and thus forms a broad shallow concavity, almost as if the insect had been stretched (Fig. 51) (1,0).
The other adaptation is for the groove to be much deeper and present as a U-shaped furrow (Fig. 52)
(0,1).
23 Propodeal spiracle. In the majority of ophionines and members of the out-groups the propodeal
spiracle is oval or even subcircular, with the longest axis 4 or less times as long as the shortest. A few
ophionines have very large, very elongate spiracles which are 8 or more times as long as broad. This is
considered to be an apomorphic development.
24-1, 24-2 Anterior transverse carina of the propodeum. The presence of this carina, like the other
propodeal carinae, is a plesiomorphic feature of ichneumonids as they are complete in fossils
(Townes, I913b). The reduction of this carina is a progressive feature and has been coded as follows -
complete (0,0), present only centrally (1,0), absent (1,1).
25-7, 25-2 Posterior transverse carina of the propodeum. Coded as for character 24 for the same reasons.
26 Longitudinal propodeal impression. The plesiomorphic condition for ichneumonids is to have the
dorsum of the propodeum more or less flat. In a few ophionines a deep longitudinal impression is
present and this is considered to be an apomorphic feature.
27-7, 27-2, 27-3 Posterior transverse carina of the mesosternum. This carina is usually complete in
campoplegines and cremastines and thus a complete carina is here considered plesiomorphic for the
Ophioninae. The carina is usually lost at two points on either side of the midline, before the mid coxae,
so that central and lateral vestiges remain (1,0,0). Sometimes the central vestige is completely lost so
only lateral traces remain (1,1,0). In a few species reduction seems to have occurred by loss of only the
central part so two broad lateral portions remain (0,0,1).
28-1, 28-2 Lateromedian longitudinal carina of the propodeum. Coded as for character 24 for the same
reasons.
29-7, 29-2 The thyridium. This is a moderately large indentation found on the second gastral tergite. In its
plesiomorphic condition it is close to the anterior margin of the tergite (0,0), but in a number of
ophionines it can be seen to have been displaced posteriorly, leaving a scar between itself and the
tergal margin (0,1). In a few taxa the thyridium is absent (1,0).
30 Epipleuron of tergite 2. This is a difficult feature to determine the polarity of as within the out-groups
both conditions occur widely. In most ophionines it is upturned, and in a few otherwise specialized
taxa it is pendant. The latter condition is tentatively considered to be apomorphic.
31 Profile of tergite 2. Tergite 2 is laterally compressed in ophionines and members of the closer out-groups
(Fig. 2). In profile it is much longer than posteriorly deep. In a few ophionines it is quadrate and this is
considered to be an apomorphic development.
32 Position of spiracles on tergite 1. The plesiomorphic condition for ophionines and members of the
out-groups is for the spiracles to be at or behind the level of the margin of the sternite. In a few taxa the
spiracles are situated before the sternite margin, a presumed apomorphic condition (1).
33 Presence of an umbo on tergite 2. The umbo is a convex area on the midline at the anterior margin of
tergite 2. It is typically present in many ophionines and members of the out-groups. The apomorphic
state (1) is where this structure has been lost.
34 Ovipositor length. The length of the ovipositor varies a great deal between ichneumonid taxa. Probably
the ancestral condition for the family was long, but almost every evolutionary lineage shows
reduction. Virtually all ophionines have short straight ovipositors resembling those of many campo-
plegines (Townes, 1970), and thus this condition is considered to be plesiomorphic. The long
ovipositors found in a few taxa are considered to be apomorphic features (1).
35 Ovipositor sheath. The ovipositor sheaths (valvulae 3) of ichneumonids are almost always slender, just
wide enough to enclose the ovipositor. In a few ophionines the sheaths are very stout. This is
considered to be an apomorphic development (1).
36-7, 36-2, 36-3 Position oflm-cu in relation to Cu la in the fore wing. In most ichneumonids these two veins
are quite widely separated and are often at least as far apart as 0-75 of the length of Cu lb . With some
reservation this is considered to be the plesiomorphic state (0,0,0). In many ophionines these veins are
PHYLOGENY OF THE OPHIONINAE
77
separated by about 0-5 times the length of C lb (1,0,0), in a number of species by about 0-25 times
(1,1,0), whilst in a very few instances the veins are practically contiguous (1,1,1).
37 Length of second discal cell. The plesiomorphic condition for ichneumonids generally, including
ophionines, is for the length of the second discal cell (measured along C la ) to exceed 1-10 times the
length of the first subdiscal cell (measured along Cui). In a number of ophionines the second discal cell
is unusually short, being less than the length of the first subdiscal cell. This is considered to be an
apomorphic feature (1).
38 Presence of a ramellus. The ramellus on Im-cu in the fore wing is apparently a remnant of the vein that
divided the first discal from the submarginal cell (the two are confluent in all extant ichneumonids
though present in fossil forms (Townes, 19736)). The retention of this stub is tentatively considered to
be a plesiomorphic feature, but I have some doubts about the polarity assignment for similar reasons
to those stated in 47-1 below.
39 Base o>/Rs+2r in fore wing. In the majority of ophionines and members of the out-groups this vein is
emitted from the pterostigma at about 30 to the fore margin of the wing; the base of the vein is straight
(Fig. 3). In some ophionines the angle of emission is greater (40+) and the vein is rather sharply bent
basally (Fig. 4). This is considered to be an apomorphic feature (1).
40 Shaft o/Rs+2r in fore wing. The plesiomorphic condition for Hymenoptera generally is for this vein to
be centrally straight; this condition is found widely in out-groups and many ophionines (Fig. 3). In
some ophionines the vein is markedly sinuous just before the centre and this (Fig. 6) is considered to
be an apomorphic development (1).
41 Position of confluence o/Rs and R! in fore wing. In virtually all ophionines and many members of the
out-groups these veins meet at the extreme distal apex of the wing so the marginal cell is very long. In a
few ophionines the confluence of these veins is more proximal on the fore margin of the wing, away
from the tip, so the marginal cell is correspondingly shorter (Fig. 4). This is considered to be an
apomorphic feature (1).
42-1, 42-2 Shape of pterostigma. In most ichneumonids the pterostigma is quite broadly triangular and this
plesiomorphic condition prevails in most species in the out-groups. In some ophionines the ptero-
stigma is still broad (0,0) but in many it is not abruptly narrowed distally but evenly tapered (1 ,0) (Figs
4, 6). In a few the pterostigma is much more slender and elongately tapered (1,1) (Fig. 5).
43-1, 43-2 Position of distal abscissa of Cu } in hind wing. It is very difficult to assign polarity to this
character as all stages from Ci close to M to Cui close to L4 can be found in members of the
out-groups. Furthermore, in most Campopleginae this vein is lost. However, in some of the more
primitive ichneumonids the usual condition is for this vein to be closer to M than to L4 . This is
tentatively assumed to be the plesiomorphic condition (0,0). The derived states, treated as a
transformation series are: intermediate between M and \A (1,0) and closer to IA than to M (1,1).
44-1, 44-2 Shape o/Rs in hind wing. The plesiomorphic condition of this vein in ichneumonids generally,
including many ophionines, is for it to be straight (0,0). In many ophionines it is quite distinctly
Figs 3-6 Fore wings: 3, Ophion; 4, Simophion; 5, Stauropoctonus; 6, Enicospilus.
78 I. D. GAULD
concave whilst in a few taxa it is exceptionally curved so that the distal portion is parallel to the fore
margin of the wing. These two derivations are treated as a transformation series and coded 1,0 and 1,1
respectively.
45-1, 45-2 Glabrous area in fore wing. All primitive ichneumonids, most taxa in the out-groups, and
virtually all lower Hymenoptera have the membrane of the fore wing uniformly hirsute close to vein
Rs+2r. This is treated as the plesiomorphic condition (0,0). In most ophionines a small glabrous area
is present in the discosubmarginal cell near where the pterostigma emits Rs+2r. This is considered
apomorphic (1,0). A further specialization is for this area to be very extensive, reaching at least 0-3 of
the way along Rs+2r (1,1).
46 Alarsclerites. The possession of alar sclerites is an unusual apomorphic feature of certain ophionines.
Other than in this subfamily these sclerites are only found in two species of ichneumonid (and then in
different positions). Elsewhere in the Hymenoptera alar sclerites are apparently only found in some
pepsine pompilids (M. C. Day, pers. comm.).
47-1, 47-2 Shape of 7m-cu in fore wing. In the lower Hymenoptera this vein (which in fact is almost
certainly not just \m-cu but a composite of this and other vein parts) is angled centrally so the
antero-proximal side of the 2nd discal cell is angulate at about 90. In many ophionines this angulation
is present (Fig. 3) and often accompanied by a ramellus (see character 38). This is here tentatively
considered to be the plesiomorphic condition (0,0). In many species this vein is evenly curved (1,0),
whilst in some it is further modified by being sinuous (1,1). I have certain reservations about
the polarity of 47-1 as the evenly curved condition is the widespread condition amongst the out-
groups.
48-1, 48-2 Length of3r-m in the fore wing. In the majority of more primitive ichneumonids 3r-m is longer
than the abscissa of M between 2m-cu and 3r-m. In many ophionines this is also the case and it is here
considered to be the plesiomorphic state (0,0). In some ophionines 3r-m is shorter, 0-75-0-50 times as
long as M (1,0), whilst in a very few it is extremely reduced, 0-35 or less times the length of M (1,1).
49 Position of abscissa ofCuj between 7m-cu and Cu la . This vein is positioned at about 90 to the axis of the
fore wing in most ichneumonids and virtually all lower Hymenoptera. In some ophionines it is
specialized in being turned so that it is almost parallel to the wing axis (1).
50 Position of cu-a in fore wing. The plesiomorphic cdndition of this vein in ichneumonids is for it to
subtend an angle of about 90 to Cu^. In a few ophionines it is strongly oblique, subtending an angle of
50-60 (1).
51-1, 51-2 Flange on fore tibial spur. The plesiomorphic state for ichneumonids is the possession of a
membranous flange on almost the entire length of the fore tibial spur, immediately behind the
microtrichial 'comb' (Fig. 48) (0,0). In some ophionines this is reduced to about 0-3 of the length of the
spur (1,0), whilst in others it is entirely lost (Fig. 49) (1,1).
52-7, 52-2 Form of hind trochanter. The possession of a simple margin to the trochanter is plesiomorphic
for ichneumonids (0,0). In a few ophionines a small marginal tubercle is present (1,0), whilst in some
cases this may be long, curved and pointed (1,1).
53 Dorsal margin of hind trochanter. This is also unspecialized in most ichneumonids, but some ophionines
are unusual in having a specialization - the margin extended as a flange over the articulation of the
trochantellus (1).
54 Hind tibial spurs. In many ophionines and members of the out-groups these spurs are flattened
internally and bear a long fringe of close hairs. In some ophionines they are specialized in being
cylindrical and bearing only scattered hairs (1).
55 Penultimate distal hamulus. The unspecialized condition in Hymenoptera is for the hamuli to be of
similar size and shape. In some ophionines the penultimate hamulus is specialized in being very much
longer and more coiled than its fellows (1).
56 Grouping of distal hamuli. The plesiomorphic condition for Hymenoptera is for the row of hamuli to be
fairly evenly spaced. In a few ophionines they are arranged in two groups, an apparently specialized
condition (1).
57 Number of hamuli distally. Most primitive Hymenoptera have from 9 to about 15 distal hamuli and, as
this condition is widespread in primitive ichneumonids and many ophionines, it is considered
plesiomorphic in context of this study. Some ophionines have only 4 or 5 distal hamuli and this
reduction is considered to be an apomorphic condition (1).
58-1, 58-2 Shape of hind tarsal claw. In most members of the out-groups and many primitive ichneumonids
the tarsal claws are fairly evenly curved (Fig. 45) and this condition, where it occurs in ophionines is
considered to be plesiomorphic (0,0). In Ophioninae the claw may be modified in one of two ways -
either by being longer and straighter (1,0) or by being almost geniculate (0,1).
59 Colour of inter ocellar area. In most Hymenoptera this area is concolorous with the vertex. In a few
PHYLOGENY OF THE OPHIONINAE 79
ophionines it is sharply chromatically contrasted, and this is tentatively considered an apomorphic
feature (1).
60 Distal pectinal tooth of hind tarsal claw. The claws of ophionines possess a sinuous row of pectinal teeth
or pectinae (Gauld & Mitchell, 1978) forming a comb-like internal surface. This is apparently an
apomorphy of the subfamily, so in the present context such a claw may be considered a plesiomorphy.
In some ophionines the claw is specialized in having the distal pectinal tooth on the outer surface of the
main tooth (Fig. 46) (1).
61-1, 61-2 Taper of mandibular teeth. In most primitive ichneumonids the mandible is weakly tapered so
that the distal apex is more than 0-5 times as broad as the base. This is considered to be the
plesiomorphic condition where it occurs in ophionines (0,0). Some species have the mandible more
strongly tapered so the apex is 0-4-0-5 times as wide as the base (1,0) whilst a few are exceptional in
having the apex <0-3 times as wide as the base (1,1).
62 Thickness of base of Rs+2r. The plesiomorphic condition for Ichneumonidae is for this vein to be
slender at its junction with the pterostigma. In some ophionines it is specialized in being broadened so
that close to the pterostigma it is more than twice its central thickness (1).
63 Presence of laterotergite 1 of gaster. In primitive ichneumonids a distinct laterotergite is present
associated with tergite 1 ; vestiges of this may be found by partial dissection in many ophionines. Some
are specialized in having lost this structure (1).
64 Presence of a pectinal comb on both sides of male claw. The plesiomorphic condition for ophionines is
for a single sinuous pectinal comb to be present on the tarsal claws; some males are specialized in
having this comb present as a double row of teeth, surrounding a flattened central area (1) (Fig. 47).
Preliminary remarks on the relative values of the characters
For analytical purposes all characters have been treated initially as having equal numerical
weight. However, a practising taxonomist does not value characters equally. Intuitively he is
more likely to consider some to be more important than others. 'Good' characters are likely to
include many of the more bizarre morphological adaptations (Gauld & Mound, 1982). At the
other extreme, characters may be considered to have very little value as indicators of phylo-
genetic affinity. This is often for one of three reasons.
(a) The apomorphic state of the character may involve the disappearance of a structure and, as
it is obvious that certain structures are frequently lost in parallel in many distantly related
lineages (e.g. the occipital carina has been lost in some taxa independently in virtually all
ichneumonid subfamilies), one suspects that an absence in two closely related taxa could also be
the result of parallel reduction.
(b) There may be some doubt concerning polarity. This is particularly true in cases where one
postulates that a particular state is an apomorphy for the subfamily, but advocates reversal in a
few exceptional taxa (e.g. character 12, ocellar size).
(c) Even if the apomorphic condition involves the development of a particular structure, the
character may be judged to be unstable, as examination of out-groups suggests it has been
derived independently in numerous evolutionary lineages, or is correlated with a particular size
range or extreme habitat occupied by a particular species. For example, very large ichneu-
monids often have similarly coarse reticulate propodeal sculpture, whilst eremic species have
short, stout antennae.
The third of these reservations is discussed for individual characters in a subsequent section,
but it is appropriate to mention a and b here. Those of the characters used in this study which
involve an apomorphic 'loss' are listed below with an asterisk(*), whilst those in which the
polarity assignment is suspect are denoted by a prime mark('). This convention is used
extensively in the cladograms presented in the following sections. For comparative purposes a
list is given below.
1-1*, 1-2*, 2, 3, 4-1, 4-2*, 5-1, 5-2, 6-1, 6-2, 7-1', 7-2', 8-1*, 8-2*, 9, 10, 11, 12*, 13, 144', 14-2',
15-1, 15-2, 16-1*', 16-2*', 17, 18-1, 18-2, 19*, 20-1, 20-2, 21-1, 21-2, 22-1, 22-2, 23, 24-1*, 24-2*,
25-1*, 25-2*, 26, 27-1*, 27-2*, 27-3*, 28-1*, 28-2*, 29-1, 29-2*, 30', 31, 32, 33*, 34', 35, 36-1',
36-2', 36-3, 37, 38*', 39, 40, 41, 42-1, 42-2, 43-1', 43-2', 44-1, 44-2, 45-1, 45-2, 46, 47-1', 47-2,
80
I. D. GAULD
48-1, 48-2, 49, 50, 51-1*, 51-2*, 52-1, 52-2, 53, 54, 55, 56, 57*, 58-1, 58-2, 59', 60, 61-1, 61-2, 62,
63*, 64.
The phylogenetic analysis
An initial attempt was made to construct a Wagner tree, but with repeated runs it became
obvious that there exists a large number of equally or almost equally (less than 1% difference in
overall length) short trees, many of which exhibit quite different topologies. As there is no basis
on which to prefer one of these, and there is no way of knowing whether all the shortest trees
have been computed, Wagner analysis was initially rejected in favour of a more detailed study of
characters and taxa using the programs LEQU.BAS and FOURS. From these it was possible to
gain some indication of the 'shape' of the primary data set. Tentative groups of taxa recognized
initially were reanalysed using Wagner and compatibility methods.
Preliminary analysis of data
Analysis using LeQuesne test program, LEQU.BAS
Using the program LEQU.BAS the primary data matrix (Appendix 2) of 51 selected ophionine
taxa was analysed (see Table 6). Of the 95 characters used, six, 8-2, 11, 28-1, 34, 49 and 53,
Table 6 Results of LeQuesne test on the primary data matrix (Appendix 2). Conventions follow table 2.
Note that a character with a single incidence of one or other state logically cannot fail the test and is
represented by .
Character number: incompatibilities observed expected 0/E ratio
1.1
71
72.6
0.98
1.2
54 57.1 0.95
2_
44 73.2
0.60
1
44 71.3
0.62
4.1
53
57.5
0.92
4.2
53 42.4 1.25
5.1
64 76.5
0.84
5.2
46 61.6
0.75
6.1
60
65.4
0.92
6.2
31 65.4 0.47
7.1
63 75.2
0.84
7.2
60 80.7
0.74
8.1
11
28.5
0.39
8.2
- -
9_
55 66.1
0.83
1
34 28.5
1.19
_U
-
-
li
68 74.8 0.91
11
52 66.1
0.79
14.1
65 73.8
0.88
14.2
66
79.8
0.83
15.1
73 81.2 0.90
15.2
62 75.2
0.82
16.1
73 50.8
0.45
16.2
77
77.5
0.99
17
65 71.3 0.91
18.1
69 75.2
0.92
18.2
69 72.3
0.95
11
65
66.1
0.98
20.1
78 79.8 0.98
20.2
68 80.5
0.84
21.1
52 65.9
0.79
21.2
11
28.2
0.39
22.1
65 80.9 0.80
22.2
61 76.3
0.80
il
57 73.2
0.78
24.1
72
81.1
0.89
24.2
76 80.5 0.94
25.1
15 28.1
0.53
25.2
72 75.8
0.95
26
49
57.8
0.85
27.1
77 80.7 0.95
27.2
77 80.0
0.96
27.3
12 27.5
0.44
28.1
-
-
28.2
69 73.2 0.94
29.1
50 50.7
0.99
29.2
79 81.1
0.97
30
68
74.8
0.91
3i
69 69.0 1.00
li
67 71.3
0.94
33
69 79.7
0.87
34
-
-
35_
55 73.2 0.75
36.1
71 64.4
1.10
36.2
76 78.8
0.96
36.3
58
69.6
0.83
37
78 82.2 0.95
38
53 62.5
0.85
39
78 79.1
0.99
40
47
71.3
0.66
11
64 62.5 1.02
42.1
66 72.3
0.91
42.2
61 77.3
0.79
43.1
73
73.8
0.99
43.2
78 81.2 0.96
44.1
82 81.2
1.01
44.2
70 68.0
1.03
45.1
69
76.3
0.90
45.2
67 80.9 0.83
46
56 76.2
0.74
47.1
46 61.6
0.75
47.2
74
77.4
0.96
48.1
73 81.2 0.90
48.2
61 72.2
0.84
49
-
50
31
28.5
1.09
51.1
61 68.1 0.90
51.2
62 74.0
0.84
52.1
51 70.6
0.72
54
57
74.8
0.76
55_
33 57.8 0.57
56
30 28.5
1.05
1Z.
74 80.8
0.92
58.1
67
76.9
0.87
58.2
27 28.1 0.96
59
78 81.7
0.95
60
19 28.5
0.67
61.1
70
80.5
0.87
61.2
72 77.3 0.93
62
75 80.3
0.93
63
74 81.9
0.90
64
5
28.5
0.18
Grand total- 2598 observed. 3007.1 expected. Overall 0/E ratio = 0.86
Ranking ratio of scoring characters
64 8.1 21.2 27.3 16.1 6.2 25.1 55 2 3 40 60 52.2 52.1 46 7.2
13 42.2 21.1 22.2 22.1 15.2 14.2 45.2 9 36.3 5.1 7.1 51.2 48.2
58.1 14.1 24.1 51.1 48.1 15.1 63 45.1 30 12 17 42.1 57 6.1 18
28.2 24.2 1.2 37 25.2 27.1 59 18.2 47.2 43.2 58.2 27.2 36.2 29
29.1 43.1 16.2 31 44.1 41 44.2 56 50 36.1 10 4.2
5.2 47.1 35 54 23
20.2 26 38 33 61.1
1 4.1 61.2 62 32
2 20.1 1.1 19 39
PHYLOGENY OF THE OPHIONINAE
81
showed only a single discordant state. In the case of 28-1 and 49 the majority state was the
derived condition, whilst in the remainder it was the plesiomorphic condition that predomi-
nated. As a minimum of two discordant character-states is logically required for a failure, it
follows none of these eight can fail; they are shown as .
The remaining 89 characters have two or more discordant states and therefore can potentially
fail the test. Each character was paired with every other one (except for the alternatives n. 1 with
.2) and of the 7,855 comparisons made 2,598 resulted in failures. A total of 3,007 failures was
expected, so although slightly better than random (O/E value = 0-86), the results suggest a very
high incidence of homoplasy. For individual characters a considerable variation in O/E ratio was
observed. Characters 6-2, 84, 16-1, 21-2, 25-1, 27-3 and 64 all scored better than 0-55, whilst at
the other extreme eight characters, 4-2, 10, 36-1, 41, 44-1, 44-2, 50 and 56 all scored worse than
1-00.
Examination of the labels matrix (Appendix 4) showed that two taxa, 909 and 922, had been
labelled exceptionally frequently (scoring 89 and 85 labels respectively). Six further species, 906,
908, 921, 923, 942 and 951, were labelled between 45 and 67 times, whilst the rest were labelled
less than 42 times. Sixteen taxa were labelled less than ten times and three, 912, 913 and 933 were
not labelled at all.
As would be expected, the elimination of bad characters (i.e. those with an exceptionally high
O/E score) produced, in general, a progressive improvement in the O/E ratios of the surviving
characters. However, the average ratio improved very slowly. With 20 informative characters
remaining it was still 0-34. This clearly shows that not all the failures of the 'better' characters
were due to clashes with the worst characters; the majority of characters seem to be somewhat
homoplastic! Complete elimination of homoplasy (as advocated by LeQuesne, 1972) results in
the removal of virtually all characters. The largest cliques contain only ten compatible
characters. Several of these have only a very low information value (i.e. have only two or three
derived states), but five more informative characters (2, 6-2, 15-1, 15-2 and 22-2) were found to
be universal to all the larger cliques. All of these are features that have been used previously in
ichneumonid higher classification (Cushman, 1947; Townes, 1971). These characters support
the cladogram shown in Fig. 7. Only the group supported by character 2 corresponds with any
previously recognized grouping of taxa, the Thyreodon group of Cushman (1947).
905 909
910-911
916-918
923 948
938 939
941
906 927
935 937
901 904
907 908
921
912 915
919 920
928 934
940 945
"
922
924 926
932 934
942 944
946 947
15.2
15.1
22.2
Fig. 7 Cladogram supported by largest clique of compatible characters.
82 I. D. GAULD
The rate at which the O/E values of individual characters improved (after removal of bad
characters) varied tremendously. For example, after removal of the eight worst characters
(Table 7), character 29-1, initially the seventy-eighth placed (and the twelfth one scheduled for
elimination) progressively improved (with the elimination of each worse character) to sixty-
seventh. Character 59, initially placed as sixty-sixth, steadily worsened its position to eighty-
third ; it would in fact be the tenth character eliminated . If the ten worst characters were removed
en bloc they would have been 4-2, 10, 36-1, 50, 56, 44-2, 41, 44-1, 31 and 16-2. By stepwise
removal of the worst character they would be 4-2, 10, 50, 36-1, 44-1, 41, 39 and 59. All further
'cleaning-up' of data was done by stepwise removal of characters.
Table 7 Results of a LeQuesne test on the primary data matrix after the progressive removal of the
highest scoring characters (4-2, 10, 50, 36-1, 44-2, 56, 44-1, 41). Conventions follow table 2; non-
informative characters omitted.
Character number: incompatibilities observed expected O/E ratio
1.1
63 67.7
0.93
1.2
49 53.8
0.91
2
38 68.4
0.56
I
40 66.7
0.60
4.1
47 54.4
0.86
5.1
58 71.1
0.82
5.2
43 57.9
0.74
6.1
53 61.3
0.86
6.2
26 6U3
0.42
7.1
57 69.9
0.82
7.2
56 74.6
0.75
8.1
9 26.9
0.33
9_
51 62.0
0.82
li
62 69.8
0.89
11
48 62.0
0.77
14.1
60 68.8
0.87
14.2
59 73.8
0.80
15.1
66 75.0
0.88
15.2
56 69.9
0.80
16.1
20 47.9
0.42
16.2
71 72.0
0.99
\]_
61 66.7
0.91
18.1
65 70.0
0.93
18.2
62 67.4
0.92
li
59 62.0
0.95
20.1
70 73.8
0.95
20.2
63 74.4
0.85
21.1
49 61.8
0.79
21.2
9 26.7
0.34
22.1
60 74.7
0.80
22.2
54 70.9
0.76
23
51 68.4
0.75
24.1
68 74.9
0.91
24.2
69 74.4
0.94
25.1
15 26.5
0.57
25.2
66 70.6
0.94
26
47 54.4
0.86
27.1
71 74.5
0.95
27.2
71 73.9
0.96
27.3
10 25.9
0.39
28.2
64 68.4
0.94
29.1
44 47.7
0.92
29.2
72 74.9
0.96
30
63 70.0
0.90
21
63 64.6
0.98
32
61 66.7
0.91
11
64 73.9
0.87
li
52 68.4
0.76
36.2
71 73.8
0.96
36.3
54 65.7
0.82
37
71 75.9
0.94
38
52 58.7
0.89
39
72 73.4
0.98
40
44 66.7
0.66
42.1
60 67.4
0.89
42.2
53 71.7
0.74
43.1
66 68.8
0.96
43.2
72 74.9
0.96
45.1
63 70.9
0.89
45.2
60 74.7
0.80
46
53 70.9
0.75
47.1
43 57.9
0.74
47.2
69 71.8
0.96
48.1
67 75.0
0.89
48.2
57 67.4
0.85
51.1
55 63.7
0.86
51.2
56 68.9
0.81
52.1
47 66.0
0.71
52.2
42 58.0
0.72
54
51 69.8
0.73
55
31 54.4
0.57
57
68 74.8
0.91
58.1
60 71.5
0.84
58.2
26 26.5
0.98
59
73 75.6
0.97
60
17 26.9
0.63
61.1
65 74.4
0.87
61.2
65 71.7
0.91
62
68 74.4
0.91
63
68 75.7
0.90
64
4 26.9
0.15
Grand total- 2179 observed, 2613.3 expected. Overall O/E ratio = 0.83
Ranking ratios of scoring characters
64 8.1 21.2 27.3 16.1 6.2 2 25.1 55 3 60 40 52.1 52.2 54 42.2 5.2 47.1 23 46 7.2
35 22.2 13 21.1 14.2 15.2 22.1 45.2 51.2 7.1 5.1 36.3 9 58.1 48.2 20.2 51.1 4.1 26
6.1 33 14.1 61.1 15.1 38 12 45.1 42.1 48.1 63 30 61.2 24.1 57 1.2 62 32 17 18.2
29.1 24.1 18.1 1.1 37 25.2 28.2 20.1 19 27.1 43.1 47.2 43.2 27.2 29.2 36.2 59 31
58.2 39 16.2
It is apparent from the generally high O/E values that homoplasy is a common feature in the
Ophioninae. Any attempt to produce a character set without homoplasy (a clique) necessitates
excluding the great majority of characters, including some of the most informative ones. This is
unacceptable for two reasons.
(1) A large amount of information would have to be discarded leaving many unresolvable
situations. As 2N t 2 apomorphies are necessary to fully resolve N t taxa, a minimum of 100
compatible characters would be needed to fully resolve the ophionine data. Consequently,
without homoplasy, even the full set of 95 characters is insufficient for full resolution. Excluding
autapomorphies for definition of terminal taxa, a minimum of 49 characters is necessary for
PHYLOGENY OF THE OPHIONINAE 83
resolution, but in the present case it can be seen that even discarding half the characters, the
resultant data set would still be extremely homoplastic.
(2) It is important to realize that just because a character is homoplastic and has been derived
independently in two (or more) lineages, this does not mean that it is not of phylogenetic value.
Such a character can still be a genuine synapomorphy for species in two different groups of taxa,
and many minimally homoplastic characters are still highly regarded by traditional taxonomists.
For example, Mason (1981) states that one of the major synapomorphies of the Braconidae (the
probable sister-group of the Ichneumonidae) is the fusion of the second and third gastral
tergites. Approximately six of the 60,000 species of Ichneumonidae also have tergites 2 and 3
fused. This parallelism in a few specialized ichneumonids does not eliminate the character as a
useful apomorphy for defining Braconidae in a phylogenetic sense, it merely reduces the value of
the character as an infallible means of identifying braconids. If it were possible to consider all
species of the two taxa Ichneumonidae and Braconidae, it is likely that the O/E value of this
character would be very low, but if the analysis were of all Parasitica, and a few taxa from each
family were selected as examples of the range of morphological variation, it is quite possible that
one of the six anomalous ichneumonids would be included, giving the character a much larger
O/E value, and consequently risking its exclusion. In the present study certain species of
Enicospilus, Leptophion and Laticoleus were deliberately selected to show the range of
morphological variation in the genus, so it is highly likely that these will contribute to the high
level of homoplasy in the way outlined above.
In the LeQuesne analysis eight characters were eliminated stepwise until none of the remainder
scored worse than 1-00. The particularly high O/E value of these characters strongly suggests
they are particularly homoplastic, and consequently of no real value in phylogenetic reconstruc-
tion. Considered from a biological point of view it certainly seems these features are homo-
plastic, but even so several seem to be of use in defining possible holophyletic groups.
Considered in order of 'worst' first these characters are as follows.
4-2*: reduction of upper tooth of mandible. The occurrence of the derived state of this feature is
limited in the matrix to three species, Rhynchophion flamipennis (908), Sicophion pleuralis
(922) and Enicospilus unidens (946). Amongst the Ophioninae in general, this feature is found in
the apomorphic state in a group of Madagascan and southern Indian species of Enicospilus
(Gauld & Mitchell, 1978; 1981) that seem to be related to E. unidens, an apparently unrelated
Melanesian species (E. interruptus Szepligeti) and in an Indian Ophion species. Thus although
the character has phylogenetic value in some cases (the definition of the unidens group) in others
it has none. Presumably the apomorphic condition found in three taxa in the matrix is the result
of parallel derivation, a conclusion that is strongly suggested by the very high label values
(Appendix 4). Amongst out-groups this character is of similar sporadic occurrence (cf. Townes,
1970), sometimes defining an apparent clade and other times being characteristic of a single
species. Structurally it is a simple adaptation for biting through fibrous cocoons and as such it is
not surprising that it has apparently arisen independently in many different evolutionary
lineages.
10: elongation of maxillae. This too is a low information value character as the apomorphic
condition occurs in just two taxa in the matrix, Rhynchophion flamipennis (908) and Sicophion
pleuralis (922). Amongst other ophionines it is limited in occurrence to some Neotropical
species of Eremotylus. All of these taxa are restricted to the drier areas of Central and South
America and the structural development seems to be an adaptation to drinking from certain
flowers. Similar structural modification can be observed in many other ichneumonids in similar
habitats, including many Tersilochinae, Cremastinae and Anomaloninae. In arid areas of
south-western Australia species of three closely related cremastine genera, Trathala, Temelucha
and Pristomerus, all have identically modified maxillae. Only familiarity with the world
cremastine fauna enables one to recognize this as evolutionary convergence rather than
considering it indicative of a holophyletic group. These observations strongly suggest that
particularly intense selection pressures in a particular area may elicit the development in
84 I. D. GAULD
parallel, of apparently identical apomorphic features in closely related lineages. This may
explain the similarity between the taxa in the matrix.
50: oblique position of cu-a in fore wing. Like the preceding two characters this is of low
information content as the derived condition occurs in just two taxa in the matrix, Lepiscelus
distans (909) and Barytatocephalus mocsaryi (927). Amongst other ophionines it is limited to
occurring in one or two species of Ophion. In the sister-group to the Ophioninae, the
Campopleginae, the derived condition apparently is useful for defining some genera (e.g.
Cymodusa, see Townes, 1970) but the character has little apparent phylogenetic significance in
the Ophioninae.
36-1': vein Im-cu separated from Cu la by less than 0-75, but more than 0-5 of the length ofCu lb .
The majority of taxa have this character present in the presumed derived state. Only Lepiscelus
distans (909), Stauropoctonus occipitalis (917), Riekophion emandibulator (920), Ophionopsis
nigrocyaneus (921), Sicophion pleuralis (922), Rictophion nebulifer (937) and Enicospilus
nephele (944) are plesiomorphic for this feature. It is not found in the plesiomorphic state in any
other Stauropoctonus species, in most other Enicospilus species nor any other Riekophion
species. If the state were a genuine symplesiomorphic feature of these taxa the 1 state must
have been derived in parallel in at least species of Riekophion, Enicospilus and Stauropoctonus
(assuming they are natural genera) and therefore the 1 state is not synapomorphic for all species
so scored. Alternatively, if the 1 state is a genuine synapomorphy for all 'other taxa', the state
must represent independent reversals in some taxa. Either way the character is homoplastic.
The same argument can be applied even if the polarity is reversed, unless one were to accept the
otherwise unsupported clade of 909+917+920+921+922+937+944.
44-2: strongly curved Rs in hind wing. In the matrix the apomorphic state of this character is
found in Simophion calvus (905), Lepiscelus distans (909), Laticoleus curvatus (915), Abancho-
gastra hawaiiensis (925), Leptophion maculipennis (928), Euryophion adustus (936), Rictophion
nebulifer (937) and Ophion luteus (939). Elsewhere in the subfamily all species of Ophion (ca 150
species) have Rs strongly curved. The genus Abanchogastra is monotypic, but only some species
of Leptophion, Enicospilus, Euryophion and Laticoleus possess the apomorphic state as do
some species of Eremotylus and Ophiogastrella. In both Leptophion and Laticoleus this feature
seems to define holophyletic species-groups, but there is no evidence to support the idea that it
could be a synapomorphy of all the taxa listed above. Clearly it must have been derived
independently in many ophionine lineages, but although a homoplastic feature, it is interesting
to note that it still is quite characteristic of certain species-groups and even genera, suggesting
that once the apomorphic state has been derived, it is unlikely to undergo reversal.
56: separation of hamuli into two groups. This feature occurs in the matrix only in two
Afrotropical species, Lepiscelus distans (909) and Laticoleus curvatus (915). The latter is the
only species in its genus with this particular modification and Lepiscelus is monobasic (Gauld &
Mitchell, 1978). The apomorphic condition is also found in some central Asian Ophion species.
The only feature these taxa seem to have in common is that they inhabit rather dry areas. The
character appears to have little phylogenetic value.
41: fore wing with R^ not reaching to tip. The apomorphic condition is found in Thyreodon
atricolor (901), Simophion calvus (905), Orientospilus melasma (906), Ophionopsis nigro-
cyaneus (921) and Sicophion pleuralis (922). It is not present in many other Thyreodon species so
it is unlikely that the derived condition in 901 is synapomorphic with that of other species. The
exclusion of 901 from the analysis actually resulted in a worsening of the O/E value of this
character from 1-02 to 1-12, suggesting it is homoplastic in other taxa. The apomorphic state of
this character is found in all species of the genera Simophion, Orientospilus and Ophionopsis,
suggesting this feature, once derived, is not likely to undergo reversal.
44-1: bowed vein Rs in hind wing. The apomorphic state of this character is found in all taxa
listed under 44-2 above and additionally in 14 other taxa, all from different genera. Elsewhere
the feature is of sporadic occurrence in Ophiogastrella, Aulophion, Leptophion, Laticoleus and
PHYLOGENY OF THE OPHIONINAE 85
Enicospilus, as well as being present in all species of Dictyonotus, Hellwigiella, Sclerophion,
Eremotylus, Xylophion and Rhopalophion. This suggests parallel development in many dif-
ferent evolutionary lines.
The high O/E values of several of these characters were surprising as almost all represent
structural specializations (not 'loss' apomorphies or dubious polarity assignments). Several (4-2,
41 and 44-2) do apparently have considerable phylogenetic value at certain hierarchical levels
and really only two (36-1 and 44-1) are likely to be considered by a traditional taxonomist as
category 3 'bad' characters. Character 10 almost certainly would intuitively be regarded as a
'good' character (vide Cushman, 1947). It is interesting to note that two taxa (909 and 922) were
cited for at least half of the eight worst characters, as having contributed to the high level of
homoplasy . These taxa, also recognizable by their high label scores, were subsequently shown to
be amongst the more difficult to place, and small differences in their position in the order of taxa
submitted to the WAGNER program, produced large changes in the resulting cladogram.
Another estimate of incidences of homoplasy in the primary data may be obtained from
perusal of the labels matrix (Appendix 4). Notably high scores, indicative of homoplastic
derivation of the apomorphic state in two or more taxa, can be observed for character 1-2 (taxa
909, 923 and 949), 4-1 (906, 909), 4-2 (908, 922, 946), 10 (908, 922), 26 (906, 923), 29-1 (922,
937), 50 (909, 927), 56 (909, 915), 58-2 (942, 951) and 60 (928, 942). The shared plesiomorphic
condition for character 38 in taxa 941 and 943 was also highly labelled, suggesting one or other
has undergone reversal. Of these 11 characters, six (4-2, 10, 50, 56, 58-2 and 60) are highly
labelled for all apomorphic states (two or three in each case), suggesting these are not useful
characters for phylogenetic reconstruction. Rather suprisingly all are 'gain' apomorphies. Four
(4-2, 10, 50 and 56) are amongst the eight characters eliminated in cleaning up the LeQuesne
matrix; one other (58-2) has a higher than average O/E score, but surprisingly one (60) has a
reasonably low O/E value.
Analysis using FOURS program
Using the FOURS program a shared derived character matrix (Appendix 3) was computed. This
shows the number of derived characters common to each pair of taxa and has values in the range
of 3-37. Multiple linkage clusters drawn for taxa with 29+ shared derived characters are shown
in Fig. 8. Four groups are apparent. The largest includes 17 species (912-5, 919, 924, 928-30,
932-4, 942-3, 945-7), the last eight of which form a particularly closely knit group. Taxon 925
and the reciprocal nearest neighbours 911 and 910 are less strongly associated with this group.
This complex contains most of the species of Enicospilus and Leptophion, all Dicamptus,
Laticoleus and Pamophion, together with Ophiogastrella and representatives of two of the three
endemic Hawaiian genera, Pycnophion and Abanchogastra. It is noteworthy that Enicospilus
nephele (944) and Leptophion tetus (931) are not associated. Each of these taxa share less
derived characters with their supposed congeners than their congeners share with other taxa,
although in each case the species' nearest neighbours are its congeners (Table 13).
Taxa 916, 917 and 918 (the Stauropoctonus group) form a cluster with 30-34 derived
characters in common. This group seems to be associated with the Enicospilus group quite
closely as 918 shares 28 derived characters with 947, 943, 942 and 919.
Taxa 935, 936 and 937 (the Euryophion group) cluster with 30-32 shared derived characters.
Taxa 901-4 (the Thyreodon group) form a cluster sharing 28-32 derived characters. Also
associated with this cluster are the reciprocal nearest neighbours 907 and 921 (Dictyonotus I
Ophionopsis).
When linkages of more than 22 shared derived characters are considered (Fig. 9), the majority
serve to consolidate the clusters already formed. The Stauropoctonus group and Ophiogastrella
are repeatedly linked to the Enicospilus cluster (both having in excess of 60% of possible
linkages to the latter) as does taxon 931. Taxa 926 and 944 are less strongly but uniquely
associated with this group and taxon 949 also links to this group. The Thyreodon, Euryophion
and Dictyonotus/ Ophionopsis groups, together with taxon 908, form a second robust cluster
with only Euryophion showing any linkage to non-group taxa. Taxon 909 is somewhat
86
I. D. GAULD
o
I
<0
TD
O
O
0\
(S
u
<u
60
J
I
BC
E
PHYLOGENY OF THE OPHIONINAE
87
Fig. 9 Multiple linkage clusters drawn for all taxa with 22 or more derived characters in common. As
virtually complete linkages occur between taxa enclosed within the boxes, these linkages are not shown
so as to avoid confusion.
intermediate linking to both the Enicospilus and Euryophion groups but shares most derived
features with taxa in the Stauropoctonus group.
Other taxa occupy a more ambivalent position. Taxon 906 is intermediate between Enico-
spilus and Euryophion, whilst 905 appears to be fairly closely associated with Euryophion
although its nearest neighbour is taxon 906. Taxon 927 is also intermediate. The remaining
species, 920, 922, 923, 938-41, 948 and 950-51, are not closely associated with other taxa. To
summarize therefore, two major groups are discernible, the Enicospilus/ Stauropoctonus com-
plex containing taxa 909-919, 924-926, 928-934, 942-947 and 949, and the Thyreodonl
Euryophion complex with taxa 901-904, 907-908, 921 and 935-937. Taxa 920, 922, 923,
938^941, 948 and 950-951 are not associated at all whilst 905, 906 and 927 are intermediates
between the major clusters. It is interesting to note that, excluding the last three taxa, the
remainder could be classified phenetically (if one counted shared plesiomorphies of which 920,
922, 938-941, 948 and 950-951 have a considerable number in common) into three groups
approximately corresponding to the Ophion (plesiomorphic group) , Thyreodon and Enicospilus
groups of Cushman (1947).
Phylogenetic analyses of groups of ophionine taxa
The Euryophion group (taxa 935, 936, 937)
It can be seen from Appendix 3 that these taxa form a group with 30+ shared derived characters,
and no more than 26 derived characters in common with any other taxon. Using these three as
the fixed taxa (3F option) of FOURS, and trying them against all other taxa, nothing broke them
up. The three taxa were found to share 29 apomorphies (2, 6-1, 7-1, 16-1, 16-2, 22-2, 24-1, 24-2,
25-1, 25-2, 27-1, 28-1, 28-2, 31, 32, 38, 39, 42-1, 42-2, 43-1, 44-1, 454, 47-1, 49, 51-1, 51-2, 58-1,
61-1 and 62). Five of these apomorphies (2, 7-1, 22-2, 42-2 and 58-1) have average or lower O/E
values (0-60, 0-84, 0-80, 0-79 and 0-87 respectively) and are shared by relatively few other taxa in
the matrix. Of these five apomorphies only one (7-1) has questionable polarity. No character is a
88
I. D. GAULD
unique apomorphy of the group, but the four apomorphies accepted above suggest this group is
monophyletic.
There are three possible arrangements of these taxa (Fig. 10). Taxa 935 and 936 have been
considered to be congeneric (in Euryophion) whilst 937 is usually placed in a separate genus,
Rictophion (Townes, 1971; Gauld & Mitchell, 1978). Rictophion has been separated from
Euryophion by two unique apomorphies, loss of thyridia and possession of 3-segmented palps.
Cladogram A supporting this arrangement is the weakest of the three as it involves only a single
informative character, 36-1 which has both a dubious polarity assignment and a very high O/E
score. Both B and C also involve some poor scoring characters (19 and 44-2 both have O/E
values of more than 0-95). Arrangement B is only supported by 'loss' apomorphies, though one
(8-1) has a low O/E value (0-39). However, it has a rather high label score. The characters
supporting C are all fairly high scoring (O/E values 0-76+) but do not involve either apomorphic
loss or dubious polarity assignment. On balance therefore, C seems to be the preferable
arrangement, suggesting that Euryophion, as currently recognized, is paraphyletic with respect
to Rictophion. This conclusion is not surprising as originally Rictophion was separated from
Euryophion on the basis of two autapomorphies. Rictophion ikuthana (937), the only species in
the genus, can be regarded as a specialized species of Euryophion. These species, and all the
remainder in the genus, are further analysed below (p. 136).
935 936 937
935 937 936
936 937 935
8.1*
44. 2
54
A B
Fig. 10 The three possible dichotomous arrangements of Euryophion latipennis (935), E. adustus (936)
and Rictophion ikuthana (937).
The Thyreodon group (taxa 901, 902, 903, 904)
These four taxa, all species currently placed in the genus Thyreodon (Townes & Townes, 1966),
form a closely knit group. Using the FOURS 3F option any combination of these species
remained as a group when tested against any fourth taxon. The Thyreodon species share 27
apomorphies (2, 6-2, 7-1, 14-1, 17, 22-2, 23, 24-1, 24-2, 25-1, 25-2, 27-1, 28-1, 28-2, 29-2, 30, 33,
36-1, 38, 424, 42-2, 47-1, 49, 51-1, 51-2, 54 and 61-1). Seven of these apomorphies (2, 6-2, 7-1,
22-2, 23, 42-2 and 54) have both lower than average O/E scores (0-60, 0-47, 0-84, 0-80, 0-78, 0-79
and 76 respectively) and are shared by relatively few other taxa. Only one, 7-1, involves
dubious polarity assignment; the remainder are gain apomorphies suggesting the group is
PHYLOGENY OF THE OPHIONINAE
89
monophyletic. Considering the possible phylogenetic relationships of these taxa to each other,
the following result was obtained using the FOURS program (Underwood, 1982).
TAXA NOS.- 901
X
X'
Y
XYZ'
XY'Z
X'Y'Z'
902
903
904
1100:
12*
13
0011:
36-
2
'
1010:
21-
1
1110:
26
58-1
1101:
37
0111:
9
The most favoured arrangement X,XYZ' is shown in Fig. 11. This involves three forward
parallelisms 21-2 (in 901 and 903), 37 (in 901/2 and 904) and 36-2 (in 903 and 904) and one
reversal (9 in 901). If the dubious polarity of 36-2 were to be reversed this character would fit the
cladogram. It is noteworthy that some authors (e.g. Ashmead, 1900; Cushman, 1947) have
placed taxon 904 and its relatives in a separate genus, Athyreodon.
904
12*, 13
26, 58.1
Fig. 11 The arrangement of Thyreodon species as supported by the largest number of compatible
characters.
The DictyonotuslOphionopsis group (taxa 907, 921)
These two taxa are reciprocal nearest neighbours. They share 30 apomorphies (2, 6-2, 7-1, 12,
13, 14-1, 16-2, 18-2, 21-1, 21-2, 22-2, 23, 24-1, 24-2, 25-1, 25-2, 28-1, 28-2, 30, 33, 38, 424, 42-2,
44-1, 47-1, 49, 51-1, 51-2, 54 and 58-1). Ten of these (2, 6-2, 7-1, 21-1, 21-2, 22-2, 23, 42-2, 54 and
58-1) have both average or lower O/E values (0-60, 047, 0-84, 0-79, 0-39, 0-80, 0-78, 0-79, 0-76
and 0-87 respectively) and are shared by relatively few other taxa in the matrix. One character,
90 I. D. GAULD
21 '2, is a unique apomorphy of this group. The primary label matrix (Appendix 4) suggests
characters 16-1, 36-1 and 61-1 (0 scores in taxon 921) may have undergone reversal, whilst
character 41-2(1 score in 921) may have been derived in parallel in 921 and other taxa.
The Thyreodon + Euryophion + Dictyonotus complex (the above three groups)
These three groups of taxa appear to be associated (see shared derived character matrix,
Appendix 3). Phylogenetically they also seem to form a distinctive clade, sharing 16 apomorphic
features (2, 7-1, 22-2, 24-1, 24-2, 25-1, 25-2, 28-1, 28-2, 38, 42-1, 42-2, 47-1, 49, 51-1 and 51-2).
Four of these apomorphies (2, 7-1, 22-2 and 42-2) have both lower than average O/E values
(0-60, 0-84, 0-80 and 0-79 respectively) and are shared by very few other taxa. Character 2 is
elsewhere only found in the apomorphic state in taxon 908, character 7-1 is apomorphic for taxa
908, 940 and 946, character 22-2 is apomorphic for taxa 906, 908, 927 and 950, whilst character
42-2 is otherwise only apomorphic for taxa 908, 909, 916, 917 and 918. Examination of the labels
matrix (Appendix 4) shows that for character 7-1, species 940 and 946 are labelled more than
twice as frequently as any other taxon. Similarly for character 22-2, species 906 and 950 are fairly
highly labelled. These data suggest parallel derivation of the apomorphic condition of these
characters in taxa 906, 940, 946 and 950.
Three other characters are found in the apomorphic condition in all except one taxon in the
group . Of these ,61-1 has a high label score for taxon 921 , suggesting the apparent plesiomorphic
condition in this taxon may be a reversal. Character 54 is present in the apomorphic condition in
all taxa in this group except 935; elsewhere it is only found, in the data matrix, in the derived
state in three taxa, 908, 941 and 950. The higher values of these last two species suggest the
apomorphic state may have been derived in parallel in them. The third of the three characters,
58-1, is usually found in the derived state in Thyreodon, so may be presumed to have undergone
reversal in 904. Elsewhere it is only found in the apomorphic state in taxa 908, 909, 927, 941 and
950. Taxon 909 has a high label score for this character, suggesting homoplasy. It seems
reasonable to treat tentatively all these three characters as apomorphies of the group. This
interpretation is certainly the most parsimonious, for in each case the alternative to single
postulated reversals must be the advocacy of multiple forward parallelisms.
One other taxon, 908, seems to be consistently associated with this group. Its nearest
neighbours all lie within the group (Table 8), and in cluster analysis it readily associated with the
group. It shares 18 of the 19 apomorphies of the group and is discordant only in having character
Table 8 The Thyreodonl Euryophion complex showing nearest neighbours with numbers of shared
derived characters. The rows give the five (or more if several are equal fifth) nearest neighbours of each
taxon listed in the left-hand column. The columns show the number of times a species is cited as a near
neighbour. It can be seen that taxon 908 occupies an intermediate position between Thyreodon (901-4)
and Euryophion (935-7), having nearest neighbours in both. The Thyreodon and Euryophion groups
have only one near neighbour in common (904-936). Only species in the latter group share large
numbers of derived characters in common with non-group taxa (the 'others' column). It is interesting to
note that taxon 907, the least autapomorphic in this group (see p. 89), is cited as a near neighbour by all
other taxa.
901
902
903
904
907
921
908
935
936
937
901
902
903
904
907
921
908
935
936
937
Others
+
32
30
28
30
27
32
+
30
29
28
26
30
30
+
29
27
25
25
~
28
29
29
+
26
26
30
28
27
+
30
28
27
26
25
30
+
27
25
25
25
28
27
+
25
25
27
24
25
+
30
31
(23- 905, 906, 918)
26
26
30
+
32
(26- 946)
24
24
27
31
32
+
(25- 905; 24- 906)
PHYLOGENY OF THE OPHIONINAE 91
28-2 plesiomorphic. However, this character is highly labelled for this taxon suggesting a
reversal. In FOURS analyses taxon 908 repeatedly broke into the group. For example -
TAXANOS.-901 921 936 908
X
1100:
21-1 30
41
X'
0011:
16-1
Y'
0101:
18-2
Y
1010:
29-2 37
Z'
0110:
9
XYZ'
1110:
13 28-2
XY'Z
1101:
6-2 12
14-1 23
X'YZ
1011:
27-1 36-
1 61-1
X'Y'Z'
0111:
16-2 31
32 44-1
TAXA NOS. - 901
907 936
908
X
1100:
204 21-
1 30
X'
0011:
31 32
Y
1010:
37
Y'
0101:
18-2
Z'
0110:
43-1
XYZ'
1110:
13 28-2
29-2
XY'Z
1101:
6-2 12
14-1 23
X'YZ
1011:
27-1
X'Y'Z'
0111:
16-1 16-
2 44-1
33
33
In both these cases exclusion of taxon 908 from the group, the XYZ' option, is not the most
parsimonious solution. Accordingly, taxon 908, Rhynchophion flammipennis , is included for
analysis in this group.
Considering only the ten taxa of this group, 35 characters are informative for examining
intra-group relationships. A LeQuesne test on this subset of data yielded the results shown in
Table 9. Progressive removal of the five worst characters (59, 36-1, 27-2, 9 and 39) markedly
altered the ranking of the 'best' scoring characters (i.e. those with an O/E value of 0-6+) ; the
overall O/E ratio improved from 0-69 to 0-52 (Table 10). The data matrix was re-ordered with
characters arranged in increasing value of O/E ratio (Tables 11, 12) before and after removal of
the worst characters. Considering only the characters with an O/E value greater than 0-55 shows
that there are two competing patterns. Characters 16-2, 44-1, 31 and 32 unite the Euryophion
group (935-7) with the Dictyonotus group (907, 921) and 908. This arrangement is contradicted
by characters 14-1, 23, 33 and 6-2, which all favour uniting the Dictyonotus group with
Table 9 Results of a LeQuesne test on the Thyreodonl Euryophion data set. Conventions as in Table 2.
Character number: incompatibilities observed expected O/E ratio
6.1:
11
22.9
0.48
6.2:
11
22.9
0.48
8.1:
7
16.2
0.42
1 :
30
26.9
1.11
\2_ :
18
27.9
0.65
11 :
26
27.9
0.93
14.1:
11
23.6
0.47
16.1:
18
26.0
0.69
16.2:
11
26.0
0.42
17_ :
12
26.9
0.45
18.2:
24
26.9
0.89
11 '
17
23.6
0.72
20.1:
11
16.2
0.68
21.1:
18
26.4
0.68
21.2:
9
15.6
0.58
23 :
11
23.6
0.47
26 :
10
23.6
0.42
27.1:
9
15.7
0.57
27.2:
28
23.1
1.21
29.2:
22
26.9
0.82
30 :
20
23.6
0.85
li :
15
27.9
0.54
li :
15
27.9
0.54
JJ3 :
11
23.6
0.47
36.1:
21
15.7
1.34
36.2:
23
23.1
1.00
1Z. :
25
26.9
0.93
li :
17
26.9
0.63
41_ :
14
16.2
0.87
43.1:
19
26.9
0.71
44.1:
12
26.4
0.45
44.2:
9
15.6
0.58
45.1:
11
23.6
0.47
59 :
23
16.2
1.42
62 :
11
23.6
0.47
Grand total- 280 observed, 406.3 expected. Overall O/E ratio = 0.69
Ranking ratio of scoring characters
16.2 26 8.1 17 44.1 62 33 45.1 14.1 23 6.1 6.2 31 32 27.1 21.2 44.2 39 12 20.1
21.1 16.1 43.1 19 29.2 30 41 18.2 37 13 36.2 9 27.2 36.1 59
92
I. D. GAULD
Table 10 Results of a LeQuesne test on the ThyreodonlEuryophion data set after progressive removal of
five highest scoring characters (59, 36-1, 27-2, 9, 39). Conventions as in Table 2.
Character number: incompatibilities observed expected 0/E ratio
8.1: 3 13.9 0.22 \2_ : 14 24.1 0.58
16.1: 13 22.3 0.58 16.2: 7 22.3 0.31
^9 : 13 20.3 0.64 20.1: 10 13.9 0.72
^3_ : 6 20.3 0.30 ^6_ : 6 20.3 0.30
30 : 16 20.3 0.79 31_ : 11 24.1 0.46
36.2: 20 20.3 0.98 39 : 12 23.2 0.52
44.1: 8 22.7 0.35 44.2: 6 13.4 0.45
6.1: 6 19.6 0.31
^3 : 23 24.1 0.96
l]^ : 8 23.2 0.34
21.1: 13 22.7 0.57
27.1: 7 13.9 0.50
32 : 11 24.1 0.46
41^ : 11 13.9 0.79
45.1: 6 20.3 0.30
6.2: 6 19.6 0.31
14.1: 6 20.3 0.30
18.2: 22 23.2 0.95
21.2: 7 13.4 0.52
29.2: 19 23.2 0.82
23 : 6 20.3 0.30
43.1: 14 23.2 0.60
62 : 6 20.3 0.30
Grand total- 158 observed, 303.4 expected. Overall 0/E ratio = 0.52
Ranking ratio of scoring characters
8.1 45.1 62 23 26 33 14.1 6.1 6.2 16.2 17 44.1 44.2 31 32 27.1 39 21.2 21.1 12
16.1 43.1 19 20.1 30 41 29.2 18.2 13 36.2
Thyreodon (901^). Characters 6-1, 8-1, 17, 26, 45-1 and 62 clash with neither of these
combinations and so, in this context, are uninformative. The very best characters in the original
35 character set (Table 11) favour the former grouping whilst in the 'cleaned-up' 30 character set
(Table 12) the best characters favour Thyreodon + Dictyonotus. Of the eight characters in the
two contradictory groups, two, 16-2 and 33, are 'loss' apomorphies and perhaps therefore the
apomorphic state is particularly likely to be subject to parallel development. Even disregarding
these one is still left with a tie with three characters supporting each arrangement, so it is
necessary to consider further characters. Of the characters having an initial O/E value of average
or better (16-2-16-1 in Table ll)two, 12 and 21-1, suggest uniting Thyreodon with Dictyonotus if
one postulates minimal reversal in Thyreodon. Character 27-1 suggests uniting Thyreodon with
Euryophion and taxon 908, whilst 21-2 is an autapomorphy of taxa 907 and 921. Character 39
suggests linking a single Thyreodon species with Euryophion. Character 20-1 is apomorphic in
only two taxa, 901 and 907.
Considering all characters the most parsimonious arrangement appears to be to treat
Euryophion as the first branch. By postulating minimal homoplasy (i.e. single reversals or
parallelisms) 12 characters can be made to support this arrangement. Only three, 30, 43-1 and
44-2, have an O/E value below average, and only one, 33, is a 'loss' apomorphy. This
arrangement involves the following homoplasy (and these are considered to be predictable from
the Thyreodon labels matrix (Appendix 5) if, for a particular character, the score obtained by
the presumed discordant species is higher than that obtained by other species in the group) - 12,
reversal in 903/4 (not predictable); 30, reversal in 908, parallelism in 935 (predictable); 39,
Table 11 The Thyreodon/ Euryophion complex with all 35 informative characters ranked according to
their initial O/E value (lowest first). All characters to the left of 27-1 have an O/E value of 0-55 or less.
16.2 8.1 44.1 33 14.1 6.1 31 27.1 44.2 12 21.1 43.1 29.2 41 37 36.2 27.2 59
26 17 62 45.1 23 6.2 32 21.2 39 20.1 16.1 19 30 18.2 13 9 36.1
901
902
903
904
908
907
921
937
935
936
0101
0101
0101
000
1
,
1
1
1
011010
1
011010
1
011010
1
011010
1 1
011011
1 1
011010
1 1
011011
1 1
100101
10111
00011
00011
1
1 1 1
1 1
1101
1001
1111
0101
1
1
000
11101
11111
1 1
11100
1001
000
1 1 1
1 1
1
PHYLOGENY OF THE OPHIONINAE
93
Table 12 The Thyreodonl Euryophion complex with the 30 'best' informative characters ranked by O/E
value computed after progressive removal of five highest scoring characters (see Table 10). All
characters to the left of 21- 1 have an O/E value of 0-55 or less.
901
902
903
904
908
907
921
937
935
936
26
23
1 1
1
1
27.1
1 C
1
1
1
1
1
13.1
C
C
C
C
1
1 C
C
1 1
20.1
1 1
1111
1111
00111
2 13
18.2
1
1
1
1 1
parallelism in 904 (predictable); 43-1, parallelism in 907 (predictable); 44-2, parallelism in 921
and reversal in 935 (not predictable).
In attempting to resolve the sister-lineage of Euryophion, the Thyreodonl Dictyonotusl
Rhynchophion branch, it is apparent that no character is a unique apomorphy of either
Rhynchophion + Dictyonotus, Rhynchophion + Thyreodon or Dictyonotus + Thyreodon.
Character 30 suggesting the latter is paralleled in taxon 935, 16-2 and 44-1 suggesting Rhyn-
chophion + Dictyonotus are paralleled in Euryophion as is character 27-1 which favours uniting
Rhynchophion + Thyreodon. In any compatibility cladogram (e.g. Fig. 12) these three taxa
must remain as an unresolved trichotomy.
935/7
17
2, 7.1, 22.2, 24.1*, 24.2*. 25.1*, 25.2*. 28.1*.
38*. 42.1, 42.2, 47.1, 49, 51.1*. 51.2*
Fig. 12 Cladogram showing inter-relationships of taxa in the Thyreodon/ Euryophion complex as
supported by the largest compatible clique.
94
I. D. GAULD
To resolve these data further it was necessary to use the WAGNER option of PHYLIP. The
shortest rooted tree obtained involved 83 transformation steps for the 45 characters showing
both and 1 states for this group of taxa (that is the 35 characters shown in the LeQuesne matrix
(Table 11) plus 1-1,4-2, 8-2, 10, 18-1, 29-1, 36-3, 43-2, 47-2 and 57 which are autapomorphies of
various individual taxa). To obtain this tree (Fig. 13) 21 characters (1-1, 4-2, 6-1, 6-2, 8-2, 10,
14-1, 17, 184, 21-2, 23, 26, 29-1, 33, 36-3, 43-2, 44-2, 45-1, 47-2, 57 and 62) are presumed to have
been uniquely derived, 11 characters (8-1, 12, 16-2, 18-2, 20-1, 27-1, 39, 41, 43-1, 44-1 and 59)
have undergone two transformations (i.e. have either been derived in parallel or have
903 904
936 937 907 921
36.2
16.1, 30, 36.1
Fig. 13 Most parsimonious cladogram for taxa in the Thyreodonl Euryophion complex. Black squares
represent uniquely derived characters; circles, apomorphic features derived independently in two
lineages; diamonds, in three lineages; squares, in four lineages, or involving four transformations.
PHYLOGENY OF THE OPHIONINAE 95
undergone reversal), 12 characters (13, 16-1, 19, 21-1, 27-2, 29-2, 30, 31, 32, 36-1, 36-2 and 37)
have undergone three transformations, whilst one (9) underwent four. It is noteworthy that the
cladogram shows only one of a number of competing cladograms with the same arrangement of
taxa and the same number of transformation steps, but different positioning of the homoplastic
characters. For example, character 30 is shown as a group apomorphy and is postulated as
undergoing reversal in the stem 936/7, and in taxon 908, but it would be equally parsimonious to
have suggested that 30 was derived in parallel in stems 901/4, 907/21 and taxon 935. Biologically
neither is implausible.
The Stauropoctonus group (Taxa 916, 917, 918)
These three taxa share 29 apomorphies (1-1, 5-1, 5-2, 14-2, 16-1, 16-2, 18-2, 21-1, 22-1, 25-1,
28-1, 28-2, 29-2, 33, 37, 38, 39, 42-1, 42-2, 45-1, 47-1, 49, 51-1, 51-2, 52-1, 52-2, 61-1, 61-2 and
62). Six of these apomorphies (5-1, 5-2, 14-2, 21-1, 42-2 and 52-2) have both lower than average
O/E values (0-84, 0-75, 0-83, 0-79, 0-79 and 0-71 respectively) and are shared by relatively few
other taxa in the matrix.
Within this group all three possible arrangements of taxa are supported: 916 + 917 by 25-2*
and 47-2; 916 + 918 by 1-2*, 36-1', 36-2', 43-1' and 63; 917 + 918 by 57*. All of these eight
informative characters have high LeQuesne test failure rates (0-9+) and most are either loss
characters or have dubiously assigned polarity, making the choice between arrangements
difficult. The most parsimonious would obviously be 916 + 918. Currently 916 and 917 are
placed together in Stauropoctonus whilst 918 is placed in a separate taxon, Aulophion (Cush-
man, 1947; Townes, 1971). Aulophion has traditionally been separated from Stauropoctonus by
the possession of two autapomorphies, loss of the posterior transverse carina of the meso-
sternum and absence of the epicnemial carina. This suggests Stauropoctonus may be a
paraphyletic assemblage. For the present these taxa are treated as an unresolved trichotomy,
but their inter-relationship is discussed further below (p. 146). It is sufficient now to state that
these three taxa seem to form a closely knit and apparently holophyletic group.
The Enicospilus + Stauropoctonus complex (Taxa 909-919, 924-926, 928-934, 942-947)
This, the largest apparent group of ophionine taxa in the matrix, contains 28 species. (N.B.
Taxon 949 which associates with this group in the cluster analysis is exceptional in having the
plesiomorphic state of a number of characters usually found in the apomorphic condition in this
group (e.g. 25-2, 28-2, 42-1, 51-1, 51-2 and 63). For the present this taxon is excluded from the
complex and its position discussed later in the work.) Their nearest neighbours are shown in
Table 13. These taxa share four apomorphic features (16-1, 25-1, 28-1 and 51-1). A further 13
characters (22-1, 25-2, 28-2, 33, 36-1, 38, 42-1, 43-1, 45-1, 47-1, 49, 51-2 and 63) are present in the
apomorphic state in all except one, two or three taxa. Most of these characters are labelled,
often highly, for the species in this group that show the plesiomorphic condition (e.g. 38 for
taxon 943 and 47-1 for taxa 944 and 945), suggesting a reversal may have occurred. The
exceptions are 22-1 and 33 which are plesiomorphic for taxa 910 and 911. Both are unlabelled in
the primary label matrix (Appendix 4). Of the 17 characters listed above as possible apomor-
phies of this complex, all except 22-1, 33 and 63 occur in the apomorphic condition in a large
number of other ophionine taxa in the matrix. Individually therefore, they are not good
characters for defining the group. Character 49 is plesiomorphic only in the solitary taxon 926,
suggesting 49 may be an apomorphy of all ophionines, and 926, an otherwise specialized species,
may have undergone a reversal. Character 16-1 is apomorphic for all taxa except 901-3 and 921;
character 25-1 for all except 948 and 951 ; character 25-2 for all except 918, 920, 922, 923, 938-41,
945, 948, 949 and 951; character 28-1 for all except 939; character 28-2 for all except 908, 920,
923, 928, 938-41, 949 and 951; character 36-1 for all except 909, 917, 920, 921, 922, 937 and 944;
character 38 for all except 938, 939, 941, 943, 948 and 951; character 42-1 for all except 905, 922,
925, 938-41 and 948-50; character 43-1 for all except 901^, 908, 909, 917, 921-22, 925 and 941;
character 45-1 for all except 901-4, 907-8, 921, 923, 925-26, 931, 941 and 948; character 474 for
all except 938, 939, 943, 944, 948 and 951; character 51-1 for all except 920, 922, 939-41 and
949-51; and character 51-2 for all except 920, 922, 932, 938-41 and 948-51. Ten of these
characters (25-1, 25-2, 28-1, 28-2, 38, 42-1, 474, 49, 514 and 51-2) have been stated above (p.
96 I. D. GAULD
Table 13 The EnicospiluslStauropoctonus complex showing nearest neighbours with numbers of shared
derived characters. Conventions as in Table 8. Particularly striking are taxa 944 (an Enicospilus species)
and 931 (a Leptophion species). Both cite their congeners as nearest neighbours, but neither is cited as
near neighbours by its congeners.
926 925 924 947 946 945 944 943 942 934 933 932 919 915 914 913 912 931 930 929 928 918 917 916 909 910 911 Others
926 + -- 25 25 27 -- -- 24 27 - ,-
925 + 27 28 30 28 -- -- 24 24
924 25 27 + 27 29 29
947 + 35 31 -- 31 34 30 30 30
946 35 + -- -- 37 32 32 -- 32
945 31 31 + -- 28 31 28 28 28
944 24 26 24 + 26 24
943 31 30 28 + 31 28
942 34 37 31 31 + 31
934 30 32 31 + -- 30 -- 30
933 29 29 29 29 + -- 29
932 28 32 30 30 - + -- -- 29 -
919 29 29 29 29 29 + 32 33 30 29 30
915 32 + 36 33 34 31
914 32 33 36 + 33 33
913 29 29 33 33 + 31
912 - 30 34 33 31 + 29
931 - 27 27 27 + -- 27 27
930 28 29 28 29 -- 28 -- + 29 30
929 30 29 -- 29 29 + 29
928 30 31 29 30 29 +
918 28 - 28 28 - 28 - - + 30 34
917 24 24 24 -- 30 + 31 25 --
916 27 27 34 31 + 27 -- --
909 24 24 24 26 25 27 + -- -- 24-936
910 26 - -- 26 28 28 26 26 -- 26 + 30
911 -- 28 - 28 30 27 30 +
90) to be apomorphies of the ThyreodonlEuryophion complex, suggesting they are characters
that unite the EnicospiluslStauropoctonus and ThyreodonlEuryophion generic complexes.
Character 16- 1 could also be considered an apomorphy of both these major groups if one were to
postulate reversal in 901-3 and 921.
Character 22-1 is found in the apomorphic state outside the EnicospiluslStauropoctonus
complex only in taxa 920, 940, 949 and 951; character 33 in taxa 901^, 907-8, 920-22 and 949
and character 63 in 905, 906 and 923. Character 33 is a loss apomorphy and has been cited above
as an apomorphy of the ThyreodonlDictyonotus lineage. It is perhaps not unreasonable to
postulate parallel loss in the EnicospiluslStauropoctonus, the ThyreodonlDictyonotus and the
920, 922 and 949 lineages. However, this feature, the loss of the umbo, is not apomorphic for
taxa 910 and 911 included above in the Enicospilus group. It is biologically rather unlikely that
the umbo would be lost then redeveloped, so this feature favours placing Ophiogastrella
(910-11) primitive with respect to the rest of the group. Such a position is also favoured by
character 22-1. The apomorphic condition, a lengthened anterior part of the propodeum, is
rather unlikely to have undergone reversal.
It is interesting to note that character 22 has two alternative derived states, 22-1 or 22-2. The
derived state 22-2 is an apomorphic feature of the ThyreodonlEuryophion complex and is only
found elsewhere in taxa 906 and 927. The alternative derived state characteristic of the
EnicospiluslStauropoctonus group (less Ophiogastrella) is also found in very few other taxa (see
above). The plesiomorphic condition (0,0) is found in taxa 905, 910, 911, 922, 923, 938, 939, 941
and 948. The character seems to be important in defining two major sister-lineages of the
subfamily.
PHYLOGENY OF THE OPHIONINAE
97
Using the CLIQUE option of PHYLIP eight cliques were found with 12 or more informative
characters. These were
A (12)
B(12)
C(12)
D(12)
E(13)
F(12)
G(12)
H(12)
ADB
18-1, 18-2, 21-1, 22-1, 27-3, 33, 35, 36-3, 42-2,61-1, 61-2, 64]
15-1, 15-2, 18-1, 18-2, 21-1, 22-1, 27-3, 33, 35, 36-3, 42-2, 64]
17, 18-1, 18-2,21-1, 22-1, 27-3, 31, 33, 39, 42-2, 60, 64]
15-1, 17, 18-1, 18-2, 21-1, 224, 27-3,31, 33, 42-2, 60, 64]
12, 14-2, 17, 18-2, 21-1, 22-1, 27-3, 31, 33, 39, 42-2, 60, 64]
12, 15-1, 17, 18-2, 21-1, 22-1, 27-3, 31, 33,42-2, 55, 64]
13, 18-1, 18-2, 21-1, 22-1, 33, 35, 36-3, 42-2, 61-1, 61-2, 64]
13, 15-1, 15-2, 18-1, 18-2, 21-1, 22-1, 33, 35, 36-3, 42-2, 64]
H = (18-2, 21-1, 22-1, 33, 42-2, 64}
Av. O/E = 0-76
Av. O/E = 0-75
Av. O/E =0-78
Av. O/E = 0-77
Av. O/E = 0-78
Av. O/E = 0-76
Av. O/E = 0-79
Av. O/E = 0-78
The best clique in the sense of the largest is E but this has a lower average O/E value (0-78) than
several others. B has the lowest average O/E value with 0-75, closely followed by A and F with
0-76. The remainder have higher values. The cladograms based on the largest clique (E) and the
best O/E scoring clique (B) are shown in Figs 14, 15. The cladogram produced from clique B
associated the various taxa fairly well with their congeners. Species of Enicospilus (942-47),
Dicamptus (932-34) and the Hawaiian genera (924-26) (regarded by Townes, 1971, as derived
Enicospilus) cluster together, except for E. cionobius (945) which is excluded by having the
plesiomorphic condition of 15-2 (possibly a reversal). Laticoleus (912-15) and Leptophion
(928-30) species form a separate cluster as do the Stauropoctonus group (916-18) plus
Lepiscelus (909). The cladogram derived from clique E has more confusing groupings, uniting
some species of Enicospilus and Laticoleus (character 17) or defining a clade containing
Dicamptus neavei, Leptophion tetus and Abanchogastra hawaiiensis (character 12). The heter-
ogeneous collection of species united by character 14-2 includes representatives of Dicamptus,
Enicospilus, Laticoleus and Leptophion; other species of these genera are excluded. The largest
clique, E, is therefore not considered particularly useful in this case; clique B appears to be a
better indicator of relationship. It is noteworthy that clique B is more informative (s = 0-681)
than clique E (s = 0-632).
Fig. 14 Cladogram for taxa in the Enicospilus/ Stauropoctonus complex based on the largest clique (E).
98
I. D. GAULD
Fig. 15 Cladogram for taxa in the EnicospiluslStauropoctonus complex based on the favoured clique (B) .
It is notable that the intersect of these cliques includes several characters previously
considered to be good indicators of phylogenetic relationship (Cushman, 1947; Townes, 1971).
Three (18-2, 21-1 and 42-2) support the group 909 + 916-18, whilst two others (22-1 and 33)
place Ophiogastrella (910-11), primitive to the other taxa. Character 64 is an autapomorphy of
Ophiogastrella.
A LeQuesne test was undertaken on the EnicospiluslStauropoctonus data set and the O/E
values are given in Table 14. Stepwise elimination of all characters scoring worse than 1-00 (a
total of 15) resulted in considerable cleaning up. The overall ratio improved from 0-89 to 0-74.
Amongst the scoring characters the most striking changes in rank occurred to 1-2 which rose
from thirty-fifth to eleventh position, and 62 which dropped from thirty-first place to position
forty-three. Amongst the best scoring characters 42-2 and 18-2 rose from fourteenth and
sixteenth positions to seventh and eighth positions (Table 15). Using the S option of FOURS the
data set was reorganized with characters ranked as per O/E value after removal of the 15 most
discordant characters (Table 16). It can be seen that the characters with the lowest O/E values,
22-1 and 33, exclude Ophiogastrella (910-11) whilst 64 is an autapomorphy of this genus. A
number of other characters may be postulated as apomorphies of Ophiogastrella, though all
apparently have been derived in parallel elsewhere. Characters 18-2 and 42-2 support the
Stauropoctonus group -I- 909, whilst 21-1 supports just the former. Character 1-2 is incompatible
with 21-1. Initially it had a poor O/E score (0-93 compared with 0-54 for 21-1) but progressive
elimination of the poorest characters produced rapid 'clean-up' until with 15 characters
eliminated it scored 0-57 compared with 0-34 for 21-1. With only the 18 best characters left in the
matrix both taxa scored equally 0-13. A considerable number of slightly homoplastic characters
(e.g. 5-1, 5-2, 37, 52-1) support 21-1 but no other character supports 1-2 and it is for this reason
that one would prefer the arrangement supported by the former character.
Several other characters support the Stauropoctonus + 909 group but necessitate postulating
parallel derivation in other places . These include 1 1 (parallelism in 925 and 945) , 39 (parallelism
in 930) ,61-1 (parallelism in 924-26 and 942-47) . Many of these parallelisms are also suggested in
the reduced labels matrix (Appendix 6), e.g. taxon 930 for character 39 acquires almost half of its
total number of labels and taxa 925 and 945 are quite highly labelled for character 1-1.
PHYLOGENY OF THE OPHIONINAE
99
Table 14 Results of a LeQuesne test on the EnicospiluslStauropoctonus data set. Conventions as in Table
2.
Character number: incompatibilities observed expected 0/E ratio
1.1
40 42.7 0.94
1.2
27 29.2
0.93
1
33
47.6
0.69
4.1:
39 35.8
1.09
5.1
45 50.5 0.89
5.2
41 42.3
0.97
6.1
28
19.9
1.40
7.2:
39 51.5
0.76
li
27 29.7 0.91
11
10 19.9
0.50
14.2
43
51.5
0.84
15.1:
37 42.3
0.88
15.2
39 50.0 0.78
16.2
48 40.1
1.20
17
25
29.7
0.84
18.1:
11 19.7
0.56
18.2
27 35.5 0.76
11
27 19.9
1.35
20.1
38
34.9
1.09
20.2:
46 50.1
0.92
21.1
16 29.7 0.54
22.1
5 19.9
0.25
24.1
49
48.1
1.02
24.2:
50 44.7
1.12
25.2
26 19.9 1.30
27.1
41 46.2
0.89
27.2
41
46.2
0.89
27.3:
11 19.1
0.58
29.2
34 43.2 0.79
30
22 19.9
1.10
li
19
19.9
0.95
33 :
5 19.9
0.25
35
28 47.6 0.59
36.1
33 28.6
1.15
36.2
44
44.3
0.99
36.3:
25 38.8
0.64
11
50 49.0 1.02
39
36 40.1
0.90
40
34
47.6
0.71
42.2:
27 35.8
0.75
43.1
32 29.0 1.10
43.2
50 48.3
1.03
44.1
55
48.2
1.14
44.2:
43 35.0
1.23
45.1
30 29.0 1.04
45.2
45 49.4
0.91
46
42
50.2
0.84
47.1:
15 19.5
0.77
47.2
52 50.5- 1.03
48.1
47 50.5
0.93
48.2
39
46.6
0.84
52.1:
45 48.1
0.94
52.2
40 42.3 0.95
li
25 35.8
0.70
56
28
19.9
1.40
5_7 :
35 35.8
0.98
59
48 51.7 0.93
60
17 19.9
0.85
61.1
35
50.7
0.69
61.2:
40 50.5
0.79
62
36 40.1 0.90
64
5 19.9
0.25
Grand total- 1035 observed, 1161.2 e
xpected.
Overall 0/E
ratio
= 0.89
Ranking ratio of scoring
characters
22.1 33 64 13 21.1
18.1 27.3 35 36
.3 61.1
3 55 40
42.2
7.2 18.
2 47.1
15.2 29.
2
61.2 14.2 46 48.2 17
60 15.1 27.1
27.2 5.1
39 62 12
45.2
20.2
1.2 59
48.1 52.
1
1.1 52.2 31 5.2 57
36.2 24.1 37 47
.2 43.2
45.1 20.1
4.1
43.1 30
24.2
44.1 36.1
16.2 44.2 25.2 19 6.1 56
Table 15 Results of a LeQuesne test on the EnicospiluslStauropoctonus data set after progressive
removal of 15 highest scoring characters (56, 19, 6-1, 25-2, 44-2, 16-2, 24-2, 44-1, 30, 57, 24-1, 43-2, 36-1,
45-1, 36-2). Conventions as in Table 2.
Character number: incompatibilities observed expected 0/E ratio
1.1
26 33.2
0.78
1.2
13 23.0 0.57
3
23
36.9
0.62
4.1
27 28.3
0.96
5.1
33 38.7
0.85
5.2
29 32.8 0.89
7.2
27
39.7
0.68
li
20 23.5
0.85
11
7 15.8
0.44
14.2
32 39.7 0.81
15.1
23
32.8
0.70
15.2
27 38.3
0.70
17
18 23.5
0.76
18.1
9 15.5 0.58
18.2
13
28.0
0.46
20.1
24 27.4
0.88
20.2
32 38.4
0.83
21.1
8 23.5 0.34
22.1
1
15.8
0.06
27.1
30 35.5
0.85
27.2
30 35.5
0.85
27.3
6 14.9 0.40
29.2
25
33.7
0.74
li
14 15.8
0.89
33
1 15.8
0.06
35
20 36.9 0.54
36.3
17
31.4
0.54
37
36 37.9
0.95
39
22 31.4
0.70
40
25 36.9 0.68
42.2
13
28.3
0.46
43.1
22 23.5
0.93
45.2
34 38.7
0.88
46
32 38.7 0.83
47.1
10
15.3
0.65
47.2
38 38.8
0.98
48.1
33 38.7
0.85
48.2
29 35.9 0.81
52.1
33
37.0
0.89
52.2
30 32.8
0.91
55
20 28.3
0.71
5!
36 39.9 0.90
60
14
15.8
0.89
61.1
24 38.9
0.62
61.2
28 38.7
0.72
62
29 31.4 0.92
64
1
15.8
0.06
Grand total -
522 observed 708.4 expected Ove
rail 0/E rat
io =
0.74
Ranking ratio of
scoring
characters
22.1 33 64 21
.1 27.3
13 42.2 18.2 36.3 35
1.2 18.1
61.1
3 47.1
40 7.2 39 15
1
15.2 55 61.2
29.2 17
1.1 14.2 48.2 46 20.2
27.1 27.2
12
5.1 48
1 20.1 45.2 5
2
31 60 52.1 59 52.2 62 43.1 37 4.1 47.2
100 I. D. GAULD
Table 16 The EnicospiluslStauropoctonus complex data set reordered according to O/E values of the
characters given in Table 15. Characters to the left of 17 have an O/E value of average (0-74) or less.
22.1 64 27.3 42.2 36.3 1.2 61.1 47.1 7.2 15.1 55 29.2 1.1 48.2 20.2 27.2 5.1 20.1 5.2 60 59 62 37 47.2
33 21.1
13
18.2
35
18.1
3
40
39
15.2 61.2
11
14.2
46 27
,1
12
48.1
4b
,1
31
52.1
52
.2
43.1
4.
1
924
11001
1
1
1
1
1
1010
000
1
1 1
e
o
1 1
926
11001
1
1
1
1
1011
1
1
000
I
1 1
1
e
e
1
926
11000
1
1
1
1
1
1010
1
1
\
1
1
e
1
1
947
11000
1
1
1
1011
1
1 1
1 1
1
1 1
1
1
1
1
1
1
946
11000
1
1
1
1
1
1011
1
1 1
1
1 1
1
1
1
1 1
945
11000
1
1
1 1
1
0111
1
1 1
1
1 1
1
I
1
1
i
944
11000
1
1
1
1001
000
1
1
1
e
1
1
943
1 1 00-0
1
1
1
1011
1
1 1
1
1
1
1
1
1
1
1
1
e
942
11000
1
1
1
1
1
1011
1 1
1 1
1
1 1
1
1
1
1 1
934
11000
1 1
1 1
1
1001
1
1 1
1 1
1
1
1
1
e
933
11000
1
1 1
1
1001
1
1 1
1
1
1
1
1
1
932
11000
1
1 1
1
1001
1
1
1
1 1
1
1
1
1 1
919
11000
1 1
1
1
0001
1
1
1
1
1 1
1
1
1 1
1
931
11000
1
1
1 1
1
1
0001
000
1
1
1
1
1
1
930
11000
1
1 1
1
1 1
0101
1
1
1
1
1
1 1
929
11000
1
1
1
1
1
0101
1
1
1
1
1
1 1
9?8
11000
1
1 1
1
1
0101
1
1
1
1
1
1
1 1
1
915
11000
1
1 1
1
1
0000
1
1
1 1
1
1 1
1
1 1
1
1 1
1
914
11000
1
1 1
1
1
0000
1 1 1
1
1 1
1
1 1
1
1
1 1
1
913
11000
1
1
1
1
0000
1
1
1 1
1
1 1
1
1
1 1
912
11000
1
1 1
1
1
0000
1
1 1
1
1 1
1
1
1
1 1
909
11000
1
1
1
1
1
1
"1 1
1
1
000
1
1
1
1
918
11010
1
1
1
1
1
1
0011
1
1
1
1
1
1
1
1
1
1
1 1
917
11010
1
1
1
1
1
0011
1
1
000
1
1
1
1
1
1
1
1
916
11010
1
1
1
1
1
1
0011
1
1
000
1
1
1 1
1
1
1 1
1
911
00100
1
1
0001
1
1 1
1
1 1
1 1
1
1 1
910
00100
1
1
0001
1
1 1
1
1 1
1 1
1 1
Character 15-1 is compatible with those characters that suggest treating 910 + 911 and 909 +
916-18 as the first two branches. This unites all the other taxa to form a residual group that
includes Enicospilus (942-47), Dicamptus (932-34), Leptophion (928-31), Laticoleus (912-15)
and Pamophion (919). This is rather a difficult group to resolve. Two of the best characters, 13
and 27-3, are incompatible. Character 35 suggests uniting Laticoleus (912-15) with Leptophion
(928-3 1 ) , and this is partially supported by character 3 . No character unites all the taxa that show
the plesiomorphic condition of 35. Three characters, 5-1, 61-1 and 61-2, suggest uniting the
Hawaiian genera (924-26) with Enicospilus (942-47). Only one taxon, 944, is at variance for a
single character, 61-2, and this has a high label value suggesting that a reversal to the
plesiomorphic state may have occurred. In several places in the data set it is apparent that similar
cases of a single taxon being at variance with its congeners is observable. In the case of 944, it is
one of only about ten Enicospilus species out of nearly a thousand to have the plesiomorphic
condition of character 61-2. Similarly 945 has the plesiomorphic state of 15-2. Both of these taxa
do not appear to be primitive in any other features and as they are thus unlikely to be ancestral to
all other taxa in the genus it is probable that the apparently unspecialized condition is a reversal.
Character 15-2 favours uniting the Hawaiian genera, Enicospilus and Dicamptus (932-34) and
this is partially supported by character 40 although it is necessary to postulate reversal in the
Hawaiian genera. Character 46 is similar. Character 7-2 is contradictory, suggesting a group
comprising 912-15, 919 and 928-34. No characters satisfactorily separate Laticoleus (912-15)
from Leptophion (928-31) though 27-1, 27-2 and 55 suggest some separation. Character 3
suggests that Pamophion belongs to the Laticoleus I Leptophion lineage.
PHYLOGENY OF THE OPHIONINAE
101
Using the CLIQUE option of PHYLIP on the data set with the basal taxa (910-11, 909,
916-18) removed revealed the existence of four cliques with eight or more informative
characters. These are
A (8)
B(9)
C(8)
D(9)
4-1, 5-1, 18-1, 27-3, 35, 36-3, 61-1, 61-2]
4-1, 5-1, 27-1, 27-2, 27-3, 35, 36-3, 61-1, 61-2]
4-1, 54, 13, 18-1, 35, 36-3, 61-1, 61-2]
4-1, 5-1, 13, 27-1, 27-2, 35, 36-3, 61-1, 61-2]
Av. O/E = 0-81
Av. O/E = 0-83
Av. O/E = 0-86
Av. O/E = 0-87
A n B C D = (4-1, 5-1, 35, 36-3, 61-1, 61-2}
These sets support very similar cladograms; the favoured one, with the lowest average O/E, is
shown in Fig. 16, but all are similar in showing two large groups (912-15 + 928-31 and 924-26 +
942-47) and leaving taxa 919 and 932-34 unresolved.
18.1
61.2
I
5.1
61.1
Fig. 16 Cladogram for taxa in the EnicospiluslStauropoctonus complex less taxa 909-911 and 916-8,
based on the favoured clique (A).
Using the WAGNER option of PHYLIP several attempts were made to construct the shortest
rooted tree possible. The minimum length tree (Fig. 17) necessitated 242 transformation steps.
Both parsimony and compatibility analyses have certain similarities. Both taxa 909 + 916-18
formed a group primitive to most other taxa whilst Enicospilus (942-47) and Dicamptus
(932-34) are amongst the most specialized taxa and separate from Laticoleus (912-15) and
Leptophion (928-31). However, the Wagner parsimony method usually separated the Hawaiian
taxa (924-26) as a relatively primitive, discrete group but placed Ophiogastrella (910-11) well in
the Enicospilus/ Dicamptus/ Leptophion/ Laticoleus complex. Compatibility methods suggested
the reverse. To assess the relative merits of these alternative arrangements it is necessary to
evaluate the characters upon which they are based. As mentioned above, the exclusion of taxa
910 and 911 is based on characters 15-1, 22-1 and 33 which are present in the apomorphic
condition in all other taxa. The former two are striking structural modifications and whilst the
latter is a loss apomorphy , it is an unusual reduction (when considered for the family as a whole).
There is no apparent functional reason why these characters should be linked, so their
congruence can be viewed as strong evidence for excluding 910-11 as the first branch. However,
910 and 911 do have a number of apomorphies in common with many other taxa in the group,
102
I. D. GAULD
IO CO
Fig. 17 Minimum length cladogram produced by Wagner parsimony analysis of the Enicospilusl
Stauropoctonus complex. This tree requires 242 transformation steps.
especially the Laticoleusl Leptophion complex (e.g. characters 7-2, 27-1, 27-2, 20-1) which it
must be assumed has been derived in parallel if Ophiogastrella is indeed the most primitive
branch.
The Wagner analysis united all taxa excluding the Stauropoctonus/909 complex on the basis of
characters 15-1, 36-2 and 43-1. The former has been postulated as having undergone reversal in
the stem 910 + 911 so cannot be considered a group apomorphy, whilst 36-2 is highly
homoplastic in any favoured arrangement, and in the minimum length tree necessitates six
transformations that include being developed, lost and subsequently redeveloped. In fact, this
character is also present in the apomorphic condition in some Stauropoctonus (916, 918), and it
is not stretching credibility to consider it an apomorphy of the entire group here being analysed,
and postulate reversal in taxa 909 and 917. This hypothesis involves only a single additional
transformation step, and would seem biologically more feasible than the gain-loss-gain scheme
favoured by the Wagner analysis. The character itself, relative position of two wing veins, has
not been used in higher classification, though it is of considerable use in separating species
(Gauld & Mitchell, 1978), and has doubtfully assigned polarity. It is also amongst the very worst
characters suggested by the original LeQuesne test (see Table 6). The third character, 43-1, is
found widely in the apomorphic condition throughout the Ophioninae. Either it has been
PHYLOGENY OF THE OPHIONINAE 103
derived in parallel in the Euryophion group, in most Ophion and related taxa and in certain
members of the present group, or it is an apomorphy of the Ophioninae and the apparently
plesiomorphic state of most Stauropoctonus group taxa represents a further apomorphy.
Initially the polarity of this character was considered to be tentatively assigned (p. 77).
Elsewhere in the Ichneumonidae, although reliance is often placed on character 43-1 in generic
keys, it is variable in most higher taxa (cf. Pimplinae in Townes, 1969), suggesting it is a highly
homoplastic feature.
Other characters involved in placing Ophiogastrella high up the tree are 7-2, 20-2, 27-1, 27-2,
37, 43-2, 45-2, 48-1 and 59. Of these, 7-2 and 59 are highly homoplastic, gain-loss-gain characters
which seem biologically implausible, and they also have dubious polarity assignment. Character
45-2 is postulated as having undergone reversal in the stem 910 + 911 so there remains five
reasonably robust apomorphies uniting Ophiogastrella with the most specialized taxa (20-2,
27-1, 27-2, 43-2 and 48-1). Considered individually, all of these characters are rather 'weak'. The
remarks made above about 43-1 can also be applied to 43-2; the loss apomorphies 27-1 and 27-2
have identical state distributions in this group and should perhaps be considered as a single
feature, the loss of the posterior mesosternal transverse carina. This carina has been lost,
presumably independently, in other ophionine evolutionary lines (e.g. Ophion, some Lepto-
phion, some of the Stauropoctonus group) and has commonly been lost in many ichneumonid
evolutionary lineages outside the subfamily under consideration. As a consequence it is not a
character which seems to unite convincingly Laticoleus (912-5) with Ophiogastrella. It is
interesting to note that there is at least some degree of correlation between the presence or
absence of this carina and the type of habitat occupied. In groups where it is usually absent (e.g.
the Phygadeuontinae), species inhabiting very wet areas have the carina complete. In the
subfamily mentioned this includes common species of Paraphylax, Amauromorpha and
Apsilops found in Old World rice padi. In the Anomaloninae the two closely related genera
Therion and Heteropelma differ in this feature; the former is characteristically found on dry open
areas, whilst most Heteropelma species occur in more humid woodlands (Gauld, 1976). In the
Ophioninae the carina is always present in those species found in humid rain forests (e.g. most
Enicospilus, Leptophion and Dicamptus) but is incomplete in species favouring drier, more
exposed habitats (e.g. some Australian Leptophion, most Ophion). Both Ophiogastrella species
and Laticoleus seem to favour drier forest habitats than Leptophion.
Character 20-2, the development of scutellar carinae, has probably been derived indepen-
dently in several evolutionary lineages (e.g. some species of Ophion and Leptophion (Gauld &
Mitchell, 1981)), but it also appears to have undergone reversal in other groups (e.g. in taxon
944 and also Enicospilus arduus (Gauld & Mitchell, 1978)). It would be only slightly less
parsimonious to suggest that the presence of complete scutellar carinae is an apomorphy of the
entire Enicospilus I Stauropoctonus complex, and to suggest that it has been lost in a few lineages.
Certainly the shared presence of these carinae is not a convincing character for uniting
Ophiogastrella with other genera.
The remaining character, 48-1, is a difference in relative lengths of wing veins and its
apomorphic state is found scattered throughout the subfamily.
To summarize therefore, the decision on where to position Ophiogastrella hinges on two
character sets, one of three unusual and fairly convincing apomorphies (15 1 , 22- 1 and 33) versus
five widely distributed, and thus unconvincing, apomorphies (20-2, 27-1, 27-2, 43-2 and 48-1).
The Wagner method, in attempting to minimize tree length, opts for including Ophiogastrella
amongst the most derived species, whilst the compatibility method favours treating it as the most
primitive taxon, thus eliminating homoplasy in three characters (22-1, 33 and 15-1) as the other
characters are mutually incompatible.
The position of the Hawaiian genera (924-6) also presents a problem. In the parsimony
dendrogram (Fig. 17) the five characters defining the stem 924 + 925 + 926 (5-1, 15-2, 36-3, 61-1
and 61-2) also define at least part of the Enicospilus + Dicamptus branch (932-4 + 942-7)
suggesting the Hawaiian genera could be placed on the larger branch, as in the compatibility
dendrogram. The parsimony method has excluded the Hawaiian genera on the basis of
characters 7-2, 20-2, 40, 43-2, 45-2, 46, 48-1 and 59. Of these 7-2 and 59 are the biologically
104 I. D. GAULD
implausible gain/loss/gain characters with dubiously assigned polarity, so the real choice has to
be made between the two character sets 5-1, 15-2, 36-3, 61-1 and 61-2 versus 20-2, 40, 43-2, 45-2,
46 and 48-1.
Considering the first set first, characters 5-1, 61-1 and 61-2 are all mandible characters. The
apomorphic state of all three is elsewhere only found in taxa 916-8 and 942-7 (excluding 944).
Although this character set has almost certainly been derived independently in Stauropoctonusl
Aulophion and Enicospilus it is not found elsewhere in the Ophioninae. It is of very rare
occurrence in other parts of the family, and where it does occur it is usually characteristic of a
group of genera e.g. the Orthocentrinae (Townes, 1971) and the XanthopimplalEchthromorpha
group of Pimplini (Townes, 1969). Character 15-2, a completely concealed spiracular sclerite, is
found in virtually all species of Enicospilus and Dicamptus. Elsewhere it is found in the derived
condition only in the Hawaiian genera, in a few species of Leptophion and in the aberrant
monotypic genus Sicophion (922). The final character in this set, 36-3, an extreme reduction in
the length of a wing vein, is only found in a very few taxa, notably (in the matrix) the Hawaiian
genera, some Enicospilus, Euryophion and Xylophion. Elsewhere in the Ophioninae it only
occurs in some Dicamptus and one aberrant Laticoleus.
The second and alternative character set (20-2, 40, 43-2, 45-2, 46 and 48-1) has partially been
considered above where characters 20-2, 43-2 and 48-1 have been suggested to be rather poor
indicators of phylogenetic affinity. Two of the Hawaiian genera show the derived state of 48-1,
so it would be equally parsimonious to postulate that a reversal had occurred in one Hawaiian
genus, as opposed to parallel derivation in the subgroup of two Hawaiian taxa and the major line
in the larger group. The remaining characters all refer to the fore wing, in particular the
antero-distal part of the discosubmarginal cell and the adjacent vein, Rs+2r. The apomorphic
conditions of characters 40, 45-2 and 46 are the most characteristic features of Enicospilus/
Dicamptus, though the same combination is also found in a very few other taxa (e.g. Riekophion
species). Several species of Enicospilus may have one or more characters in the plesiomorphic
state (e.g. the E. senescens species-group (Gauld & Mitchell, 1978)), but the trace features
found in one or two species suggests this condition represents a reversal. This suite of characters
does seem therefore to be good evidence for excluding the Hawaiian taxa from the Enicospilus
lineage. However, the plesiomorphic condition of all three of these characters occasionally is
found in some Enicospilus species inhabiting oceanic islands, and in these cases it can be seen
that a reversal has occurred as related taxa still have the apomorphic states. For example, the
three species E. vidus, E. ditor and E. donor form a closely interrelated group on the Galapagos
Islands (Gauld & Carter, 1983) . They belong to the E. capensis species-group and probably have
evolved from a migrant South American species, all of which have the apomorphic condition for
characters 40, 45-2 and 46, as has E. vidus. In E. donor they all are plesiomorphic, whilst in E.
ditor intermediate conditions exist. This strongly suggests reversal has occurred. A similar case
can be demonstrated for Hawaiian Enicospilus (see Cushman, 1944). The majority of Enicospi-
lus species are both synchronous and sympatric with numerous congeners. The three characters
are developed in such a way as to form a distinctive, species-specific pattern (especially the exact
shape of the sclerite, character 46) (see figs 384-641 in Gauld & Mitchell, 1981). In areas where
few species occur, such as deserts, there is considerable variation within a species in the exact
expression of these characters. Possibly their reversal on oceanic islands is facilitated by a
reduction in selection pressure that favours uniformity in areas of high possible interspecific
interaction. It is possible that the apparent plesiomorphic condition of these characters in the
Hawaiian genera is also a reversal, though in this case no definite close relatives are known.
The remaining character in the second set, 48-1, refers to the relative lengths of veins 3r-m
and M . The apomorphic condition has previously been used only to characterize species and
both states can be found in most genera, suggesting it is a rather homoplastic feature.
To sum up, once again the Wagner method positioned a group on the basis of the larger set
and again biological evidence suggests that perhaps this is incorrect and the position indicated by
the compatibility method is more plausible.
The parsimony analysis united all species of Enicospilus (942-47) into a single clade but
Dicamptus was paraphyletic with respect to this. This is not surprising as traditionally Enico-
PHYLOGENY OF THE OPHIONINAE
105
spilus is separated from Dicamptus on account of its specialized mandibles; Dicamptus is
recognized only by the specialized features it has in common with Enicospilus plus the
possession of unspecialized mandibles, a plesiomorphic feature (Townes, 1971). The compati-
bility method proved to be less successful at producing groups corresponding with recognized
genera, probably because Enicospilus and Leptophion are polythetic taxa (Gauld & Mound,
1982).
The relative positions of Leptophion, Laticoleus and Pamophion are less clear. The Wagner
method suggested Leptophion and Laticoleus form separate groups, though both are charac-
terized by the apomorphic condition of 35, and with one exception, 37. The compatibility
method suggests Laticoleus + Leptophion form a group defined by character 35 and these two
taxa, plus Pamophion, form a group defined by character 3.
To conclude this section, on balance, secondary evidence suggests Ophiogastrella and the
Hawaiian genera are better placed by the compatibility method whilst Enicospilus is most
successfully aggregated by parsimony analysis. A 'hybrid' compromise cladogram was post-
ulated that required 249 transformation steps (Fig. 18). This compromise arrangement groups
various congeneric species together, thus to some extent corroborating the initial assumption
that ophionine genera are not polyphyletic groups.
The relationship of the component genera of the Enicospilus/ Stauropoctonus complex can be
illustrated most clearly if the highly homoplastic characters are omitted, the relative positions of
"O X5
o
^O ^Q ^^ ^Q
NO N 5^ 00
00 K) GO
Fig. 18 Hybrid cladogram derived from subjective evaluation of parsimony and compatibility analyses of
Enicospilus/ Stauropoctonus data set. This arrangement requires 249 transformation steps.
106
I. D. GAULD
16-1, 25-1. 25-2, 28-1. 361, 38,42-1,
43-1, 45-1, 471, 49, 51-1,51-2,63'
Fig. 19 Cladogram showing putative phylogenetic relationships between the genera of the Enico-
spiluslStauropoctonus complex as supported by the least homoplastic characters.
PHYLOGENY OF THE OPHIONINAE 107
species within genera disregarded (and the object of this part of the analysis is to establish the
phylogenetic relationships of the genera) and single exceptional character scores, that are
almost certainly incidences of homoplasy (e.g. 15-2 in taxon 945; 61-2 in taxon 944), ignored.
This simplified cladogram together with supporting characters is illustrated in Fig. 19. Essenti-
ally it is the same arrangement as Fig. 18.
This cladogram suggests that Stauropoctonus is paraphyletic with respect to Aulophion,
Leptophion is paraphyletic with respect to Laticoleus and Enicospilus is paraphyletic with
respect to the PycnophionlAbanchogastralBanchogastra lineage.
The major groups and their relationship with the unassigned taxa
The inter-relationship of the two major generic complexes
The Enicospilus/ Stauropoctonus complex has been defined above as a holophyletic group on the
basis of 17 apomorphies (16-1, 224, 25-1, 25-2, 28-1, 28-2, 33, 36-1, 38, 42-1, 43-1, 45-1, 47-1, 49,
51-1, 51-2 and 63) (though Ophiogastrella is primitive in respect of two, 22-1 and 33) and the
Thyreodon/Euryophion complex has similarly been defined by a possible 19 apomorphies (2,
7-1, 22-2, 24-1, 24-2, 25-1, 25-2, 28-1, 28-2, 38, 42-1, 42-2, 47-1, 49, 51-1, 51-2, 54, 58-1 and 614).
Ten of these apomorphic features (25-1, 25-2, 28-1, 28-2, 38, 42-1, 47-1, 49, 51-1 and 51-2) are
common to both sets, suggesting a sister group relationship. Of the apomorphies remaining for
defining the Thyreodon/Euryophion lineage, three (24-1, 24-2 and 61-1) occur extensively in
various species-groups and species in the Enicospilus/ Stauropoctonus complex, thus weakening
their credibility as characters defining the former lineage. Character 42-2 has apparently been
derived in parallel in the Stauropoctonus line, but the remaining five features (2, 7-1, 22-2, 54
and 58-1) are robust characters that strongly support the monophyly of the Thyreodonl
Euryophion lineage. Of the six remaining apomorphies defining the Enicospilus/ Stauropoctonus
lineage (16-1, 22-1, 33, 36-1, 45-1 and 63) two (22-1 and 33) exclude Ophiogastrella; the
apomorphic state of 33 has also been derived in parallel in the Thyreodon line. The derived
states of characters 16-1 and 36-1 occur in most species of the Thyreodon/Euryophion complex,
suggesting they are group apomorphies, whilst the derived state of 45-1 also occurs, in parallel,
in Euryophion. Character 63 is the single convincing apomorphy supporting the monophyly of
the Enicospilus/ Stauropoctonus complex. A cladogram showing the putative phylogeny of these
taxa is shown in Fig. 20.
The position of the unassigned taxa
The position of the remaining taxa (905, 906, 920, 922, 923, 927, 938-41 and 948-51) can now be
considered in relation to the robust arrangement of the majority of ophionine taxa presented
above. To reduce the data matrix to more manageable proportions a number of hypothetical
taxonomic units (HTUs) were constructed to represent genera reasonably placed in the earlier
part of the work. Scores were assigned to these HTUs by marking each character with the
condition found in the majority of species of the genus (including all taxa examined in earlier
taxonomic studies (Gauld, 1977; Gauld & Mitchell, 1978, 1981) but not included in the primary
data matrix). This method of scoring, rather than deducing a hypothetical ancestor, was adopted
because of the high incidence of homoplasy. This idiosyncracy is unlikely to alter any cladistic
arrangement significantly, whereas the assumption that the group-ancestor has a particular
character present in the plesiomorphic condition, based on occasional incidence of the occur-
rence of the plesiomorphic (and possibly reversed) condition in some members of the group,
could result in considerable alteration to the position assigned to the group. The HTUs utilized
are 960 (Laticoleus), 961 (Ophiogastrella), 962 (Leptophion), 963 (Stauropoctonus), 964
(Thyreodon/ Dictyonotus) , 965 (Euryophion), 967 (Dicamptus), 968 (Enicospilus). [N.B. There
is no 966.] The character scores of these HTUs are given in Appendix 7.
The results of a LeQuesne test on taxa 905, 906, 920, 922, 923, 927, 938-41, 948-51 and 960-68
are given in Table 17. The overall O/E value is strikingly high (0-90), indicating the extremely
homoplastic nature of this data set. Perusal of the labels matrix (Appendix 8) shows that a large
108
I. D. GAULD
16.1*', 25.1*. 25.2*. 28.1*. 28.2*, 36. 1'^
38*', 42.1, 47.1', 49, 51.1*. 51.2*B
Fig. 20 Putative phylogenetic inter-relationships between groups of taxa in the Enicospilus/Stauropocto-
nus and ThyreodonlEuryophion complexes.
number of characters have been labelled highly for certain taxa; this will be discussed below.
Wagner analysis of this data set yielded a number of equally short cladograms with quite
different topologies (Figs 21-24). A large number of other trees that were only slightly less
parsimonious were also generated (Gauld, unpublished data). Compatibility analysis yielded a
single largest clique of 11 informative characters (2, 3, 7-1, 17, 25-1, 26, 38, 40, 41, 47-1 and 52-1)
which supports the cladogram shown in Fig. 25. Although differing in many topological details,
certain features are common to the dendrograms. In all of the Wagner trees the two major
complexes emerged as the most derived groups. Five taxa (938, 939, 941, 948 and 951) occupy
the most primitive position in three of the dendrograms and are amongst the eight most primitive
taxa in the other two, suggesting these species are amongst the most primitive of all ophionines.
This group of species comprises examples of the genera Xylophion, Ophion, Agathophiona,
Rhopalophion and Sclerophion, Gauld (1979) suggested these might represent a holophyletic
group, the Ophion subgroup, characterized by having a slender Rs+2r, which is also straight
basally, having an incomplete posterior transverse carina of the mesosternum, having a
subapically impressed clypeus, having Im-cu usually with a ramellus, and always angled, and
having the anterior area of the propodeum 'occluded'. However, the holophyly of the group is
open to question. The shape of Rs+2r (corresponding to characters 39, 40 and 62), of Im-cu (38
PHYLOGENY OF THE OPHIONINAE 109
Table 17 Results of LeQuesne test on data set comprising HTUs and unplaced taxa (905, 906, 920, 922,
Character
number
: incompatibilities
observ
ed expected 0/E ratio
1.1:
44
45.4
0.97
1.2:
38
37.7
1.01
2_ :
15
25.6
0.59
5.1:
35
38.2
0.92
6.1:
30
38.2
0.79
7.1:
35
37.4
0.94
1 :
35
38.2
0.92
\2_ :
36
38.2
0.94
14.1:
50
45.1
1.11
15.1:
53
58.9
0.90
15.2:
30
37.5
0.80
16.2:
56
57.6
0.97
18.1:
54
60.2
0.90
18.2:
35
37.4
0.94
11 :
30
38.2
0.79
20.2:
53
58.6
0.90
22.1:
45
60.0
0.75
22.2:
49
50.1
0.98
24.1:
54
60.8
0.89
24.2:
44
50.1
0.88
25.1:
14
25.1
0.56
26 :
24
25.6
0.94
27.1:
53
58.6
0.90
27.2:
57
61.0
0.93
29.1:
37
37.5
0.99
29.2:
52
58.9
0.88
30 :
42
38.2
1.10
Ji :
45
45.9
0.98
33_ :
50
60.9
0.82
35 :
47
45.9
1.02
36.2:
59
56.5
1.05
36.3:
28
37.3
0.75
37 :
53
60.9
0.87
39 :
61
59.6
1.02
40 :
22
38.2
0.58
11 :
44
45.9
0.96
42.2:
37
37.4
0.99
43.1:
44
37.4
1.18
43.2:
53
60.2
0.88
44.2:
34
37.4
0.91
45.1:
48
45.0
1.07
45.2:
47
58.7
0.80
47.1:
22
45.4
0.48
47.2:
34
37.7
0.90
48.1:
59
56.9
1.04
51.1:
57
58.6
0.97
51.2:
55
60.7
0.91
52.1:
22
25.6
0.86
55 :
20
25.6
0.78
1Z. =
56
59.6
0.94
58.1:
42
51.1
0.82
61.1:
51
58.7
0.87
61.2:
41
45.0
0.91
62 :
56
57.6
0.97
^ : 26 38.2 (J.68
7.2: 47 61.0 0.77
14.2: 46 54.1 0.85
IT_ : 24 25.6 0.94
20.1: 60 60.0 1.00
23 : 43 51.1 0.84
25.2: 58 61.2 0.95
28.2: 57 59.6 0.96
31^ : 28 25.6 1.09
36.1: 25 24.9 1.00
38 : 34 51.1 0.67
42.1: 61 60.2 1.01
44.1: 59 61.2 0.96
46 : 33 45.9 0.72
48.2: 32 37.5 0.85
54 : 39 45.9 0.85
j>9 : 45 59.6 0.76
63 : 52 60.9 0.85
Grand total- 1528 observed, 1705.7 expected. Overall 0/E ratio = 0.90
Ranking ratios of scoring characters
47.1 25.1 40 2 38 3 46 22.1 36.3 59 7.2 55 6.1 19 45.2 15.2 33 58.1 23 54 14.2
48.2 63 52.1 61.1 37 24.2 43.2 29.2 24.1 18.1 15.1 47.2 20.2 27.1 51.2 44.2 61.2
5.1 9 27.2 18.2 7.1 17 26 57 12 25.2 28.2 41 44.1 1.1 16.2 62 51.1 22.2 32 29.1
42.2 20.1 36.1 1.2 42.1 39 35 48.1 36.2 45.1 31 30 14.1 43.1
and 47) and the form of the propodeum (22) are symplesiomorphies and therefore not
admissable for defining a holophyletic group. Both the incomplete posterior transverse carina of
the mesosternum (equivalent to 1 scores for 27-1 and 27-2) and the impressed clypeus
(corresponding to a 1 score for character 7-2) are apomorphic features found in many other taxa.
In many of the Wagner trees generated the five taxa formed a clade defined by characters 18-1
and 44-1 in addition to 7-2 and 27-2 (27-1 is most parsimoniously considered an apomorphy of
the subfamily) . All of these characters are highly homoplastic so it is not surprising that the
compatibility cladogram leaves these taxa as an almost unresolved paraphyletic assemblage.
Four further taxa, 920 (Riekophion emandibulator) , 922 (Sicophion pleuralis) , 940 (Eremo-
tylus boguschi) and 950 (Hellwigiella nigripennis) have been placed in the same group as the
Ophion subgroup (Townes, 1971; Gauld, 1979). The undescribed genus (949) would, on
Townes' (1971) criteria, also belong to this group. There is no evidence in any analysis that
suggests that these taxa and the Ophion subgroup constitute a holophyletic clade. The definition
of the group as adopted by previous authors rests on a plesiomorphic feature, possession of a
membranous flange on the fore tibial spur, suggesting this 'group' is in fact a paraphyletic
assemblage. In Wagner analyses 920, 922, 940, 949 and 950 may be positioned primitively with
respect to most other taxa (e.g. Figs 22, 24) but some members may occasionally be united with
other taxa. Taxa 949 and 950 are sometimes placed in much more derived positions, the former
within the Enicospilusl Stauropoctonus complex (e.g. Fig. 23) and the latter as the sister-lineage
to the ThyreodoniEuryophion complex (e.g. Fig. 21). Hellwigiella (950) is placed as the
sister-group to the ThyreodoniEuryophion complex on the basis of a number of shared, rather
striking, derived characters. These characters (which may be shared by several or all species in
the complex) are adaptations to a diurnal eremic existence, and include possession of small ocelli
(12), a shorter stouter flagellum (14-1), pendant epipleuron 2 (30), cylindrical hind tibial spurs
(54) and long weakly curved claws (58-1). Many or all of these features are also found in the
110
I. D. GAULD
3> ' i
t ( )
O
Fig. 21 283 transformation step cladogram derived from Wagner analysis of HTUs plus unplaced taxa.
apomorphic condition in other diurnal eremic ophionines including an undescribed species of
Ophion from Australia (in ANIC) and Agathophiona species from Mexico. In another group of
ichneumonids, the Anomaloninae, similar differences occur between closely related eremic and
non-eremic organisms (e.g. Gravenhorstia (Erigorgus) and Gravenhorstia (Gravenhorstia)
species (Gauld, 1976)). The occurrence of so many apomorphic states of these characters
together in day-flying species strongly suggests the characters should be considered as a
character suite and perhaps accorded less taxonomic weight. Unlike members of the Thyreodonl
Euryophion complex, Hellwigiella shows the plesiomorphic state of important characters used
for defining the group (2, 7-1, 44-2). Furthermore, a position near to the base of the common
PHYLOGENY OF THE OPHIONINAE
111
-~ -<
Fig. 22 283 transformation step cladogram derived from Wagner analysis of HTUs plus unplaced taxa.
stem of the two major complexes is suggested by the fact that Hellwigiella shows the plesiomor-
phic states of characters 16-1, 42-1, 52-1 and 52-2.
The position of the new genus (949) is rather perplexing as the taxon exhibits a number of
derived features shared with the more specialized taxa, especially those in the Enicospilusl
Stauropoctonus complex (e.g. 15-1, 16-1, 22-1, 33, 45-1) yet at the same time lacks one major
apomorphy of the group (63) . Taxon 949 also shows the plesiomorphic state of many of the stem
characters, including 25-2, 28-2, 42-1, 51-1 and 52-2. It does, however, share a small number of
derived features (including 1 1 , 15 1 , 16- 1 , 33 and 45-1) with another enigmatic taxon , Sicophion
pleuralis (922). Sicophion, like taxon 949, is primitive in a surprising array of other features and
it seems plausible that these two taxa have a sister-group relationship and represent a primitive
group that evolutionarily converged with the Enicospilus group, possibly as a result of similar
selection pressures. The striking phenetic resemblance between 949 and some species of
Leptophion is perhaps more understandable when one considers these taxa have apparently
evolved in mid altitude tropical cloud forests - 949 in the Neotropics and Leptophion in South
East Asia.
Taxon 920, Riekophion emandibulator , is often placed with 967 and 968 (Dicamptus and
112
I. D. GAULD
co m GO
Fig. 23 284 transformation step cladogram derived from Wagner analysis of HTUs plus unplaced taxa.
Enicospilus) in compatibility analyses on the basis of sharing the derived states of characters 40
and 46 (e.g. Fig. 25). In other respects Riekophion does not appear to be at all closely related to
either Dicamptus or Enicospilus (Gauld, 1977) and the high label scores obtained by the species
for the derived states of these characters (Appendix 8) strongly suggests parallel derivation in
two lineages.
PHYLOGENY OF THE OPHIONINAE
113
C/> i i i i
Fig. 24 284 transformation step cladogram derived from Wagner analysis of HTUs plus unplaced taxa.
Accepting that the taxa of the Ophion 'genus group' (920, 922, 938-41, 948-51) are the most
primitive ophionines, then the most parsimonious arrangement of these taxa with respect to all
other groups is that presented in Fig. 26. All other ophionine taxa, 922 + 949 + 920 and 950 +
940 remain as an unresolved trichotomy. The Ophion genus-group (i.e. the Ophionini of
Townes, 1971) is therefore a paraphyletic grade, an assemblage of less-specialized ophionines.
The 'group' possibly comprises three apparently holophyletic clades, the Ophion-group (=
Ophion subgroup of Gauld, 1979), the Eremotylus group (940, 950) and the Sicophion group
(920, 922, 949). The clade, 'all other ophionine taxa', comprises the two groups discussed above
(the Thyreodon/Euryophion complex and the Enicospilusl Stauropoctonus complex) and four
unplaced taxa, 905, 906, 923 and 927. The position of these four taxa can now be considered in
relation to the fairly rigid structure derived above.
Taxa 905 and 906 (Simophion calvus and Orientospilus melasmd) share 27 derived characters
(Appendix 3); 906 also has a large number in common with other taxa, particularly 947 (28), 946
114
I. D. GAULD
O O O "O -O O O "O O O -O
E:ao*<>KJK)p!OpOi
O CO ' O K3 NJ C*>
Oi
*_zt
"O ^O O O *O O
a ^ O O K) |v
(Ji O SI 00
40
t
47.1'
8g ^
25.1*
Fig. 25 Cladogram based on largest compatible set of characters obtained from HTUs plus unplaced taxa.
(27), 942 (26), 918, 915, 914, 911 and 910 (25). Taxon 905 is less closely associated with other
taxa but shows some affinity with the Euryophion group (937 (25), 936 (24) and 935 (23)). Taxa
905 and 906 share several unusual apomorphic features including 6-1 (elsewhere only found in
the derived state in 909, 935-37 and 946), 19 (elsewhere only found in 908, 918, 935 and 937), 23
(elsewhere only found in the DictyonotuslThyreodon lineage) and 41 (elsewhere only found in
922 and 923). Taxa 905 and 906 also share all the stem characters of the Enicospilusl
Stauropoctonus + Thy reodonl Euryophion lineage, except that 905 is plesiomorphic for 42-1.
This suggests that they may belong near the other groups, possibly as a sister-species pair. Of the
features defining the two major complexes these taxa share only one, an apomorphy of 63,
suggesting placement near the base of the Enicospilusl Stauropoctonus lineage. It is noteworthy
that in earlier Wagner analyses (e.g. Figs 22-24) 905 and 906 generally were split up, 905 often
grouping with 923 , 964 or 965 , whilst 906 usually was united with 960 or 961 . If the two taxa were
entered together first in the data file, then they were not separated , but remained as a distinct
clade(Fig.21).
Taxon 927 (Barytatocephalus mocsaryi) shares 24 derived characters with taxon 936, 23 with
914 and 946 and 22 with 913. It shares all the stem features with the major complexes and
additionally has the apomorphic condition for 22-2 and 58- 1 , suggesting it belongs near the base
of the Thy reodonl Euryophion lineage. It does not have the major apomorphies of the
Enicospilusl Stauropoctonus branch (that is 22-1, 33 and 63).
Taxon 923 (Prethophion latus) is an enigmatic Neotropical taxon that has relatively few
characters in common with any other species. The largest number, 22, are shared with 918 and it
has 21 in common with 906. It does have all the 'gain' stem apomorphies in common with other
higher taxa, but is plesiomorphic for 25-2, 28-2 and 45. It does not have any of the apomorphies
defining the Thyreodonl Euryophion complex, but is apomorphic for character 63, suggesting it
belongs near the base of the Enicospilusl Stauropoctonus stem. It is most parsimonious to treat
Prethophion latus as the sister species of the 905 -I- 906 lineage (on the basis of the derived state
of characters 27-1 and 27-2), but this association must be regarded as very tenuous.
PHYLOGENY OF THE OPHIONINAE
115
941 951 939 948 938 950 940 922 949 920 o
51.1
Fig. 26 Most parsimonious arrangement of 'Ophion genus-group' and all other taxa. Conventions as in
Fig. 13.
The cladogram showing the preferred arrangement of taxa as described above is shown in Fig.
27. It is interesting to note that this user defined cladogram is one step more parsimonious than
any obtained by Wagner analysis.
116
I. D. GAULD
o . .
Fig. 27 Cladogram derived from subjective evaluation of character complexes involved in grouping taxa
in cladograms presented in Figs 21-25. It is noteworthy that this, at 242 steps, is a more parsimonious
arrangement than the best obtained by Wagner analysis, though this is a fortuitous event.
Discussion of suggested phytogeny
A putative phylogeny of the Ophioninae is shown in Fig. 27. Parts of this cladogram seem to be
fairly robust, but other groups are only supported by weak homoplastic characters. The overall
arrangement, placing the Ophion and Eremotylus groups in primitive positions and having a
bifurcated evolutionary line, does seem reasonably sound. It corresponds with increasing
morphological complexity, particularly in the form of the fore tibial spur (character 51), the
region of the spiracular sclerite (character 15), the modification of the propodeum (characters
22, 25) and the loss of umbo (character 33). What little is known about the structure of the final
instar larvae supports this arrangement. The cephalic capsules of the larvae of Enicospilus,
Dicamptus, Euryophion and Thyreodon are clearly more specialized than those of Ophion in
having a modified hypostoma (Short, 1978). The larvae of Euryophion and Thyreodon are
similar in having numerous setae (11+) on each part of the maxillary lobe median to the
hypostomal spur and adjoining the maxillary palp. Dicamptus and Enicospilus have eight or less
setae in this position. Certain venational characters also support the phylogeny proposed. These
include modification of Im-cu and broadening of the base of Rs+2r (characters 38, 47 and 62).
PHYLOGENY OF THE OPHIONINAE 117
However, in almost all adult characters it is necessary to postulate parallelism. Reduction of the
tibial spur membrane has occurred independently in Xylophion and the main evolutionary line,
and specialization of the propodeum has occurred in both the main lineage and Eremotylus. It is
quite disconcerting to see the degree of evolutionary convergence that has occurred in separate
lineages (such as the development of alar sclerites and a sinuous Rs+2r in Riekophion and
DicamptuslEnicospilus and the general similarity of structure between the undescribed Costa
Rican genus and species of Leptophiori) . No amount of rearrangement of lineages could possibly
remove any but a small proportion of such homoplasy. Perhaps the development of such
similarities is due to some underlying adaptation of the genotype not visually manifested (the
underlying synapomorphies of Saether, 1979). It is difficult to explain otherwise (unless one
postulates reticulate evolution - 'ancient hybrids' giving introgression) how such an unusual
feature as alar sclerites has appeared, apparently independently in at least six evolutionary
lineages of Ophioninae (Afrophion, Sclerophion, Leptophion, Laticoleus, Riekophion and
DicamptuslEnicospilus). Similar examples of such unusual parallelisms can be found in the
modification of the penultimate distal hamulus, the projecting pecten of the hind tarsal claw and
the development of a spine on the hind trochantelli.
The Ophion group of genera (Ophion, Alophophion, Sclerophion, Afrophion, Rhopa-
lophion, Agathophiona, Xylophion) is defined by the possession of the apomorphic states of
four rather homoplastic characters, the loss of the posterior transverse carina of the meso-
sternum (27-2), the possession of a diagonal mesopleural furrow (18-1), possession of an
impressed clypeal margin (7-2) and having Rs in the hind wing at least slightly curved (44-1). Not
all species of these genera necessarily possess all these apomorphies; Afrophion species are
exceptional in having both a blunt clypeal margin and a straight Rs whilst some Rhopalophion
species have the mesopleural furrow obsolescent. Although this genus-group is amongst the
most primitive of ophionines, its position at the base of the phylogenetic tree rests on the
assumptions that characters 38 and 47-1 (shape of Im-cu and presence of a ramellus) are cor-
rectly polarized. As mentioned above (pp. 77, 78) the polarity of both features is questionable
and if it could be demonstrated that the assigned polarity is incorrect then these two features
would become convincing apomorphies supporting the holophyly of the Ophion group. This
change in polarity would necessitate placing the Ophion group a little higher up the evolutionary
tree, possibly as the sister lineage to the Eremotylus group; these two groups share the
apomorphic states of characters 18-1 and 44-1.
The holophyly of the Eremotylus group is supported by the angulate base of Rs+2r (39),
which is somewhat thickened (62), and by the form of the clypeus. In most species it is slightly
flared outwards in profile (7-1) and often slightly concave. The mandibles are a subtly different
shape to those of virtually all other ophionines but this feature is very difficult to define.
The Sicophion group is a rather tenuous association of primitive taxa. The holophyly of the
group is supported by the apomorphic states of characters 33 and 45-2. Apart from possession of
certain primitive features such as the plesiomorphic state of character 51-2, and more extensive
propodeal carination, all of these taxa appear to be highly specialized, sharing a variety of
apomorphic features with other taxa placed higher up the phylogenetic tree. More evidence of
the holophyly of this group needs to be obtained before it can confidently be accepted.
The holophyly of the Thyreodon group is supported by the specialized form of the anterior
part of the propodeum (character 22-2) and by the form of the claw (character 58-1). The
inclusion of Barytatocephalus within this clade does seem reasonable. In addition to the
characters used in the analysis, two other features suggest Barytatocephalus belongs here. Vein
3r-m in the fore wing of species of this genus forms a more obtuse angle with Rs+2r than is the
case with most other ophionine taxa. Most other members of this clade have a similar
specialization. At the base of the hind wing, in the anal cell, a vestige of a vein (? 2A) is often
discernible. In most ophionines this trace, when observable, can be seen to be fairly near to and
parallel with the hind margin of the wing; in both Barytatocephalus and other members of the
Thyreodon complex (but not Euryophion) this vein trace is remote from the wing margin and
close to and parallel with \A. Townes (1971) considered Barytatocephalus to be a derivation of
Enicospilus, but there is little evidence for this supposition. The exposed spiracular sclerite, the
118 I. D. GAULD
general head shape and form of the propodeum suggest Barytatocephalus should not be included
in the Enicospilus lineage.
The Enicospilus group is the largest complex and one of the most difficult to define. Its
holophyly is supported by a single apomorphy, loss of vestigial first laterotergites (character 64),
but the included taxa do resemble each other in possessing a large number of derived features in
common. Within this group five subordinate lineages are recognizable, the Orientospilus,
Ophiogastrella, Stauropoctonus, Leptophion and Enicospilus subgroups. Each of these is
apparently a holophyletic group. These are the various groupings recognized in the classification
proposed below.
A classification of the Ophioninae
Hennig (1966) pointed out the existence of two sorts of monophyletic taxa - holophyletic and
paraphyletic groups. Traditional taxonomists may accord either group the status of a super-
generic rank, though cladists do not recognize-paraphyletic taxa (Farris, 1979; Carpenter, 1982).
Whilst there are many cases of small paraphyletic taxa that can be satisfactorily amalgamated
with a holophyletic taxon to form a slightly larger holophyletic taxon (e.g. Aulophion +
Stauropoctonus), there are other cases where dogmatic adherence to cladistic tenets produces
impractical results. For example, if the Hawaiian genera were incorporated into Enicospilus to
give a single holophyletic taxon, the resultant genus would contain an extraordinary range of
morphological and probably biological diversity. I concur with Martin (1981) in believing that a
classification, as opposed to a phylogeny, is a compromise between known or hypothesized
inter-relationships and nomenclatural convenience. A good classification has at least two
important facets: (a) it facilitates generalizations to be made about the component taxa, and (b)
it is predictive. Farris (1979) and Mickevich (1978) both contend that a purely phylogenetic
classification is an optimum as it is both more predictive and more stable than one that is not
entirely phylogenetic. Whilst I agree with these authors that this is often the case, I do not think it
is always so, especially where there are very unequal rates of evolution in different lineages.
Consider, for example, the ichneumonid subfamily Mesochorinae. Five genera, Cidaphus,
Astiphromma, Mesochorus, Stictopisthus and Plectochorus each show progressive morphologi-
cal complexity. All are quite large genera and are functionally useful, but only Plectochorus
appears to be holophyletic. Each of the other four taxa is apparently paraphyletic with respect to
the genera that are more specialized than it (Townes, 1971; Gauld, 19846). Although not a strict
phylogenetic classification, this arrangement of the species (which is based on adult morpho-
logical features) was found to have predictive value for larvae (see Short, 1978) which seem to
corroborate the idea that these taxa are a nesting paraphyletic series. I personally see no way of
establishing a strict phylogenetic classification for such a group, other than by lumping all the
genera, an action that would result in a much less informative and less useful classification than
the present one.
The existence of such nesting paraphyletic groups would seem to be a corollary of the
punctuational model of evolution. Stanley (1979) argues convincingly that many major adaptive
changes have occurred very rapidly in isolated populations - thus a new organizational level
(higher taxon) may have arisen from a small population of a pre-existing (and, if the rate of
phyletic evolution is small, virtually unchanging) species placed in a different higher taxon.
To return to the present example. If the Hawaiian genera have resulted from rapid radiation
from an Enicospilus ancestor into niches hitherto, for enicospilines, unexploited, it is fairly
unlikely that any prediction made from study of the biology of continental Enicospilus species
will apply to Pycnophion. Similarly, any generalizations made about Enicospilus are likely to
have to be prefaced by 'except in some Hawaiian species'. Consequently the purposes of
prediction and generalization would both be better served by recognizing the Hawaiian taxa as
distinct genera, even though Enicospilus thus becomes a paraphyletic taxon.
A second area of contention relevant to the present work concerns the reality and recognition
of polythetic superspecific taxa. The currently recognized ophionine genera have been accepted
initially as monophyletic groups, though in several instances it has been mentioned that a
PHYLOGENY OF THE OPHIONINAE
119
number are polythetic (Cushman, 1947). Some authors (e.g. L0vetrup, 1973) tend to dismiss
polythetic taxa as the product of unsatisfactory discrimination. Whilst I accept that this is an
explanation, I dispute that it is the most obvious one. Most cladistic studies reveal that
homoplasy is a common phenomenon. If a data set is highly homoplastic, then logically it follows
that some phylogenetically 'real' clades will only be definable in a disjunct (sensu Hull, 1965)
way. Imposing a monothetic classification on such groups would result in the erection of a
multiplicity of new genera, many of which would be monobasic. The resultant classification will
have little predictive value as a high proportion of new taxa are likely to necessitate new genera
for their accommodation (Gauld & Mound, 1982).
The following classification is suggested for the subfamily Ophioninae. It is not strictly
cladistic, as it recognizes some paraphyletic taxa; however it is a fair approximation of the
presumed phylogeny. Formal tribes are not recognized, but these could be used for the various
genus-groups. It should be noted that some very small genera were not incorporated in the
cladistic analysis (these are denoted by an asterisk), but they were closely studied in an earlier
work (Gauld, 1979). This has facilitated their placement in the present system.
Subfamily OPHIONINAE
Ophionidae Shuckard, 1840
Enicospilini Townes, 1971
OPHION genus-group
Op/iion Fabricius, 1798
Alophophion Cushman, 1947*
Sclerophion Gauld, 1979 rfitC
Afrophion Gauld, 1979*
AgathophionaWestwood, 1882
Rhopalophion Seyrig, 1935
Xylophion Gauld, 0-979 / ? y
SICOPHION genus-group
Riekophion Gauld, 1977
Sicophion Gauld, 1979
Janzophion gen. n.
EREMOTYLUS genus-group
Eremoty/usFoerster, 1869
Trophophion Cushman, 1947*
Hel/wig/e/faSzepligeti, 1905
THYREODON genus-group
BarytatocephaIusSchu\z, 1911
Euryophion Cameron, 1906
Rictophion Townes, 1971 Syn. n.
Dictyonotus Kriechbaumer, 1894
Ophionopsis Tosquinet, 1903
Syn. n.
IWiync/iop/iJonEnderlein, 1912
T/rjreodonBrulle, 1846
ENICOSPILUS genus-group
Orientospilus subgroup
Prethophion Townes, 1971
Simophion Cushman, 1947
Orientospilus Morley, 1912
Ophiogastrella subgroup
OphiogastrellaBrues, 1912
Stauropoctonus subgroup
Lepiscelus Townes, 1971
Stauropoctonus Br nuns, 1889
Aulophion Cushman, 1947 Syn. n.
Leptophion subgroup
Pamophion Gauld, 1977
Laticoleus Townes, in Townes &
Townes, 1973
Leptophion Cameron, 1901
Enicospilus subgroup
P/campfusSzepligeti, 1905
Enicospilus Stephens, 1835
Pycnophion Ashmead, 1900
Banchogastra Ashmead, 1900
Abanchogastra Perkins, 1902
Zoogeographic discussion
Present day distribution
The approximate numbers of species per area are given in Table 18. For the purposes of this
discussion Melanesia, Australia and New Zealand are considered as a single region, the
Australo-Pacific, whilst for purposes of comparison Hawaii, with three endemic and one
cosmopolitan genera, is ignored.
Altogether 20 of the 32 ophionine genera (62-5%) are restricted to a single zoogeographic
region. This is a surprisingly high proportion compared with many other ichneumonid subfami-
lies. For example, only about 30% of pimpline genera are endemic to a single region (Townes,
120 I. D. GAULD
Table 18 Approximate numbers of species of various ophionine genera in each zoogeographic area.
PRETHOPHION
ORIENTOSPILUS
OPHIOGASTRELLA
ABANCHOG ASTRA
BANCHOGASTRA
M O
O H
> M
50 -- 25 1
5 10 30 50
- - 25
- - 1
i
ALOPHOPHION
SCLEROPHION
AFROPHION
AGATHOPHIONA
RHOPALOPHION
XYLOPHION
RIEKOPHION
SICOPHION ~ ~ " 2
TiMyflPHTDN 1
EREMOTYLUS 8 ~ 1 5 20
TROPHOPHION
HELLWIGIELLA 1
BARYTATOCEPHALUS 2
EURYOPHION ~ 7 1
DICTYONOTUS 1 2 2
RHYNCHOPHION ~ ~ 3 1
40 3
LEPISCELUS ~ 1
STAUROPOCTONUS 1 1 2 1 1 4
PAMOPHION -- -- -- 1
LATICOLEUS 10
LEPTOPHION 11 17 7
DICAMPTUS 13 12 1 5
ENICOSPILUS 50 150 180 150 50 2 20 100
PYrNflPHTflN 3
1969) and a similar proportion of genera of other subfamilies are likewise restricted. The
majority of the restricted ophionine genera are small taxa with three or fewer species; only
Alophophion, Ophiogastrella and Laticoleus are larger than this and they have ten or more
species each.
Six of the more widely distributed genera occur in only two regions: Rhynchophion and
Thyreodon (Neotropical/Nearctic), Euryophion and Orientospilus (Afrotropical/Oriental),
Leptophion (Oriental/ Australo-Pacific) and Simophion (Palaearctic/Nearctic). Two genera
occur in just three regions, Dictyonotus (Palaearctic/Oriental/Afrotropical) and Dicamptus
(Afrotropical/Oriental/Australo-Pacific), whilst Eremotylus occurs in four regions (Nearctic/
Palaearctic/Oriental/Neotropic). Ophion occurs in all regions except the Afrotropical and
Stauropoctonus in all but the Nearctic (though it is absent from the African mainland). Only
Enicospilus is truly cosmopolitan with quite large numbers of species in all areas.
PHYLOGENY OF THE OPHIONINAE 121
Considered from a geographical standpoint the Neotropical region contains both the most
genera (12) and has the highest degree of generic endemicity (50%). The Afrotropical and
Oriental regions both contain ten genera, but the former has a far higher degree of endemicity
(40% compared with 10%). The Australo-Pacific region contains eight genera, three of which
(37%) are endemic whilst the Palaearctic, with a similar number of genera, has only two
endemics (25%). The Nearctic is the most impoverished with seven genera, one of which (14%)
is endemic. The faunal similarity (at generic level) between the regions is shown in Table 19.
Table 19 The generic faunal affinities between major zoogeographic regions (Hawaii is excluded). The
values at the intersects of rows and columns V may be defined as:
, _ number of genera common to areas X and Y
total number of genera present in areas X and Y
It is noteworthy that contiguous or recently contiguous areas have a significantly higher value for V than
do areas that are not contiguous or that have only been contiguous in the distant past (more than 40
million years ago), suggesting that dispersal may be a more important event than vicariance in the
biogeographical history of the subfamily.
o r 1 o f r 1
PALAEARCTIC 0.36 0.25 0.23 0.38 0.20
AFROTROPICAL 0.06 0.10 0.20 0.43
ORIENTAL 0.21 0.22 0.38
AUSTRALO-PACIFIC 0.15 0.18
NEOTROPICAL 0.36
Hypothesized evolutionary history of the group: a scenario
Primitive ichneumonids are known from the Upper Cretaceous (80-90mya) (Townes, 1973ft)
and the family may have originated at the beginning of the Cretaceous (Rodendorf & Rasnitsyn,
1980), possibly from a protoichneumonoid ancestor such as the Praeichneumonidae (Rasnitsyn,
1983). The age of the subfamily Ophioninae is not known but it is presumed that it post-dates the
primitive Upper Cretaceous groups. A fossil Ophion species is known from from the Lower
Oligocene (35-40mya) in France and Rodendorf (1962) states that the Ophioninae (?sensu lato,
i.e. including the Campopleginae) dates back to the Palaeocene (55-60mya). Possibly the group
radiated around the beginning of the Tertiary some 65-70 million years ago.
At this time the continents were still in close proximity to one another ( Audley-Charles et al. ,
1981; Owen, 1981). Australia was connected to Antarctica and an archipelagic connection
probably existed between this landmass and South America. Europe and North America were
contiguous and South America was not widely separated from Africa. The most primitive
lineage of the Ophioninae (the Ophion genus-group) may have been widespread at this time, as
would have been its sister-group, the stem group of the remaining Ophioninae. The Sicophion
lineage became isolated in the southern continent of Australia/ Antarctica/South America at an
early stage whilst its sister-lineage, the ancestor of the Eremotylus, Enicospilus and Thyreodon
lineages, remained in Laurasia or Africa. The ancestor of the Eremotylus lineage probably
differentiated in Laurasia whilst the ancestor of the Enicospilus I Thyreodon radiated in Africa. I
suggest that early offshoots of this line must have spread to South America, either by flying
122 I. D. GAULD
across the widening South Atlantic, or by dispersal through Laurasia and across a water gap to
South America. Some kind of barrier to the spread of organisms to and from South America is
suggested by the high degree of generic endemicity and the fact that the Neotropical species of
Enicospilus seem to belong to very few species-groups compared with those of other tropical
areas. I suggest that the less vagile Thyreodon lineage reached South America much later,
possibly from Asia via the Bering Straits. Probably the radiation of the Enicospilus lineage
resulted in widespread extinction of members of the Ophion group, thus leaving the isolated
specialized groups extant today. Ophion almost certainly has undergone secondary radiations
giving rise to species complexes in South America (derived from the Nearctic) and Australia
(derived initially from the Palaearctic fauna spreading through the Oriental region, where it has
subsequently become isolated on mountain tops) (Gauld, 19840). Except for Xylophion, a
remnant of the early widespread Ophion group-distribution, and Riekophion, a remnant of the
southern radiation of the Sicophion lineage, the ophionine fauna of the Australo-Pacific has
largely been derived from the Oriental region (as has been observed for other groups, e.g.
Wilson, 1959; Gupta, 1962), though New Guinea has served as the epicentre of a considerable
radiation, particularly in the case of Enicospilus , over 100 endemic species of which occur on the
island.
Although the above scenario is largely speculative it is the most parsimonious interpretation
of the present distribution in relation to the postulated phylogenetic history of the group. Other
scenarios would involve postulation of more widespread extinctions, more transoceanic disper-
sive events or necessitate postulating a much earlier origin for the subfamily.
The subfamily Ophioninae
In the following section the various genera and new synonymies are discussed in some detail.
Keys have not been given here to genera as any attempt to produce a key to world genera would
involve the usage of rather difficult couplets to allow for evolutionary parallelism between
different species-groups in similar habitats in different zoogeographic regions. Practical keys to
genera on a regional basis already exist. The Afrotropical region is covered by Gauld & Mitchell
(1978) and the Indo- Australian region by Gauld & Mitchell (1981). This latter key will also
suffice for the Palaearctic region. A key to the Neotropical genera is currently in preparation and
this will suffice for the Nearctic. Townes (1971) also offers reasonable generic keys, but these are
now rather dated, particularly his key to the 'Ophionini' (see Gauld, 1979).
The OPHION genus-group
This group contains seven genera, Afrophion, Agathophiona, Alophophion, Xylophion, Scler-
ophion, Rhopalophion and Ophion. The first six are holophyletic groups but Ophion is probably
paraphyletic with respect to them (see p. 125).
The group is characterized by the absence of a posterior mesosternal transverse carina,
usually by possession of a diagonal mesopleural furrow, generally by possession of an impressed
clypeal apex and by having Rs in the hind wing usually at least slightly curved and often very
strongly bowed. Virtually all species in this genus-group have Im-cu in the fore wing centrally
angled somewhat, and usually have a distinct ramellus present at this point.
As Ophion is apparently the paraphyletic stem-group from within which all other genera in
this group have arisen, and the genus is primarily a Holarctic taxon, it seems probable that this
group originated in the temperate north. It is probable that at some period it was present in most
regions and has gradually disappeared from equatorial regions leaving isolated relicts in South
Africa (Afrophion), Australia (Xylophion), Madagascar (Rhopalophion} and Patagonia
(Alophophion). The possibility that there has been repeated expansion into and extinction
within the tropics is suggested by the presence of some groups of Ophion species on isolated
mountains in South East Asia, New Guinea and South America, and by the occurrence of
distinctive Ophion species-complexes in Australia and New Zealand (Gauld, 19840).
PHYLOGENY OF THE OPHIONINAE 123
AFROPHION Gauld
Afrophion Gauld, 1979: 79. Type-species: Ophion nubilicarpus Tosquinet, by original designation.
Mandibles stout, not twisted, subequally bidentate, barely narrowed apically; outer mandibular surface
convex, with strong proximal concavity. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in
profile weakly convex with apical margin sharp but not clearly impressed; clypeus in anterior aspect weakly
convex. Ocelli large, the posterior ones very close to the eyes; frontal carina absent; occipital carina
complete, ventrally joining hypostomal carina. Antennae moderately long, 1-3 times length of fore wing.
Pronotum unspecialized; spiracular sclerite exposed; notauli vestigial; epicnemial carina complete;
mesopleural furrow strong, diagonal. Scutellum moderately convex, quite narrow and not carinate
laterally; posterior transverse carina of mesosternum absent except laterally. Propodeum with anterior
area either occluded or extremly short, the remainder of the propodeum rather abruptly declivous, with
irregular vestiges of carinae and rather coarsely microreticulate. Fore wing with pterostigma broad;
marginal cell elongate ; Rs+2r very broad and slightly angled before joining pterostigma; anterior corner of
discosubmarginal cell glabrous, in one species extensively so and with a detached alar sclerite (Gauld &
Mitchell, 1978); Im-cu with small ramellus, centrally angled. Hind wing with Rs from straight to weakly
curved. Fore tibial spur with a membranous flange behind macrotrichial comb; mid and hind trochantelli
unspecialized; inner hind tibial spur flattened, with a margin of close fine hair; hind tarsal claws
unspecialized, those of male slightly more closely pectinate than those of the female. Gaster slender;
tergite 2 elongate, with thyridia very weak, close to anterior margin; umbo distinct; epipleuron up- turned.
Ovipositor sheath slender; male with gonosquamae very large, ploughshare-like.
Afrophion is restricted to the extreme south of Africa. The genus contains two species, the type-species
and A. hynnis (Gauld & Mitchell). It is distinguished from Ophion by the form of the propodeum, the
unique structure of the male genitalia and by the basally incrassate Rs+2r in the fore wing.
AGATHOPHIONA Westwood
Agathophiona Westwood, 1882: 19. Type-species: Agathophiona fulvicornis Westwood, by monotypy.
Mandibles twisted 5-10, barely tapered, with lower tooth slightly the longer; outer mandibular surface
slightly convex, with a strong proximal concavity. Maxillary palp 5-segmented, labial palp 4-segmented;
mouthparts exceptionally specialized in that the distal part of the maxilla projects below the clypeus and
the labial glossae project by a distance almost equal to length of hind wing; clypeus in profile rather flat,
margin blunt; clypeus in anterior aspect broad, very weakly convex. Ocelli quite small, the posterior ones
separated from eye by about their own minimum diameter; frontal carina absent or present but weak;
occipital carina dorsally complete or narrowly obsolescent centrally, ventrally obsolescent, not joining the
very weak hypostomal carina. Antennae rather short and quite stout, distal segments quadrate. Pronotum
unspecialized; spiracular sclerite exposed; notauli short but strongly impressed near anterior margin;
epicnemial carina more or less complete; mesopleural furrow virtually absent. Scutellum moderately
strongly convex, not laterally carinate; posterior transverse carina of mesosternum present only laterally as
vestiges. Propodeum with anterior area occluded, rather deeply impressed; propodeal carinae vestigial;
posterior area striate or coriaceous. Fore wing with pterostigma moderately broad; marginal cell slender;
Rs+2r more or less straight, not thickened before joining pterostigma; discosubmarginal cell without a
glabrous anterior area; Im-cu generally with only a trace of a ramellus, rather evenly but quite strongly
curved. Hind wing with Rs quite strongly curved. Fore leg with tibial spur bearing a membranous flange
behind the macrotrichial comb; mid and hind trochantelli unspecialized; inner hind tibial spur cylindrical,
with scattered hairs marginally; hind tarsal claws long and weakly curved. Gaster rather stout; tergite 2 in
profile only slightly longer than posteriorly deep; thyridia small and close to anterior margin; epipleuron
up-turned. Female with subgenital plate enlarged, as long as tergite 2 and medially notched; ovipositor
sheath unspecialized.
Agathophiona is a monobasic genus; the type-species occurs in Mexico. A. fulvicornis is diurnally active
(Townes, 1971). Structurally this is one of the most specialized of all ophionine genera. The remarkable
mouthparts are presumably an adptation to feeding from certain types of flowers but the function of the
rather unusually modified female terminalia is not known.
124 I. D. GAULD
ALOPHOPHION Cushman
Alophophion Cushman, 1947: 439. Type-species: Ophion chilensis Spinola, by original designation.
Mandibles not twisted, weakly narrowed apically, subequally bidentate; outer mandibular surface flat,
usually punctate and hirsute. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile
weakly convex, margin often impressed, sometimes very narrowly so; clypeus in anterior aspect weakly
convex. Ocelli generally large, the posterior ones close to the eyes; frontal carina absent; occipital carina
absent. Antennae moderately to very long, at least 1-3 times length of fore wing. Pronotum unspecialized
or mediodorsally somewhat flattened and quite long; spiracular sclerite exposed; notauli present on
anterior part of mesoscutum; epicnemial carina generally strong; mesopleural furrow distinct, diagonal,
extending from episternal scrobe to near subalar prominence. Scutellum very weakly convex, usually
narrow and not carinate laterally; posterior transverse carina of mesosternum absent except for lateral
vestiges. Propodeum with anterior area occluded, transverse and often lateromedian longitudinal carinae
discernible, often almost complete; posterior area smooth or rugulose. Fore wing with pterostigma broad;
marginal cell long; Rs+2r slender, curved near proximal 0-3 before joining pterostigma near centre;
dicosubmarginal cell with glabrous area anterior; Ira-cu generally centrally angled, sometimes with a short
ramellus which is directed more anteriorly than that of Ophion. Hind wing with Rs curved. Fore tibial spur
with a membranous flange behind macrotrichial comb; mid and hind trochantelli unspecialized; inner hind
tibial spur flattened, with a margin of long close hairs; hind tarsal claws unspecialized. Gaster moderately
slender; tergite 2 in profile elongate, thyridia oval, separated from anterior margin of tergite by its own
length or less, umbo distinct; epipleuron up-turned. Ovipositor sheath narrow.
Alophophion is a moderately large genus that is restricted to South America. The majority of species
occur in southern Chile and Patagonia. Cushman, when describing Alophophion as a distinct genus,
commented that it barely warranted generic distinction from Ophion. Whilst it is undoubtedly very close to
Ophion, the combination of characters exhibited by the group clearly separate it as a holophyletic lineage.
This lineage is characterized by the following apomorphies: occipital carina entirely absent; Rs+2r joining
pterostigma near centre; first subdiscal cell stouter than normal; ramellus, when present, directed more
anteriorly than that of other ophionines.
OPH/OJVFabricius
Ophion Fabricius, 1798: 210, 235. Type-species: Ichneumon luteus L., by subsequent designation, Curtis,
1836: 600.
Paniscus Schrank, 1802: 316. Type-species: Ichneumon luteus L., by monotypy.
Psylonychia Szepligeti, 1905: 21. [Nomen nudum.]
Stenophthalmus Szepligeti, 1905: 23. Type-species: Stenophthalmus algiricus Szepligeti, by subsequent
designation, Viereck, 1914: 137. [Homonym of Stenophthalmus Becker, 1903.]
Pachyprotoma Kohl, 1906: 223. Type-species: Ophion (Pachyprotoma) capitatus Kohl, by monotypy.
Australophion Morley, 1912: 4, 30. Type-species: Ophion peregrinus Smith, by monotypy.
Neophion Morley, 1912: 4, 30. Type-species: Neophion crassus Morley, by subsequent designation,
Viereck, 1914: 100.
Apatophion Shestakov, 1926: 262. Type-species: Apatophion mirsa Shestakov, by original designation.
Platophion Hellen, 1926: 13. Type-species: Platophion areolaris Brauns, by subsequent designation,
Cushman, 1947: 475.
Potophion Cushman, 1947: 476. Type-species: Potophion caudatus Cushman, by original designation.
Psylonychia Cushman, 1947: 476. [Unavailable name, proposed in synonymy.]
Apomesus Townes, 1971: 54. Type-species: Apomesus longiceps Townes, by original designation.
Mecetron Townes, 1971: 60. Type-species: Stenophthalmus choaspese Uchida, by original designation.
Mandibles not twisted, from very weakly to moderately narrowed distally, generally subequally bidentate
or with upper tooth slightly the longer; outer mandibular surface more or less flat, except for basal
concavity, moderately punctate. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile
weakly to moderately convex, margin impressed, acute; clypeus in anterior aspect weakly convex, rarely
truncate or even slightly concave. Ocelli usually large, the posterior ones separated from eyes by less than
their own diameters; frontal carina absent; occipital carina usually complete dorsally, ventrally not
reaching the hypostomal carina, rarely with occipital carina mediodorsally obsolescent, or in a few species
with it entirely absent. Antennae generally of moderate length, in a few deserticolous species the flagellum
short and with central segments quadrate. Pronotum unspecialized; spiracular sclerite usually completely
exposed; notauli weak but discernible on anterior 0-2 of mesoscutum; epicnemial carina generally strong
and well developed on mesopleuron; mesopleural furrow strongly impressed, extending diagonally from
PHYLOGENY OF THE OPHIONINAE 125
episternal scrobe towards subalar prominence. Scutellum weakly to moderately convex, usually carinate
only on anterior 0-2 or less, rarely with lateral longitudinal carinae complete to posterior margin; posterior
transverse carina of mesosternum usually present only laterally as vestiges. Propodeum with anterior area
occluded except for a small semicircular depression centrally; propodeal carina variously developed, at
most with both transverse carinae and the lateromedian longitudinal carinae complete, enclosing an area
superomedia, in the most exceptional cases with all carinae only vestigial; posterior area usually rather
smooth. Fore wing with pterostigma moderately to very stout; marginal cell very long and slender; Rs+2r
usually virtually straight, slender, barely broadened before joining pterostigma, in some species evenly
broadened to join pterostigma; discosubmarginal cell with glabrous area in anterior corner; Im-cu usually
with a well-developed ramellus, rarely with stub-like indication of this vein present; Im-cu centrally
strongly geniculate. Hind wing with Rs from weakly to very strongly curved. Fore tibial spur with
membranous flange behind the macrotrichial comb; mid and hind trochantelli unspecialized; inner hind
tibial spur flattened, with a margin of close long hairs; hind tarsal claws unspecialized, those of males
usually more closely pectinate than those of the female. Gaster moderately slender; tergite 2 in profile
moderately to strongly elongate, usually with thyridia large, close to anterior margin; umbo well
developed; epipleuron up-turned. Ovipositor sheath usually slender, in one species-group exceptionally
stout.
Ophion is a very large genus that is widely distributed throughout the world except for the Afrotropical
region. In tropical South America and South East Asia few species occur and these are restricted to high
altitudes (Gauld & Mitchell, 1981). Ophion as currently defined is almost certainly a paraphyletic
assemblage; Agathophiona, Rhopalophion, Sclerophion, Xylophion, Alophophion and Afrophion are
phenetically highly divergent (Gauld, 1979), holophyletic species-groups which have probably arisen from
within Ophion. A number of other phenetically less divergent species-groups were included by Gauld
(1979) within Ophion, but as the phylogenetic relationship of the taxa becomes better understood it will be
necessary to erect a number of additional genera in order to establish a system of holophyletic taxa.
Previous authors (e.g. Morley, 1912; Cushman, 1947; Townes, 1971) have attempted to remove other
species-groups from Ophion and treat these as separate genera, but in each case the group removed has not
been holophyletic as comparatively few of the subtropical and eremic Ophion species are known. Recent
collecting has greatly increased our knowledge of these insects but more work needs to be undertaken
before it is possible to fully resolve this genus. The major species-groups of Ophion are outlined below; all
except the unsatisfactorily resolved luteus-group are holophyletic but I have avoided treating these groups
as distinct genera until the luteus anathema can be resolved. As each of these groups appears to be defined
by a series of unique apomorphies it is not possible to suggest their phylogenetic inter-relationship.
Ophion peregrinus species-group. This species-group contains all the described New Zealand Ophion
species (Townes et al, 1961). Its monophyly is suggested by the possession of a mediodorsally
lengthened pronotum, having Rs+2r joining the pterostigma distal to the pterostigma base and having
the longitudinal lateral carina of the propodeum with a raised ridge diverging towards the propodeal
spiracle. In O. peregrinus this is only weakly developed but in other taxa this ridge is frequently very
strong and continuous with the carina, and the front part of the carina is absent so the apparent carina
curves anteriorly towards the spiracle where it terminates. O. peregrinus is apparently the most
primitive taxon as it retains most of the propodeal carinae and has a strongly geniculate Im-cu; the
other taxa have only a vestige of the anterior carina (the lateromedian and posterior transverse carina
are lacking) and frequently have a reduced ramellus and more evenly curved Im-cu. Most also have a
- fairly straight Rs in the hind wing.
O. bicarinatus species-group. This complex contains O. bicarinatus Cameron, O. facetious Gauld &
Mitchell, O. gerdius Gauld & Mitchell, O. horus Gauld & Mitchell and all of the described Australian
species (Gauld, 1977). The group is characterized by possession of a proximally broadened Rs+2r
which is slightly curved before reaching the pterostigma. The European species, O. minutus, may
belong to this group. The five Australian species form a monophyletic subgroup characterized by the
black interocellar area. In the Oriental region the species of the bicarinatus group are restricted to
mountains in Sri Lanka, India, Burma, Malaysia, Taiwan and Sumatra. Their present fragmented
distribution and their paraphyletic nature with respect to the Australian species-group suggests that
this group was more widespread throughout the Indo- Australian region in the past, probably when the
climate was cooler (Gauld, 1984a).
O. cronus species-group. This monobasic group is restricted to montane New Guinea (Gauld & Mitchell,
1981). The group is characterized by having a very sparsely hirsute discosubmarginal cell and a very
short 3r-m. It does not appear to be closely related to any other Indo- Australian species.
126 I. D. GAULD
O. caudatus species-group. This group is characterized by the elongate head shape; the eyes are more oval
than normal and the labium is specialized in having the prementum extended far beyond the insertion
of the labial palps (Gauld & Mitchell, 1981). This group contains O. caudatus (Cushman), O. silus
Gauld & Mitchell, O. longiceps (Townes), O. ascus Gauld & Mitchell, O. sumptions Gauld &
Mitchell, O. mastrus Gauld & Mitchell and an undescribed species from Sulawesi (BMNH). All these
species occur in mountains in the Oriental region. An undescribed species from high altitude in Peru
(TC) has a similarly modified head and like other members of the O. caudatus group it has large wings,
elongate trochantelli and virtually obscured mesopleural punctures. This Neotropical species clearly
seems to belong in this species-group, but it is possible that the apomorphies defining the group are a
suite of characters facilitating existence at high altitude and that the Neotropical and Oriental lineages
have undergone morphological convergence.
O. areo/ar/s species-group. This group is characterized by the loss of occipital carina, possession of a
quadrate scutellum and a somewhat broadened ovipositor sheath, and by having a characteristic
pattern of propodeal carina in which the area superomedia is more or less discernible, the posterior
transverse carina is often complete and the anterior transverse carina is absent except centrally. This
group contains O. areolaris Brauns, O. ocellaris Ulbricht and O. fuscomaculatus Cameron. The first
two are western Palaearctic species whilst O. fuscomaculatus has an eastern Palaearctic range that
extends into the higher mountains of the Oriental region.
O. similis species-group. This group contains the Palaearctic species O. similis (Szepligeti), O. mirsa
(Shestakov), O. buchariensis Meyer and two undescribed Nearctic species. All are stout insects with
short, compact gasters, rather convex, irregularly sculptured propodea and somewhat shorter
antennae than is usual for species of this genus. The ocelli and eyes are moderately small so the
orbital-ocellar distance and the malar space are broader than is normal for Ophion species. All are
apparently diurnally active and most are associated with dry areas.
O. dentatus species-group. This group contains O. dentatus Smith, O. turcomanicus Szepligeti and O. virus
Gauld & Mitchell. The group is characterized by possession of unusually long, fairly slender
mandibles, having long, weakly curved claws and possessing numerous spines on hind tarsal segments
1-3. Species of this group are widely distributed throughout the eastern Palaearctic region, particu-
larly Central Asia. Several also extend into the drier parts of the Indian subcontinent.
O. luteus species-group. This large group contains all of the other described species of the genus from the
Palaearctic, Nearctic and Neotropical regions. It is apparently a paraphyletic group and I can only
define it in terms of plesiomorphic features (i.e. absence of apomorphic characters exhibited by other
species-groups). With more study it may be possible to subdivide this group into several holophyletic
species-complexes but little is yet known about either the eastern Palaearctic or the Nearctic species.
RHOPALOPHION Seyrig
Rhopalophion Seyrig, 1935: 49. Type-species: Rhopalophion curvus Seyrig (= Ophion discinervus
Morley), by original designation.
Mandibles not twisted, weakly narrowed apically, more or less equally bidentate; outer mandibular surface
slightly convex with distinct proximal concavity. Maxillary palp 5-segmented, labial palp 4-segmented;
clypeus in profile weakly convex, margin slightly impressed and reflexed or sometimes almost blunt;
clypeus in anterior aspect weakly convex or truncate. Ocelli moderately small to large, the posterior ones
separated from the eyes by less than 0-8 times their own maximum diameter; frontal carina absent;
occipital carina complete, joining hypostomal carina close to base of mandible. Antennae of moderate
length, at least 1-2 times as long as fore wing. Pronotum unspecialized; spiracular sclerite exposed; notauli
virtually absent; epicnemial carina present, extending onto pleuron; mesopleural furrow absent or very
indistinct. Scutellum weakly to moderately convex, laterally carinate for its entire length; posterior
transverse carina of mesosternum absent except for lateral vestiges. Propodeum with anterior area
occluded; anterior and posterior transverse carinae usually complete, rarely with the former vestigial, very
rarely all carinae absent; longitudinal carinae not developed; posterior area finely alutaceous to smooth.
Fore wing with pterostigma moderately broad; marginal cell long; Rs+2r almost straight, not or only very
slightly expanded before joining pterostigma; discosubmarginal cell anteriorly broadly glabrous from base
of Rs+2r to ramellus; Im-cu very angulate, with an extremely long ramellus that reaches at least 0-6 of
distance to Rs&M. Hind wing with Rs virtually straight or very weakly curved. Fore tibial spur with
membranous flange behind macrotrichial comb reaching 0-5-0-7 of length of spur; mid and hind
trochantelli unspecialized; inner hind tibial spur moderately flattened, with a fringe of long close hairs;
PHYLOGENY OF THE OPHIONINAE 127
hind tarsal claws unspecialized, those of male far more closely and finely pectinate than those of the female.
Gaster slender; tergite 2 in profile moderately long with thyridia oval and close to anterior margin; umbo
quite well developed; epipleuron up-turned. Ovipositor sheath slender.
Rhopalophion is a small genus with three described species, discinervus which is widespread throughout
the Afrotropical mainland, and divergens and parallelus which are restricted to Madagascar (Delobel,
1975). R. divergens and R. parallelus are apparently sister-species and constitute a clade defined by the
following apomorphic features, possession of small ocelli, possession of subquadrate lower face and
possession of a relatively long malar space. This clade is the sister- lineage to R. discinervus which is defined
by possession of a very angulate Im-cu and a distally swollen ramellus (Gauld & Mitchell, 1978). The
females of discinervus are unusual amongst ophionines in that the distal tarsal segment of the female has a
lateral projection. This is similar to one found in a species of Ophiogastrella (Cushman, 1947). Rhopa-
lophion species are easily distinguished from other taxa in the Ophion genus-group on account of their
unique venation, total lack of propodeal longitudinal carinae and long fore tibial spur.
SCLEROPHION Gauld
Sclerophion Gauld, 1979: 77. Type-species: Pleuroneurophion longicornis Uchida, by original desig-
nation.
Mandibles stout, not twisted, barely narrowed apically; outer mandibular surface weakly convex.
Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile weakly convex, margin not
impressed; clypeus in anterior aspect weakly convex. Ocelli large, the posterior ones separated from eyes
by less than 0-2 of their minimum diameter; frontal carina absent; occipital carina complete, ventrally
reaching hypostomal carina. Antennae very long and slender, more than 1-7 times length of fore wing.
Pronotum unspecialized; spiracular sclerite exposed; notauli weak but discernible; epicnemial carina
complete, strong; mesopleural furrow weak, diagonal, with upper end rather broad and shallow. Scutellum
moderately convex, carinate laterally on at least its anterior 0-4; posterior transverse carina of meso-
sternum absent except as lateral vestige. Propodeum with anterior area occluded except medially where it
is semicircular; anterior and posterior transverse carinae usually complete, other carinae weak; posterior
area usually rather smooth. Fore wing with pterostigma moderately slender; marginal cell elongate; Rs+2r
thickened and evenly curved before joining pterostigma; discosubmarginal cell with an extensive glabrous
area anteriorly, the distal margin of this area with a thickened corneous mark; Im-cu with long ramellus,
centrally geniculate. Hind wing with Rs very weakly curved. Fore tibial spur with a membranous flange
behind macro trichial comb; mid and hind trochantelli unspecialized; inner hind tibial spur flattened, with
margin of close, long hairs; hind tarsal claws unspecialized. Gaster slender; tergite 2 very elongate, with
thyridia elliptical, close to anterior margin; umbo well developed; epipleuron up-turned. Ovipositor
sheath slender.
Sclerophion species may be recognized by their characteristic fore wing venation, particularly the
thickened Rs+2r and the corneous mark in the discosubmarginal cell. The genus contains two species, the
type-species and S. uchidai Gauld & Mitchell. These occur in the mountains at the eastern Palaearctic/
Oriental interface (Gauld & Mitchell, 1981).
XYLOPHION Gauld
Xylophion Gauld, 1979: 77. Type-species: Ophion xylus Gauld, by original designation.
Mandibles not twisted, weakly to moderately tapered, subequally bi dentate; outer mandibular surface flat,
punctate, with proximal concavity. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile
convex, margin impressed, acute; clypeus in anterior aspect weakly convex. Ocelli large, the posterior ones
separated from eyes by their own minimum diameter or less; frontal carina absent; occipital carina strong,
ventrally often reaching hypostomal carina, sometimes ventrally obsolescent. Antennae of moderate
length or long, usually at least 1-5 times length of fore wing. Pronotum more or less unspecialized, though
one species has trace of a median transverse crest; spiracular sclerite exposed; notauli quite weak but
usually distinct on anterior 0-2 of segment, rarely absent; epicnemial carina strong, reaching onto pleuron
above level of lower corner of pronotum ; mesopleural furrow weak , diagonal . Scutellum weakly convex , at
most carinate laterally on anterior 0-2; posterior transverse carina of mesosternum present only laterally as
vestiges. Propodeum with anterior area more or less occluded except centrally where it is discernible as a
semicircular area; anterior transverse carina usually more or less complete, posterior transverse carina
from present laterally as vestiges to almost complete; longitudinal carinae vestigial; posterior area rather
smooth. Fore wing with pterostigma large and triangular; marginal cell long; Rs+2r evenly broadened and
curved before joining pterostigma; discosubmarginal cell with a large glabrous area in anterior corner;
128 I. D. GAULD
lra-cu with a short ramellus, centrally angulate. Hind wing with Rs strongly curved. Fore tibia with a
membranous flange extending from 0-1-0-3 of its length behind macrotrichial comb; mid and hind
trochantelli unspecialized; inner hind tibial spur flattened, with a margin of long, close hairs; hind tarsal
claws of female unspecialized, of male unique in having a central gap between an inner and outer row of
pectinal teeth. Gaster slender; tergite 2 in profile very elongate with thyridia elliptical, separated from
anterior margin by its own length or less; umbo strongly developed; epipleuron up-turned. Ovipositor
sheath slender.
Xylophion is a small genus restricted to the Australian tectonic plate. Three species, one of which is
undescribed, occur in south-eastern Australia. One of these, X. xylus, also occurs more widely throughout
Australia and has been found in the New Guinea highlands (Gauld, 1984a).
Xylophion is characterized by the venation, particularly the possession of a very short 3r-w, and the
reduced flange on the fore tibial spur. A particularly striking, and hitherto unrecognized, autapomorphy of
the genus is the form of the male tarsal claw. Instead of having a sinuous row of pectinal teeth present (as is
normal for ophionines), Xylophion males have a marked discontinuity in the centre of the row between the
teeth on the inner and outer edges of the claw. In its least derived form only the central gap exists
(Xylophion species 1, BMNH) but the other two species are specialized further in different ways. X. ketus
has the inner and outer marginal rows extended laterally so that there is a central overlap between the two
parallel ends of the rows of teeth. In X. xylus there is a less pronounced overlap but the distal portion of the
claw is flattened and the terminal tooth reduced so the pectinal row forms a 'fence' around the distal end of
the claw.
The SICOPHION genus-group
This group comprises three genera, Sicophion and Janzophion from montane tropical South
America and Riekophion from Australia. The species in this complex exhibit an unusual
combination of plesiomorphic and apomorphic features. All possess a well-developed membra-
nous flange on the fore tibial spur though none has an umbo on tergite 2 and often the upper
corner of the pronotum is expanded to partially occlude the spiracular sclerite. It is suggested
here that these three genera comprise a distinct lineage which is derived with respect to the
Ophion group, and that the Sicophion group represents a separate southern radiation of the
ophionines. However, as mentioned above (see p. 117), if the polarity of characters 38 and 47-1
has been misinterpreted then the Ophion lineage may represent a more derived group than the
Sicophion lineage. If this were the case then it is possible that the Sicophion group is a collection
of relict genera whose present southern distribution can be explained by Darlington's (1965)
'glove hypothesis'.
JANZOPHION gen. n.
Type-species: Janzophion nebosus sp. n.
Mandibles twisted perhaps 5, slightly tapered distally, with upper tooth slightly the longer; outer
mandibular surface with a diagonal line of hair extending from upper proximal corner to near centre.
Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile convex, margin impressed, in
anterior aspect with margin weakly convex. Ocelli very large, the hind ones more or less contiguous with
eyes; frontal carina absent; occipital carina absent. Antennae very long and slender, about 1-5 times length
of fore wing. Pronotum unspecialized; spiracular sclerite partially concealed; notauli vestigial; epicnemial
carina present laterally and ventrally; mesopleural furrow indistinct. Scutellum weakly convex, laterally
carinate only on anterior 0-1-0-3; posterior transverse carina of mesosternum complete. Propodeum with
anterior area long; anterior transverse carina and usually posterior transverse carina complete, often with a
median longitudinal carina present; propodeum otherwise rather smooth. Fore wing with pterostigma
moderately broad; marginal cell very long; Rs+2r bowed, thickened before joining pterostigma; discosub-
marginal cell with glabrous area near anterior corner, but anterior to this is narrow hirsute region; \m-cu
without a ramellus, proximally fairly straight, distally strongly bowed. Hind wing with Rs weakly curved;
marginal cell proximally glabrous; penultimate distal hamulus longer than its fellows, the distal one slightly
shorter but longer than the proximal ones. Fore tibial spur with a membranous flange behind the
macrotrichial comb; mid and hind trochantelli unspecialized; inner hind tibial spur flattened, with margin
of close long hairs; hind tarsal claw unspecialized. Gaster slender; tergite 2 in profile very elongate, with
thyridia remote from anterior margin; umbo vestigial; epipleuron up-turned. Ovipositor sheath slender.
PHYLOGENY OF THE OPHIONINAE 129
Janzophion is a distinctive genus which bears a very strong phenetic similarity to the Old World genus
Leptophion. Unlike Leptophion, Janzophion species lack the occipital carina, and have a well-developed
membranous flange present on the fore tibial spur, behind the macrotrichial comb.
Janzophion nebosus sp. n.
Fore wing length 14-16 mm.
Head slightly more elongate than normal for ophionines; lower face polished, 1-4-1 -5 times as long as
broad; malar space 0-50-0-55 times basal mandibular width; head strongly narrowed behind eyes, occiput
mediodorsally slightly concave. Flagellum with 66-68 segments, the tenth segment about 2-0 times as long
as wide. Mesoscutum with margin slightly out-turned; scutellum finely shagreened; mesopleuron with
upper part highly polished, finely and sparsely punctate, ventrally slightly more coriaceous; metapleuron
similar. Propodeum in profile evenly declivous; propodeal spiracle joined to pleural carina by weak ridge;
most of alitrunk bearing fine pale pubescence. Fore wing with AI = 1-20-1-25; CI = 0-43-0-47; ICI =
0-61-0-70; SDI = 1-14-1-17; cu-a proximal to base of Rs&M by about 0-3 times its own length; outer hind
corner of 2nd discal cell about 90. Hind wing with about 8 distal hamuli; NI = 3-70-4-60. Legs
unspecialized; hind tarsal claws of male with slightly finer and denser pectinate comb than that of female.
Gaster slender, male with subgenital plate bearing long fine pubescence; gonosquamae quite long, dorsally
somewhat membranous.
Pale yellowish species, with interocellar area, part of mesoscutum, much of mesopleuron, metapleuron
and part of propodeum blackish; gaster with distal part of tergite 5 and tergites 6+ infuscate. Pterostigma
and Rs+2r blackish, other veins flavous; wing hyaline, proximal angle of marginal cell infumate.
REMARKS. This species has been taken at light in cloud forests between 1500 and 2350 m in Costa Rica.
Nothing is known of its biology.
MATERIAL EXAMINED
Holotype cf , Costa Rica: Alajuela Prov; Volcan Poas N. P., xii.1982 (Janzen & Hallwachs) (BMNH).
Paratypes. Costa Rica: 1 cf , same data as holotype (BMNH); 1 cf , same locality as holotype, xii.1981
(Janzen & Hallwachs} (BMNH); 1 $, Monte Verde Reserve, 1500 m, ii.1980 (Mason) (TC).
RIEKOPHION Gauld
Riekophion Gauld, 1977: 21. Type-species: Allocamptus emandibulator Morley, by original designation.
Mandibles not twisted, weakly evenly tapered or distally parallel-sided, subequally bidentate; outer
mandibular surface flat, punctate. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile
convex, margin usually blunt, in one species, impressed; clypeus in anterior aspect from weakly convex to
slightly concave. Ocelli from large, the hind ones contiguous with eyes, to rather small, the hind ones
separated from eyes by 0-5 times their minimum diameter; frontal carina absent, or weakly present in one
species; occipital carina usually complete, in one species ventrally obsolescent, not reaching hypostomal
carina. Antennae moderately long and relatively slender, at least 1-3 times length of fore wing. Pronotum
either unspecialized, or in two species mediodorsally long and flat; spiracular sclerite exposed, or in one
species with slight expansion of dorsal lobe of pronotum; notauli vestigial or absent; epicnemial carina
complete, generally curved to meet anterior margin of pleuron; mesopleural furrow very weak to quite
distinct diagonally. Scutellum weakly convex, very characteristic in being fairly narrow and barely tapered,
and medially longitudinally higher than laterally; carinae if present only on anterior 0-3 or less; posterior
transverse carina of mesosternum complete. Propodeum with anterior area moderately to very long, often
irregularly striate; propodeum usually with anterior and posterior transverse carinae complete, latero-
median ones weaker but often complete behind anterior transverse carina and usually enclosing a long
narrow area superomedia. Fore wing with pterostigma moderately to very slender; marginal cell usually
long; Rs+2r proximally sinuous, only weakly broadened and not curved abruptly to join the pterostigma,
unusual in joining stigma distal to proximal end; discosubmarginal cell anteriorly glabrous, often with
group of isolated hairs near base of Rs+2r, always bearing a detached corneous or sclerotized patch in
membrane. Im-cu weakly sinuous or fairly evenly curved. Hind wing with Rs straight or very weakly
curved. Fore tibial spur with a membranous flange behind macrotrichial comb; mid and hind trochantelli
unspecialized; inner hind tibial spur flattened, the margin with close long hairs; hind tarsal claws elongate,
with thyridia small, widely separated from anterior margin to which it is joined by a groove; umbo absent;
epipleuron up-turned. Ovipositor sheath slender; unusual in having male subgenital plate ornamented
with projections.
Riekophion is a very distinctive genus, characterized by the quadrate scutellum, unusual venation and
130 I. D. GAULD
ornamented subgenital plate (Gauld, 1977). It contains three species which occur in southern and western
Australia. The phylogenetic inter-relationships of these species are suggested in Fig. 28. Nothing is known
of the biology of these insects.
SICOPHION Gauld
Sicophion Gauld, 1979: 71. Type-species: Sicophion pleuralis Gauld, by original designation.
Mandibles stout, twisted about 25-35 and with lower tooth slightly the longer; outer mandibular surface
more or less flat, sparsely punctate. Maxillary palp 5-segmented, labial palp 4-segmented; maxillae
elongate; clypeus in profile convex, apical margin not impressed, in anterior aspect truncate. Ocelli very
large, the posterior ones close to margin of eye; frontal carina absent; occipital carina mediodorsally
incomplete, ventrally obsolescent, not reaching hypostomal carina. Antennae very long and slender, at
least 1-4 times length of fore wing. Pronotum unspecialized; spiracular sclerite virtually completely
occluded by enlarged flap formed from hind corner of pronotum; notauli absent; epicnemial carina strong,
extending onto mesopleural furrow; mesopleural furrow strong, extending from episternal scrobe forward.
Scutellum quite convex, narrow, carinate only on anterior 0-2-0-3; posterior transverse carina of
mesosternum present only laterally as vestiges. Propodeum with anterior area occluded; anterior trans-
verse carina present centrally, laterally obsolescent; posterior transverse carina present as lateral vestiges;
propodeum otherwise rather smooth, posterior area not differently sculptured from spiracular area. Fore
wing with pterostigma broad, distally abruptly narrowed; marginal cell moderately long and slender,
unique amongst ophionines in being broadest distad of centre at sinuation in distal abscissa of Rs; Rs+2r
not angled near junction with pterostigma, but somewhat broadened, somewhat angled slightly proximal
to centre; discosubmarginal cell very extensively glabrous anteriorly, the glabrous area bearing indistinct
corneous areas; Im-cu without a ramellus, rather abruptly curved. Hind wing with Rs unique in the
Ophioninae in being slightly convex. Fore tibial spur with membranous flange behind macro trichial comb;
mid and hind trochantelli very elongate, but unspecialized; inner hind tibial spur strongly flattened with a
fringe of long, close hairs; hind tarsal claws unspecialized, markedly sexually dimorphic, those of male
being far more closely pectinate than those of the female. Gaster exceptionally slender; tergite 2 elongate,
thyridia absent; umbo absent; epipleuron very narrow, pendant. Ovipositor sheath slender, ovipositor
unique in being proximally angled and without a subapical notch.
Sicophion is a small Neotropical genus with a single described Bolivian species (Gauld, 1979). In the
BMNH is a short series of a putative second species collected in Costa Rica by Janzen & Hallwachs. These
differ from S. pleuralis in having a more extensive glabrous area in the discosubmarginal cell and a less
obviously 'bent' Rs+2r. There are subtle differences in colour and head shape also.
Sicophion is one of the most distinctive of ophionine genera. The sinuous Rs in the fore wing, the slightly
convex Rs in the hind wing and the basally angulate, acutely pointed ovipositor with no subapical notch are,
amongst ophionines, unique autapomorphies of the genus. Their very slender form with extensive
semi-matt black coloration is a typical feature of ophionines from higher elevations (e.g. Enicospilus
ruwenzorius Gauld & Mitchell) and the rather elongate head shape is also found in some upper montane
species (e.g. Ophion longiceps (Townes), Gauld & Mitchell, 1981). The presumed loss of the dorsal notch
is a feature that occurs in several genera of ichneumonids with more slender ovipositors (e.g. Parania in the
Anomaloninae, Gauld, 1976).
Sicophion species are only known to occur at mid and high elevation in the Neotropical region.
Specimens have been taken between 1600 and 3000 m. Nothing is known of their host ranges.
The EREMOTYLUS genus-group
This group contains three genera, Eremotylus, Trophophion and Hellwigiella. The last two are
monobasic and probably really represent single phenetically highly divergent species of Ere-
motylus. The majority of species of all genera are eremic organisms.
The Eremotylus genus-group is characterized by the fore wing venation; Rs+2r is strongly
geniculate and thickened before joining the pterostigma whilst \m-cu is usually fairly evenly
arcuate. The clypeus is usually flat or out-flared and the margin is not subapically impressed.
EREMOTYLUS Foerster
Eremotylus Foerster, 1869: 150. Type-species: Ophion marginatus Gravenhorst (= Anomalon margina-
tum Jurine), by subsequent monotypy, Thomson, 1888: 1193.
PHYLOGENY OF THE OPHIONINAE
131
Fig. 28 Suggested phylogenetic inter- relationships of species of Riekophion. The apomorphic characters
supporting this arrangement are: 1, mandible slender; 2, ocelli small; 3, occipital carina mediodorsally
broadened; 4, body extensively black; 5, pronotal lobe expanded; 6, pronotum medio-dorsally flat and
lengthened; 7, presence of hair patch in fenestra; 8, cubital index small; 9, distal sclerite lost; 10, ocelli
grossly enlarged; 11, scutellum quadrate; 12, possession of alar sclerites; 13, metapleuron inflated; 14,
Rs+2r sinuous; 15, Rs+2r joining pterostigma near centre; 16, male subgenital plate ornamented.
Camptoneura Kriechbaumer, 1901: 23. Type-species: Ophion marginatus Gravenhorst (= Anomalon
marginatum Jurine), by subsequent designation, Viereck, 1914: 27. [Junior homonym of Camptoneura
Agassiz, 1846.]
Genophion Felt, 1904: 123. Type-species: Genophion gilletti Felt (= Ophion costale Cresson), by original
designation.
Camptoneuroides Strand, 1928: 52. [Replacement name for Camptoneura Kriechbaumer.]
Clistorapha Cushman, 1947: 450. Type-species: Ophion subfuliginosus Ashmead, by original designation.
Boethoneura Cushman, 1947: 450. Type-species: Boethoneura arida Cushman, by original designation.
Chilophion Cushman, 1947: 450. Type-species: Ophion abnormum Felt, by original designation.
Chlorophion Townes, 1971: 55. Type-species: Chlorophion vitripennis Townes, by original designation.
132 I. D. GAULD
Mandibles not twisted, usually quite long, moderately strongly narrowed, usually subequally bidentate or
with upper tooth slightly the longer; outer mandibular surface flat or slightly convex, punctate. Maxillary
palp 5-segmented, labial palp 4-segmented; clypeus in profile more or less flat or even slightly out-flared,
margin never impressed; clypeus in anterior aspect truncate or slightly concave. Ocelli usually moderately
large to large, the posterior ones separated from the eyes by less than 0-5 their own maximum diameter, or
in a few species with the ocelli small and the posterior ones separated from the eyes by more than their own
maximum diameter; frontal carina absent or rarely very faintly indicated; occipital carina complete, rarely
ventrally somewhat obsolescent. Antennae moderately long, at least 1-2 times length of the fore wing.
Pronotum unspecialized; spiracular sclerite exposed; notauli vestigial or absent; epicnemial carina usually
complete; mesopleural furrow usually rather weak but discernible, faintly indicated between episternal
scrobe and subalar prominence. Scutellum weakly convex, usually carinate 0-3 to 0-8 of its length, rarely
without carinae; posterior transverse carina of mesosternum from complete to present only laterally as
vestiges. Propodeum generally with anterior area short but not occluded, in some species moderately long;
anterior transverse carina usually distinct, at least centrally, sometimes complete, rarely absent; posterior
transverse carina present laterally as vestiges or absent, other carinae generally absent; posterior area from
smooth and polished to rugulose. Fore wing with pterostigma moderately slender; marginal cell long and
slender, rarely rather short; Rs+2r abruptly geniculate and thickened near pterostigma; discosubmarginal
cell with a small glabrous area in anterior corner; Im-cu without a ramellus, arcuate or very weakly
sinuous. Hind wing with Rs from almost straight to strongly curved. Fore tibial spur with a membranous
flange behind macrotrichial comb; mid and hind trochantelli unspecialized; inner hind tibial spur generally
flattened with a marginal fringe of close hairs; hind tarsal claws usually unspecialized, sometimes slightly
longer and less evenly curved than normal. Gaster slender; tergite 2 in profile elongate with thyridia
separated from fore margin by its own length or a little more, characteristically at the posterior end of a
weak trough; umbo usually distinct, rarely weak; epipleuron up-turned. Ovipositor sheath slender.
Eremotylus is a moderately large genus containing about 35 species, most of which occur in the drier
regions around the Mediterranean, the Middle East, Central Asia, the south-western United States and
northern Mexico. Isolated species have a wider distribution; E. subfuliginosus occurs in the north-eastern
part of the U.S.A., whilst the morphologically very specialized species E. marginatus is not uncommon
throughout much of western Europe. E. vitripennis occurs in the drier parts of southern South America,
whilst E. perdix occurs in the Indian subcontinent (Gauld & Mitchell, 1981). The Palaearctic species were
monographed recently by Horstmann (1981) who recognized eight species. The New World species are
extremely poorly known. Virtually nothing is known about the host ranges of species of Eremotylus. The
only reliable host record to hand is of one European species (E. curvinervis Kriechbaumer) which has been
reared from a species oiDryobota Lederer (Lepidoptera: Noctuidae) (Seyrig, 1926).
Some authors (e.g. Cushman, 1947; Townes, 1971) divided the species of Eremotylus between a number
of small genera characterized mostly by differences in the development of carinae, particularly the
transverse mesosternal carina. Gauld (1979) pointed out that despite these differences, all species share a
large number of features and seem to comprise a natural group. Horstmann (1981) accepted this treatment.
The present study has reinforced the author's opinion that this is a natural group; it is definable on the basis
of several apomorphies including the characteristically modified Rs+2r, the arcuate or slightly sinuous
Im-cu and the unoccluded anterior propodeal area. The characteristic clypeus and rather slender, longish
mandibles are useful confirmatory characters.
The present disjunct distribution of the genus suggests that at one time the range of the genus must have
been wider. Further discussion of zoogeography is best left until the New World species are better known.
HELLWIGIELLA Szepligeti
Hellwigiella Szepligeti, 1905: 23. Type-species: Hellwigiella nigripennis Szepligeti, by subsequent desig-
nation, Viereck, 1914: 67.
Mandible twisted about 15, not appreciably narrowed, with upper tooth slightly shorter than the lower and
bearing a pronounced very sharp ventral flange; outer mandibular surface punctate, with a strong proximal
concavity. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile quite convex centrally,
with margin reflexed, quite thin; clypeus in anterior aspect almost truncate. Ocelli small, the posterior ones
separated from eye margin by more than their own maximum diameter; frontal carina absent; occipital
carina strong, ventrally joining hypostomal carina. Antennae short and stout, central segments transverse,
the flagellum barely longer than the fore wing. Pronotum mediodorsally quite long, otherwise unspecial-
ized; spiracular sclerite exposed; notauli absent; epicnemial carina strong, complete; mesopleural furrow
vestigial. Scutellum quite convex, laterally carinate about 0-8 of its length; posterior transverse carina of
mesosternum complete or obsolescent centrally. Propodeum with anterior area quite short, impressed as a
PHYLOGENY OF THE OPHIONINAE 133
deep U-shaped groove; anterior transverse carina more or less complete, other carinae indistinct, posterior
area reticulate. Fore wing with pterostigma moderately stout; marginal cell quite long and slender; Rs+2r
abruptly geniculate and slightly broadened before joining pterostigma; discosubmarginal cell with a very
small glabrous area anteriorly; Im-cu evenly curved, usually without a ramellus. Hind wing with Rs evenly
bowed. Fore tibial spur with a membranous flange behind macrotrichial comb; mid and hind trochantelli
unspecialized; inner hind tibial spur almost cylindrical, with a fringe of sparse hairs; hind tarsal claw quite
long and weakly curved. Gaster moderately slender, tergite 2 about twice as long as deep with thyridia
separated from anterior margin by about their own length; umbo distinct; epipleuron membranous,
pendant in most individuals. Ovipositor sheath slender.
Hellwigiella contains what is, I believe, a single Mediterranean species, though it is sculpturally and
chromatically rather variable and has been divided into two (Szepligeti, 1905). Hellwigiella is morphologi-
cally extremely distinctive on account of the characteristic clypeus and mandible, and the form of the
propodeum. The majority of apomorphic features characterizing the genus are obviously adaptations to a
diurnal existence and most are paralleled in other genera with similar habits. It is possible that Hellwigiella
is merely a specialized diurnal offshoot of Eremotylus, but if this were so then it is necessary to postulate
that the specialized invaginated condition of the anterior area of the propodeum (character 22-2) has been
derived from the partially elongated condition of Eremotylus (character 22-1). In the preceding phylo-
genetic analysis it has been suggested that the apomorphic states of these characters represent alternative
specializations, so it is possible that Eremotylus and Hellwigiella have a sister-group relationship. Nothing
is known of the biology of Hellwigiella.
TROPHOPHION Cushman
Trophophion Cushman, 1947: 447. Type-species: Trophophion tenuiceps Cushman, by original desig-
nation.
Mandibles barely twisted, quite stout, weakly narrowed, subequally bidentate; outer mandibular surface
weakly convex. Maxillary palp 5-segmented, labial palp 4-segmented; maxilla and labium elongate,
projecting below mandibles; clypeus in profile flat, in anterior aspect almost truncate. Ocelli small, the
posterior ones separated from the eye by more than their own maximum diameter; frontal carina absent;
occipital carina complete dorsally, ventrally incomplete. Antennae short, slightly clavate, barely longer
than fore wing. Pronotum unspecialized; spiracular sclerite exposed; notauli absent; epicnemial carina
complete; mesopleural furrow indistinct. Scutellum weakly convex, without lateral carinae; posterior
transverse carina of mesosternum present only laterally as vestiges. Propodeum with anterior area short
but not occluded; anterior transverse carina present medially, posterior one present laterally as vestiges,
other carinae obsolescent; posterior area polished, punctate. Fore wing with pterostigma moderately
slender; marginal cell moderately short; Rs+2r thickened and curved to join pterostigma; discosubmar-
ginal cell with glabrous area anteriorly; Im-cu evenly curved, without a ramellus. Hind wing with Rs
bowed. Fore tibial spur with a membranous flange behind macrotrichial comb; mid and hind trochantelli
unspecialized; inner hind tibial spur weakly flattened, fringed with long fine hairs; tarsal claws rather long
and weakly curved. Gaster moderately stout and rather more cylindrical than that of most ophionines;
tergite 2 in profile less than twice as long as posteriorly deep, with thyridia separated from anterior margin
by about its own length; umbo rather weak; epipleuron up-turned. Female with tergites 3-7 shallowly
notched medially, with subgenital plate large, strongly sclerotized and centrally notched; ovipositor and
sheath rather stout.
Only a single species of Trophophion is known. This occurs in the drier south-west of the U.S. A. Its host
is unknown. Trophophion is recognizable on account of the specializations of the head and mouthparts and
the female gaster. In other respects it is very similar to some Eremotylus species, and it is almost certainly a
specialized offshoot of this genus. However, I have hesitated to synonymize Trophophion until some
details of its biology and host relationship are known. If these are very different from those of Eremotylus
then I suggest if be left as a separate genus even though this leaves Eremotylus as a paraphyletic taxon.
Additional material from the dry parts of the U.S. A. and northern Mexico might help in elucidating the
phylogenetic position of this bizarre species.
The THYREODON genus-group
This group contains five genera, Barytatocephalus, Euryophion, Thyreodon, Rhynchophionand
Dictyonotus. The first of these is included in the complex on the basis of three apomorphies, the
possession of an impressed anterior propodeal area (22-2), long, weakly curved tarsal claws, and
134
I. D. GAULD
absence of any transverse carinae on the propodeum. Barytatocephalus lacks the tilted mandibu-
lar axis, slender pterostigma and cylindrical hind tibial spurs of other genera, and thus seems to
constitute the most primitive branch of the lineage. The remaining genera form a holophyletic
group, and this may be divided into two holophyletic subgroups, one comprising Thyreodon,
Dictyonotus and Rhynchophion, the other containing just Euryophion. The former subgroup is
defined by the possession of a pointed clypeus, a very short and stout flagellum and an
exceptionally elongate propodeal spiracle. A further possible apomorphy of this lineage is the
possession of an evenly hirsute anterior part of the discosubmarginal cell (assuming that
presence of a small glabrous area is an apomorphy of the subfamily). Euryophion has a slightly
concave clypeus and a somewhat thickened and usually proximally curved Rs+2r. These two
lineages seem to be biologically distinct; species of theThyreodon subgroup have only been
recorded as parasites of Sphingidae (Gauld & Mitchell, 1978; 1981; Carlson, 1979), whilst
Euryophion species attack Eupterotidae and Saturniidae.
The close relationship between Dictyonotus, an Old World genus, and Rhynchophion and
Thyreodon, primarily Neotropical genera, suggests they may have had a common ancestor that
was widely distributed throughout the Nearctic and eastern Palaearctic regions. D. purpuras-
cens, which could well be rather similar to the group ancestor, is currently widely distributed in
the eastern Palaearctic, occurring as far north as 50 (Townes et al. , 1965) , so a slight extension of
its range northwards would have permitted migration across the Bering Strait. Movement from
Asia to America is postulated on the belief the group has had an Old World origin, a suggestion
favoured by the present distribution of the most primitive member of the group (Barytato-
cephalus) and the sister-lineage of the Thyreodon subgroup (Euryophion).
The exact relationship between the genera Thyreodon, Rhynchophion and Dictyonotus is
unclear, as slight evidence in the form of a very few highly homoplastic characters can be
marshalled for placing Thyreodon as the sister-group of either Rhynchophion or Dictyonotus, or
S
$
w.
10
16
Fig. 29 Suggested phylogenetic inter-relationships of Dictyonotus species. The length of the lines is
proportional to the number of apomorphies characterizing it. The apomorphic characters supporting
this arrangement are: 1, presence of a metapleural tubercle; 2, striation present on gena; 3, hind tarsus
flattened; 4, Im-cu sinuous; 5, body densely pubescent; 6, tergite 2 posteriorly deeper than long; 7,
malar space longer than basal mandibular width; 8, occipital carina mediodorsally broadened; 9, petiole
depressed; 10, tergite 2 posteriorly at least as deep as long; 11, petiolar spiracles anterior to margin of
sternite; 12, presence of metanotal swelling; 13, reduction in number of hamuli; 14, development of
strongly punctate scutellum; 15, thyridia remote from anterior margin of tergite; 16, complete posterior
transverse carina of mesosternum (? a reversal); 17, presence of tubercle on metanotal margin; 18,
epipleuron 2 pendant.
PHYLOGENY OF THE OPHIONINAE 135
even Rhynchophion + Dictyonotus (see p. 93). What is clear is that both Rhynchophion and,
more particularly, Thyreodon, are characterized by a string of apomorphies (as is nigrocyaneus),
whilst purpurascens may well have changed very little from the form of the group ancestor (Fig.
29). One wonders what adaptation acquired by the Thyreodon lineage has allowed it to radiate
so markedly in the Neotropics and give rise to a complex of 30 or more quite closely related
species attacking sphingids, whilst in Asia, Dictyonotus, although presented with a very similar
diversity of potential hosts (Rothschild & Jordan, 1903), has apparently failed to radiate
appreciably.
BARYTATOCEPHALUS Schulz
Barycephalus Brauns, 1895: 43. Type-species: Barycephalus mocsaryi Brauns, by subsequent designation,
Viereck, 1914: 19. [Homonym of Barycephalus Guenther, I860.]
Barytatocephalus Schulz, 1911: 23. [Replacement name for Barycephalus Brauns.]
Mandible not twisted, very weakly narrowed, with upper tooth a little stouter but no longer than the lower;
outer mandibular surface with a strong proximal concavity, and with a weak trace of a diagonal groove.
Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile convex, margin impressed, acute;
clypeus in anterior aspect weakly convex. Ocelli small, the posterior ones remote from the eyes; frontal
carina absent, though a small trace may be discernible in some specimens; occipital carina complete,
ventrally reaching to hypostomal carina. Antennae quite short, about 1-3 times as long as fore wing, central
segments slightly elongate. Pronotum unspecialized; spiracular sclerite exposed; notauli absent; epic-
nemial carina complete, present on mesopleuron; mesopleural furrow weak but discernible as a short
diagonal impression extending from episternal scrobe towards subalar prominence. Scutellum convex,
carinate laterally only on anterior 0-4; posterior transverse carina of mesosternum complete. Propodeum
convex, with anterior area impressed as a deep groove; propodeal carinae absent; posterior area coarsely
punctate. Fore wing with pterostigma moderately slender; marginal cell moderately long; Rs+2r slender,
proximally almost straight; discosubmarginal cell with a small glabrous area at anterior corner, remainder
of cell sparsely hirsute; \m-cu somewhat irregularly convex, without a ramellus. Hind wing with Rs
virtually straight; genus unusual in having distal abscissa of Cu\ very weak and the first abscissa of Cu\ and
cu-a forming an almost straight line, oblique, so anterior corner of first submarginal cell is about 50. Fore
tibial spur without membranous flange behind macrotrichial comb; mid and hind trochantelli simple; inner
hind tibial spur somewhat flattened (though less so than in most taxa in this subfamily), with a fringe of
moderately long hairs; hind tarsal claws long, weakly curved. Gaster moderately slender; tergite 2 in
profile with thyridia close to anterior margin, elongate, umbo vestigial; epipleuron up-turned. Ovipositor
slender.
Barytatocephalus is a small genus containing a single species that is widely distributed throughout the
eastern Mediterranean region and the southern U.S.S.R. Nothing is known of its biology.
DICTYONOTUS Kriechbaumer
Dictyonotus Kriechbaumer, 1894a: 198. Type-species: Dictyonotus melanarius Kriechbaumer (=
Thyreodon purpurascens Smith), by monotypy.
Aglaophion Cameron, 1903: 131. Type-species: Aglaophion flavinervis Cameron, by monotypy.
Ophionopsis Tosquinet, 1903: 389. Type-species: Ophionopsis fulvipes Tosquinet (= Ophionopsis nigro-
cyaneus Tosquinet), by subsequent designation, Viereck, 1914: 106. Syn. n.
Hybopleurax Enderlein, 1912: 624. Type-species: Hybopleurax sumatranum Enderlein (= Thyreodon
purpurascens Smith), by monotypy.
Hypselogastrina Enderlein, 1918: 217. Type-species: Hypselogastrina saliina Enderlein (= Ophionopsis
nigrocyaneus Tosquinet), by original designation. Syn. n.
Coracophion Shestakov, 1926: 260. Type-species: Coracophion manganicolor Shestakov (= Thyreodon
purpurascens Smith), by monotypy.
Mandibles stout, not twisted, barely narrowed distally, equally bidentate or with lower tooth slightly the
longer; outer mandibular surface more or less flat, with scattered hairs. Maxillary palp 5-segmented, labial
palp 4-segmented; clypeus in profile flat, with margin often slightly out-flared, in anterior aspect convex
with a median obtuse point. Ocelli small, the hind ones separated from the eye margins by at least their own
minimum diameter; frontal carina present; occipital carina complete, ventrally usually not joining the
hypostomal carina. Pronotum unspecialized; spiracular sclerite exposed; notauli absent; epicnemial carina
more or less complete, always present on mesopleuron; mesopleural furrow weakly impressed, horizontal,
136 I. D. GAULD
extending from episternal scrobe to near upper end of epicnemial carina. Scutellum moderately convex,
with longitudinal lateral carinae present only on anterior end; posterior transverse carina of mesosternum
complete; metanotum exceptional in being produced into a blunt prominence above upper end of spiracle.
Propodeum with anterior area short, impressed as a deep groove; propodeal carinae absent or with vestiges
of lateromedian ones discernible; posterior area convex, rugose-reticulate. Fore wing with pterostigma
slender; marginal cell of moderate length; Rs+2r very slightly bowed proximally, not appreciably
broadened; Im-cu evenly curved to somewhat sinuous, without a ramellus; discosubmarginal cell evenly
hirsute anteriorly. Hind wing with Rs weakly curved or almost straight. Fore tibial spur without a
membranous flange behind macro trichial comb; mid and hind trochantelli unspecialized; inner hind tibial
spur cylindrical, without a row of long marginal hairs; hind tarsal claws long and weakly curved. Caster
stout to very stout; tergite 2 in profile from about 1-5 times as long as posteriorly deep to less than 1-0 times
as long as deep, with thyridia oval, separated from anterior margin by about its own diameter or more;
umbo absent; epipleuron pendant. Ovipositor sheath slender.
This genus contains four species, D. purpurascens (Smith), D. flavinervis (Cameron) from the Oriental
and eastern Palaearctic regions and D. nigrocyaneus (Tosquinet) comb. n. and D. setus (Gauld & Mitchell)
comb. n. from the Afrotropical region. The last two were formerly placed in Ophionopsis, a genus
originally proposed to accommodate the morphologically aberrant nigrocyaneus. D. setus is much more
similar to the main Dictyonotus line than is nigrocyaneus, suggesting the two genera should be united
(Gauld & Mitchell, 1978). The Dictyonotus lineage is distinguished only by one weak apomorphy (Fig. 29),
and purpurascens in turn by a further one, suggesting this species may resemble the ancestor of this genus,
and also the ancestors of Rhynchophion and Thyreodon.
The hosts of Dictyonotus are apparently the larvae of Sphingidae (Townes et al. , 1965 ; Gauld & Mitchell ,
1978; 1981).
EURYOPHION Cameron
Euryophion Cameron, 1906: 83. Type-species: Euryophion nigripennis Cameron, by monotypy.
Eurycamptus Morley, 1912: 27. Type-species: Ophion latipenne Kirby, by subsequent designation,
Viereck, 1914: 57.
Thoracophion Roman, 1943: 22. Type-species: Thoracophion ventrator Roman (= Ophion latipenne
Kirby), by monotypy.
Primophion Townes, 1971: 65. Type-species: Primophion adustus Townes, by original designation.
Rictophion Townes, 1971: 66. Type-species: Euryophion nebulifer Morley (= Cymatoneura ikuthana
Kriechbaumer), by original designation. Syn. n.
Mandibles large, not twisted, weakly narrowed apically, subequally bidentate. Maxillary palp 3-5
segmented, labial palp 3-4 segmented; clypeus in profile flat or with apex out-turned, in anterior aspect
usually with margin slightly concave. Ocelli small to large; frontal carina present or absent; occipital carina
usually complete, rarely dorsally incomplete. Antennae rather stout, not longer than fore wing. Pronotum
unspecialized, spiracular sclerite exposed; notauli weak or vestigial; epicnemial carina present ventrally,
sometimes laterally absent; mesopleural furrow usually vestigial. Scutellum without lateral carinae; hind
margin of metanotum at the most only weakly swollen before propodeal spiracle; posterior transverse
carina of mesosternum absent except for lateral and rarely central vestiges. Propodeum with anterior area
short, impressed as a deep trough; carinae virtually absent though their former position may be indicated
by rugosities; posterior area from punctate to finely wrinkled to coriaceous. Fore wing with pterostigma
slender; marginal cell moderately long; Rs+2r abruptly curved and thickened basally; discosubmarginal
cell without a distinct fenestra, but usually with a small glabrous area in anterior corner; Im-cu fairly
evenly curved, without a distinct ramellus. Hind wing with Rs from weakly to strongly curved. Fore tibial
spur without a membranous flange behind macro trichial comb; mid and hind trochantelli simple; inner
hind tibial spur subcylindrical or slightly flattened, often with a reduced fringe of hairs; hind tarsal claws
long and weakly curved, rarely with a small apical flange. Gaster stout, tergite 2 in profile short and deep,
generally less than 3 times as long as deep posteriorly; thyridia present or absent; umbo vestigial;
epipleuron pendant or up- turned. Ovipositor sheath slender, often concealed.
Although no character is a unique autapomorphy of this genus, the species share a large number of
apomorphic features (p. 87). They resemble each other greatly in venation, shape of the gaster and form
of the head. No other ophionines have reduced palpar segments nor do any have the claw flange found in
some more specialized Euryophion. The initial analysis suggested that Euryophion is paraphyletic with
respect to Rictophion. In order to resolve this matter further a more detailed analysis was undertaken using
all species in the two genera, and this confirmed the paraphyletic nature of Euryophion. The sole
PHYLOGENY OF THE OPHIONINAE 137
representative of Rictophion, R. ikuthana, can be considered a specialized species of Euryophion, and it is
on this basis that Rictophion is here treated as a synonym.
The more detailed analysis involved eight species, E. latipennis (201), E. adustus (202), E. ikuthana
(203), E. nigripennis (204), E. meridionalis (205), E. variegatus (206), E. vexatious (207) and E. pisinnus
(208). Eighteen characters from the original set (1-1, 8-1, 8-2, 9, 13, 18-2, 19, 27-2, 29-1, 29-2, 30, 36-1,
36-2, 37, 43-2, 54, 57 and 59) were used together with the following eight characters.
70 Labrum shape. In most ichneumonids the labrum is either semicircular or triangular, but broader
basally than medially long; some Euryophion are specialized in having the labrum longer than broad
(1).
71 Hind tarsal claw. In some Euryophion the hind claw, is specialized in having a small flange near the
apex (1).
72 Rs in hind wing. The moderately curved condition is considered plesiomorphic for Euryophion; in
some taxa it is strongly curved (1).
73 Wing patterning. The plesiomorphic condition for ophionines (and members of out-groups) is
unpatterned wings. In some Euryophion species the fore wings are distinctly patterned (1).
74 Wing ground colour. Transparent, virtually colourless wings are plesiomorphic for ophionines. The
strongly infumate condition found in some Euryophion species is considered to be a derived feature
(1).
75 Rs in fore wing. The plesiomorphic condition is for this vein to be simply arcuate. The sinuous
condition of a few species of Euryophion is considered to be derived (1).
76 Rugosity of propodeum. The plesiomorphic condition for Euryophion species and related genera
appears to be possession of at least some rugae close to the position of the vestigial carinae; the
smooth, punctate propodeum of one Euryophion species is considered to be derived (1).
77 Hairiness of ovipositor sheath. The plesiomorphic condition for most ophionines, and other
ichneumonids, is for the sheath to bear close, moderately long pubescence. The short sparse
pubescence of one species is considered to be a derived characteristic (1).
The primary data matrix obtained (Table 20) was analysed using the methods outlined in the generic
analysis. Characters 8-1, 18-2, 29-1, 29-2, 76 and 77 each have only a single derived state and therefore
make no contribution to resolving phylogenetic relationships. A LeQuesne test on the remaining data set
(Table 21) showed three characters (9, 59, 74) to be particularly homoplastic. Removal of these caused an
improvement in the overall O/E ratio from 0-62 to 0-45. The high values evident in the labels matrix
Table 20 Primary data matrix for genus Euryophion. The taxa corresponding to the numbers are given in
the text (p. 137).
1.1 8.2 13 19 29.1 30 36.2 43.2 57 70 72 74 76
8.1 9 18.2 27.2 29.2 36.1 37 54 59 71 73 75 77
201 1
1000
010001100000
11000000
202
0011
000010111011
00010100
203
1110
111100000110
00000000
204 1
1000
010001110000
11000100
205
0000
010001100101
00000000
206
0001
011000111111
00101010
207
0001
011000111011
00111011
208
0000
001000111011
10111010
Table 21 Results of LeQuesne test on Euryophion data matrix. Conventions as in Table 2.
Character number: Failures observed expected O/E ratio
1.1: 2 9.3 0.22 8.1: 8 13.2 0.60 9 : 10 9.3 1.08 _13_ : 9 13.2 0.68
^9_ : 8 9.3 0.86 27.2: 10 14.1 0.71 30 : 6 13.2 0.45 36.2: 10 13.2 0.76
37_ : 5 14.1 0.35 43.2: 12 13.2 0.91 54 : 6 13.2 0.45 57 : 8 13.2 0.60
59 : 13 13.2 0.98 70 : 2 9.3 0.22 7l_ : 5 13.2 0.38 72 : 8 13.2 0.60
.73 : 5 13.2 0.38 M. : 12 9 - 3 K29 Zi : 5 13 - 2 - 38
Grand totals- failures observed expected O/E ratio
72 116.7 0.62
Ranking ratios
1.1 70 37 71 73 75 30 54 8.1 57 72 13 27.2 36.2 19 43.2 59 9 74
138 I. D. GAULD
(Appendix 9) strongly suggest parallelism for the derived state of character 9 (in taxa 202 and 203), of
character 59 (in taxon 208 with respect to other taxa) and character 74 (in taxa 202 and 204). The single
largest compatible clique comprises nine informative characters, 1-1, 8-2, 30, 37, 54, 70, 71, 73 and 75.
These support a cladogram that resolves all taxa except 206-8 which remain as a trichotomy. Parsimony
analysis yields two equally long minimum length trees which differ in their arrangement of taxa 206-8.
There is little biological justification for preferring one or other of these two arrangements (Figs 30, 31) but
the larger number of reversals involved in Fig. 31 mitigate marginally in favour of treating taxa 206 and 207
as sister-species.
203 202 208 207 206 205 201 204
8.2
18.2
29.1
29.2
8.1, 9
127.2
36.1
13
74
43.2
36.2
74
43.2
71
73
75
27.2
37
36.2
57
72
136.1
PHYLOGENY OF THE OPHIONINAE 139
203 202 208 207 206 205 201 204
8.2
18.2
29.1
27.2
36. Ir
36.2
Figs 30, 31 Alternative cladograms showing most parsimonious arrangement of species of Euryophion.
Squares indicate autapomorphic developments; circles that an apomorphic feature has been derived in
parallel in two separate lineages; diamonds that a feature has undergone three transformations.
140 I. D. GAULD
All methods of analysis yielded similar results in suggesting that E. latipennis and E. nigripennis are very
closely related, with the former possibly the ancestor of the latter. E. meridionalis was always placed on the
sister-species to the clade latipennis + nigripennis, and these three taxa formed a rather distinct
species-group. The three taxa whose interrelationship is difficult to resolve, E. variegatus, E. vexatious and
E. pisinnus, form a very distinct clade whilst E. ikuthana and E. adustus are less closely related. Clearly it
would be unsatisfactory to place ikuthana in a separate genus without at least creating separate genera for
the variegatus species-group and the latipennis species-group. E. adustus could be incorporated into the
former or also treated as a separate genus. As these insects are essentially similar animals that form a
distinct group with respect to other ophionines, it is suggested that they be placed in a single genus,
Euryophion, but to represent the phylogeny of the group four species-groups may be recognized. The
formal classification may thus be summarized as follows.
EUR YOPHION Cameron
Rictophion Townes
latipennis species-group
latipennis (Kirby)
nigripennis Cameron
meridionalis (Morley)
ikuthana species-group
ikuthana (Kriechbaumer)
adustus species-group
adustus (Townes)
variegatus species-group
variegatus Gauld & Mitchell
pisinnus Gauld & Mitchell
vexatious Gauld & Mitchell
Euryophion is primarily an Afro tropical genus with one species, E. vexatious, inhabiting southern India
(Gauld & Mitchell, 1978; 1981). This taxon is one of the most derived in the genus, and it is very closely
related to two African species. This suggests that Euryophion may only have recently become established
in the Oriental region. The data strongly suggest an African origin for the genus.
RHYNCHOPHION Enderlein
Rhynchophion Enderlein, 1912: 630. Type-species: Rhynchophion odontandroplax Enderlein, by original
designation.
Mandibles stout, not twisted, barely narrowed distally, with lower tooth somewhat longer than the upper;
outer mandibular surface more or less flat, with scattered hairs. Maxillary palp 5-segmented, labial palp
4-segmented; maxilla and labium elongate, projecting below apex of clypeus by about a distance equal to
half the length of the head; clypeus in profile flat, with margin often slightly out-flared, in anterior aspect
with median obtuse point. Ocelli small, the hind ones separated from the eye by at least their own
diameter; frontal carina absent or present but weak; occipital carina complete, ventrally not joining
hypostomal carina. Antennae short and stout, not as long as fore wing, and centrally with segments
transverse. Pronotum unspecialized; spiracular sclerite exposed; notauli obsolescent; epicnemial carina
present only ventrally, not extending onto mesopleuron; mesopleural furrow weakly impressed, hori-
zontal. Scutellum convex, laterally carinate only at extreme anterior end; posterior transverse carina of
mesosternum present only laterally as vestiges. Propodeum with anterior area short, present as a deep
transverse groove; propodeal carinae absent or with lateral longitudinal ones present as vestiges posterior-
ly; posterior area convex, rugose-reticulate grading to punctate. Fore wing with pterostigma very slender;
marginal cell quite short; Rs+2r slightly bowed proximally, barely widened near pterostigma; discosub-
marginal cell uniformly hirsute; \m-cu fairly evenly curved, without a ramellus. Hind wing with Rs very
weakly bowed. Fore tibial spur without a membranous flange behind macrotrichial comb; mid and hind
trochantelli simple; inner hind tibial spur cylindrical, without a fringe of long close hairs; hind tarsal claw
long, weakly curved. Gaster stout; tergite 2 in profile about as deep posteriorly as long, sometimes deeper,
with thyridia close to anterior margin; umbo absent; epipleuron up- turned. Ovipositor sheath slender.
Rhynchophion is a small genus restricted to the southern part of the U.S. A. and the Neotropical region.
Three species have been described, but these may be variants of a single species. They differ principally in
the colour of the wings and antennae.
PHYLOGENY OF THE OPHIONINAE 141
THYREODON Bm\\6
Thyreodon Brulle, 1846: 150. Type-species: Thyreodon cyaneus Brulle, by subsequent designation,
Hooker, 1912: 107.
Athyreodon Ashmead, 1900: 87. Type-species: Athyreodon thoracicus Ashmead (= Ophion atriventris
Cresson), by original designation.
Tipulophion Kriechbaumer, 19016: 75. Type-species: Tipulophion gigas Kriechbaumer (= Ophion
atriventris Cresson), by monotypy.
Macrophion Szepligeti, 1905: 32. Type-species: Macrophion ornatus Szepligeti (= Ophion atriventris
Cresson), by subsequent designation, Viereck, 1912: 640.
Oleter Shestakov, 1926: 259. Type-species: Oleter selenaction Shestakov (= Thyreodon laticinctus
Cresson), by original designation.
Mandibles not twisted, massive, weakly to moderately narrowed distally, usually fairly evenly bidentate,
or with lower tooth slightly the longer; outer mandibular surface flat except for a deep proximal concavity.
Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile from flat to slightly out-flared
ventrally, generally with margin sharp; clypeus in anterior aspect with a median apical tooth. Ocelli usually
small, the posterior ones widely separated from the eyes, or in a few species with ocelli very large, almost
contiguous with eyes; frontal carina present, often very strong; frons medially unusual in many species in
being quite strongly raised between the antennal bases; occipital carina present, usually complete,
sometimes not reaching hypostomal carina ventrally. Antennae short and stout, central flagellar segments
transverse. Pronotum mediodorsally specialized in having anterior and posterior margins raised to form
flanges or transverse crests; spiracular sclerite exposed; notauli often strong, generally extending the entire
length of the mesoscutum, frequently with crests across anterior ends; epicnemial carina complete, present
on mesopleuron; mesopleural furrow absent. Scutellum rather small, moderately convex, either without
lateral carinae, or, at most, with carinae present on anterior 0-4; posterior transverse carina of meso-
sternum present as lateral and central vestiges. Propodeum highly modified, with central part strongly
swollen so anterior part is in a very deep groove, as is the spiracle, and with the metapleuron appearing
unusually small ; propodeal carinae absent, but usually with postero-dorsal surface of propodeum bearing a
deep longitudinal concavity; posterior area otherwise from almost smooth to very strongly reticulate. Fore
wing with pterostigma slender; marginal cell from short to moderately long; Rs+2r usually only slightly
curved and weakly broadened before joining pterostigma; discosubmarginal cell without a glabrous area
anteriorly; Im-cu usually evenly curved, without a ramellus. Hind wing with Rs more or less straight. Fore
tibial spur without a membranous flange behind the macrotrichial comb; mid and hind trochantelli simple;
inner hind tibial spur cylindrical, without a pronounced marginal fringe of long close hairs; hind tarsal claws
long and weakly curved. Gaster from moderately stout to long and slender; tergite 2 in profile from slightly
to very elongate, with thyridia widely separated from anterior margin; umbo absent; epipleuron pendant,
but often creased anteriorly. Ovipositor sheath slender.
Thyreodon is a New World genus with about 25 described species. The majority are diurnally active, and
frequently seen feeding from flowers. These species are generally black or brightly coloured, and have
small ocelli whereas a few nocturnal species are predominantly brown and have large ocelli. It is perhaps an
interesting behavioural 'hangover' from their nocturnal ancestry, that predominantly diurnal species are
quite frequently nocturnally active as well and are taken at light.
The centre of diversity of the genus appears to be in northern South America. One species extends as far
north as Canada.
The ENICOSPILUS genus-group
This is the largest genus-group in the subfamily, containing both the greatest number of genera
and species; it also contains some of the most morphologically specialized of all ophionines. The
14 genera in this group can be divided into five subgroups (the Orientospilus, Ophiogastrella,
Stauropoctonus, Leptophion and Enicospilus subgroups) which correspond with the main
evolutionary lineages apparent in the group. The holophyly of this group is supported by a single
apomorphy, the loss of the vestigial first laterotergites, which is not a particularly convincing
reason for uniting the component genera. However, all share a number of other features,
including the complete atrophy of the flange on the fore tibial spur, though this feature is shared
with the Thyreodon genus-group. It is possible that the Enicospilus group may be paraphyletic
with respect to the Thyreodon group but this relationship could not be resolved further with the
142 I. D. GAULD
characters and material at hand. The majority of genera in this group (that is the Stauropocto-
nus, Leptophion and Enicospilus subgroups) clearly constitute a holophyletic group (see Fig.
19); the Ophiogastrella subgroup would seem to be the sister-lineage to these. The position of
the Orientospilus subgroup is less clear, but it seems reasonably placed as the most primitive
extant lineage of the group.
This genus-group is primarily a Pan-tropical complex of genera. Only isolated species of
Simophion, Stauropoctonus, Dicamptus and Enicospilus occur in the Holarctic region. The
hosts of the majority of species appear to be tree- and shrub-feeding lepidopterous larvae,
though only a small percentage of all the species has been reared.
The ORIENTOSPILUS subgroup
This group contains three genera, Orientospilus, Simophion and Prethophion. The systematic position of
the latter genus is far from clear as it is united with Orientospilus and Simophion solely on the basis of a
weak character - an incomplete posterior transverse carina of the mesosternum - and its present position is
really only the most parsimonious possible. Prethophion does share certain features with Thyreodon - it
has a slightly twisted mandibular axis, though not as pronounced as that of species in the Thyreodon
lineage, and a similarly stout gaster; like species of the Thyreodon lineage the anterior margin of the tegula
is not reflexed (possibly a derived feature). However, the plesiomorphic form of the anterior part of the
propodeum and tarsal claws seem to exclude Prethophion from the Thyreodon group as here defined.
Furthermore, the first laterotergite is lost and the mandibles are specialized in a quite different way, and
much more closely resemble the derived form found in Orientospilus.
Simophion and Orientospilus are more convincingly related; both have a similarly modified clypeus,
have a basally angulate Rs+2r in the fore wing, have lost the epicnemial carina laterally and have a short
marginal cell.
ORIENTOSPILUS Morley
Orientospilus Morley, 1912: 6. Type-species: Orientospilus individuus Morley, by subsequent designation,
Morley, 1913: 378.
Mandibles not twisted, long, strongly and evenly tapered, with upper tooth much longer than the lower;
outer mandibular surface flat, sparsely hirsute. Maxillary palp 5-segmented, labial palp 4-segmented;
clypeus in profile flat with margin out-flared; in anterior aspect concave. Ocelli moderately large, the
posterior ones separated from the eyes by 0-3-1-3 times their minimum diameter; frontal carina absent;
occipital carina complete, though ventrally not reaching hypostomal carina. Antennae moderately stout,
the distal segments quadrate, the flagellum barely longer than the fore wing. Pronotum mediodorsally with
anterior and posterior margins reflexed, raised as keels; spiracular sclerite exposed; notauli absent;
epicnemial carina present ventrally, not reaching above level of lower corner of pronotum; mesopleural
furrow vestigial or absent. Scutellum quadrate, weakly to moderately convex, with lateral longitudinal
carinae complete; posterior transverse carina of mesosternum present only laterally as vestiges. Prop-
odeum with anterior area short, deeply impressed as groove; anterior transverse carina from complete to
obsolescent, other carinae absent; posterior area coarsely reticulate, usually medially concave. Fore wing
with pterostigma of moderate breadth, evenly narrowed distally; marginal cell short; Rs+2r curved and
broadened before joining pterostigma; discosubmarginal cell with a glabrous area anteriorly; Im-cu
evenly arcuate without a ramellus. Hind wing with Rs from virtually straight to curved abruptly proximally.
Fore tibial spur without a membranous flange behind macro trichial comb; mid and hind trochantelli
unspecialized; hind inner tibial spur more or less flattened, with margin of long close hairs; hind tarsal claw
unspecialized. Gaster slender; tergite 2 in profile elongate, with thyridia obsolescent, or if discernible,
small and quite close to anterior margin; umbo present; epipleuron usually up-turned, rarely, in some
individuals, the posterior part is pendant. Ovipositor sheath slender.
This small genus contains three described species, capitatus Gauld & Mitchell from southern and west
Africa, melasma Townes from Madagascar and individuus Morley from eastern peninsular India.
Structurally these are very similar insects. The Madagascan species is the least specialized; both the Indian
and African species have lost the anterior transverse carina of the propodeum and have a shorter, stouter
and apically more setaceous flagellum and have a broader malar space suggesting they are sister-species
(Fig. 32). The Madagascan species may well be the ancestor of the other two.
PHYLOGENY OF THE OPHIONINAE
143
Fig. 32 Putative phylogenetic arrangement of species of Orientospilus . This cladogram is supported by
the following apomorphic features: 1 , malar space very wide; 2, ocelli widely separated from eyes; 3, Rs
in hind wing curved proximally; 4, flagellum with proximal segments almost quadrate; 5, anterior
transverse carina of propodeum absent; 6, malar space moderately wide, at least 0-4 times basal
mandibular width.
PRETHOPHION Townes
Prethophion Townes, 1971: 74. Type-species: Prethophion latus Townes, by original designation.
Mandibles very slightly twisted, very strongly narrowed, subequally bidentate; outer mandibular surface
slightly convex, with a proximal concavity. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus
in profile almost flat, margin centrally blunt, clypeus in anterior aspect virtually almost truncate. Ocelli
large, the posterior ones more or less touching eyes; frontal carina absent; occipital carina entirely absent.
Antennae moderately slender, barely longer than fore wing but with central segments clearly elongate.
Pronotum unspecialized; spiracular sclerite exposed; notauli vestigial; epicnemial carina present on
mesosternum, present but very weak on mesopleuron; mesopleural furrow very distinct, oblique,
extending from episternal scrobe to subalar prominence, the speculum posterodorsal to this carina strongly
inflated. Scutellum convex, laterally carinate only on extreme anterior end; posterior transverse carina of
mesosternum absent except for lateral vestiges. Propodeum short and abruptly declivous posteriorly;
anterior area unspecialized; anterior transverse carina complete; blunt tubercles (which are possibly
vestiges of posterior transverse carina) discernible; posterior area concave, very finely alutaceous. Fore
wing with pterostigma moderately slender; marginal cell slender; Rs+2r proximally not broadened, almost
straight; discosubmarginal cell without a glabrous area anteriorly; Im-cu fairly evenly curved, without a
ramellus. Hind wing with Rs strongly curved. Fore tibial spur without a membranous flange behind
macrotrichial comb; mid and hind trochantelli not specialized; inner hind tibial spur flattened, with a
margin of long close hairs; hind tarsal claw unspecialized. Gaster quite stout, centrally almost cylindrical;
144 I. D. GAULD
tergite 2 in profile elongate, with thyridia close to anterior margin; umbo low but distinct; epipleuron
up-turned. Ovipositor sheath stout.
This enigmatic genus contains a single species which occurs at low altitudes in Peru and Bolivia. Nothing
is known of its biology.
SIMOPHION Cushman
Simophion Cushman, 1947: 446. Type-species: Simophion excarinatus Cushman, by original designation.
Mandibles fairly evenly tapered, not or barely twisted, subequally bidentate or with lower tooth slightly the
shorter; outer mandibular surface punctate, proximally concave. Maxillary palp 5-segmented, labial palp
4-segmented; clypeus in profile flat, apically sometimes slightly out-flared, in anterior aspect concave.
Ocelli moderately large, the posterior ones separated from eye by 0-2-0-7 times their maximum diameter;
frontal carina absent; occipital carina complete, though generally not reaching hypostomal carina
ventrally. Antennae moderately long and slender, about 1-4-1-6 times length of fore wing. Pronotum
unspecialized; spiracular sclerite exposed; notauli vestigial; epicnemial carina present ventrally, not
reaching onto mesopleuron laterally; mesopleural furrow weak, diagonal, extending from episternal
scrobe to near subalar prominence. Scutellum moderately convex, narrow, without lateral carinae;
posterior transverse carina of mesosternum present only laterally as vestiges. Propodeum with anterior
area short, striate slightly centrally; anterior transverse carina absent as are other propodeal carinae, the
propodeum being narrowed distally; marginal cell short; Rs+2r very abruptly angled before joining
pterostigma, its extreme end abruptly broadened; discosubmarginal cell very sparsely hirsute, with a
glabrous area anteriorly; Im-cu curved or sinuous, without a ramellus. Hind wing with Rs very strongly
bowed. Fore tibial spur without a membranous flange behind macrotrichial comb; mid and hind
trochantelli unspecialized; inner hind tibial spur flattened, with a margin of long hairs; hind tarsal claws
unspecialized. Gaster slender; tergite 2 in profile elongate, with thyridia obsolescent, or if present quite
large and relatively close to anterior margin of tergite; umbo distinct; epipleuron up-turned. Ovipositor
sheath slender.
This moderately small genus is restricted to the 'Mediterranean' type biomes of the northern hemi-
sphere. Two species occur in the Middle East and Central Asia (Horstmann, 1981), whilst five or six occur
in the south-west of the United States (Townes, 1971). Similar disjunct distributions occur in several
genera of lower Aculeata (M. C. Day, pers. comm.) and in the myrmicine ant genus Messor (Bolton,
1982). Amongst ophionines Eremotylus has a similar, though slightly more extensive distribution.
The OPHIOGASTRELLA subgroup
This group comprises a single Neotropical genus, Ophiogastrella, which was treated as a member of the
Ophion genus-group by Cushman (1947), but as a member of the Enicospilini by Townes (1971). Townes &
Townes (1973) considered it to be closely related to Laticoleus.
Structurally Ophiogastrella shows a remarkable combination of primitive and derived features. The
specialized fore tibial spur and first laterotergite suggest it belongs close to the base of the Enicospilus
lineage, but the well-developed umbo, exposed spiracular sclerite and unspecialized anterior part of the
propodeum suggest it is more primitive than many other genera. These features, together with the lack of
an impressed clypeal apex, externally flat mandible and slender ovipositor sheath suggest it is not at all
closely related to Laticoleus. The peculiar male claws are an autapomorphy of this taxon.
Ophiogastrella is confined to the southern part of Central America and northern South America. There
is no evidence to suggest the group has ever occurred elsewhere.
OPHIOGASTRELLA Brues
Ophiogastrella Brues, 1912: 201. Type-species: Ophiogastrella maculithorax Brues, by original desig-
nation.
Brachyscenia Enderlein, 1921: 36. Type-species: Brachyscenia nigriventris Enderlein, by original desig-
nation.
Mandibles not twisted, evenly but only moderately narrowed apically, subequally bidentate; outer
mandibular surface flat except for small proximal concavity, centrally with scattered hairs. Maxillary palp
5-segmented; labial palp 4-segmented; clypeus in profile flat or weakly convex, margin blunt or sharp,
never impressed; clypeus in anterior aspect weakly convex, straight or very slightly concave. Ocelli large,
the posterior ones generally very close to the eyes; frontal carina absent; occipital carina complete,
ventrally (as genal carina) usually sharply angled and complete to hypostomal carina. Antennae of
PHYLOGENY OF THE OPHIONINAE 145
moderate length, 1-2-1-5 times the length of the fore wing, rarely slightly longer. Pronotum unspecialized;
spiracular sclerite exposed; notauli absent; epicnemial carina complete; mesopleural furrow absent or
weakly impressed, diagonal, extending from episternal scrobe to near subalar prominence. Scutellum
weakly to moderately convex, with lateral longitudinal carina extending about 0-8 of its length; posterior
transverse carina of the mesosternum vestigial. Propodeum with anterior area short, unspecialized;
anterior transverse carina from complete to absent, other carinae at most present only as vestiges;
posterior part of propodeum generally rather smooth. Fore wing with pterostigma quite large and broad;
marginal cell long; Rs+2r curved or slightly angled proximally, from slender to moderately broadened;
discosubmarginal cell with a small glabrous area anteriorly; Im-cu very strongly and evenly curved,
without a ramellus. Hind wing with Rs from almost straight to strongly bowed. Fore tibial spur without a
membranous flange behind the macrotrichial comb; mid and hind trochantelli simple or with a blunt tooth
near distal margin; inner hind tibial spur flattened, with a fringe of close hairs; hind tarsal claws of female
unspecialized, of male flattened with pectinal comb present on both sides (Fig. 47). Gaster moderately
slender; tergite 2 in profile elongate, with thyridia remote from anterior margin; umbo quite well
developed; epipleuron generally upturned. Ovipositor sheath slender.
Ophiogastrella is a small genus containing about 10 species, of which only three are described (Townes &
Townes, 1966). They are restricted to the Neotropical region from Costa Rica to about 10S in Brazil, and
seem to be associated with areas which have a pronounced dry season.
The species are rather similar to each other, and best separated by differences in the length of 3r-w,
shape of Rs in the hind wing and sculpture of the propodeum. There are also more subtle differences in the
shape of the clypeus, and the head.
The STAUROPOCTONUS subgroup
This subgroup contains two genera, Stauropoctonus (including as a synonym Aulophiori) and Lepiscelus.
The exposed spiracular sclerites of species in this complex suggest it is one of the more primitive of
enicospiline lineages. It is also one of the taxonomically most distinctive as its members are characterized
by an incomplete occipital carina, a transverse mesopleural furrow, slender pterostigma, basally thickened
and bent Rs+2r and a very sinuous Im-cu. Most species have Im-cu and Cw la basally widely separated,
and all have the mid and hind trochantelli specialized.
The group contains few species. The most primitive Stauropoctonus species and the solitary Lepiscelus
occur in the Old World, suggesting an Old World origin for the group.
LEPISCELUS Townes
Lepiscelus Townes, 1971: 73. Type-species: Lepiscelus gracile Townes (= Eremotylus distorts Seyrig), by
original designation.
Mandibles twisted about 5, proximally strongly narrowed, distally parallel-sided with upper tooth about
twice as long as the lower; outer mandibular surface more or less flat. Maxillary palp 5-segmented, labial
palp 4-segmented; clypeus in profile weakly convex, margin narrowly acute, laterally blunted, short,
exposing labrum; clypeus in anterior aspect truncate or even slightly concave. Ocelli large, the hind ones
very close to the eyes; frontal carina absent; occipital carina entirely absent. Antennae very long and
slender, about 1-7 times length of the fore wing. Pronotum mediodorsally rather long, flat; spiracular
sclerite exposed; notauli very short but often strongly impressed in part; epicnemial carina weak, laterally
becoming obsolescent at level of lower corner of pronotum; mesopleural furrow very weakly impressed,
transverse. Scutellum almost rectangular, carinate laterally only at extreme anterior end; posterior
transverse carina of mesosternum complete. Propodeum with anterior area moderately long, dorsally
without carinae, with posterior area finely wrinkled to rugulose. Fore wing with pterostigma slender;
marginal cell long; Rs+2r abruptly curved and broadened before joining pterostigma; discosubmarginal
cell anteriorly glabrous; Im-cu very strongly sinuate. Hind wing with Rs strongly curved; hamuli arranged
in two groups, the proximal group comprising two which are longer and flatter than those in the distal group
of three. Fore tibial spur without a membranous flange behind the macrotrichial comb; mid and hind
trochantelli extended apically as a broad flange over the proximal end of the femur; inner hind tibial spur
flattened, with a margin of long, close hairs; hind tarsal claws quite long, those of male more closely
pectinate than those of the female. Gaster slender; tergite 2 in profile elongate, with thyridia remote from
anterior end; umbo vestigial; epipleuron up-turned. Ovipositor sheath slender; male with gonosquama
unusually long.
Lepiscelus contains a single species which is widespread, but apparently rather uncommon, throughout
the Afrotropical region between latitude 10N and 20S. The most distinctive feature of this genus is the
146 I. D. GAULD
flanged mid and hind trochantelli which are, amongst the Ichneumonidae, a unique feature of Lepiscelus
distans. Nothing is known of the biology of this insect, but the form of the mandibles and head shape are
reminiscent of some species of Enicospilus that either inhabit arid areas or are active in the dry season.
STAUROPOCTONUS Brauns
Stauropoctonus Brauns, 1889: 75. Type-species: Ophion bombycivorus Gravenhorst, by monotypy.
Stauropodoctonus Morley, 1913: 375. [Unjustified emendation.]
Nipponophion Uchida, 1928: 201. Type-species: Nipponophion variegatus Uchida (= Ophion bombyci-
vorus Gravenhorst), by monotypy.
Aulophion Cushman, 1947: 458. Type-species: Aulophion bicarinatus Cushman, by original designation.
Syn. n.
Mandibles twisted about 85, evenly, but quite strongly narrowed apically, subequally bidentate. Maxillary
palp 5-segmented, labial palp 4-segmented; clypeus in profile weakly convex, in anterior aspect with
margin slightly convex. Ocelli large, the posterior ones close to eye margins; frontal carina absent; occipital
carina usually absent, partially present in one species. Antennae very long and slender, sometimes up to 2
times length of fore wing. Pronotum unspecialized, spiracular sclerite exposed; notauli vestigial or absent;
epicnemial carina from present and complete to absent; mesopleural furrow strong, extending from
episternal scrobe to upper end of prepectal carina (or the corresponding position if this carina is absent), in
some species impressed as a deep groove. Scutellum moderately convex to convex, either without lateral
carinae or with carina incomplete; posterior transverse carina of mesosternum from complete to reduced to
lateral vestiges. Propodeum with anterior area long; anterior transverse carina usually complete, the
posterior one sometimes discernible; posterior area from weakly rugulose to reticulate. Fore wing with
pterostigma moderately slender; marginal cell long; Rs+2r abruptly curved and slightly thickened basally;
discosubmarginal cell with a glabrous area anteriorly, which may be expanded along Rs+2r; Im-cu
without a ramellus, either evenly bowed or sinuous. Hind wing with Rs from almost straight to weakly
bowed. Fore tibial spur without a membranous flange behind macrotrichial comb; mid and hind
trochantelli specialized in having the outer distal margin produced into a strongly decurved sharp spine;
inner hind tibial spur flattened, with a marginal fringe of long hairs; hind tarsal claws unspecialized. Gaster
slender; tergite 2 in profile very elongate, with the thyridia remote from anterior margin; umbo vestigial;
epipleuron pendant or up-turned. Ovipositor sheath slender.
In the analysis of the genera Stauropoctonus appeared to be paraphyletic with respect to Aulophion. The
latter genus has been separated from Stauropoctonus by possession of two apomorphic features (Cushman,
1947; Townes, 1971). No apomorphies have ever been suggested for Stauropoctonus, though the clade
Stauropoctonus + Aulophion is one of the most clearly defined in the subfamily. It is characterized by the
following combination of apomorphies (though none is actually unique to the clade) - at least partial loss of
occipital carina; presence of projections on mid and hind trochantelli; presence of impressed transverse
furrow on mesopleuron; presence of strongly twisted mandibles; having Rs+2r in the fore wing basally
angled. In addition to these features the head and mesoscutal profile of species of this clade are
characteristic, though these 'apomorphies' are so subtle I have not been able to code them. To further
resolve the relationship between Aulophion and Stauropoctonus more data were analysed.
The following taxa were included - Aulophion sp. 1 (? bicarinatus) (301), Aulophion sp. 2 (302),
Stauropoctonus bombycivorus (303), 5. torresi (304), 5. townesorum (305) and S. occipitalis (306). This is
all the species in the group except for one (or possibly two) Aulophion species which are very closely
related to taxon 302. To represent the range of interspecific variation in the group the following characters
from the primary set were utilized- 1-2, 19, 20-1, 24-1, 25-2, 27, 30, 36, 43-1, 43-2, 44-1, 45-2, 47-2, 57 and
59. (It is noteworthy that 27 and 36 are composites of 27-1 and 27-2 and 36-1 and 36-2 respectively; these
characters showed identical state distribution over the data and therefore were treated as single characters
to avoid excess weighting.) In addition to these fifteen, three further characters were used.
80 Median carina of propodeum. This apomorphic feature (1) is found only in a few isolated Neotropical
ophionines. The plesiomorphic condition, no median carina, is found in virtually all ophionines and
members of the various out-groups.
81 Metapleural sculpture. In most out-group taxa, most ophionines and several species of this group the
metapleuron is smooth and finely punctate. Some Stauropoctonus species are specialized in having
this region coarsely rugose (1).
82 Position of cu-a with respect to Rs&M in fore wing. The plesiomorphic condition for ophionines
apparently is for cu-a to be proximal to the base of Rs&M. A few Stauropoctonus are specialized in
having these veins opposite (1).
PHYLOGENY OF THE OPHIONINAE
147
Table 22 Primary data matrix for genus Stauropoctonus. The taxa corresponding to the numbers are
given in the text (p. 146).
1.2 20.1 25.2 30 43.1 44.1 47.2 59 81
19 24.1 27 36 43.2 45.2 57 80 82
301
302
303
304
305
306
.47.2
36r
45.2
47.2
57
I-
1
80
81, 82
119, 27.1, 30
143.1, 44.1
k45.2
1.2, 43.2
25.2
Fig. 33 Favoured cladogram showing putative phylogenetic arrangement of species of Stauropoctonus.
This is based on the largest compatible set often informative characters and requires 25 transformation
steps.
148
I. D. GAULD
The primary data matrix for taxa 301-6 is shown in Table 22. Parsimony and compatibility analyses,
using techniques outlined above, yielded two 'best' alternative hypotheses of phylogenetic relationship
(Figs 33, 34). The cladograms have a large number of features in common. Firstly, both suggest
Stauropoctonus is paraphyletic with respect to Aulophion, and this in turn suggests that Aulophion can be
treated as a synonym of Stauropoctonus , as has been formally proposed above. Secondly, both suggest that
i
19, 27.1, 30
i 43.1, 44.1
45.2
|47.2r, 57
1.2, 43.2
25.2 , 47.2
Fig. 34 Alternative cladogram for species of Stauropoctonus. This is the most parsimonious arrangement
of taxa discovered and requires 24 transformation steps. (Symbols as for Figs 30-31.)
PHYLOGENY OF THE OPHIONINAE 149
Aulophion is both a holophyletic clade and the most derived lineage in the group. Thirdly, both place 5.
ocdpitalis as the most primitive taxon in the group. The two cladograms differ in their arrangement of 5.
bombycivorus, S. torresi and 5. townesorum and these different arrangements depend on the alternative
compatible sets of characters 36 and 45-2 versus 59, 81 and 82. Character 36 has been demonstrated to be
both unreliable and of dubious polarity, whilst 45-2 can be regarded as a loss apomorphy (loss of hairs on
wing membrane). Character 59 (colour of interocellar area) is also far from a convincing apomorphy, but
81 and 82 strongly suggest that torresi and townesorum are sister-species. S. bombycivorus is rather difficult
to place, but could be either the ancestor of, or a close relative of the ancestor of both the 'Aulophion'
species-group and torresi/townesorum.
The more primitive species of this genus (ocdpitalis and bombycivorus) are restricted to the Old World,
the former to the mountains of Madagascar and the latter to the southern part of the Palaearctic, except in
Asia where it extends north into Kamchatka and the Kurile Islands (Townes, Momoi & Townes, 1965).
Possibly the ancestor of these taxa was widely distributed in the Old World at one time, and the
Madagascan survivor is a relict of this. There are clear indications in the Madagascan fauna of an ancient
Palaearctic affinity amongst the Hymenoptera. The symphytan taxon Cephidae is represented on the
island and in the Holarctic only (Benson, 1935), and a number of ichneumonid genera have rather similar
distributions, e.g. Neliopisthus, Euceros (Townes, 1969: Barren, 1978). As the widespread Palaearctic
species bombycivorus may be almost directly ancestral to the two other species-groups, it may at one time
have also occurred in the Nearctic region. The torresi/townesorum group is Malesian/Melanesian, with the
former species constituting part of the characteristically intrusive, Indo-Papuan faunal element in northern
Australia (Gauld, 19840). The distribution of torresi, in transcending both Wallace's and Weber's lines,
suggests its current range is the result of a dispersive rather than a vicariance event. The Neotropical
species form a rather distinctive clade. The most northerly extent of their distribution seems to be Costa
Rica where one species occurs in lower montane (1000-1500 m) forest, whilst none has been found south of
25S. Perhaps the ancestor of this group reached South America from the north during the Miocene when
an archipelagic connection existed between the two continents (Rich & Rich, 1983). This evolutionary
'scenario' is based on the supposition that a Stauropoctonus occurred in the United States. At present there
is no evidence for this; the scenario presented is merely the most parsimonious interpretation of data.
Other scenarios would necessitate advocating much more widespread extinction.
The LEPTOPHION subgroup
This complex of genera is characterized by the form of the mandibles, which have a more or less discernible
impressed groove extending diagonally, the convex clypeus with an impressed acute margin, the very long
slender antennae and, for most species, a somewhat broadened ovipositor sheath. The venation of all
species is rather similar in having Rs+2r basally broadened, and often angulate before joining the
pterostigma; Im-cu is either sinuate or evenly curved.
This subgroup contains three genera, Leptophion, Laticoleus and Pamophion which are difficult to
resolve as the holophyly of Leptophion cannot convincingly be demonstrated. Pamophion, a monobasic
taxon, is holophyletic and is apparently the sister-lineage of Leptophion + Laticoleus. The latter is a
holophyletic group, but it may have arisen from within the former. The only apomorphy that can be
postulated for Leptophion is the presence of a specialized penultimate hamulus. (The apomorphies shown
in Fig. 19 are subject to much exception and parallelism in some species of Laticoleus, and reference to the
overall classification suggests that presence of a complete posterior transverse carina of the mesosternum is
plesiomorphic for this complex - though some taxa have partially lost this carina.) The assumption that the
specialized hamulus is a group apomorphy necessitates speculating that reversal has occurred in several
taxa, but there is some evidence that this has happened. For example, L. tetus has a slight indication of
some enlargement of the hamulus and its sister-species, L. yampus, has it clearly specialized, though not as
large as that of L. iochus, the sister-species of yampus + tetus. L. ankylosus and L. eithos are highly
specialized species (without a modified hamulus) which are closely related to L. vernalis which has a long
penultimate hamulus, and Leptophion species 1 (BMNH) which has it only slightly modified. These four
species apparently belong to the maculipennis lineage, the remainder of which all have a highly specialized
hamulus. A tentative phylogeny for this subgroup is proposed in Fig. 35. Laticoleus is retained as a separate
genus, but its validity needs to be reassessed as more material becomes available for study and the
phylogeny of the species of Leptophion is better understood.
Judging from the present distribution of this group (Fig. 36) it is most parsimonious to postulate an Old
World origin for this complex of genera. The more primitive taxa are restricted to Melanesia or Australia
which suggests an Australo-Melanesian origin for this group. However, if Laticoleus is truly primitive with
respect to Leptophion, then an alternative hypothesis is that this group was once widely spread in the Old
LEPTOPHION
MACULIPENNIS GROUP
A
Fig. 35 Cladogram showing putative phylogeny of groups of species in the Leptophion subgroup. Black
squares represent apomorphic features, white plesiomorphic; diagonally divided squares indicate a
mixture of the two states in the component species of a particular lineage. The characters that support
this arrangement are: 1, loss of specialized hamulus; 2, Rs+2r emitted from centre of pterostigma; 3,
Im-cu very sinuous; 4, specialized hind tarsal claws; 5, loss of posterior transverse carina of mesoster-
num; 6, presence of a 'hair brush' on mandible; 7, genal carina evanescent before joining hypostomal
carina; 8, lengthened penultimate distal hamulus; 9, exceptional broad ovipositor sheath that in profile is
almost quadrate; 10, long malar space; 11, ovipositor sheath at least moderately broad, in profile more
than 0-4 times as deep as long; 12, marginal cell of hind wing at least partially glabrous; 13, anterior area
of propodeum elongate; 14, epipleuron 2 pendant; 15, hind trochantellus marginally produced into blunt
tooth; 16, posterior transverse carina of mesosternum present.
PHYLOGENY OF THE OPHIONINAE
151
Fig. 36 The geographical distribution of taxa of the Leptophion subgroup.
World tropics and has suffered considerable extinction, leaving only a rather specialized primitive lineage
in the Afrotropical region and a few relicts east of Wallace's line. Gauld (1984a) suggested that the anici
and iochus groups of Leptophion, in Australia, have had a northern origin, and that once their ancestors
had adapted sufficiently to cross the rain forest/savannah interface (Taylor, 1972) then they had ample
opportunity to give rise to radiations in species-poor Australia (Gauld, 1984a). The most specialized group
of Leptophion, the maculipennis group, is best represented in the more easterly parts of Indonesia and
Melanesia. It is probable that this species-group arose in this area and a few species have spread west to the
Asian mainland.
LATICOLEUS Townes
Laticoleus Townes, in Townes & Townes, 1973: 358. Type-species: Coiloneura unicolor Szepligeti, by
original designation.
Mandibles not twisted, weakly narrowed apically, generally subequally bidentate; outer mandibular
surface with subbasal swelling weak or well developed, with a diagonal groove extending from the upper
proximal corner to between bases of teeth. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus
in profile convex, margin impressed, in anterior aspect with margin weakly convex. Ocelli generally
moderately large, the posterior ones separated from the eyes by 0-2-1-0 times their minimum diameter;
frontal carina absent; occipital carina complete, the lower part (genal carina) reaching to hypostomal
carina. Antennae very long and slender, at least 1-5 times length of fore wing. Pronotum usually
unspecialized, or in one Madagascan species with anterior margin mediodorsally expanded and curved
back; spiracular sclerite partially concealed; notauli vestigial or absent; epicnemial carina complete, weak
or even absent on mesopleuron; mesopleural furrow undeveloped. Scutellum weakly convex, with lateral
longitudinal carinae present, more or less complete; posterior transverse carina of the mesosternum
incomplete, present as lateral or rarely central vestiges. Propodeum with anterior area moderately long,
striate or smooth; anterior transverse carina present or absent, other carinae usually absent; posterior area
from rugose to virtually smooth. Fore wing with pterostigma moderately stout; marginal cell long; Rs+2r
152 I. D. GAULD
conspicuously thickened near pterostigma, often abruptly curved in proximal part; discosubmarginal cell
with a glabrous area in anterior corner, this area generally quite large, and in one species, bearing a weak
sclerite; Im-cu without a ramellus, usually fairly evenly bowed, in a few species slightly sinuous. Hind wing
with Rs from almost straight to strongly curved, with marginal cell proximally, at least narrowly glabrous;
distal hamuli unspecialized, or with proximal three enlarged. Fore tibial spur without a membranous flange
behind macrotrichial comb; mid and hind trochantelli unspecialized or with a weak tooth projecting
distally; inner hind tibial spur flattened with a margin of close long hairs; hind tarsal claws usually
unspecialized, rarely with very coarse pectinae. Gaster slender; tergite 2 in profile very elongate, with
thyridia well removed from anterior margin; umbo absent; epipleuron up-turned. Ovipositor sheath
exceptionally broad.
Laticoleus is a moderately small Afrotropical genus containing 11 described species (Gauld & Mitchell,
1978) . The majority occur in Madagascar and east Africa. Nothing is known of the biology of species of this
genus.
LEPTOPHION Cameron
Leptophion Cameron, 1901: 227. Type-species: Leptophion longiventris Cameron, by monotypy.
Spilophion Cameron, 1905: 124. Type-species: Spilophion maculipennis Cameron, by monotypy.
Coiloneura Szepligeti, 1905: 35. Type-species: Coiloneura melanostigma Szepligeti (= Leptophion lon-
giventris Cameron), by subsequent designation, Viereck, 1914: 35.
Mandibles not twisted, weakly narrowed apically, generally subequally bidentate; outer mandibular
surface often with a basal swelling, and with a diagonal groove extending from near upper corner to
between bases of teeth, sometimes with this groove bearing a brush of long hairs, other times with the
groove very weak. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile convex, margin
impressed acute, in anterior aspect usually with margin weakly convex, rarely slightly concave. Ocelli
generally large, the posterior ones close to the margins of the eyes, or rarely in some Australian examples,
with ocelli smaller; frontal carina absent; occipital carina either complete or obsolescent at extreme ventral
end so the carina fails to join the hypostomal carina. Antennae in most species very long and slender, at
least 1-5 times the length of the fore wing, in some Australian species shorter, barely longer than the fore
wing. Pronotum unspecialized; spiracular sclerite partially concealed, in a few Malesian taxa almost
completely covered; notauli vestigial or absent; epicnemial carina complete; mesopleural furrow usually
undeveloped. Scutellum moderately convex, from without lateral carinae to with these carinae complete;
posterior transverse carina of mesosternum complete, interrupted before mid coxae, or rarely absent.
Propodeum with anterior area long and usually striate; anterior transverse carina usually complete, rarely
absent, the other carinae usually not discernible; posterior area from strongly rugose to smooth and
shining. Fore wing with pterostigma moderately broad; marginal cell long; Rs+2r conspicuously thickened
near pterostigma, usually abruptly curved or angled near base; discosubmarginal cell usually with a
glabrous area in anterior corner, this area usually large and extending 0-2 of the way down Rs+2r, rarely
with a weak alar sclerite, sometimes secondarily with micro trichia obscuring part of glabrous area; Im-cu
usually without a ramellus, either evenly bowed or, more usually, sinuous. Hind wing with Rs from almost
straight to very strongly bowed, with marginal cell adjacent to this vein at least narrowly glabrous; distal
hamuli unspecialized or with penultimate hamulus long and coiled. Fore tibial spur without a membranous
flange behind macrotrichial comb; mid and hind trochantelli unspecialized; inner hind tibial spur flattened,
with a fringe of long close hairs; hind tarsal claws unspecialized or with distal pectinal tooth projecting
beyond the apex of the claw. Gaster usually slender, tergite 2 in profile from moderately to very elongate,
with thyridia remote from anterior margin; umbo absent; epipleuron up-turned. Ovipositor sheath
moderate, broad.
Leptophion is a moderately large Indo-Pacific genus containing 30 described species (Gauld, 1977;
Gauld & Mitchell, 1981). The majority are restricted to lower montane forests on the islands to the east of
the Sunda Shelf. Many are endemic to a particular island and recent collecting has yielded an undescribed
species from Sulawesi (BMNH). Most species are only known from relatively few individuals.
The genus is divisible into four species-groups, the largest of which, the maculipennis species-group, may
be subdivided into three species-complexes.
Key to species-groups and complexes of Leptophion
1 Anterior transverse carina of mesosternum incomplete; propodeum often with anterior trans-
verse carina absent anici species-group
PHYLOGENY OF THE OPHIONINAE 153
- Anterior transverse carina of mesosternum complete; propodeum almost always with anterior
transverse carina complete 2
2 Hind tarsal claw simple ; genal carina usually joining hypostomal carina 3
- Hind tarsal claw with distal pectinal tooth projecting apically; genal carina usually not reaching
hypostomal carina (maculipennis species-group) 4
3 Mandible with a dense brush of long stout hair on outer surface iochus species-group
- Mandible with scattered hairs on outer surface longicornis grade-group
4 Fore wing with Rs+2r emitted from near centre of pterostigma; anterior corner of discosubmar-
ginal cell partially hirsute eithos species-complex
- Fore wing with Rs+2r emitted from proximal end of pterostigma; anterior corner of discosub-
marginal cell glabrous 5
5 Penultimate distal hamulus unspecialized; fore wing with Im-cu from evenly curved to weakly
sinuous radiatus species-complex
- Penultimate distal hamulus long and coiled; fore wing with Im-cu moderately to strongly
sinuous maculipennis species-complex
longicornis grade-group
This paraphyletic assemblage contains two species, L. longicornis (Szepligeti) and L. bakeri (Cheesman).
The group can only be characterized by symplesiomorphies. It is restricted to New Guinea and the New
Hebrides.
anici species-group
This holophyletic group contains three species, L. anici Gauld, L. antennatus (Morley) and L. unical-
caratus Gauld. The group is characterized by the loss of the posterior transverse carina of the mesoster-
num; species tend to have less propodeal sculpture than most Leptophion species, and usually the
transverse carina is lost. The anici species-group is restricted to Australia and New Caledonia where
species occur in drier habitats than mos't Leptophion.
iochus species-group
This holophyletic group contains three Australian species, L. iochus Gauld, L. yampus Gauld and the
aberrant L. tetus Gauld. The group is characterized by possession of a brush of hair on the outer surface of
the mandible; they are generally more robust species with more densely pubescent wings than most
Leptophion species.
maculipennis species-group
This holophyletic group is characterized by the presence of a modified hind tarsal claw; the majority of
species have infumate marks in the proximal corner of the marginal cells and in most the genal carina does
not join the hypostomal carina. This is the largest species-group and may be subdivided into three
apparently holophyletic species-complexes.
eithos species-complex. This group contains four species, L. eithos Gauld & Mitchell, L. ankylosus
Gauld & Mitchell, L. vernalis Gauld & Mitchell and Leptophion species 1 (BMNH). It is characterized by
the highly modified fore wing venation in which Rs+2r is emitted near the centre of the pterostigma. This
complex is restricted to Western New Guinea, Sulawesi and the intervening islands.
radiatus species-complex. This group contains six species, L. radiatus (Uchida), L. pterospilus Gauld &
Mitchell, L, vechti Gauld & Mitchell, L. lavellai Gauld & Mitchell, L. cheesmanae Gauld & Mitchell and
L. illustrious Gauld & Mitchell. It is characterized by the possession of unspecialized hamuli; most species
have a rather evenly curved Im-cu and very few have Rs in the hind wing appreciably curved. This group is
widely distributed from the Solomon Islands to the Continental Asian mainland.
maculipennis species-complex. This group contains 13 species, L. maculipennis (Cameron), L. pubes-
cens Gauld & Mitchell, L. gobius Gauld & Mitchell, L. kus Gauld & Mitchell, L. juxtus Gauld & Mitchell,
L. magus Gauld & Mitchell, L. townesi Gauld & Mitchell, L. alleni Gauld & Mitchell, L. nodus Gauld &
Mitchell, L. samari Gauld & Mitchell, L. samuelsoni Gauld & Mitchell, L. quorus Gauld & Mitchell and
L. longiventris Cameron. These are the characteristic Leptophion species of Cushman (1947) and Townes
(1971) in that they possess both a specialized hind tarsal claw and modified penultimate hamulus. Most
species have Im-cu very strongly sinuous and Rs in the hind wing strongly bowed. This is the most
widespread of all groups with species throughout the Indo-Pacific region. One species has colonized
tropical Australia (Gauld, 1984a).
154 I. D. GAULD
PAMOPHION Gauld
Pamophion Gauld, 1977: 28. Type-species: Pamophion sorus Gauld, by original designation.
Mandibles not twisted, weakly narrowed apically, subequally bidentate; outer mandibular surface with
subbasal swelling, and with a diagonal groove. Maxillary palp 5-segmented, labial palp 4-segmented;
clypeus in profile convex, margin impressed, acute, in anterior aspect weakly convex. Ocelli large, the
posterior ones close to the eye margins; frontal carina absent; occipital carina complete except at extreme
ventral end. Antennae long and slender. Pronotum unspecialized; spiracular sclerite partially concealed;
notauli vestigial; epicnemial carina quite strong, present on mesopleuron; mesopleural furrow unde-
veloped. Scutellum weakly convex, carinate laterally most of its length; posterior transverse carina of
mesosternum incomplete, usually discernible as a discontinuous ridge. Propodeum with anterior area
rather short, striate; anterior transcarina complete, other carinae absent, posterior area rugose-reticulate.
Fore wing with pterostigma moderately broad; marginal cell long; Rs+2r conspicuously thickened near
pterostigma, proximally curved; discosubmarginal cell broadly glabrous anteriorly; Im-cu without a
ramellus, moderately sinuous. Hind wing with Rs straight; marginal cell evenly hirsute; hamuli unspecial-
ized. Fore tibial spur without a membranous flange behind the macro-trichial comb; mid and hind
trochantelli with a distinct blunt distal tooth; inner hind tibial spur flattened, with a fringe of long, close
hairs; hind tarsal claws unspecialized. Gaster slender; tergite 2 in profile very elongate, with thyridia
moderately close to anterior margin; umbo absent; epipleuron pendant, narrow. Ovipositor sheath
slender, unspecialized.
A single species, P. sorus, is known to occur in Queensland. It is apparently the most primitive species in
this genus-group. Nothing is known of its biology.
The ENICOSPILUS subgroup
This group contains five genera, Dicamptus, Enicospilus, Pycnophion, Banchogastra and Abanchogastra.
The last three are endemic Hawaiian taxa and probably constitute a monophyletic clade which almost
certainly arose from within Enicospilus. They are so phenetically divergent, however, that it would be
quite impractical to include them within Enicospilus, an otherwise structurally uniform genus of over 700
species.
The relationship of Enicospilus to Dicamptus is not clear; the latter may be paraphyletic with respect to
the former though subtle differences in venation, alar sclerite form, sculpture and body shape suggest that
Dicamptus is actually holophyletic.
The Enicospilus group is characterized by a number of apomorphic features including having the
spiracular sclerite concealed, having an elongate anterior propodeal area, having a rather sinuous Rs+2r
and generally having alar sclerites. Individual species may be exceptional in one or two of these features.
Dicamptus, clearly the least specialized genus in this complex, is restricted to the Old World, and
Enicospilus is apparently most diverse in the Old World tropics, suggesting a palaeotropical origin for the
group. Preliminary study of Neotropical species of Enicospilus suggests that very few large species-groups
occur in South America; several of these also occur in the Nearctic region. The Australian Dicamptus and
Enicospilus have apparently been derived from immigration from South East Asia (Gauld, 1984o).
The Hawaiian genera seem to be a holophyletic group. This is attested by the total lack of alar sclerites,
possession of a straight, rather slender Rs+2r and loss of the last 0-3 or so of the lateral scutellar carinae.
There are a number of other similarities in sculpture and exact position of wing veins that further supports
this clade.
The inter-relationship of these three genera may be misrepresented in the cladogram (Fig. 37).
Considering only the characters in the primary data matrix five (16-2, 27-3, 37, 48-1 and 49) support
Pycnophion + Abanchogastra whilst three (13, 42-2, 43-1) support the group Pycnophion + Banchogastra.
None supports the third combination. Of the first five characters 16-2 has almost certainly undergone
reversal in Banchogastra as the character is apomorphic in the greater part of the presumed ancestral
lineage of the species. The plesiomorphic state is otherwise generally only found in primitive ophionines
and (again as a reversal) in a few diurnal specialized species (Banchogastra has not been taken at light and
thus may well be diurnally active). Character 27-3, a centrally interrupted posterior mesosternal carina, is
only found in the primary data set in Pycnophion and Abanchogastra, but this carina is weak centrally in
Banchogastra and may even be absent narrowly in some specimens. A medioventrally evanescent
mesosternal carina can therefore be considered an apomorphy of the Hawaiian genera. Character 49,
position of Cu\ in the fore wing, is plesiomorphic only for Banchogastra. The derived condition must
therefore be considered to be an apomorphy of the subfamily and the anomalous inclination of this vein in
Banchogastra is presumably a further specialization, perhaps resulting from the very close proximity of the
PHYLOGENY OF THE OPHIONINAE
155
Fig. 37 Putative phylogenetic arrangement of Hawaiian genera in relation to Enicospilus moea Chees-
man. This cladogram is supported by the following apomorphic features: 1, petiolar spiracles anterior to
margin of sternite 1; 2, tergite 2 depressed; 3, epipleuron 2 pendant; 4, further reduction in ocellar size;
5, possession of an elongate ovipositor; 6, loss of hair in discosubmarginal cell centrally; 7, development
of a stout gaster; 8, possession of large thyridia close to anterior margin of tergite; 9, metapleuron
inflated; 10, epicnemial carina medioventrally incomplete; 11, possession of a short, rounded prop-
odeum; 12, presence of incipient frontal carinae; 13, possession of inflated hind trochanters; 14, loss of
occipital carina mediodorsally; 15, upper tooth of mandible compressed; 16, increase in torsion of
mandible; 17, reduction in size of ocelli; 18, loss of fenestra; 19, development of fine, granulate thoracic
sculpture; 20, loss of alar sclerites; 21, reduction in length of second abscissa of Cui in fore wing; 22,
medially evanescent posterior transverse carina of mesosternum ; 23 , straight and slender Rs +2r; 24, loss
of posterior part of lateral carina of scutellum.
bases of \m-cu and Cw la . In some individuals this abscissa of Cui is occluded. The remaining two
apomorphies (37 and 48-1) are venational features which are usually rather variable. The three apomor-
phies uniting Pycnophion with Banchogastra are slightly more convincing, particularly 42-2, the ptero-
stigma shape. Furthermore, Pycnophion and Banchogastra resemble each other in many other specialized
features. Both have smaller ocelli than normal, have a medioventrally interrupted epicnemial carina, have
156 I. D. GAULD
a strongly inflated metapleuron and a rather short anterior propodeal area and possess rather inflated
trochanters. The thyridia of tergite 2 are large and close to the anterior margin (? a reversal) and the gaster
is stouter than most other ophionines. The first segment of the gaster is very much broader and shorter than
the corresponding segment in other enicospilines. In view of these marked similarities it would seem that
Pycnophion + Banchogastra constitute a distinct clade.
A Marquesan species, Enicospilus moea Cheesman, shares a number of apomorphies with the Hawaiian
genera, including possession of a straight, rather slender Rs+2r, having a medioventrally obsolescent
posterior mesosternal carina and having only the anterior 0-7 of the scutellum carinate. Like the Hawaiian
genera the abscissa of Cui between Im-cu and Cw la is very short and cu-a is well proximal to the base of
Rs&M. This species does, however, possess fenestra but this, rather than any of the extant Hawaiian
species, may be closest to the base of the Pycnophion/ Banchogastra/ Abanchogastra evolutionary line (Fig.
37).
ABANCHOGASTRA Perkins
Abanchogastra Perkins, 1902: 141. Type-species: Abanchogastra debilis Perkins, by monotypy.
Mandibles twisted about 45, strongly and evenly narrowed with teeth subequal, slightly depressed; outer
mandibular surface almost flat. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile
weakly convex, margin blunt; clypeus in anterior aspect weakly convex. Ocelli large, the posterior ones
separated from eye by 0-1-0-2 times their own diameter; frontal carina absent; occipital carina present,
mediodorsally interrupted, ventrally weak but joining hypostomal carina. Antennae incomplete. Pro-
notum unspecialized; spiracular sclerite concealed by pronotal flange; notauli absent; epicnemial carina
weak but present on mesopleuron; mesopleural furrow absent. Scutellum weakly convex, carinate laterally
about 0-7 of its length; posterior transverse carina of mesosternum centrally obsolescent. Propodeum with
anterior area long, striate, carinae absent, posterior area coriaceous. Fore wing with pterostigma of
moderate width; marginal cell long; Rs+2r straight and slender; discosubmarginal cell uniformly hirsute;
Im-cu evenly curved, without a ramellus. Hind wing with Rs weakly curved. Fore tibial spur without a
membranous flange behind macrotrichial comb; mid and hind trochantelli unspecialized; inner hind tibial
spur flattened, hind tarsal^claws unspecialized. Gaster slender; tergite 2 very elongate, thyridia elliptical,
remote from anterior margin, umbo absent; epipleuron up-turned. Ovipositor sheath slender.
This genus contains a single Hawaiian species which differs strikingly from any Enicospilus in the fore
wing venation. Nothing is known of its biology.
BANCHOGASTRA Ashmead
Banchogastra Ashmead, 1900: 87. Type-species: Banchogastra nigra Ashmead, by original designation.
Mandibles twisted about 25, strongly and evenly narrowed apically with upper tooth a little longer than the
lower; outer mandibular surface with strong proximal concavity, remainder of surface virtually flat but with
diagonal tract of dense hair. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile
weakly convex, margin sharp, not impressed at all; clypeus in anterior aspect with margin very weakly
convex. Ocelli small, the posterior ones separated from eyes by more than their own minimum diameter;
frontal carina very weak but discernible; occipital carina complete, ventrally not quite reaching hypo-
stomal carina. Antennae moderately slender, about 1-2 times length of fore wing. Pronotum unspecialized;
spiracular sclerite completely occluded by pronotal flange; notauli weak; epicnemial carina strong, present
on mesopleuron but medioventrally interrupted; mesopleural groove virtually absent. Scutellum convex,
carinate laterally about 0-8 of its length; posterior transverse carina of mesosternum complete, or slightly
weak medioventrally. Propodeum with anterior area moderately short but clearly discernible; anterior
transverse carina complete; posterior transverse carina vestigial, others absent; posterior area rugose.
Fore wing with pterostigma moderately slender; marginal cell long; discosubmarginal cell evenly hirsute;
Rs+2r straight, expanded slightly immediately before joining pterostigma; Im-cu evenly arcuate, ramellus
absent; unusual in having Im-cu and Cw la basally contiguous or united. Hind wing with Rs almost straight.
Fore tibial spur without a membranous flange behind macrotrichial comb; mid and hind trochantelli short
and stout, otherwise unspecialized; inner hind tibial spur slightly flattened, with a fringe of fine hairs; hind
tarsal claws long but closely pectinate. Gaster stout; tergite 2 depressed, barely longer than deep, not
longer than broad; thyridia large, oval, close to anterior margin; umbo absent; laterotergite narrow,
pendant. Ovipositor sheath short, slender.
Banchogastra is a small genus containing two Hawaiian species (Townes, 1971). Nothing is known of
their biology.
PHYLOGENY OF THE OPHIONINAE
157
Figs 38-^t3 Stereoscan photographs of ophionines: 38-41, mandibles of (38) Ophion sp.; (39) Ophio-
gastrella sp.; (40) Enicospilus sp.; (41) Leptophion alleni Gauld & Mitchell. 42, 43, hind corner of
pronotum and spiracular sclerite of (42) Ophion sp. ; (43) Enicospilus sp.
158
I. D. GAULD
Figs 44-49 Stereoscan photographs of Ophioninae: 44, mandibles of Thyreodon sp. 45-47, hind tarsal
claws of (45) Ophion sp. ; (46) Leptophion sp. $ ; (47) Ophiogastrella sp. cf . 48, 49, fore tibial spurs of
(48) Ophion sp.; (49) Enicospilus sp.
PHYLOGENY OF THE OPHIONINAE
159
Figs 50-52 Stereoscan photographs of dorsal region of posterior part of alitrunk of: 50, Ophion sp. ; 51,
Enicospilus sp.; 52, Thyreodon sp.
DICAMPTUS Szepligeti
Dicamptus Szepligeti, 1905: 21. Type-species: Dicamptus giganteus Szepligeti, by monotypy.
Mandibles not twisted, generally very weakly narrowed, almost equally bidentate; outer mandibular
surface weakly convex with proximal concavity, sometimes with pronounced proximal swelling and
diagonal hirsute groove. Maxillary palp 5-segmented, labial palp 4-segmented; clypeus in profile from flat
to convex or even pyramidal, usually with margin impressed and acute; clypeus in anterior aspect truncate
or weakly convex. Ocelli generally large, the posterior ones usually virtually contiguous with eyes except in
a few Afro tropical species; frontal carina absent; occipital carina usually complete, ventrally joining
hypostomal carina. Antennae moderately to extremely elongate, from slightly longer than fore wing to 2-0
times fore wing length and with upwards of 90 flagellar segments. Pronotum unspecialized; spiracular
sclerite virtually completely concealed by pronotal lobe; notauli vestigial or absent; epicnemial carina
usually complete, often strong; mesopleural furrow generally absent. Scutellum weakly to moderately
convex, with lateral longitudinal carinae strong, usually complete; posterior transverse carina of meso-
sternum complete. Propodeum with anterior area long, striate; anterior transverse carina usually complete
and other carinae obsolescent or absent; posterior area usually coarsely reticulate. Fore wing with
pterostigma fairly slender; marginal cell very long; Rs+2r broadened and variously sinuate before joining
pterostigma; discosubmarginal cell with a large glabrous fenestra near anterior corner, this fenestra
bearing at least one sclerite (proximal one); Im-cu from arcuate to sinuous, usually without a ramellus.
Hind wing with Rs virtually straight. Fore tibial spur without a membranous flange behind macrotrichial
comb; mid and hind trochantelli unspecialized; inner hind tibial spur flattened, with a margin of long close
hairs; hind tarsal claws unspecialized. Gaster slender; tergite 2 very elongate, with thyridia remote from
anterior margin; umbo absent; epipleuron up-turned. Ovipositor sheath slender.
Dicamptus is a moderate-sized genus with 27 described species and two undescribed species (ANIC;
BMNH) in the Palaeotropical region. A few species extend into temperate areas, D. fuscicornis (Erichson)
reaches Tasmania and D. nigropictus (Matsumura) occurs northwards as far as Japan and Korea.
Two quite distinct species-groups are recognizable. The pulchellus group comprises gracile species with a
very small cubital index (0-55 or less), a short vein 3r-m (less than 0-5 of the length of M between 2m-cu
and 3r-ra) and a fairly well-developed mandibular groove. This group contains D. collessi Gauld, D.
uptoni Gauld and Dicamptus species 1 (ANIC) which are Australian endemics, D. indicus Nikam and D.
fuscicornis which are widespread Indo- Australian species, D. isshikii, an Oriental species and the
Afrotropical species D. crassellus (Morley), D. xhosa Delobel, D. betsileo Delobel, D. pellucidus
(Kriechbaumer), D. seyrigi Delobel, D. townesi Delobel and D. pulchellus (Morley).
The remaining species (the giganteus group) are in general very much larger insects with a larger cubital
index, a longer 3r-m and stouter mandibles. This may well be a paraphyletic grade-group, as almost
160 I. D. GAULD
certainly the characters defining it are plesiomorphic for the genus. The majority of species in this group
occur in areas with a pronounced dry season , particularly parts of Africa and India. Some of the very largest
species (reticulatus (Cameron), nigropictus (Matsumura) and giganteus Szepligeti) occur in montane
forest, a very wet habitat. None of the species in the giganteus group occur east of Weber's line. In New
Guinea they seem to have been replaced by some exceptionally large species of Enicospilus (E. enormous
Gauld & Mitchell species complex).
ENICOSPILUS Stephens
Enicospilus Stephens, 1835: 126. Type-species: Ophion merdarius Gravenhorst sensu Stephens (=
Ichneumon ramidulus L.), by subsequent monotypy, Stephens, 1845.
Henicospilus Agassiz, 1846: 138. [Unjustified emendation.]
Allocamptus Foerster, 1869: 150. Type-species: Ophion undulatus Gravenhorst, by subsequent desig-
nation, Thomson, 1888: 1189.
Dispilus Kriechbaumer, 18946: 309. Type-species: Ophion (Dispilus) natalensis Kriechbaumer, by
monotypy.
Pleuroneurophion Ashmead, 1900: 86. Type-species: Pleuroneurophion hawaiiensis Ashmead, by original
designation.
Cymatoneura Kriechbaumer, 1901a: 22. Type-species: Ophion undulatus Gravenhorst, by subsequent
designation, Viereck, 1914: 8.
Pterospilus Kriechbaumer, 1901c: 156. Type-species: Ophion (Enicospilus) dubius Tosquinet, by subse-
quent designation, Viereck, 1914: 126. [Junior homonym of Pterospilus Rondani, 1856.]
Trispilus Kriechbaumer, 1901c: 156. Type-species: Ophion (Enicospilus) trimaculatus Tosquinet (=
Henicospilus seminiger Szepligeti), by monotypy.
Metophion Szepligeti, 1905: 28. Type-species: Metophion bicolor Szepligeti, by subsequent designation,
Viereck, 1914: 94.
Ceratospilus Szepligeti, 1905: 28. Type-species: Ceratospilus biroi Szepligeti, by monotypy.
Atoponeura Szepligeti, 1905: 34. Type-species: Atoponeura concolor Szepligeti (= Enicospilus ato-
poneurus Cushman), by monotypy.
Ophiomorpha Szepligeti, 1905: 34. Type-species: Ophion curvinervis Cameron (= Enicospilus cameronii
Dalla Torre), by subsequent designation, Hooker, 1912: 134. [Junior homonym of Ophiomorpha
Nilsson, 1836.]
Cryptocamptus Brethes, 1909: 230. [Unnecessary replacement name for Allocamptus Foerster.]
Eremotyloides Perkins, 1915: 530. Type-species: Eremotyloides orbitalis Ashmead, by monotypy.
Amesospilus Enderlein, 1918: 222. Type-species: Ophion unicallosus Snellen, by original designation.
Schizospilus Seyrig, 1935: 79. Type-species: Schizospilus divisus Seyrig, by original designation.
Mandibles twisted from 10 to 90, weakly to very strongly narrowed, from equally bidentate to with upper
tooth conspicuously the longer, rarely with lower tooth the longer; outer mandibular surface usually with a
proximal concavity, the remainder either almost flat or with a diagonal hirsute groove. Maxillary palp
5-segmented, labial palp 4-segmented; clypeus in profile from flat to nasute, with margin blunt or acute or
acute and subapically impressed; clypeus in anterior aspect usually weakly convex, rarely truncate or
joncave, in one Neotropical species with an indication of a median tooth. Ocelli usually large, generally
with the posterior ones very close to or contiguous with the eyes, in a few species with the ocelli smaller;
frontal carina absent; occipital carina usually complete, sometimes mediodorsally obsolescent or inter-
rupted. Antennae usually more than 1-5 times as long as the fore wing. Pronotum usually unspecialized, in
some Madagascan species with median transverse crests; spiracular sclerite almost always completely
occluded by flange of pronotum; notauli vestigial or absent; epicnemial carina usually well developed,
present on mesopleuron, usually reaching to level of lower corner of pronotum; mesopleural furrow
usually absent. Scutellum from almost flat to strongly convex, almost always with strong lateral carinae
virtually entire, rarely with these carinae short; posterior transverse carina of mesosternum usually
complete. Propodeum with anterior area moderately long to long, generally striate; anterior transverse
carina of propodeum usually present, almost always complete; posterior transverse carina usually absent,
rarely present laterally; posterior area usually reticulate or rugose, sometimes finely irregularly wrinkled,
in a few species concentrically striate, very rarely the posterior area almost smooth or punctate. Fore wing
with pterostigma moderately broad to quite slender; marginal cell long; Rs+2r almost always broadened
and variously sinuate before joining the pterostigma; discosubmarginal cell usually with a glabrous fenestra
adjacent to the vein Rs+2r, this fenestra often bearing one or more detached sclerites; Im-cu usually
arcuate or sinuous, rarely somewhat angulate; ramellus usually absent, rarely in one or two species
well-developed. Hind wing with Rs straight or weakly bowed. Fore tibial spur without a membranous
PHYLOGENY OF THE OPHIONINAE 161
flange behind macrotrichial comb; mid and hind trochantelli usually simple, rarely with a marginal tooth;
inner hind tibial spur flattened, with margin of close long hairs; hind tarsal claws various, most usually
unspecialized and slightly sexually dimorphic, sometimes extremely sexually dimorphic, the male having a
very fine pectinal comb, in a few taxa with the claws strongly geniculate, occasionally with the claws
asymmetrical, in other taxa with the distal pectina projecting beyond the apex of the true claw, and very
rarely, in a few eremic species, with the pectina reduced. Gaster from moderately stout to very slender;
tergite 2 usually very long and slender, rarely posteriorly almost as deep as long, almost always with
thyridia remote from anterior margin; umbo vestigial or absent; epipleuron usually up- turned, pendant in a
few (or one) Neotropical species. Ovipositor sheath slender; ovipositor usually short, straight, in isolated
taxa it may be upcurved or decurved, and very rarely it may be straight but project well beyond the apex of
the gaster.
Enicospilus is an extremely large genus, most species of which occur in the tropics. The major centres of
radiation appear to be New Guinea, where there are about 200 endemic species, and Madagascar, which
has about half that number. The Philippines, Hawaii and Zaire river basin are other areas of pronounced
endemicity. In the tropics, the greatest diversity of species seems to occur at mid-elevation in the cloud
forests (moss forests) between 1200 and 2000 m. A number of species are restricted to areas that have a
pronounced dry season (e.g. Enicospilus capensis), and a few species seem to" inhabit deserts (e.g. E.
psammus). Many species are capable of sustained flight over great distances (e.g. across the Tasman Sea
from Australia to New Zealand) and several occur on almost all South Pacific archipelagos as far east as the
Tuamotus. Most small oceanic islands apparently have several species of Enicospilus present, and on the
Micronesian islands six of the 33 recorded ichneumonids are Enicospilus (Townes, 1958). Outside the
tropics there are notably fewer species. Scaramozzino (1983) records only ten from Italy whilst Viktorov
(1957) knew of only 15 species from the Soviet Union. Five nominal species are recorded from Britain, but
probably only four species occur there (Fitton et al., 1978).
Certain generalized distribution patterns can be observed in the genus. New Guinea and Madagascar
have large numbers of endemic species in endemic species-groups; for example eight of the 24 species-
groups recognized as occurring in the Afrotropical region are restricted to Madagascar, and these eight
groups contain 35 species (Gauld & Mitchell, 1978). The Philippines, Hawaii and South America each
contain very few species-groups, but these may be very large. Relatively few of the numerous other
species-groups are endemic to a single zoogeographic area. Several of the species-groups that are
widespread in South East Asia have their greatest diversity in New Guinea (e.g. the xanthocephalus and
tremulus species-groups, Gauld & Mitchell, 1981). Virtually no widespread South East Asian groups have
a localized centre of diversity outside Melanesia; most species seem to be widespread with isolated local
endemics (e.g. theflavicaput species-group, Gauld & Mitchell, 1981). Many widespread South East Asiaa
groups are also represented in mainland Africa (e.g. the capensis and antefurcalis species-groups). A
number of species-groups are either endemic to mainland Africa (e.g. the babaultii, biimpressus and
rubens species-groups) or are most diverse in Africa (e.g. the dolosus species-group). Most species
occurring in the Palaearctic region belong to species-groups that are well-represented in the Nearctic
region, but neither region appears to have any endemic species-groups, nor is any species-group endemic
to the Holarctic region. The Enicospilus species of Australia are virtually all either widespread South East
Asian species or are endemics derived from South East Asia (Gauld, 1984a). New Zealand has no endemic
species, but shares two with Australia. A distinct faunal region is apparent which comprises part of the
Mediterranean basin, most of the Middle East and extends eastwards into Central Asia and south-
eastwards into north-west India. Although the fauna of this area is poorly known, it seems that at least one
species-group may be endemic to the region (the przewalskii species-group).
Gauld & Mitchell (1978; 1981) outline a very large number of species-groups and several of these are
refined by Gauld (1984a). Large numbers of other species are currently unplaced; these may belong to less
clearly definable groups, or they may be aberrant members of existing groups, or they may represent
numerous monobasic species-groups. The phylogenetic inter- relationships of most of these species-groups
are very difficult to assess as most are definable on the basis of a number of autapomorphies. Few share
obvious specializations with other species-groups.
Repeated attempts have been made to subdivide Enicospilus into a number of smaller genera
(Kriechbaumer, 1901c; Szepligeti, 1905; Seyrig, 1935) but none of these subdivisions has endured, largely
for two reasons - the authors proposing the separation have very limited experience of the range of
morphological diversity afforded by the genus, and the characters used to effect separation are superficial
differences, usually in the number of alar sclerites. The clearest demonstration of the great variability of
alar sclerite form and number can be seen between the closely related species of the unidens species-group,
a group definable by several autapomorphies. The proximal sclerite is present in all species but only E.
unidens has a central sclerite; the distal sclerite is present in E. unidens, E. akainus and E. mirax but absent
162 I. D. GAULD
in E. gonidius and E. amygdalis. The most important features for recognizing apparently 'natural'
groupings seem to be the form of the tarsal claws (including the degree of sexual dimorphism), the structure
of the mandibles and clypeus, the presence or absence of the proximal sclerite and the structure of the male
genitalia.
Little is known of the biology of species of Enicospilus. The majority of host records refer to species
parasitizing larvae of Noctuidae, Geometridae, Lymantriidae or Saturniidae. A few species attack
pyralids. Many of the larger species that attack saturniid larvae spin their cocoon within the host cocoon,
but other species spin a cocoon that is not enclosed by that of the host. Most species seem to attack larvae
that are free-living, tree-leaf-feeding caterpillars, but the few with longer ovipositors seem to attack larvae
mining stems (e.g. E. terebrus). It must be stressed that the hosts of the majority of tropical species are not
known.
PYCNOPHION Ashmead
Pycnophion Ashmead, 1900: 87. Type-species: Pycnophion molokaiensis Ashmead, by monotypy.
Mandibles twisted about 20, evenly tapered, with upper tooth broader but of about equal length to the
lower tooth; outer maridibular surface slightly concave, sparsely pubescent. Maxillary palp 5-segmented,
labial palp 4-segmented; clypeus in profile moderately convex, margin blunt; clypeus in anterior aspect
weakly convex. Ocelli moderately large, the posterior ones separated from the eye by about 0-5 times their
own diameter; frontal carina very weak; occipital carina complete, mediodorsally with a depression.
Antennae moderately long, about 1-3 times the length of the fore wing. Pronotum slightly flattened
mediodorsally; spiracular sclerite covered by pronotal flange; notauli vestigial; epicnemial carina strong,
curved to nearly reach the anterior margin of the pleuron above the level of the lower corner of the
pronotum; mesopleural furrow absent. Scutellum convex, laterally carinate at least 0-6 of its length;
posterior transverse carina of the mesosternum centrally obsolescent. Propodeum with anterior area
moderately short but clearly discernible; anterior transverse carina complete, at least centrally, the
posterior one vestigial, the other absent; posterior area finely coriaceous. Pterostigma moderately slender;
marginal cell long; Rs+2r virtually straight, evenly but abruptly widened before joining pterostigma;
discosubmarginal cell with an ill-defined glabrous area anteriorly, the entire cell very sparsely hirsute;
Im-cu fairly evenly arcuate, ramellus absent. Hind wing with Rs straight. Fore tibial spur without a
membranous flange behind macro trichial comb; mid and hind trochantelli unspecialized; inner hind tibial
spur somewhat flattened, with a fringe of long hairs; tarsal claws unspecialized. Gaster moderately stout;
tergite 2 in profile, slightly longer than posteriorly deep, with thyridia large, oval and close to anterior
margin; umbo vestigial; epipleuron up-turned. Ovipositor exceptional in being very long and up-curved,
reaching beyond apex of gaster by at least length of tergites 2-5.
Pycnophion is a small genus containing three Hawaiian species (Townes etal., 1961). It is one of the most
atypical of ophionine genera and females superficially resemble campoplegines or phygadeuontines. One
species, P. fuscipennis Perkins, has been reared as fuparasite of the larvae of Hyposmocoma chilonella
(Lepidoptera: Cosmopterygidae) (Swezey, 1931) which are borers in the pith of stems of Rubus, Acacia
and a variety of other woody plants.
Acknowledgements
Much of this work was undertaken as part of a PhD thesis under the supervision of Dr Garth Underwood. I
am extremely grateful for his advice at all stages of the study. I would like to thank Dr J. Felsenstein for
providing a copy of his program PHYLIP, and the staff of the computing centre at the City of London
Polytechnic for their assistance with computational problems. I am grateful to many of my colleagues at the
British Museum (Natural History) for their opinions on various problems and comments on various drafts,
though the conclusions expressed do not necessarily reflect their points of view; in particular I would like to
thank Dr A. D. Austin, Mr B. Bolton, Mr M. C. Day, Dr M. G. Fitton, Dr J. D. Holloway, Dr L. A.
Mound, Dr G. S. Robinson, Mr R. I. Vane- Wright and Dr P. Whalley. I am grateful to the following for
lending me particularly interesting specimens from the collections in their care: Dr D. Kasparayan, Dr H.
Townes and Dr L. Zombori, and I am particularly grateful to all those entomologists, too numerous to
mention, who have sent me ophionines from all over the world. Finally I would like to thank Mr J. Carter
for taking the stereoscan pictures and Ms P. A. Mitchell for typing the manuscript.
PHYLOGENY OF THE OPHIONINAE 163
References
Achterberg, C. van 1976. A preliminary key to the subfamilies of the Braconidae (Hymenoptera).
Tijdschrift voor Entomologie 119: 33-78.
1984. Essay on the phylogeny of Braconidae (Hymenoptera: Ichneumonoidea). Entomologisk
Tidskrift 105: 41-58.
Agassiz, J. L. R. 1846. Nomendator zoologicus, Index Universalis. 1135 pp. Soloduri.
Ashmead, W. H. 1900. Classification of ichneumon-flies or the superfamily Ichneumonoidea. Proceedings
of the United States National Museum 23: 1-220.
Audley-Charles, M. G., Hurley, A. M. & Smith, A. G. 1981. Continental movements in the Mesozoic and
Cenozoic. In Whitmore, T. C. (Ed.), Wallace's Line and Plate Tectonics. 91 pp. Oxford.
Barron, J. R. 1978. Systematics of the World Eucerotinae (Hymenoptera, Ichneumonidae) Part 2.
Non-Nearctic species. Naturaliste Canadien 105: 327-374.
Benson, R. B. 1935. On the genera of Cephidae, and the erection of a new subfamily Syntexidae. Annals &
Magazine of Natural History (10) 16: 535-553.
Bolton, B. 1982. Afrotropical species of the myrmicine ant genera Cardiocondyla, Leptothorax, Melisso-
tarsus, Messor and Cataulacus (Formicidae). Bulletin of the British Museum (Natural History) (Ento-
mology) 45: 307-370.
Brauns, S. 1889. Die Ophionoiden. Archiv des Vereins der Freunde der Naturgeschichte in Mecklenburg 43:
58-100.
1895. Descriptions specierum novarum Ichneumonidarum e fauna Hungarica. Termeszetrajzi
Fuzetek 18: 42-49.
Brethes, J. 1909. Hymenoptera Paraguayensis. Anales del Museo National de Historia Natural de Buenos
Aires 19: 225-256.
Brock, J. P. 1982. A systematic study of the genus Ophion in Britain. Tijdschrift voor Entomologie 125:
57-97.
Brues, C. T. 1912. Brazilian Ichneumonidae and Braconidae obtained by the Stanford expedition. Annals
of the Entomological Society of America 5: 193-228.
Brulle, M. A. 1846. In Lepeletier de Saint-Fargeau, A., Histoire naturelle des insectes 4. Hymenopteres.
vii+680pp. Paris.
Butler, P. M. 1982. Directions of evolution in the mammalian dentition. In Joysey, K. A. & Friday, A. E.
(Eds), Problems of Phylogenetic Reconstruction, xi+442 pp. London.
Cameron, P. 1901. On Hymenoptera collected in New Britain by Dr Arthur Willey. Proceedings of the
Zoological Society of London 1901: 224-248.
1903. Descriptions of new genera and species of Hymenoptera taken by Mr Robert Shelford at
Sarawak, Borneo. Journal of the Straits Branch of the Royal Asiatic Society 39: 89-181.
1905. On the phytophagous and parasitic Hymenoptera collected by Mr Ernest Green in Ceylon.
Spolia Zeylanica 3: 67-144.
1906. Descriptions of new species of parasitic Hymenoptera chiefly in the collection of the South
African Museum, Cape Town. Annals of the South African Museum 5: 17-186.
Carlson, R. W. 1979. Family Ichneumonidae. In Krombein, K. V., Hurd, P. D., Smith, D. R. & Burks, B.
D. , Catalog of Hymenoptera in America North of Mexico 1: 1-1198.
Carpenter, J. M. 1982. The phylogenetic relationships and natural classification of the Vespoidea
(Hymenoptera). Systematic Entomology 7: 11-38.
Clausen, C. P. 1940. Entomophagous Insects. 688 pp. New York.
Curtis, J. 1836. British Entomology 13: plates 578-625. London.
Cushman, R. A. 1926. Some types of parasitism among Ichneumonidae. Proceedings of the Entomological
Society of Washington 28: 25-51.
1944. The Hawaiian species of Enicospilus and Abanchogastra (Hymenoptera, Ichneumonidae).
Proceedings of the Hawaiian Entomological Society 12: 39-56.
1947. A generic revision of the ichneumon-flies of the tribe Ophionini. Proceedings of the United
States National Museum 96: 417-482.
Darlington, P. J. 1965. The biogeography of the southern end of the world, vii+236 pp. Cambridge,
Massachusetts.
Davies, R. G. 1981. Information theory and character selection in the numerical taxonomy of some male
Diaspididae (Hemiptera: Coccoidea). Systematic Entomology 6: 149-178.
Davies, R. G. & Boratyhski, K. L. 1979. Character selection in relation to the numerical taxonomy of some
male Diaspididae (Homoptera: Coccoidea). Biological Journal of the Linnean Society 12: 95-165.
164 I. D. GAULD
Day, W. H. E. 1983. Computationally difficult problems in phylogenetic systematics. Journal of Theoretical
Biology 103: 429-438.
Delobel, A. 1975. Deux nouveaux Rhopalophion Seyrig appartenant a la faune malgache. Bulletin de la
Societe entomologique de France 80: 43-47.
Enderlein, G. 1912. Beitrage zur Kenntnis aussereuropaischer Ichneumoniden II. Ophionoiden. Der
Gattung Thyreodon und ihre Verwandten. Zoologischer Anzeiger 39: 624-632.
- 1918. Ichneumoniden. In Michaelson, W., Beitrage zur Kenntnis der Land und Susswasserfauna
Deutsch-Sudwestafrikas 2(4). 22 pp. Jena.
1921. Beitrage zur Kenntnis aussereuropaischer Ichneumoniden. Stettiner Entomologische Zeitung
82: 1-45.
Fabricius, J. C. 1798. Entomologia systematica emendata et aucta . . . adjectis synonymis, locis obser-
vationibus, descriptionibus. Supplementum. 572pp. Halfniae.
Farris, J. S. 1969. A successive approximation approach to character weighting. Systematic Zoology 18:
374-385.
1979. The information content of the phylogenetic system. Systematic Zoology 28: 483-519.
Felsenstein, J. 1981 . A likelihood approach to character weighting and what it tells us about parsimony and
compatibility. Biological Journal of the Linnean Society 16: 183-196.
1982. Numerical methods for inferring evolutionary trees. Quarterly Review of Biology 57: 379-404.
Felt, E. P. 1904. Nineteenth report of the state entomologist, 1903. Beneficial insects. Bulletin of the New
York State Museum 76: 97-125.
Fitton, M. G. & Gauld, I. D. 1976. The family-group names of the Ichneumonidae (excluding the
Ichneumoninae). Systematic Entomology 1: 247-258.
Fitton, M. G., Graham, M. W. R. de V., Boucek, Z. R. J., Fergusson, N. D. M., Huddleston, I., Quinlan,
J. & Richards, O. W. 1978. Kloet and Hincks, a check list of British Insects. Part 4: Hymenoptera.
Handbooks for the Identification of British Insects 11(4): 1-159.
Foerster, A. 1869. Synopsis der Familien und Gattungen der Ichneumoniden. Verhandlungen des
Naturhistorischen Vereins der Preussischen Rheinlande und Westfalens 25: 135-221.
Friday, A. E. 1982. Parsimony, simplicity and what actually happened. Zoological Journal of the Linnean
Society 74: 329-335.
Gaffney, E. S. 1979. An introduction to the logic of phylogeny reconstruction. In Cracraft, J. & Eldredge,
N. (Eds), Phylogenetic Analysis and Palaeontology, ix+233 pp. New York.
Gauld, I. D. 1976. The classification of the Anomaloninae (Hymenoptera: Ichneumonidae). Bulletin of the
British Museum (Natural History) (Entomology) 33: 1-135.
1977. A revision of the Ophioninae (Hymenoptera: Ichneumonidae) of Australia. Australian Journal
of Zoology, Supplementary Series 49: 1-112.
1979. An analysis of the classification of the Ophion genus-group (Ichneumonidae). Systematic
Entomology 5: 59-82.
- 1983. The classification, evolution and distribution of the Labeninae, an ancient southern group of
Ichneumonidae (Hymenoptera). Systematic Entomology 8: 167-178.
1984a. The Australian Ophioninae (Insecta; Hymenoptera): a historical biogeographic study.
Journal of Biogeography 11: 269-288.
19846. An introduction to the Ichneumonidae of Australia. 413 pp. London.
Gauld, I. D. & Carter, J. M. 1983. The Ophioninae of the Galapagos Islands (Hymenoptera: Ichneumoni-
dae). Journal of Natural History 17: 145-155.
Gauld, I. D. & Huddleston, T. 1976. The nocturnal Ichneumonoidea of the British Isles, including a key to
genera. Entomologist's Gazette 27: 35-49.
Gauld, I. D. & Mitchell, P. A. 1978. The taxonomy, distribution and host preferences of African parasitic
wasps of the subfamily Ophioninae. 287 pp. Slough.
1981. The taxonomy, distribution and host preferences of Indo-Papuan parasitic wasps of the subfamily
Ophioninae. 611 pp. Slough.
Gauld, I. D. & Mound, L. A. 1982. Homoplasy and the delineation of holophyletic genera in some insect
groups. Systematic Entomology 7: 73-86.
Guise, A., Peacock, D. & Gleaves, T. A. 1982. A method for identification of parallelism in discrete
character sets. Zoological Journal of the Linnean Society 74: 293-303.
Gupta, V. K. 1962. Taxonomy, zoogeography and evolution of Indo- Australian Theronia (Hymenoptera:
Ichneumonidae). Pacific Insects Monograph 1: 1-142.
Hellen, W. 1926. Beitrage zur Kenntnis der Ichneumoniden Finlands. 2 Subfam. Ophioninae und
Anomaloninae. Acta Societatis pro Fauna et Flora Fennica 56: 3-27.
Hennig, W. 1966. Phylogenetic Systematics. xiii+263 pp. Urbana, Illinois.
PHYLOGENY OF THE OPHIONINAE 165
Hooker, C. W. 1912. The ichneumon-flies of America belonging to the tribe Ophionini. Transactions of the
American Entomological Society 38: 1-176.
Horstmann, K. 1981. Die Palaarktischen Arten der Gattungen Eremotylus Forster und Simophion
Cushman (Hymenoptera, Ichneumonidae). Entomofauna 2: 415^32.
Hull, D. L. 1965. The effect of essentialism on taxonomy: two thousand years of stasis 1 . British Journal for
the Philosophy of Science 15: 314-326.
Kluge, A. G. 1976. Phylogenetic relationships in the lizard family Pygopodidae: an evaluation of theory,
methods and data. Miscellaneous Publications, Museum of Zoology, University of Michigan 152: 1-72.
Kohl, F. F. 1906. Ichneumonidae. In Penther, A. & Zederbauer, E., Ergebnisse einer naturwissenschaft-
lichen Reise zur Erdschias-Dagh (Kleinasien). Annalen des Naturhistorischen (Hof) Museums Wien 20:
220-246.
Konigsmann, E. 1978. Das Phylogenetische System der Hymenopteren 3. Terebrantes (Unterordnung
Apocrita). Deutsche Entomologische Zeitschrift (N.F.) 25: 1-55.
Kriechbaumer, J. 1894a. In Sickmann, F., Beitrage zur Kenntnis der Hymenopteren-Fauna des nordis-
chen China. ZoologischeJahrbiicherSystematikS: 195-236.
18946. Hymenoptera Ichneumonidae. Berliner Entomologische Zeitschrift 39: 297-318.
1901a. Bemerkungen iiber Ophioniden. Zeitschrift fur Systematische Hymenopterologie und Dipter-
ologiel: 18-24.
1901ft. Bemerkungen iiber Ophioniden. Zeitschrift fur Systematische Hymenopterologie und Dipter-
ologie 1:73-79.
1901c. Ueber die Gattungen der von Tosquinet in seiner Ichneumonides d'Afrique beschrieben
Ophionarten. Zeitschrift fur Systematische Hymenopterologie und Dipterologie 1: 155-156.
LeQuesne, W. J. 1969. A method of selection of characters in numerical taxonomy. Systematic Zoology 18:
201-205.
1972. Further studies on the uniquely derived character concept. Systematic Zoology 21: 281-288.
L0vetrup, S. 1973. Classification, convention and logic. Zoologica Scripta 2: 49-61.
Mackerras, I. M. 1962. Speciation in Australian Tabanidae. In Leeper, G. W. (Ed.), The Evolution of
Living Organisms. 459 pp. Melbourne.
Martin, R. 1981. Phylogenetic reconstruction versus classification: the case for clear demarcation.
Biologist 28: 127-132.
Mickevich, M. F. 1978. Taxonomic congruence. Systematic Zoology 27: 143-158.
Middlekauf, W. W. 1983. A revision of the sawfly family Orussidae for North and Central America
(Hymenoptera; Symphyta, Orussidae). University of California Publications in Entomology 101: 1-46.
Morley, C. 1912. A revision of the Ichneumonidae ... 1. ix+88 pp. London.
1913. The fauna of British India, including Ceylon and Burma. Hymenoptera 3, Ichneumonidae, 1
Ichneumones Deltoidei. 532 pp. London.
Moutia, L. A. & Curtois, C. M. 1952. Parasites of the moth borers of sugarcane in Mauritius. Bulletin of
Entomological Research 43: 325-359.
Nelson, G. & Platnick, N. 1981. Systematics and biogeography. Cladistics and vicariance. 567 pp. New
York.
Owen, H. G. 1981. Constant dimensions or an expanding Earth? In Cocks, L. R. M. (Ed.), The Evolving
Earth. 264 pp. London & Cambridge.
Pampel, W. 1913. Die weiblichen Geschlechtsorgane der Ichneumoniden. Zeitschrift fur Wissenschaftliche
Zoologie 108: 290-357.
Panchen, A. L. 1982. The use of parsimony in testing phylogenetic hypotheses. Zoological Journal of the
Linnean Society 74: 305-328.
Perkins, J. F. 1959. Ichneumonidae, key to subfamilies and Ichneumoninae 1. Handbook for the
Identification of British Insects 7(2ai): 1-116.
Perkins, R. C. L. 1902. Four new species and a new genus of Parasitic Hymenoptera (Ichneumonidae;
Ophioninae) from the Hawaiian Islands. Transactions of the Entomological Society of London 1902:
141-143.
1915. On Hawaiian Ophioninae (Hymenoptera, Ichneumonidae). Transactions of the Entomological
Society of London 1914: 521-535.
Pratt, V. 1972. Numerical taxonomy - a critique. Journal of Theoretical Biology 36: 581-592.
Price, P. W. 1975. The evolutionary strategies of parasitic insects and mites. 244 pp. New York.
Quinlan, J. & Gauld, I. D. 1981. Symphyta (except Tenthredinidae). Hymenoptera, new edition.
Handbooks for the Identification of British Insects 6(2a): 1-67.
Rasnitsyn, A. P. 1983. Ichneumonoidea (Hymenoptera) from the Lower Cretaceous of Mongolia.
Contributions of the American Entomological Institute 20: 259-265.
166 I. D. GAULD
Rich, P. V. & Rich, T. H. 1983. The Central American disperal route: biotic history and palaeogeography.
In Janzen, D. H. (Ed.), Costa Rican Natural History . 816 pp. Chicago & London.
Richards, O. W. 1956. Hymenoptera: introduction and key to families. Handbooks for the Identification of
British Insects 6(1): 1-94.
Rodendorf , B. B. 1962. Fossil Insects (English translation by H. Vaitactis. Unpublished - copy in BMNH).
358 pp.
Rodendorf, B. B. & Rasnitsyn, A. P. 1980. [Historical development of the insect class.] [In Russian.] Trudy
Palaeontologicheskogo Instituta 175: 1-269.
Rohlf, F. J., Colless, D. H. & Hart, G. 1983. Taxonomic congruence re-examined. Systematic Zoology 32:
144-158.
Rohlf, F. J. & Sokal, R. R. 1980. Comments on taxonomic congruence. Systematic Zoology 29: 97-101.
Roman, A. 1943. Neue Schlupfwespen aus Ostafrika. Folium Entomologicum Festschrift zum 60-
Geburtstage von Felix Bryk: 20-23.
Rothschild, W. & Jordan, K. 1903. A revision of the lepidopterous family Sphingidae. cxxxv+813 pp.
Tring.
Saether, O. A. 1979. Underlying synapomorphies and anagenetic analysis. Zoologica Scripta 8: 305-312.
Scaramozzino, P. L. 1983. II genere Enicospilus Steph. in Italia. Nota preliminare. Atti XIII Congresso
Nazionale Italiano di Entomologia 1983: 113-117.
Schrank, F. P. 1802. Fauna Boica 2(2): 977 pp. Ingolstadt.
Schulz, W. A. 1911. Zweihundert alte Hymenopteren. Zoologischen Annalen (Wurzburg) 4: 1-220.
Seyrig, A. 1926. Etudes sur les Ichneumonides (Hymenoptera) 1. Eos 2: 115-133.
1935. Hymenoptera 2, Ichneumonidae. Mission Scientifique de I'Omo 3(18): 1-103.
Shaw, M. R. 1983. On evolution of endoparasitism: the biology of some genera of Rogadinae (Braconi-
dae). Contributions of the American Entomological Institute 20: 307-328.
Shestakov, A. 1926. Tabula diagnostica et species palaearcticae. Konowia 5: 256-263.
Short, J. R. T. 1978. The final larval instars of the Ichneumonidae. Memoirs of the American Entomo-
logical Institute 25: 1-508.
Shuckard, W. E. 1840. In Swainson, W. & Shuckard, W. E. (Eds), The cabinet cyclopedia: on the history
and natural arrangement of insects. iv+406pp. London.
Sokal, R. R. & Sneath, P. H. A. 1963. Principles of numerical taxonomy, xv -1-573 pp. San Francisco.
Stanley, S. M. 1979. Mac? revolution. Patterns and Process. 332 pp. San Francisco.
Stephens, J. L. 1835. Illustrations of British Entomology Mandibulata 7. 306 pp. London.
1845. Illustrations of British Entomology Mandibulata 7. Index & list of plates. 6 pp. London.
Strand, E. 1928. Miscellanea nomenclatorica zoologica et palaeontologica. Archiv fur Naturgeschichte
92A: 30-75.
Strauch, J. R. 1984. Use of homoplastic characters in compatibility analysis. Systematic Zoology 33:
167-177.
Swezey, O. H. 1931. Some observations on the insect faunas of native forest trees in the Olinda forest on
Maui. Proceedings of the Hawaiian Entomological Society 7: 493-504.
Szlpligeti, G. V. 1905. Hymenoptera, Ichneumonidae. In Wytsman, P., Genera Insectorum 34: 1-68.
Taylor, R. W. 1972. Biogeography of insects of New Guinea and Cape York Peninsula. In Walker, D.
(Ed.), Bridge and barrier: the natural and cultural history of Torres Strait. 437 pp. Canberra.
Telenga, N. A. 1969. Origin and evolution of parasitism in Hymenoptera Parasitica and development of their
fauna in the USSR (English translation). 112 pp. Jerusalem.
Thomson, C. G. 1888. Ofversigt af de i Sverige funna arter af Ophion och Paniscus. Opuscula Entomo-
logica 12: 1185-1201.
Tosquinet, J. 1903. Ichneumonidae nouveaux. Memoires de la Societe Royale Entomologique de Belgique
10: 1^02.
Townes, H. 1958. Hymenoptera: Ichneumonidae, Stephanidae and Evaniidae. Insects of Micronesia 19:
35-87.
1969. Genera of Ichneumonidae 1. Memoirs of the American Entomological Institute 11: 1-300.
1970. Genera of Ichneumonidae 3. Memoirs of the American Entomological Institute 13: 1-307.
1971. Genera of Ichneumonidae 4. Memoirs of the American Entomological Institute 17: 1-372.
1973a. Two ichneumonids (Hymenoptera) from the early Cretaceous. Proceedings of the Entomo-
logical Society of Washington 75: 216-219.
19736. Three tryphonine ichneumonids from Cretaceous amber (Hymenoptera). Proceedings of the
Entomological Society of Washington 75: 282-287.
Townes, H., Momoi, S. & Townes, M. 1965. A catalogue and reclassification of Eastern Palaearctic
Ichneumonidae. Memoirs of the American Entomological Institute 5: 1-661.
PHYLOGENY OF THE OPHIONINAE 167
Townes, H. & Townes, M. 1966. A catalogue and reclassification of Neotropic Ichneumonidae. Memoirs of
the American Entomological Institute 8: 1-367.
1973. A catalogue and reclassification of Ethiopian Ichneumonidae. Memoirs of the American
Entomological Institute 19: 1-416.
Townes, H., Townes, M. & Gupta, V. K. 1961. A catalogue and reclassification of Indo- Australian
Ichneumonidae. Memoirs of the American Entomological Institute 1: 1-522.
Uchida, T. 1928. Zweiter Beitrage zur Ichneumoniden Fauna Japans. Journal of the Faculty of Agriculture
of Hokkaido Imperial University 21: 177-297.
Underwood, G. 1982. Parallel evolution in the context of character analysis. Zoological Journal of the
Linnean Society 74: 245-266.
Viereck, H. L. 1912. Contributions to our knowledge of bees and ichneumon-flies including descriptions of
twenty-one new genera and fifty-seven new species of ichneumon-flies. Proceedings of the United States
National Museum 42: 613-648.
1914. Type-species of the genera of ichneumon-flies. Bulletin of the United States National Museum
31: 1-186.
Viktorov, G. A. 1957. Species of the genus Enicospilus Stephens in USSR. Entomologicheskoe Obozrenie
36: 179-210.
Watrous, L. E. & Wheeler, Q. D. 1981. The out-group comparison method of character analysis.
Systematic Zoology 30: 1-11.
Westwood, J. O. 1882. Descriptions of new or imperfectly known species of Ichneumones Adsciti.
Tijdschrift voor Entomologie 25: 2-48.
Wilson, E. 0. 1959. Adaptive shift and dispersal in a tropical ant fauna. Evolution 13: 122-144.
168
I. D. GAULD
Appendix 1 The ophionine taxa used in the cladistic study.
901 Thyreodon atricolor (Olivier)
902 Thyreodon laticinctus Cresson
903 Thyreodon fulvescens Cresson
904 Thyreodonflamminiger (Morley)
905 Simophion calvus Viktorov
906 Orientospilus melasma Townes
907 Dictyonotus purpurascens (Smith)
908 Rhynchophion flammipennis (Ashmead)
909 Lepiscelus distans (Seyrig)
910 Ophiogastrella sp. 1 (BMNH)
91 1 Ophiogastrella sp. 2 (BMNH)
912 Laticoleus unicolor (Szepligeti)
913 Laticoleus pronotalis Gauld & Mitchell
914 Laticoleus spilus Gauld & Mitchell
915 Laticoleus curvatus Delobel
916 Stauropoctonus bombycivorus (Gravenhorst)
917 Stauropoctonus occipitalis Gauld & Mitchell
918 Aulophion sp. 1 (BMNH)
919 Pamophion sorus Gauld
920 Riekophion emandibulator (Morley)
921 Ophionopsis nigrocyaneus Tosquinet
922 Sicophion pleuralis Gauld
923 Prethophion latus Townes
924 Pycnophion molokaiensis Ashmead
925 Abanchogastra hawaiiensis (Ashmead)
926 Banchogastra nigra Ashmead
927 Barytatocephalus mocsaryi (Brauns)
928 Leptophion maculipenhis (Cameron)
929 Leptophion anici Gauld
930 Leptophion pterospilus Gauld & Mitchell
931 Leptophion tetus Gauld
932 Dicamptus neavei Gauld & Mitchell
933 Dicamptus giganteus Szepligeti
934 Dicamptus fuscicornis (Erichson)
935 Euryophion latipennis (Kirby)
936 Euryophion adustus (Townes)
937 Rictophion ikuthana (Kriechbaumer)
938 Xylophion xylus (Gauld)
939 Ophion luteus (L.)
940 Eremotylus boguschi (Meyer)
941 Agathophionafulvicornis Westwood
942 Enicospilus tremulus Gauld & Mitchell
943 Enicospilus spathius Gauld & Mitchell
944 Enicospilus nephele Gauld & Mitchell
945 Enicospilus cionobius Gauld & Mitchell
946 Enicospilus unidens Seyrig
947 Enicospilus mahalonius Gauld & Mitchell
948 Rhopalophion discinervis (Morley)
949 Janzophion nebosus sp. n.
950 Hellwigiella nigripennis Szepligeti
951 Sclerophion uchidai Gauld & Mitchell
Hypothetical taxa
960 Laticoleus ancestor
961 Ophiogastrella ancestor
962 Le-'ophion ancestor
963 Stauropoctonus ancestor
Nearctic
Neotropical
Neotropical
Neotropical
Palaearctic
Afrotropical
Oriental/Eastern Palaearctic
Neotropical
Afrotropical
Neotropical
Neotropical
Afrotropical
Afrotropical (Madagascar)
Afrotropical
Afrotropical
Palaearctic
Afrotropical (Madagascar)
Neotropical
Australian
Australian
Afrotropical
Neotropical
Neotropical
Hawaiian
Hawaiian
Hawaiian
Palaearctic
Oriental
Australian/New Caledonian
Oriental
Australian
Afrotropical
Oriental
Oriental/ Australian
Afrotropical
Afrotropical
Afrotropical
Australian/Papuan
Palaearctic
Palaearctic
Neotropical
Oriental
Oriental
Oriental
Oriental
Afrotropical
Afrotropical (Madagascar)
Afrotropical
Neotropical
Palaearctic
Oriental/Eastern Palaearctic
964 Thyreodonl Dictyonotus ancestor
965 Euryophion ancestor
966 Dicamptus ancestor
967 Enicospilus ancestor
PHYLOGENY OF THE OPHIONINAE 169
Appendix 2 Primary data matrix of 51 selected ophionines and 95 characters. indicates a presumed
plesiomorphic condition, 1 a presumed apomorphic state.
4.1 5.1 6.1 7.1 8.1 9 11 13 14.2 15.2
3 4.2 5.2 6.2 7.2 8.2 10 12 14.1 15.1 16.1
901
1
1
1
1
1
1
902
1
1
1
1
1
1
1
903
1
1
1
1
1
904
1
1
1
1
1
1
905
1
1
906
1
1
1
1
907
1
1
1
1
1
1
1
908
1
1
1
1
1
1
1
1
909
1
1
1
1
1
1
910
1
1
911
1
1
912
1
1
1
1
1
913
b
1
1
1
1
914
1
1
1
1
915
1
1
1
1
1
916
1
1
1
1
1
1
917
1
1
1
1
1
918
1
1
1
1
1
1
919
1
1
1
1
1
920
1
921
1
1
1
1
1
1
1
922
1
1
1
1
1
1
1
1
923
1
1
1
1
924
1
1
1
1
1
925
1
1
1
1
1
1
926
1
1
1
1
1
1
927
1
1
1
1
928
1
1
1
1
1
929
1
1
1
1
930
1
1
1
1
1
931
1
1
1
1
1
932
1
1
1
1
1
1
933
1
1
1
1
1
934
1
1
1
1
1
935
1
1
1
1
1
1
936
1
1
1
1
1
1
937
1
1
1
1
1
1
1
938
1
1
939
1
1
940
1
1
941
1
1
1
1
942
1
1
1
1
1
943
1
1
1
1
1
1
1
944
1
1
1
1
945
1
1
1
1
1
1
946
1
1
1
1
1
1
1
947
1
1
1
1
1
1
1
948
1
1
1
949
1
1
1
1
1
1
950
1
1
1
1
951
1
p
1
1
170 I. D. GAULD
Appendix 2 - cont.
16.2 18.1 19 20.2 21.2 22.2 24.1 25.1 26 27.2 28.1 29.1
17 18.2 20.1 21.1 22.1 23 24.2 25.2 27.1 27.3 28.2 29.2
901
1
1
1
1
1
1
1
1
1
1
1
1
1
1
902
1
1
1
1
1
1
1
1
1
1
1
1
1
903
1
1
1
1
1
1
1
1
1
1
1
1
1
1
904
1
1
1
1
1
1
1
1
1
1
1
905
1
1
1
1
1
1
1
1
1
1
1
1
1
906
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
907
1
1
1
1
1
1
1
1
1
1
1
1
1
1
908
1
1
1
1
1
1
1
1
1
1
1
1
909
1
1
1
1
1
1
1
1
1
910
1
1
1
1
1
1
1
1
1
1
911
1
1
1
1
1
1
1
1
1
1
1
912
1
1
1
1
1
1
1
1
1
1
913
1
1
1
1
1
1
1
1
1
1
1
1
1
914
1
1
1
1
1
1
1
1
1
1
1
1
915
1
1
1
1
1
1
1
1
1
1
1
1
916
1
1
1
1
1
1
1
1
1
917
1
1
1
1
1
1
1
1
1
1
918
1
1
1
1
1
1
1
1
1
1
1
919
1
1
1
1
1
1
1
1
1
920
1
1
1
1
1
1
921
1
1
1
1
1
1
1
1
1
1
1
1
922
1
1
1
1
1
1
1
1
1
1
923
1
1
1
1
1
1
1
1
924
1
1
1
1
1
1
1
1
925
1
1
1
1
1
1
1
1
1
1
1
926
1
1
1
1
1
1
1
927
1
1
1
1
1
1
1
1
1
1
928
1
1
1
1
1
1
1
1
929
1
1
1
1
1
1
1
1
1
1
1
930
1
1
1
1
1
1
1
1
931
1
1
1
1
1
1
1
1
1
932
1
1
1
1
1
1
1
1
1
1
933
1
1
1
1
1
1
1
1
934
1
1
1
1
1
1
1
1
1
935
1
1
1
1
1
1
1
1
1
1
936
1
1
1
1
1
1
1
1
1
1
937
1
1
1
1
1
1
1
1
1
1
1
1
1
938
1
1
1
1
1
1
1
939
1
1
1
1
940
1
1
1
1
1
1
941
1
1
1
1
1
1
1
942
1
1
1
1
1
1
1
1
1
943
1
1
1
1
1
1
1
1
944
1
1
1
1
1
1
1
945
1
1
1
1
1
1
1
1
946
1
1
1
1
1
1
1
1
1
1
1
947
1
1
1
1
1
1
1
1
1
1
1
1
948
1
1
1
1
1
1
1
1
1
949
1
1
1
1
1
1
1
950
1
1
1
1
1
1
1
1
1
1
951
1
1
1
1
1
1
1
PHYLOGENY OF THE OPHIONINAE 171
Appendix 2 - cont.
30 32 34 36.1 36.3 38 40 42.1 43.1 44.1 45.1 46 47.2
31 33 35 36.2 37 39 41 42.2 43.2 44.2 45.2 47.1
901
1
1
1
1
1
1
1
1
1
902
1
1
1
1
1
1
1
1
903
1
1
1
1
1
1
1
1
904
1
1
1
1
1
1
1
1
1
1
905
1
1
1
1
1
1
1
1
1
1
906
1
1
1
1
1
1
1
1
1
1
907
1
1
1
1
1
1
1
1
1
908
1
1
1
1
1
1
1
1
1
909
1
1
1
1
1
1
1
1
1
1
910
1
1
1
1
1
1
1
1
1
1
911
1
1
1
1
1
1
1
1
1
912
1
1
1
1
1
1
1
1
1
1
1
913
1
1
1
1
1
1
1
1
1
1
914
1
1
1
1
1
1
1
1
1
1
1
1
1
915
1
1
1
1
1
1
1
1
1
1
1
1
1
916
1
1
1
1
1
1
1
1
1
1
1
1
917
1
1
1
1
1
1
1
1
1
918
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
919
1
1
1
1
1
1
1
1
1
1
1
1
1
920
1
1
1
1
1
1
1
1
1
1
1
921
1
1
1
1
1
1
1
1
1
1
1
922
1
1
1
1
1
1
1
923
1
1
1
1
1
1
1
1
1
924
1
1
1
1
1
1
1
1
1
1
1
925
1
1
1
1
1
1
1
1
1
926
1
1
1
1
1
1
1
1
1
1
1
927
1
1
1
1
1
1
1
1
1
928
1
1
1
1
1
1
1
1
1
1
1
1
1
1
929
1
1
1
1
1
1
1
1
1
1
1
1
930
1
1
1
1
1
1
1
1
1
1
1
1
1
931
1
1
1
1
1
1
1
1
1
1
932
1
1
1
1
1
1
1
1
1
1
1
1
1
1
933
1
1
1
1
1
1
1
1
1
1
1
1
934
1
1
1
1
1
1
1
1
1
1
1
1
935
1
1
1
1
1
1
1
1
1
1
1
1
936
1
1
1
1
1
1
1
1
1
1
1
1
1
1
1
937
1
1
1
1
1
1
1
1
1
1
1
1
938
1
1
1
1
1
1
1
1
1
939
1
1
1
1
1
1
940
1
1
1
1
1
1
1
1
941
1
1
1
1
1
1
1
942
1
1
1
1
1
1
1
1
1
1
1
1
1
1
943
1
1
1
1.
1
1
1
1
1
1
1
944
1
1
1
1
1
1
1
1
945
1
1
1
1
1
1
1
1
1
1
1
1
946
1
1
1
1
1
1
1
1
1
1
1
1
1
1
947
1
1
1
1
1
1
1
1
1
1
1
1
948
1
1
1
1
1
949
1
1
1
1
1
1
1
1
1
1
1
950
1
1
1
1
1
1
1
1
1
1
1
951
1
1
1
1
1
1
1
1
1
1
1
172 I. D. GAULD
Appendix 2 - cont.
48
.1
49
51
.1
52
.1
53
55
57
58
.2
60
61
.2
63
48
.2
50
51
.2
52
.2
54
56
58
.1
59
61
.1
62
64
901
1
1
1
1
1
1
902
1
1
1
1
1
1
903
1
1
1
1
1
1
1
904
1
1
1
1
1
905
1
1
1
1
1
1
1
906
1
1
1
1
1
1
1
1
1
1
907
1
1
1
1
1
1
908
1
1
1
1
1
1
909
1
1
1
1
1
1
1
1
1
1
1
1
1
910
1
1
1
1
1
1
1
1
1
1
911
1
1
1
1
1
1
1
1
1
1
1
912
1
1
1
1
1
1
1
1
913
1
1
1
1
1
1
1
914
1
1
1
1
1
1
1
1
1
1
915
1
1
1
1
1
1
1
1
1
1
916
1
1
1
1
1
1
1
1
1
1
917
1
1
1
1
1
1
1
1
1
918
1
1
1
1
1
1
1
1
1
1
919
1
1
1
1
1
1
1
1
1
920
1
1
1
1
921
1
1
1
1
1
922
1
1
1
1
923
1
1
1
1
1
1
1
924
1
1
1
1
1
1
1
1
925
1
1
1
1
1
1
1
1
1
926
1
1
1
1
1
1
1
927
1
1
1
1
1
1
1
928
1
1
1
1
1
1
1
1
929
1
1
1
1
1
1
1
930
1
1
1
1
1
1
1
931
1
1
1
1
1
1
1
932
1
1
1
1
1
1
933
1
1
1
1
1
1
1
934
1
1
1
1
1
1
1
1
935
1
1
1
1
1
1
1
936
1
1
1
1
1
1
1
1
937
1
1
1
1
1
1
1
938
1
1
1
1
1
1
1
1
939
1
1
1
940
1
1
1
1
1
941
1
1
1
1
942
1
1
1
1
1
1
1
1
1
1
1
1
943
1
1
1
1
1
1
1
1
1
1
944
1
1
1
1
1
1
1
945
1
1
1
1
1
1
1
1
1
1
946
1
1
1
1
1
1
1
1
1
1
1
947
1
1
1
1
1
1
1
1
1
1
948
1
1
1
1
1
949
1
1
1
1
1
1
1
950
1
1
1
1
1
951
1
1
1
1
1
1
PHYLOGENY OF THE OPHIONINAE 173
Appendix 3 Shared derived character matrix for the 51 selected ophionines.
901 902 903 904 905 906 907 908 909 910 911 912 913 914 915 916 917 918 919 920 921 922 923
951 7 7 8 9 14 17 8 8 12 19 19 21 20 22 22 14 12 19 19 13 4 11 12
950 18 16 16 17 20 22 21 18 14 17 17 17 18 19 19 15 13 16 18 10 19 15 13
949 10 9 10 11 15 17 15 12 21 18 17 22 20 22 24 22 20 24 23 17 10 18 14
948 9 9 10 10 15 15 11 10 9 19 18 18 20 20 20 11 9 14 16 8 7 11 12
947 17 16 18 18 20 28 19 16 24 25 25 27 29 30 30 27 24 28 29 20 14 19 17
946 19 18 19 20 22 27 21 18 24 26 28 28 29 32 29 26 23 27 29 20 15 20 17
945 14 13 13 14 17 24 17 13 22 22 22 26 24 28 27 24 23 26 27 19 12 18 17
944 12 12 12 13 16 18 15 12 17 17 17 21 19 21 21 20 18 21 21 16 11 15 13
943 13 12 13 14 16 22 16 12 21 21 21 23 22 26 25 25 23 28 27 15 10 16 15
942 16 15 15 17 19 26 18 14 21 25 25 28 27 29 28 26 24 28 29 18 12 19 17
941 13 14 12 12 13 10 13 17 9 11 12 11 13 13 13 7 6 10 7 8 13 8 12
940 10 10 10 12 16 16 12 10 13 13 15 14 15 16 15 14 13 15 14 12 7 12 12
939 4 5 6 6 13 11 6 7 7 13 12 11 12 12 13 7 5 11 9 7 3 7 10
938 9 10 11 11 18 18 11 12 12 20 20 19 19 21 21 14 11 19 17 13 7 13 15
937 20 21 22 22 25 24 24 27 23 18 19 18 18 20 21 19 18 23 16 13 24 16 18
936 24 25 24 26 24 23 26 25 24 21 22 19 21 22 22 22 20 24 20 13 24 14 19
935 21 21 22 22 23 22 24 25 23 18 19 18 19 20 20 21 18 23 18 13 22 15 18
934 14 13 15 15 18 23 17 13 18 25 25 28 26 30 28 22 18 22 29 18 12 15 15
933 14 13 14 14 15 20 16 12 21 22 23 27 25 28 28 22 20 22 29 18 12 16 13
932 18 17 16 18 19 22 20 17 18 23 25 25 26 29 26 21 18 22 26 18 16 16 14
931 15 15 15 15 18 19 19 15 20 22 21 25 24 27 27 22 19 21 27 13 15 13 18
930 15 14 14 17 18 21 17 13 20 22 22 28 27 29 28 24 22 25 29 14 12 16 16
929 16 16 16 17 21 23 18 16 20 26 25 29 28 29 30 23 21 27 28 14 13 17 20
928 14 13 13 15 18 20 17 14 21 23 22 28 27 29 31 23 21 24 30 14 13 15 17
927 19 19 17 18 19 20 21 18 18 20 21- 20 22 23 21 18 17 18 21 11 16 13 15
926 15 14 13 14 15 21 17 15 17 18 19 21 20 22 21 20 18 21 22 11 14 14 16
925 17 16 16 18 20 20 19 16 22 21 21 20 23 23 23 23 22 24 21 12 14 17 16
924 16 15 14 16 18 22 18 14 18 21 20 23 24 24 23 23 21 23 23 13 13 16 16
923 13 15 16 15 19 21 14 16 16 17 16 17 18 18 19 19 14 22 15 10 12 12
922 13 11 12 12 16 17 14 15 16 13 14 18 18 18 19 16 18 21 17 12 12
921 27 26 25 23 15 15 30 27 18 12 12 12 14 15 15 16 16 16 14 9
920 11 11 12 12 12 13 12 11 15 14 15 16 16 19 18 15 13 16 16
919 16 15 16 17 17 22 18 13 21 26 26 30 28 33 32 27 24 28
.918 18 17 19 29 22 25 21 19 26 24 23 25 24 27 27 34 30
917 17 15 15 17 15 19 19 15 25 18 18 20 19 22 22 31
916 17 16 18 19 19 21 20 16 27 21 21 22 22 25 24
915 17 16 17 17 22 25 18 17 24 28 28 34 33 36
914 17 16 18 18 22 25 18 17 21 28 30 33 33
913 18 17 19 19 21 24 18 17 20 26 27 31
912 15 14 15 15 19 24 16 15 19 26 26
911 17 16 18 18 21 25 17 15 19 30
910 15 14 16 16 20 25 17 15 17
909 17 17 18 18 20 20 20 18
908 25 25 25 24 19 19 28
907 30 28 27 26 18 19
906 20 18 20 20 27
905 16 16 17 18
904 28 29 29
903 30 30
902 32
174
Appendix 3 - cont.
I. D. GAULD
924 925 926 927 928 929 930 931 932 933 934 935 936 937 938 939 940 941 942 943 944 945 946 947
951
13
11
12
13
18
20
18
15
17
18
18
12
14
12
19
13 14
950
16
16
16
18
17
17
17
17
20
15
18
20
21
19
13
9 15
949
17
18
15
14
23
22
22
20
20
22
21
16
16
15
17
10 16
948
14
15
11
15
15
18
14
16
15
13
15
12
14
11
18
12 11
947
27
28
25
20
26
25
26
24
28
29
30
19
20
18
18
9 16
946
29
30
27
23
26
26
27
24
32
29
32
22
26
21
21
9 19
945
24
23
23
18
26
25
26
23
26
28
28
17
18
16
16
9 15
944
21
19
19
14
19
20
20
18
23
22
24
15
16
16
13
6 11
943
23
23
24
16
24
24
24
21
24
27
26
15
18
15
15
9 13
942
29
28
27
21
28
27
28
23
30
29
31
17
22
17
19
9 17
941
8
11
8
12
9
12
8
10
13
6
7
13
15
14
12
9 9
940
16
15
14
16
14
15
15
14
15
13
14
14
17
13
15
10
939
9
9
6
10
11
13
9
10
9
8
9
8
11
9
14
938
15
16
12
17
16
19
15
15
18
14
17
15
18
15
937
15
16
16
19
16
19
16
15
18
15
16
31
32
936
21
22
20
24
21
22
21
19
21
17
18
30
935
16
17
16
20
16
18
17
17
18
16
17
934
23
22
22
19
26
25
27
25
30
29
933
21
19
22
17
26
25
26
24
26
932
23
24
22
21
24
25
26
23
931
21
21
21
21
27
27
26
930
23
21
20
21
30
29
929
23
22
21
21
29
928
22
22
20
20
927
21
20
18
926
25
22
925
27
948
949
950
951
17
18
13
950
14
16
949
13
15 14 18 15 13 16 19 18
8 22 19 17 25 22 23
1 15 13 9 14 17 16
8 34 31 24 31 35
19 10 37 30 26 31
7 31 28 24
5 26 23
6 31
PHYLOGENY OF THE OPHIONINAE 175
Appendix 4 Primary labels matrix derived from analysis of 5 1 selected ophionines . The figure in brackets
to the immediate right of the taxon number indicates the number of times that taxon alone is responsible
for the failure of a character in the LeQuesne test. Notably high values in any column suggest that a
particular character state shared by the respective characters is likely to be the result of parallel
development or reversal. For example, character 1-2 has high values for taxa 909 and 923; as both taxa
share an apomorphic 1 state, the labels matrix suggests this condition has been independently derived in
the two taxa, and thus should not be considered a synapomorphy uniting 909 + 923.
1.1 2
1.2
1 O
4.1 5.1 6.1 7.1 8.1 10 13 14.2 15.2 16
3 4.2 5.2 6.2 7.2 9 12 14.1 15.1 16.1
1 -9
on?
I o;
f 7) 1
1 1
qno
(11 )
i 1111 t;
ar\A
I fi) 1
?
one;
(m)
2 -- -- 1
one
1 4ft)
11 5 f\
on?
(\?\
11 1
908
909
Qin
(67) 1 3
(89) 3 11
( 6)
22 2 - 5 1 34 3 -- 2 1 2 2
2 13 7 2 -- -- 1 3 1 6
i
Q1 1
( c.\
o
Q14
( c.\
Q1 C
( 'i/i\ i i
91 11 11
Q1 fi
1 R) 1
_ 1 __ __ __ __ __ __ __ __ 1
Q1 7
(n)
qi Q
( 1Q) 9
4
Q1 Q
( ?)
1 1
O?0
1 Q\
1
Q?1
K6) 9
39. C 1 "3 9 , 1 9 .
922
Q9-3
(85) 4 1
(Ac.} t: i -J i
.... 12 6 1 1 2 34 1 1 5 2 11
1 14 11
Q?4
(n) i
1 ?
qpc
1 of)^ A
17 111
9?6
(14)
O 9 9 , 1
927
Q?R
(42) 1 1 --
(21)
10 1 1 3 1 1 1 1
911 1
ooq
( &}
q-jn
( 1}
1
q-n
1 ->\
1 1
932
(17)
2 -- 4 6 -- 1
Q'U
( 4)
4 1
QQC
1 1 q^ c o
i i 111
QTfi
f 1?) 9
1 2 - -- 4
0-37
(?a) i )
Q-30
(10)
Q-5Q
( 5)
Q4D
( fi}
941
(35)
2 3 1
94?
(46)
1 9 ? 1 1 ? . 1
943
(35)
414 3 -- 2
944
( 8)
1 1
QdK
(1?^ T.
59 ") i
946
(36)
1O9Qin ??
947
(12)
111 12'--
948
(16)
1 i i
949
(19) in
? 1
psn
(16)
.. __ ._ _- l -- -- -- 4 -- -- --
951
(46)
i i ? 2 1 2
iUNLABELLED TAXA: 912, 913, 933
176 I. D. GAULD
Appendix 4 - cont.
17 18.2 20.1 21.1 22.1 23 24.2 25.2 27.1 27.3 29.1 30
18.1 19 20.2 21.2 22.2 24.1 25.1 26 27.2 28.2 29.2 31
901 1 - -
902 6 1
903 - -
904 - - --
905 -- 4 - 2 1 3 - 1
906 2 - 1 2 1 5 6 6 1 11 1 1 8 1
907 -- 1 11 - - 1 1
908 ....*-. 7 2 3 3 1 1 8 1 2 2
909 1 1 4 - 3 2 2 1 -- 1 3 1 1 2
910 1 1
911 - - 1 1 - -
914 - - 1 1 -- - 1
915 4 1 1 1 1 1 1 1
916 - - --
917 - 1 - -- --
918 6 4 6 1 1 - 3
919 2
920 1 -- - 1 .... 2
921 1 - 2 1 4 11 1 1 2 1 2 3 -- 5
922 1 2 1 3 1 2 2 5 1 1 1 17 4 1
923 5 3 1 l 2 1 1 1 -- 3 10 4 1 --
924 - -1 1 12 1
925 1 2 12 1 --
926 - 2 1 1 1 - 5
927 1 1 1 2 3 - 1
928 -- - -- 2
929 - 1 - - 1 1
930 -- --
931 -- 1 - - -
932 - 1 - - 6
934 - - -
935 - i 1 _ l .. l .. 2 2
936 - -- - - - 1 -
937 2 2 1 10 1 1
938 1 .... l - - - i - l l
939 1 - -
940 1 1 - -
941 i ._ 1 3 e 6
942 - l 1 1 1 l 2 1 --
943 .. i ._ _ i ._ i i l ..
944 .. .. 2 -- - -
945 -- - 1 - - - - 1 -- -- -- -- -- -
946 l 2 1 -- - - 1 1 1
947 5 -- -- 5 - 2 -
948 - - - l .. - i - 15 - - - 1 - - -- --
949 .. 3 - i __ _. ._ _
950 - - 2 4 3 2 1 -- -- -- - 1 --
951 2 -- l - 1 15 1 1 1 2 -- 1 -
PHYLOGENY OF THE OPHIONINAE 177
Appendix 4 - cont.
32 35 36.2 37. 39 41 42.2 43.2 44.2 45.2 47.1 48.1
33 36.1 36.3 38 40 42.1 43.1 44.1 45.1 46 47.2
on?
1
qn-3 i
004
i
one
one i
o , c o
i i
on? i i
1 ?
i
QDR ? . 1
11-31
3 2
909 1 1 4 1 1 2
Qin
2 ...... 3 5 2 1 3
i
- 1 2 1
01 1
i
OTA 1
i
915 - - 1 1 1
qi e.
i
Q17 4 ? . 1
01 Q 1
__ __ _2 1
2
Q1Q
920 2
f.
2
921 4 9
4 11 i
1 1 R
922 1 1 3
923 3 1 6
q?4
1 9 4 i 2 1
1 1 i
1 1 11
1 4 ...... 3
i i
925 -- - - 3
926 4
...... 3 3 i i 6
? 1
4 __ __
927 2 1 1 2 1
928 - 1 l
929
1 1 1 1 1
- 1 ........ i i 2
-- 1 1 1
1
930 - 1
931
2 1 - -
i i
932 i
-3
3
9-54
1
935 i 1
1 1
935 i i 4
111
937 i 2 3 -
1 1 C 1
938 -- -- -- -- 3 -- 2
2
939 1
C
940 1
941 1 1 11 1 - 12
942 - 1 1 - - 3 -- 1
943 1 2 17
944 1 1 1
945
__ 2 3
1 2 - 1 2 1
5 i
i i
o
1 1
1111
- 6 -- 2
i 5
1 - __
946 1 1 1
947 .- __ _. ._ _. __
- 2 3 1
? ?
94fi
111
9 I 1 1
949 __ _. __ __ ._
o
1 ?
950 -- -- -- -_ -_ _-
1
951 - 3 -- - - 1 - 2
4 i - i i i -
1 1 3 2 -- --
178 I. D. GAULD
Appendix 4 - cont.
48.2 51.1 52.1 54 56 58.1 59 61.1 62 64
50 51.2 52.2 55 57 58.2 60 61.2 63
901 -- -
902 -
903 1 7 -- --
904 - 1 3 1 --
905 - -- - -- -
906 6 3 5 - 1
907 -- - - - - - 1 --
908 1 1 2 2 2 - 2
909 31 1 30 7 1 - 1 3 1 2
910 1 - 2 5
911 1 - 1 5
914 - - 6 -- -
915 i .... 3 30 - 1 1 1
916 i i 4 -
917 __ __ __ __ 3 4 __ __ __ __ __ ._ i __ 4 __
918 - 2 2 1
919
920 - -- 1 1 1
921 1 11 - - --
922 - 6 5 1 - 1 1 1 1
923 1 - . 1 4 - 1 --
924 1 -
925 6 - 1 1
926 -- - 2 -
927 31 1 2 1 1 1 2 2
928 1 - 2 1 19 1 1
929 - 4 -
930 -
931 -- 1
932 2 - 1 - 1
934 i .. __ __ __ 1 _
935 __ ._ ._ 2 - 3 -
936 - -- 2 -- 3 -- -- --
937 - 3 2 1 -
938 7 - 1
939 - 1 1
940 2 1 - - -
941 4 1 4 1 3 - 1
942 2 -- 1 1 1 27 1 19 2 2 -- 1
943 -.-... i e 7 - -- -- 2 -- 1
944 - 1 -
945 -- -- - 5 2
946 2 -- - -- 1 - - 2 2 --
947 1 - 1 -
948 -- 1 1 -- - 1
949 2 2 - 7 - 1 1 2
950 - 3 2 - 3 3 - - - --
951 1 -- 2 1 27 1 1 1 1 1 1 --
PHYLOGENY OF THE OPHIONINAE
179
CO LT) CO
i i ro
I CM i-H CO LD CO
i * CO CM CO t t CM CM
I CM IO
CM I CO
I I-H <>
CO LO CO
-H I CM .-H
IT) I-H CM
CO I 3 CM CXl T-H
r^
*" *VI
CO CM CM
CM i i co in co
CT> I CT> I
I CM I
^H I CO
CM CO CM . <
ro CM I-H i CM
ro i ro I-H , i
i cr> I-H i ,-H
i CM fi ro ID co
CM CM ro
^4 00
CM I-H co in co
CM .-H ro LO co
u
o
I
3
53
oo j^j
i< OO i < i <
I
O
I
o
o
(U
t-H f-1 C\J 00
00 r-l I-H ^H
< i <i ro
TD
_X
rt
^- r-l l-t
U
I
-J
I I I I
PHYLOGENY OF THE OPHIONINAE 181
Appendix 6 - cont.
40
43.
1
44.
1 45.1 46
47,
,2
48.2 52.2 56 59
61.1 62
42.2
43
.2
44.2 45.2
47.1
48.
1 52.1 55 57 60
61.2
64
909
QI n
1
12 10
4
2
11 3 2
1
4
1 5 3 28 - 2
1
1 2
e
c
Ql 1
O
i
1
3
c
Ql 7
3,
1
1
Ql 4
j
1
g
915
1 1
2
2
6 - 1
__ __
1
3 ._ 28
1 1 1
--
916
1
1
i
1
.. .. i ..
1
1
1 1 18 1
1 __
Ql 1
i 7
1
1
? T
918
Ql Q
1
8 -
1
5
4
1
52
7
1
2
1 2 2 1 3
1
3 3 1
1 1
Q9A.
1
1
i
1 2
i
i
1
qoc
1 1 R
1
in i ?
c
A
926
1
2
1
7 1 1
4
1 1 -
2 2 1
928
1
-
1
1
3 - - 1
1
1
1 2 17
1 1
-
qpq
1
1 1 --
I
4 1
1
2
1
i i
931
2
1
n 1
3
1 .... 3 ..
1 1 -
QQ9
2
222
? 4
i i
Q"3A
1
1 ?
942
1
1
1
1 1
1
1
1 .... i .. _. -. 17
1 1
943
4
3
2
15 1
1
5 6 1 -
.. i ..
944
3
2 1
1
1
1 2
15 1
3
__ i i .... 3 .. -
- 5
945
4
1
1
-. .. 2
1
2
-- 1 1 12 -
1 2
946
1
i i
1
1
1 11
1
1
1 11
947
1
1
3
3
1
4 i !____.. i __
1 1 5
182
I. D. GAULD
CM i-HOi-H.-HOOi-lr-l
o o o
CM O O O
o
o o o
CM
co
i-H O O O
o o o
r~.
r-l O O O
i-H O i-H
I-H O
o o
o o
o o
i-l O
o o
o o
o o o
o o o
O r-l ,-H
O .-H r-H
-4 O O
o o o
o o o
o o o
OOOOO.-HOO
OOr-HOOi-HOO
o o
o o
-H O
o o o
o o o
o o o
o o o
o o o
o o o
o o o
o o o
o o
o o
o o
O 1-1
o o
o o
I-H I-H I-H o o
O O O O .-H
O O O
o o o
o o o
o o o
o o o
o o o
,-H O .-I
o o o
o o o
o o o
O <-i
o o
-I O
1-H O
o o
o o
o o
O r*
O i-H ^H
O ^H rt
o o o
o o o
o o o
o o o
o o o
o o o
r-l O O
o o o
I-H O O
O ^-1 O
^-i O O
o o o
1-H O O
o o o
O O ^H
o o o
o o o
o o o
^H O O
o o o
o o o
tr oooooooo
OOOOOOi-H-H
CO OOOt-HOi tOO
tx
oooooooo
CM
IO
CO Or-Hl I I lOOr-Hi I
Or-HOOOOOO
10 ,-H ,-1 ^H
VO O O O
O ^H 1-H ^H
10 O O
1-H 11 1-H If)
o o
o o
o o
o o
CO i-HOt-HOOOOO
CO
IT) O O O O O O
OO
O O O O <-< I-H
O 1-H
O ^H
o o
r-< O
o o
o o
oooooooo
r
i lOi li lt-HOi-H> I IT) i li-Hi Ir-lOOOr-l
IQ
OOOOOi-HOO to OOOOOOOO
uo ,
OOOOOOOO 10 OOi-HOOOOO
a-
OOOOOOOO 10 OOOOr-H^HOO
Oi-i.-H^H.-!O.-i.-i 10 OOOOOOOO
CM
CM
OOOOOOOO tO OOOi-HOOOO
CM
1-Hi-Hi li li It-Hi-Hi-H IO Ol lOl-HOOOO
CM
OOOOOOOO 10 ^H^HrtrH^H^Hrt^H
o
i-Hi-HOOOOOO to OOOOOOOO
en
i-H i-H O CO r-4 i-H O CO ^f i-H i-H i-H t I i-H f I rH i-H
CM
oo
OOOOOOOO <* Oi-HOOOOOO
co
1-H 1-H 1-H I-H i-H 1-H 1-H 1-H ^" 1 * 1 I O CO CO O 1 I r 1
CM
T--
i-Hi-Hi-Hi-Hi-Hi-Hi-H^H 3" OOOi-HOOOO
r-
OOOOl-Hl-HOO ^- i-Hi-Hi-Hi-Hi-Hi-Hi-l!-!
IO
-HOOOi-Hi-lOO ^J" OOOOOOi-Hi-H
CM
to
OOOOr-HOOO S" i-HOOOOOi-Hi-i
1-H
tn
OOOOi i i i O O ^ i if-Hi-Hi-HOi i t i i i
Oi iCMco^rtor^oo
to to to to to to to to
cncncncncncncncn
i
a
O
ITi
ON
ON"
O i-l t-l
i-H I-H ur>
-H CM r-H
Vf) ^H ^H ,_(
00
s
<N
ON
<N
ON
i CM in <-i t-i
CO OJ CM i-H CM
VD CM r-l
t-H IT) i-l
T3
8
"S.
OJ
"cl
c
<u
X
5
03
CM i-H i-l
184
I. D. GAULD
C\J I CNJ CvJ
vo c\j co in co
CM .-H i-H
,-H CM CO I-H
evj c>o t-i
f i n cvi
oo I-H ro co
CM ,-H r-t i-H
I-H I-H CM
i I CM i I i I
.2
o
i
CO CTl O i I
PHYLOGENY OF THE OPHIONINAE
Appendix 9 Labels matrix derived from the Euryophion primary data matrix (Table 20).
1.1
19
30
37
54
59
71
73
185
75
8.1
13
27.2
36.!
>
43.;
I
57 70
72 74
oni
, fi >
_
A
1
1 ?
o
CU1
202
(36)
1 2 10
7
8 5 1
t
1
3
X
1
1
1 L.
2 2 1
c
2 6 3 12
3
203
(24)
6 10
1
1 5 3
4
2
3
3
62
1 1
204
(18)
2 2 1
1
1 2
6
1
2
2 1 2
1 12
ync.
(11)
3
- 1 4
2
3
4
5 ?
KV3
206
\ * A /
(10)
__
2
1
1
6
242
2
?n?
I n)
t.VJ /
208
\ /
(22)
1 1
5
821
_ _
1
1
1 11
3231
3
Index to generic names
Synonyms are in italics.
Abanchogastra 156
Dispilus 160
Pachyprotoma 124
Afrophion 123
Pamophion 154
Agathophiona 123
Enicospilus 160
Paniscus 124
Aglaophion 135
Eremotyloides 160
Platophion 124
Allocamptus 160
Eremotylus 130
Pleuroneurophion 160
Alophophion 124
Eurycamptus 136
Potophion 124
Amesospilus 160
Euryophion 136
Prethophion 143
Apatophion 124
Primophion 136
Apomesus 124
Genophion 131
Psylonychia 124
Athyreodon 141
Pterospilus 160
Atoponeura 160
Hellwigiella 132
Pycnophion 162
Aulophion 146
Henicospilus 160
Australophion 124
Hybopleurax 135
Rhopalophion 126
Hypselogastrina 135
Rhynchophion 140
Banchogastra 156
Rictophion 136
Barycephalus 135
Janzophion 128
Riekophion 129
Barytatocephalus 135
Boethoneura 131
Brachyscenia 144
Camptoneura 131
Camptoneuroides 131
Ceratospilus 160
Chilophion 131
Chlorophion 131
Clistorapha 131
Laticoleus 151
Lepiscelus 145
Leptophion 152
Macrophion 141
Mecetron 124
Neophion 124
Nipponophion 146
Schizospilus 160
Sclerophion 127
Sicophion 130
Simophion 144
Spilophion 152
Stauropoctonus 146
Stauropodoctonus 146
Stenophthalmus 124
Coiloneura 152
Coracophion 135
Ofeter 141
Thoracophion 136
Cryptocamptus 160
Ophiogastrella 144
Thyreodon 141
Cymatoneura 160
Ophiomorpha 160
Tipulophion 141
Ophion 124
Trophophion 133
Dicamptus 159
Ophionopsis 135
Dictyonotus 135
Orientospilus 142
Xylophion 127
British Museum (Natural History)
An introduction to the Ichneumonidae of Australia
/. D. Gauld
In the important field of biological and integrated control of pests the parasitic Hymenoptera
are of particular significance, and this work considers one of the largest families of Parasitica,
the Ichneumonidae. The group has received little attention in Australia - though it has already
been utilized successfully in curtailing the ravages caused by accidentally introduced pests. For
selective control programmes to be effective, however, a sound knowledge of the biology of
both the pest and its parasites is essential - and a sound taxonomic base is vital for the
development of such knowledge.
Ironically, considering the group's economic importance, the parasitic Hymenoptera is
amongst the least studied of any group of living organisms, and taxonomic difficulties have
presented major problems to many entomologists working with the Parasitica. An
Introduction to the Ichneumonidae of Australia will go a long way towards rectifying this
situation, being a taxonomic treatment, by genus, of the Australian ichneumonids, a
comprehensive illustrated identification guide, and a summary of all available information on
the group. It will also serve as an introduction to the biology and distribution of Australian
ichneumonids, and provide a check-list of the described species and an index to their known
hosts. It provides an important revision of ichneumonid nomenclature in order to bring the
group into line with the generally accepted principles of zoological nomenclature.
1984, 413pp, 3 maps, 580 figs. Paperback. 565 00896 X 40.00
Titles to be published in Volume 51
The ichneumon-fly genus Banchus (Hymenoptera) in the Old World
By M. G. Fitton
The phylogeny, classification and evolution of parasitic wasps of the subfamily Ophioninae
(Ichneumonidae)
By I. D. Gauld
A cladistic analysis and classification of trichodectid mammal lice (Phthiraptera: Ischnocera)
By C. H. C. Lyal
The British and some other European Eriococcidae (Homoptera: Coccoidea)
By D. J. Williams
Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk
Printed in Great Britain by Henry Ling Ltd, Dorchester
BRITISH MUSEUM
NATURAL HISTORY)
Bulletin of the 26AUGI987
PRESENTED
British Museum (Natural History)
A cladistic analysis and classification of
trichodectid mammal lice
(Phthiraptera: Ischnocera)
C. H. C. Lyal
Entomology series
VolSl No 3 31 October 1985
The Bulletin of the British Museum (Natural History), instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.
Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers are
works of reference that will remain indispensable for years to come.
Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to:
Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW75BD,
England.
World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.)
Trustees of the British Museum (Natural History), 1985
The Entomology series is produced under the general editorship of the
Keeper of Entomology: Laurence A. Mound
Assistant Editor: W. Gerald Tremewan
ISBN 565 06013 9
ISSN 0524-6431 Entomology series
Vol 51 No 3 pp 187-346
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 31 October 1985
A cladistic analysis and classification of trichodectid
mammal lice (Phthiraptera: Ischnocera)
2v
C. H. C. Lyal
Department of Entomology, British Museum (Natural History), Cromwell Road, London
SW7 5BD
Contents
Synopsis 187
Introduction 187
Acknowledgements 188
Method of systematic analysis 188
Relationship of systematic analysis and classification 189
Morphology 189
Character survey 203
Character analysis 213
Identification of apomorphic states 213
Cladistic analysis 223
Taxonomic history of Trichodectidae 234
Proposed classification 241
Descriptions of genera and subgenera 247
Bovicolinae Keler 247
Eutrichophilinae Keler 265
Dasyonyginae Keler 267
Trichodectinae Kellogg 286
Neotrichodectinae subfam. n 321
Keys to Trichodectidae 334
Key to subfamilies 335
Key to genera and subgenera 335
References 340
Index 344
Synopsis
Observations on phthirapteran morphology are analysed with particular reference to the Trichodectidae.
Problems of structure, homology and nomenclature associated with major morphological features are
briefly reviewed and solutions suggested where possible; where appropriate, the morphological terminolo-
gy is clarified. A cladistic analysis of the 351 species and subspecies of Trichodectidae is carried out using
the states of 187 characters, and the taxa are re-classified in accordance with the results and the principles of
phyletic sequencing. Twenty genera are accepted in five subfamilies. Ten of the genera are divided into 27
subgenera. This necessitates the description of one new subfamily, three new genera and four new
subgenera. Three genera are newly synonymised, eight genera and subgenera are recalled from
synonomy, and four genera are reduced to subgenera. The generic placements of 106 species and
subspecies are changed. Keys to the subfamilies, genera and subgenera are provided, and the genera and
subgenera are described.
Introduction
The Trichodectidae is a family of ischnoceran chewing lice parasitic on mammals. There are 351
described species and subspecies grouped into between 13 and 39 genera, various workers
having widely differing views on generic limits. The classification of the Trichodectidae at the
generic level is perhaps more confused than that of any other group of lice (Hopkins, 1949, 1960;
Emerson & Price, 1981) and no workable keys to genera are available. The confusion and
Bull. Br. Mus. nat. Hist. (Ent.) 51 (3): 187-346 Issued 31 October 1985
188 C. H. C. LYAL
disagreement surrounding the classification of the family persists despite a sound basis of
taxonomic knowledge at the species level, derived largely from the works of Werneck (1948,
1950), although this author did not attempt to produce any keys.
The purpose of this study is to present a classification of the species of Trichodectidae and to
describe and provide a key to the recognised genera and subgenera. The foundation of the
classification proposed below is a cladistic analysis, and for the purpose of determining character
polarity in that analysis a suitable out-group had to be identified first. Ideally this out-group
should be the sister-group of the taxon studied, and in this case would be expected to be a group
in the same suborder, the Ischnocera. However, the holophyly of the Ischnocera has not been
demonstrated (Lyal, 1985), although there is evidence that indicates the holophyly of a group
comprising most of the Ischnocera (including the Trichodectidae). The classification of the
Ischnocera is unsettled, and the sister-group of the Trichodectidae has not been identified. A
single teneral male specimen of a possible sister-group has been seen, purportedly collected
from a corvid in Austria. The specimen has the single tarsal claws of a mammal parasite, but no
other apomorphies of the Trichodectidae, although it has the same general facies as members of
the family. As out-group, therefore, the Ischnocera s.l. is used, reference being made where
appropriate to the other three suborders of Phthiraptera (Lyal, 1985).
The term 'holophyletic' is used below to refer to groups of taxa comprising a single ancestral
species (known or inferred) and all of its descendants. The more familiar term 'monophyletic' is
avoided because ambiguities in its use have caused misunderstandings in the resolution of
systematic problems.
Scale lines equivalent to 50 /am are given in most figures.
This study formed part of the results of a research project submitted to the University of
London for the degree of Doctor of Philosophy.
Acknowledgements
I am very grateful to all those who have given advice and read parts of this paper whilst in preparation,
particularly Dr R. G. Davies, University of London; Dr L. A. Mound, Dr T. Clay, Mr W. R. Dolling, Mr
D. Hollis and Mrs J. M. Palmer, British Museum (Natural History); and Dr B. Heming, University of
Alberta. I also thank the following people who were of great assistance during part of this study which was
carried out in the United States: Dr K. C. Emerson, Florida; Dr K. C. Kim, Pennsylvania State University;
Dr R. D. Price, University of Minnesota; Dr O. Flint, United States National Museum; Dr R. Traub and
Miss H. Starke, University of Maryland School of Medicine; and Dr W. A. Drew and Dr D. Peters,
Oklahoma State University.
Method of systematic analysis
Systematic analysis is carried out in order to determine 'relationships' between different taxa,
the relationships discovered being employed to study some aspect of the biology or evolution of
the taxa and/or to provide a framework for their classification. Taxa may be deemed to be
'related' according to a number of different criteria, so the type of analysis performed is dictated
by the requirements of other studies to be undertaken. In this case the systematic analysis is
preliminary to an examination of the phylogenetic linkages between Trichodectidae and their
hosts (Lyal, in prep. , a). The relationships required from the analysis are therefore phylogenetic
in nature. Although it has been suggested that phylogenetic relationships can be determined by
phenetic analysis (Sneath & Sokal, 1973), this is generally believed not to be the case, and a
phyletic analysis is required (Wiley, 1981). In this study a manual cladistic analysis was deemed
most suitable as computerized methods of sufficient power and capacity were unavailable when
the study was initiated (Lyal, 1983).
The first stage in the analysis was a character survey of all species of the family, with the aim of
discovering suitable apomorphies. As analysis proceeded some characters initially selected were
rejected, if the polarity of their states could not be determined, or if their apomorphic states
were restricted to single species.
The observations were recorded in a data matrix, but it rapidly became apparent that this
TRICHODECTID MAMMAL LICE 189
matrix was too large to be manipulated conveniently. Preliminary inspection of the taxa had led
to the recognition of sufficient apomorphies to divide the family into five or six plausible
holophyletic groups, so secondary data matrices were compiled for each of these. The smaller
number of species and apomorphies then necessary on each of these matrices enabled manual
analysis to proceed. These matrices were inclusive rather than exclusive, and taxa with low
probabilities of membership were included; as a consequence some taxa were included in more
than one matrix. The matrices were examined and taxa or holophyletic groups sharing the
greatest number of synapomorphies were united as sister-groups. This process was continued
until the cladistic relationships within each matrix were resolved as fully as possible. The full
matrix was then compiled using the largest holophyletic groups identified in the analysis of the
secondary matrices. This matrix was then analysed using the same technique and the full
cladogram drawn. The cladogram, which gave the full distribution of apomorphies, was
examined to ensure that the most parsimonious distribution of apomorphies had been achieved.
Character states were weighted by a simple 'gain' or 'loss' criterion. 'Loss' states were only
used in the analysis where 'gain' states were not available, and after 'gain' states had been used to
resolve the cladogram as fully as possible. It follows that, whereas the distribution of 'gain' states
should be maximally parsimonious over the cladogram, this is unlikely to be the case for 'loss'
states.
Relationship of systematic analysis and classification
Biological classifications afe designed not only to enable the taxa classified to be located in the
literature and in collections, but also to store information about those taxa. In many cases the
nature of this information is not clear from the classification, and taxa are grouped together on
the basis of some ill-defined combination of phenetic and phyletic relationships. The type of
information stored in the classification (i.e. the type of relationship used in its construction)
may, however, be indicated precisely, and the classification used as an efficient data-retrieval
system. This can only be done by utilisation of cladistic information alone; attempts to
incorporate phenetic or anagenetic information lower the retrieval facility (Cracraft, 1974).
The method whereby the classification was constructed from the results of the cladistic
analysis in this study was by using both rank and sequence in which taxa of equal rank are listed
(G. Nelson, 1972, 1974; Cracraft, 1974; Wiley, 1979, 1981). This process has been termed
phyletic sequencing (Cracraft, 1974). The convention employed is that within a classification
holophyletic taxa of equal rank are listed ('sequenced') so that each taxon is the sister-group of
all those taxa of the same rank (and within the same taxon of immediately higher rank) listed
below it in the classification (modified from Cracraft, 1974). Use of this criterion allows recovery
of a dichotomously-branching cladogram from the classification. However, holophyletic groups
with a trichotomous or polychotomous interrelationship must also be accommodated. Wiley
(1979) recommends that these be placed in the classification with equal rank and be noted as
'sedis mutabilis' to indicate that their order is unknown or interchangeable. It must be
remembered that this convention is for a formal classification, not for regular use in discussion.
All hierarchical groups may be sequenced, but it is probably more efficient to employ informal
groups below the subgenus level, especially with larger genera.
Morphology
The discussion below is intended to clarify morphological terms used in this study, to point out
structures of taxonomic or systematic interest, and to describe briefly the morphology of the
insects.
Head
Figures 1 and 2 depict the dorsal and ventral aspects of a generalised trichodectid head with the
regions and major features labelled with the terms used in this study.
The internal carinae of the head are strongly developed in many species, and differences in the
190
C. H. C. LYAL
dorsal preantennal
sulcus
median broadening of
clypeal marginal carina
clypeo-f rental sulcus
pulvinus
clypeal marginal carina
? frontal sulcus
f lagellum
eye
occipital ring
osculum
clypeus
clypeus
sub-genal area
genal area
temple
f rons
gula
Figs 1, 2 Terminology applied to trichodectid head, modified slightly from Symmons (1952). Mandible,
maxilla, hypopharynx and labium omitted. 1, structures of head ($); 2, regions of head ($).
degree of sclerotisation and position are useful taxonomically, particularly at the species level.
The postoccipital sulcus is marked by a greatly-developed internal carina ('occipital ring')
projecting posteriorly into the thorax; this structure is present in many other lice, but rarely so
well developed.
The postero-lateral margins of the head ('temple margins') are more or less rounded and may
be produced at the postero-lateral angles in some species. The lengths of the setae on the
postero-lateral angles may provide characters of taxonomic value. The eyes are positioned
adjacent to the antennae or more posteriorly; in some species (e.g. those in the genus
Cebidicold) the eyes occupy lateral projections.
Anterior to the antennal socket the margin of the head is produced into a variable sclerotised
conus (Fig. 1), but the trabecula (Clay, 1946) is absent. A sclerotised conical projection from the
dorsal nodus of the clypeo-frontal sulcus may be present projecting parallel to the margin of the
antennal socket. This projection is frequently present in only one sex of a species.
TRICHODECTID MAMMAL LICE 191
The anterior margin of the head may, in plan, form an unbroken arc ('circumfasciate head'),
but in most Trichodectidae it is interrupted medially by the pulvinus or the osculum (Symmons,
1952). The pulvinus is a thick bilobed pad of unsclerotised tissue developed from the clypeo-
labral suture, and is found in all Ischnocera (Fig. 1). The osculum is a median indentation of the
anterior margin of the head (Fig. 2). When the insect is at rest the pulvinus and osculum have
been observed to be applied to the hair of the host (see discussion of mandibular morphology
below). The width of the osculum is therefore probably influenced by the diameter of the hair in
the region of the host's body inhabited by the louse. In cases where the pulvinus fails to reach the
anterior margin of the head a small ventral sulcus ('ventral preantennal sulcus' of Clay, 1951)
may connect it to the margin or osculum. Many species have a small dorsal preantennal sulcus
(Clay, 1951) also interrupting the anterior margin medially.
Between the antennae and the osculum the margin of the head may, in plan, be shallowly or
broadly convex, sinuate or straight (Figs 3-6); sclerotisation along the margin (forming the
'clypeal marginal carina') may be minimal or heavy, but is usually pronounced in the median
dorsal region (Fig. 1). Much of the variation is possibly influenced by the density and texture of
the hair of the host.
Only two structures of the mouthparts need be mentioned, the mandibles and the cibarial
sclerite.
The mandibles of Trichodectidae, like those of many other Amblycera and Ischnocera, are
asymmetric. There are three apical teeth on the right mandible and two on the left; on the right
the centre tooth is generally longest, whilst on the left the posterior tooth is normally the better
developed (Figs 7-10). Mandibular asymmetry is considered by Snodgrass (1935) to be a feature
of insects that masticate their food, and the lack of asymmetry in the mandibles of some species
oiRidnus (Amblycera) is associated by B. Nelson (1972) with blood-feeding in those species.
Trichodectes canis, the only trichodectid known to take blood meals (Bouvier, 1945), has
dimorphic mandibles, as do all other Trichodectidae.
In addition to the collection and preparation of food the mandibles of Trichodectidae are
employed in anchoring the insect. When at rest, Trichodectes canis and T. metis have been
observed to enclose a hair in the pulvinus with the mandibles and whilst in this position may
completely release the grip of the tarsal claws and straighten the legs laterally from the body.
The insect is then held on only by the mandibles. In addition to the species above, specimens of
Felicola sp. collected from dried museum skins have been found in this position, and specimens
of most genera have been found preserved in alcohol clinging to single hairs by their mandibles
alone. The selective pressures on the mandibles of Trichodectidae are different, therefore, from
those operating on lice that do not have a mandibular anchoring mechanism, and possibly
precludes the assumption of monomorphy in blood-feeding species. The 'interior' face of the
right or of both mandibles may be ridged, so that when the mandibles are folded closed, the
ridges on one mandible are not covered by the other, and all or most of the exposed mandibular
area is ridged. The ridges are therefore not positioned in such a way that they can act against one
another, and it is suggested that they are developed to prevent the mandibles slipping on a hair
when clasping it (Figs 8-10). When the mandibles are closed, the right is always interior to the
left, which may have a concavity to receive it (Figs 7, 9, 10); the left mandible is not then kept by
the right from contact with the hair, and the greatest possible mandibular area is utilised. In
some species the right mandible has a basal notch on the anterior margin, which receives the tip
of the left mandible (Figs 7, 10); this feature may act in concert with the restraint of the right
mandible to 'lock' the mandibles closed about a hair. Although, as said above, none of the
Trichodectidae have monomorphic mandibles, the mandibles of Damalinia (T.) conectens (Fig.
7) are very slender and pointed, and the centre tooth of the right mandible of Dasyonyx spp. and
Eurytrichodectes spp. (Fig. 9) is long and pointed, suggesting in each case a piercing function and
thus possible haemophagy . In contrast, the mandibles of Damalinia (T.) indica (Fig. 8) are blunt
and broad, suggesting an adaptation to grinding and chewing.
The sitophore sclerite (Figs 11, 12) varies considerably in the Psocodea (Cummings, 1913;
Haub, 1967, 1972, 1973, 1977). The form of the sclerite, particularly of its posterior border, is
obscured in slide-mounted specimens, and differential inclination of the specimens may change
192
C. H. C. LYAL
Figs 3-10 Variation in trichodectid head and mouthpart structure. 3-6, $ head, dorsal aspect, of (3)
Bovicola caprae; (4) Damalinia lineata; (5) Felicola subrostratus; (6) Trichodectes zorillae. 7-10,
mandibles, dorsal aspect, of (7) Damalinia conectens; (8) Damalinia indica; (9) Eurytrichodectes
paradoxus, showing mandibles interlocking; (10) Trichodectes canis, with detail of postero-dorsal
margin of left mandible.
TRICHODECTID MAMMAL LICE
193
14
13
12
11
Figs 11-16 Trichodectid morphology. 1 1 , Damalinia neotheileri, sitophore sclerite. 12, Bovicola hemitra-
gi, sitophore sclerite. 13, Damalinia indica, right antenna of d", dorsal, arrow indicating prolongation of
antero-ventral pedicel margin. 14-16, lateral aspect of metatarsal claws of (14) Dasyonyx (D.)
smallwoodae; (15) Dasyonyx (N.) diacanthus; (16) Eurytrichodectesparadoxus.
the apparent dimensions (Haub, 1977). Trichodectidae differ from other Psocodea in having an
open posterior border to the sclerite, a character state identified by Haub (1973) as plesiomor-
phic, but which is almost certainly apomorphic. Within the family the sclerite is present in all
species and the form is fairly constant (Fig. 11), although in two species groups the posterior
arms are extended laterally (Fig. 12) and the sclerite is difficult to see. The minor variation in the
form of the sclerite at the species level is so far unexploited in taxonomic study.
The sitophore sclerite is cup-shaped, and is opposed on the dorsal surface of the sitophore by a
closely-fitting projection, the epipharyngeal crest. These two structures have been identified as a
'mortar and pestle' for grinding food (Weber, 1936) and as components of a salivary pump
194 C. H. C. LYAL
(Buckup, 1959; Keler, 1966). Rudolph (19820, 19826, 1983) demonstrates that it is a pump,
sucking water absorbed from the atmosphere onto the lingual sclerites through the Y-shaped
duct linking these and the sitophore sclerite.
The antennae are made up of a scape, pedicel and flagellum. In common with other
Thysanura-Pterygota the pedicel and flagellum are annulations formed by subdivision of a single
segment (Imms, 1939; Matsuda, 1965). Consequently, the only intrinsic muscles of the antennae
are in the scape and insert on the pedicel, moving the pedicel and flagellum as a unit; the scape
itself is moved by muscles inserted on its basal margin and arising on the anterior arms of the
tentorium. The positions of the muscle insertions on the scape and pedicel are such that the an-
tenna can be moved through 360 degrees, but there is no muscular mechanism for flexing the
flagellum. Whilst the prime function of the antenna is sensory, secondary modification in the
male has taken place in some Ischnocera (including almost all Trichodectidae) and some
Anoplura, the antennae being used to clasp the female round the abdomen during copulation
(Keler, 19380; Sikora & Eichler, 1941; Lyal, in prep., b). This development has led to an
increase in length and degree of sclerotisation of the pedicel and flagellomeres and an increase in
strength of the antennal muscles. The intrinsic muscles of the scape are enlarged and the scape
expanded; the extrinsic muscles are also enlarged and the head concomitantly broader in the
male than in the female. To facilitate free movement of the flagellum the joint between the scape
and the pedicel is broad and membranous (other than at the two articulatory points). However,
because there is no direct muscular control of the flagellum, the degree of free movement
relative to the pedicel must be limited, so that control may be effected by the intrinsic muscles of
the scape. Observations made in this study reveal that the apex of the pedicel is angled relative to
the long axis of the annulation, the longest margin of the pedicel being the antero-ventral (Fig.
13). There is very little membrane between the pedicel and the flagellum on the antero-ventral
margin, but more on the postero-dorsal, so some flexibility between the pedicel and flagellum is
possible, although limited in the anterior direction by the prolongation of the pedicel. During
copulation the male clasps the female around the top of the abdomen from underneath
(Werneck, 1936; Sikora & Eichler, 1941), the pedicel and flagellum being curved to match the
curvature of the abdomen. The antennae are raised above the head of the male and the intrinsic
adductor muscles contracted. The pedicel and flagellum are thus brought down over the
abdomen of the female, the flagellum being constrained by the form of its junction with the
pedicel. The form of the pedicel-flagellum joint is such that, should the haemolymph pressure be
reduced, the joint membrane would collapse and further contract the flagellum against the
female. However, lice are not known to have control over haemolymph pressure in their
antennae, although larvae of Lepidoptera are known to control their antennal movement
partially by this means (Matsuda, 1965). The mechanical strength and degree of possible control
of a system involving joints as described above is likely to be inversely proportional to the
number of joints in succession, with a single joint being the most efficient. The degree of
curvature attainable using three segments is sufficient to grasp the female. For these reasons
only the first flagellomere is required to take on a clasping function; the apical two flagellomeres
may be retained in a sensory capacity (most Ischnocera) or lost (Trichodectidae). In
Trichodectidae the last two flagellomeres have contracted and fused to the first flagellomere in
all males (although a small semicircular sclerotisation, probably representing the apical flagello-
mere, is present in the male of Eurytrichodectes paradoxus) and in females of the Neo-
trichodectinae, Trichodectinae, Bovicolinae and most Eutrichophilinae. The sensilla of the two
apical flagellomeres are retained on the remaining flagellomere.
The firmness with which the female is held may be increased by projections in the form of
spikes or denticles on the antennae, particularly on the flagellum. Trichodectidae are character-
ised by the possession of two modified setae apically on the male flagellum, which take the form
of sharp, stout teeth (Fig. 13).
Trichodectidae have two sensilla coeloconica and three sensilla placodea on the terminal
flagellomere, sometimes closely associated (Clay, 1970; Kim & Ludwig, 1978). In Lorisicola
malaysianus and L. mjoebergi the sensilla are in pits with tongue-like projections around them.
TRICHODECTID MAMMAL LICE
195
Thorax
Figures 17 and 18 depict a generalised trichodectid thorax, labelled with the terms used in this
study. In all Trichodectidae the rhombic sclerite anterior to the pronotum is present; the
pronotum itself generally bears two lateral sclerites, sometimes narrowly joined medially. The
mesonotum and metanotum are fused, with prominent pleural ridges extending onto them;
there may, as with the pronotum, be a medial zone of desclerotisation. Contrary to the assertion
of Mayer (1954), the lateral cervical sclerite is always present and bears two anterior setae. The
pronotum is fused to the propleuron, which in turn is fused to the prosternum. The prosternum
may extend anteriorly between the fore-coxae and be unsclerotised medially. The mesosternite,
fused to the prosternite, may also be medially divided. The mesepisternum is difficult to delimit,
but rarely extends unbroken between the metasternite and the metanotum. The metasternite, if
present, is only rarely fused to the mesosternite, and is never sclerotised medially. Posterior to
rhombic sclerite
prothorax
pterothorax
abdominal
pleurum I
prosternum
lateral cervical
sclerite
mesothoracic
spiracle
profurcal pit
mesofurcal pit
mesosternum
metasternum
postcoxale
met epimeron
Figs 17, 18 Diagrammatic representations of generalised trichodectid thorax, with terms used in this
study. 17, dorsal. 18, ventral.
196 C. H. C. LYAL
the metacoxa may be a semicircular sclerite ('postcoxale' of Matsuda, 1970). In Procavicola
(Meganarionoides) this is very heavily sclerotised and fused to abdominal pleurum II; the two
postcoxales may also fuse medially.
One, two or three setae are always present medially on the anterior margin of the thorax
posterior to the temple margin of the head. The lateral margins of the thorax also bear setae,
which may be more or less abundant. A row of setae is generally present across the posterior
dorsal margin of the prothorax and pterothorax, sometimes interrupted medially ('median
gap'). Setae are sometimes present on the dorsal disc of the prothorax and pterothorax
('anterior setae'), and medially on the mesosternum.
The legs of Trichodectidae, in common with those of most other mammal lice, terminate in a
single tarsal claw. In most species this claw is smooth ventrally, but it is toothed ventrally in
Dasyonyx (Figs 14, 15) and ridged in Eurytrichodectes (Fig. 16). In many species a small hyaline
projection, which may be pointed or blunt (Figs 14, 15), is present at the base of the claw. Mayer
(1954), in her study of Bovicola caprae, terms this structure a 'pulvillus' but, as a true pulvillus is
generally paired (Richards & Davies, 1977; Chapman, 1982), this term is inappropriate. The
projection is more probably an empodium or arolium, or even a simple basal tooth of no wider
homology. Kim & Ludwig (1978, 1982) maintain that the pulvillus and empodium do not occur
in the Phthiraptera, but Clay (1969) demonstrates the presence of an empodium in Menoponi-
dae (Amblycera), and Clay (1970) figures structures in Boopia (Amblycera) that almost
certainly are pulvilli, although she follows Keler (no reference given) in terming them euplantu-
lae (plantulae) of the second tarsomere.
Abdomen
In this study the 'true' segment number is referred to by roman numerals to distinguish it from
the apparent number. The terms 'sternite', 'pleurite' and 'tergite' are used for the sclerites of the
sternum, pleurum and tergum respectively of each segment.
Segment I is represented in Trichodectidae by the reduced tergum I only. In this respect the
family is similar to the Trichophilopteridae, but differs from all other Ischnocera, in which
segment I is absent.
Segments III-VIII bear the spiracles, if these are present; spiracles are never present on
segments I and II. Within the Trichodectidae many species have fewer than the plesiomorphic
number of six pairs of abdominal spiracles (Table 1). Loss has apparently taken place
sequentially from the posterior, so that if any spiracles are present one pair is on segment III, and
if more than one pair is present there are no intercalating segments which lack spiracles between
those segments with spiracles. There is, however, no evidence that spiracles have necessarily
been lost one pair at a time. Whilst in Procaviphilus (Meganarionoides) angolensis, P. (M.)
colobi and P. (M.) baculatus the posterior pair of spiracles only is very small and apparently in
process of being lost, in Lorisicola (L.) hercynianus and L. (L.) siamensis the posterior two pairs
are extremely small, probably non-functional, and apparently in process of being lost. In a
number of clades, sister-groups exhibit multiple discontinuities in spiracle number. The
sister-species Lorisicola (L.) mjoebergi and L. (L.) malaysianus have six and zero pairs
respectively; Felicola viverriculae and an undescribed sister-species have three and zero pairs
respectively; the Lorisicola (P.) lenicornis - wernecki clade and the sister L. (P.) acuticeps -
neoafricanus clade have four and zero pairs respectively. Variation within species can occur,
although it is generally erratic. Felicola subrostratus normally has three pairs of abdominal
spiracles; the species is widespread and found on many hosts, but on Madagascar, where the host
is Eupleres goudoti, there may be three or two pairs, and specimens exhibiting asymmetry are
present in the British Museum (Natural History) collection. Asymmetry has also been noted in
Trichodectes emeryi, one paratype of which has six spiracles on one side of the abdomen and five
on the other. Trichodectes (S.) potus is unusual in that the female has three pairs of abdominal
spiracles and the male only two, the only known example in the lice of sexual dimorphism in
spiracle number.
Most Trichodectidae have either six, three or no pairs of abdominal spiracles, other numbers
TRICHODECTID MAMMAL LICE
197
TAXA
NUMBER OF PAIRS
OF SPIRACLES
Geomydoecus , Neotrichodectes , Felicola (F. ) helogale,
F. (F.) hopkinsi, F. (S.) fahrenholzi, F. (S.) guinlei,
Lorisicola (L.) mataysianus, L. (P.) paralaticeps - mungos
clade, L. (P.) acuticeps - neoafricanus clade
Felicola (S. ) bedfordi
1
Felicola (F. ) subrostratus from Eupleres ,
Trichodectes (S. ) potus tj
2
Fnlicola (F. ) all species except helogale and hopkinsi,
3
guinlei,
Trichodectes (S. ) fallax, T. (S. ) o<?tomacuatws,
T. (S. ) potws ?
Lorisicola (P.) bengalensis - juccii clade, L. (P.)
aspidorhynchus , L. (P.) sumatrensis , L. (P.) lenicomis ,
L. (P. ) uernecki
4
Trichodectes (.Paratrichodectes)
5
Trichodectes (T.), T. (5.) all species except fallax -
potus clade, Uerneckodectes , Neolutridia , Lutridia ,
Protelicola, Lorisicola (L.) all species except
mlaysianus , Dasyonyginae, Eutrichophilinae , Bovicolinae
6
Table 1 Distribution of number of pairs of abdominal spiracles in the Trichodectidae.
being less common (Table 1). From the cladogram, reduction to five, four and one pair can be
seen to have occurred once, reduction to two and three pairs twice, and reduction to none eight
times. The loss of abdominal spiracles plainly cannot be used a priori as a taxonomic character
defining (holophyletic) genera, as proposed by Ewing (1936), but neither is it as variable as
suggested by Keler (1938) and Hopkins (1941), who treat it as a character of specific value only.
The selective advantage of this reduction is not known, but it may be an adaptation to exclude
dust from the tracheal system, or to reduce water loss.
A posterior commissure joining the two main abdominal tracheal trunks is present in
Anoplura, Rhyncophthirina, Boopiidae and Trimenoponidae (Amblycera), Philopteridae
(Ischnocera) and 'some Trichodectidae' (Harrison, 1915; Ferris, 1931). All species of
Trichodectidae examined in this study possess a posterior commissure, which is consequently
assumed to have a universal distribution throughout the family. The presence of the posterior
commissure is assumed to be plesiomorphic for the Phthiraptera (Clay, 1970).
In all lice segment IX is the genital segment, and the male genitalia open to its posterior and
the female genitalia to its anterior. Modifications of this segment are discussed in detail below.
In both sexes of Trichodectidae segments X and XI are fused. In females segment XI lies
caudally, posterior to X, but in males the modifications of segment IX have led to the
displacement of both X and XI onto the dorsal surface of the abdomen. Most male Trichodecti-
dae have the genital opening posterior or postero-dorsal. The postgenital segments are reduced
to form a small anal cone arising from the dorsal (anterior) wall of the genital chamber (Keler,
1957) (Fig. 19).
Conical projections are present either dorsoposteriorly or ventroposteriorly on the first three
198
C. H. C. LYAL
tVllp
sVII
tergum VII
tVllla
sternum VII
tergum VIII
tVlllp
gc
segment IX
p-gs'
Fig. 19 Diagrammatic representation of terminal segments of trichodectid male abdomen, with terms
used in this study, t - tergite; s - sternite; p - pleurite; ac - anal cone; gc - genital chamber; p-gs -
post-genital sclerite; sgpr- lateral rods of subgenital plate.
pleura (II, III and IV) in many Trichodectidae. The distribution of the pleural projections is
summarised in Table 2.
Male Trichodectidae, alone among Phthiraptera, possess 'lateral abdominal flecks', first
noted in Werneckiella by Moreby (1978). The flecks are small pits positioned on the antero-
dorsal angles of pleura III-VII, occasionally on II and VIII, and anterolaterally on terga III-V in
many species (Fig. 218). In Werneckiella there are small sclerites situated anteriorly and
posteriorly of the pit; in Trichodectes canis there are no sclerites bordering the pit but a small
sclerite is present at its base; in Neotrichodectes there are no sclerites associated with the pit at
all. The function of these structures is unknown.
In some species of Phthiraptera terga II and/or III of the males are ornamented with large
setae or projections. Such developments are numerous in the Trichodectidae. Many species of
Felicola have a pair of long setae medially on male tergum II (Figs 180, 187, 188), whilst
Geomydoecus (Thomomydoecus) spp., Trichodectes ovalis, T. ugandensis and an undescribed
subspecies of T. galictidis have paired 'combs' of long setae on male terga II and III (Fig. 161),
and Bovicola multispinosa and B. hemitragi have paired semicircular 'brushes' of setae on male
tergum II (Fig. 41). Damalinia ornata has sclerotised blunt projections on male terga II and III.
These specialised setae and projections may in some way assist the male to hold the female
during copulation, although in no case have the setae or projections been observed to be
damaged, as they might be expected to be should they operate against the female abdomen, and
there is no observational evidence to support the hypothesis.
Males of Neotrichodectes species have a pair of small median setae on terga II-VII , sometimes
separated by a seta of normal length (Fig. 229). This feature, not found in females or males of
any other group, is of unknown function.
Female genitalia
The female subgenital plate of Trichodectidae comprises either sternite VIII or sternites VIII
and VII fused. The posterior margin of sternum VIII forms the ventral margin of the vulva. In
many species the margin is expanded posteriad and slightly laterad (Fig. 94), and in others the
TRICHODECTID MAMMAL LICE
199
TAXA
SEX
PLEURUM
I I
D V S
PLEURUM
III
D V S
PLEURUM
I V
D V S
Seotrichodectes mephitidia
9
-
Cebidicola
3 9
* -
Procavicola - Eurytrichodectes clade, Lorisicola
mjoebergi, Felioola zeyloniaus - viverriculae clade
3 9
-f + +
Felioola bedfordi, F. congoensis, F. cooleyi, F.
cyniotis , F. decipiens, F. helogale, F. minimus,
F. liberiae - subro stratus clade, Lorisicola.
aspidorhynchus , L. aaffra, L. felis , L. hercynianus ,
L. malaysianus, L. mangos, L. spenceri , L. sumatrensis ,
L. wernecki, L. bengalensis - juocii clade
i 9
f -f
Lorisiaola acuticeps, L. africanus , L. lenicornis ,
L. neoafriaanus
<i 9
?
Felioola occidentalis , F. auadraticeps , F. vulpis
i 9
+
+
Felicola calogaleus , F. setosus
<i 9
* *
* *
Trichodectes zorillae
e?
9
f +
+ +
+ +
4- ?
Geomydoecus (G. ) most species
f 9
+
+
Geomydoecue (G. ) thomomyus - dakotensis clade
Geomydoecus (Thomomydoecue)
3
9
* *
+ -f
f +
+ * *
Table 2 Distribution of abdominal pleural projections in the Trichodectidae. For pleura II, III and IV an
indication is given whether a dorsal ('D') or ventral ('V') projection is present, and whether those
projections are sclerotised ('S'). In each case the presence state is indicated by ' + '. Very light
sclerotisations are indicated by '?'.
centre of the margin is greatly expanded into a flat lobe termed the 'subgenital lobe' (Figs 149,
153, 175). The subgenital lobe appears to have evolved at least three times in the Trichodecti-
dae, and its form is of taxonomic and systematic value. Its function, however, is not known.
To each side of the vulva is a projection from sternum VIII, the gonapophysis of segment VIII
(Lyal, in prep., b). The structure of this is variable. In all cases a basal internal apophysis is
produced, presumably as a muscle attachment (Keler, 1938a). The gonapophysis lies longitudi-
nally, parallel to the abdomen; it may curve onto the dorsal surface apically. It may be long,
slender and apically acute (Fig. 235), long and apically rectangular (Fig. 154), broad and
membranous (Fig. 243), produced into a rounded, toothed or rectangular medial lobe with an
apical 'spur' (Figs 175, 233, 236), with the spur reduced (Fig. 237) or absent (Fig. 211),
sometimes with serrations laterally (Fig. 211). Setae may be present on the gonapophyses and
the ventral vulval margin in various configurations, occasionally arising from sclerotised
tubercles. In some cases a sclerotised band links the gonapophyses and the vulval margin, but in
most cases this is not present and the connection is membranous.
The curvature of the gonapophyses about both longitudinal and lateral axes produces a
complex three-dimensional structure difficult to interpret on slide-mounted specimens. The
orientation of the gonapophysis may be altered by pressure of the coverslip during mount-
ing, especially if it has a membranous base. In addition, differences in orientation, either
200 C. H. C. LYAL
natural or caused by the mounting process, may give the impression of very different shapes.
Observations by Ferris (1951) and Murray (1957a, 19576) indicate that the gonapophyses are
used, at least in some species, to trap the hair on which the egg is to be laid, guide the egg onto it,
and mould the attaching cement. For Bovicola ovis the diameter of the hair is of importance in
determining whether the egg will be laid (Murray, 19576) and this is almost certainly detected by
the gonapophysis. There is thus selection pressure on the form of the gonapophysis relating to
the structure of the host hair.
Posterior to the vulva and sometimes covered by the subgenital lobe is the 'sub-vulval area',
which sometimes bears characteristic spines or scales.
Also posterior to the vulva is a single sclerite which, as it cannot be homologised with either
the gonangulum or sternite IX + X (Lyal, in prep., 6), is termed the 'post-vulval sclerite'
(Moreby, 1978). This sclerite may be single, medially divided, narrow and strip-like, broad and
triangular, fused to the postgenital pleurite, or absent (Figs 154, 175, 236, 237, 243).
The female genital chamber, opening at the vulva, is oval, dorso-ventrally compressed, and
lightly-sclerotised. The dorsal wall of the chamber may be heavily sculptured and bear
sclerotised spicules, ridges or spines; this sculpturing may extend on to the postgenital sterna.
The ventral wall may be similarly sculptured, although usually to a lesser extent; where the
subgenital lobe is present the sculpturing of the genital chamber may extend onto its dorsal
surface . The sculpturing of the walls of the chamber probably provides a highly f rictional surface
against which the spicular surface of the endophallus acts to provide a firm union during
copulation. The nature of the sculpturing and the distributions of spicules or spines may be
taxonomically useful at the specific or subspecific level. The common oviduct opens into the
anterior end of the genital chamber and curves sharply posteriad to lie dorsally to the chamber.
Dorsal to the vulva the common oviduct curves sharply anteriad and divides into the two paired
oviducts (Fig. 20). The genital chamber, although assuming a more or less circular cross-section
during the passage of an egg and perhaps during copulation, is, when at rest, a dorso-ventrally
gonapophysis VIII
paired oviducts
common oviduct
sternum VII
sternum VIM
genital chamber
Fig. 20 Diagrammatic three-dimensional representation of trichodectid female genital chamber and
oviducts (internal), with terms used in this study.
TRICHODECTID MAMMAL LICE 201
flattened, fairly rigid structure, the minimum width of which is governed by the diameter of the
egg. The oviduct, however, is an elastic, folded membranous tube, compressed and folded to
reduce volume, and expanding only to allow passage of the egg. The common oviduct must, at
the junction with the genital chamber, be the same diameter as the chamber. At its division into
the paired oviducts, however, it is narrow and greatly folded when at rest. Between these two
points it narrows more or less abruptly, and folds may be apparent in its walls. Price & Emerson
(1971) interpret these folds in Geomydoecus as striations of the dorsal wall of the genital
chamber, and term this apparently membranous structure the 'genital sac'. The true dorsal wall
is interpreted as the ventral wall and the true ventral wall is apparently not observed. The form
of the folds of the 'genital sac' (common oviduct), as well as its length and the width of its
junction with the genital chamber, have been utilised by Price and his co-workers as specific and
subspecific characters in their re visionary work on Geomydoecus (e.g. Price, 1974; Price &
Emerson, 1971; Price & Hellenthal, 1976; Timm & Price, 1980). The apparent constancy of the
dimensions of the oviduct is related to the size of the egg and perhaps to the restrictions in
variability of size of the genital chamber imposed by selection. The apparent constancy of the
folds is more puzzling although it is possible that tubes of identical length and diameter have an
optimum folding pattern if compressed in the same way. The wall of the common oviduct
may also be lightly sclerotised, at least near to the junction with the genital chamber. The
folds of the common oviduct have not been utilised as a taxonomic character elsewhere in the
Phthiraptera.
Blagoveshtchenski (1956) examines several species of Trichodectidae but fails to find
evidence of a spermatheca in the family. In this study examination of slide-mounted specimens
of most species has failed to reveal a spermatheca in any genus apart from problematically in
Dasyonyx, where a lightly-sclerotised sac, differing in form between species, is developed from
the wall of the common oviduct. The relatively anterior position of this sac in comparison to the
spermatheca of other Phthiraptera suggests that the two structures are not homologous. No
histological or anatomical studies have been carried out on the sac.
The male abdomen and copulation
The opening of the male genital chamber ('genital opening') is always posterior to sternum IX,
and primitively in the lice is ventrally positioned and distant from the anus, which is terminal.
During copulation the male and female are usually oriented the same way, with male ventral to
female. The tip of the male's abdomen is curved dorsally and anteriorly so that the genital
openings of the male and female meet. The male genitalia consequently enter the female genital
chamber 'upside-down' , with the ventral side of the former coming into contact with the dorsal
wall of the latter. Should sclerotised tergites be absent (as in Neotrichodectes and Geomydoecus)
or greatly reduced (as in Trichodectes galictidis and Felicold) flexion of the male abdomen during
copulation is evenly distributed along the membranous dorsal surface. If sclerotised tergites are
present and fully-developed, however, flexion must be about the sclerite-membrane-sclerite
joints of the dorsal surface, these thus functioning as 'hinges'. At each such hinge, there will be
considerable deformation of the internal structures during flexion, whilst between the hinges
there will be no deformation. Increasing the number of hinges on the dorsal surface permits
smoother curvature of the abdomen and reduces the internal deformation at each hinge. Many
male Trichodectidae have tergal sclerites divided into an anterior and a posterior plate on at
least some segments (Figs 103, 105, 136), possibly for this reason. The degree of internal
deformation may be further reduced by modification of the shape of the plates. The form of plate
which, when placed in series, causes maximum internal deformation and requires maximum
membrane area: sclerite area ratio for flexion is a simple rectangle, with anterior and posterior
margins at right angles to the long axis of the abdomen. These factors can be diminished by
introducing curvature in the anterior and posterior margins of the plates, either with both
margins parallel or opposite, producing a series of alternating biconvex and biconcave plates
(Lyal, 1983). Both these patterns are approached in the Trichodectidae (Figs 68, 103, 105). The
flexibility required of the ventral surface of the abdomen is developed, should the sterna be
202 C. H. C. LYAL
sclerotised, by increasing the lengths of the sclerites and allowing them to overlap when the
abdomen is at rest (Figs 103, 105).
Despite the increased flexibility of the abdomen achievable by modifications of the tergal and
sternal plates, the position taken by the male during copulation is compatible with a ventral
genital opening only if the abdomen is long and slender. A short, broad abdomen will not
deform sufficiently to turn back on itself and bring the male genital opening into contact with the
vulva. Such a short, broad abdominal form has been developed several times in the Phthirap-
tera, however, including at least once in the Trichodectidae. In most cases the limitation
imposed on the degree of curvature of the abdomen has been met by a posterior or even dorsal
migration of the genital opening, thus reducing the degree of curvature necessary. This
displacement has been effected by an increase in length of sterna VII, VIII and IX and a
concomitant shortening of the corresponding terga. In Trichodectidae this may lead to the
division of tergite VIII longitudinally by segment IX (Fig. 212). The re-positioning of the genital
opening has increased its proximity to the anus, and in Trichodectidae the anus is contained
within the genital chamber with the recuced segments X and XI projecting from the genital
opening (Fig. 19). The curvature required of the abdomen has been further limited in some
Trichodectidae by two distinct adaptations. In some species of Felicola segment IX is developed
into a long, slender posterior process and the genital opening is dorsal and apical (Fig. 187). The
basal apodeme and parameres are elongate and slender (Fig. 205). Most of the curvature
required is probably developed at the base of segment IX and the junction of the parameres and
the basal apodeme. Constriction of the endophallus at the latter fold is prevented by the
presence of a reduced, circular mesomeral arch, lacking an extension, which, being fused to the
endophallus, prevents the structure from being pinched shut. In some Lorisicola species
segment IX projects dorsally from segment VIII, the genital opening being dorsal (Fig. 114).
This positioning limits the degree of curvature required of the abdomen by increasing the
proximity of the genital opening to the vulva on minimum curvature.
Male terminalia
In Trichodectidae the subgenital plate is fundamentally uniform in construction, although great
superficial differences may exist between species. The plate is here considered to comprise eight
discrete elements: sternites VII, VIII, IX (and, occasionally, VI); a pair of rods (referred to as
'subgenital plate rods' or 's.g.p.r.' in this study) which lie laterally to the sternites and sometimes
fuse them together (Figs 19, 165, 212); the two pleurites of segment IX, and the post-genital
sclerite, which is of uncertain homology (Fig. 19). Any of these elements may be present,
reduced or absent, or fused to adjacent sclerites. The sternites and the post-genital sclerite may
be whole or medially divided. In cases of extreme reduction only the lateral rods may be left
(Fig. 163) or all the sclerites absent (Fig. 187). In the most complete form (Fig. 68), all the
sclerites are fused, forming a squared plate; usually there are membranous areas surrounding
the sternal setae ('perisetal gaps'), but these may be absent (Fig. 115).
The subgenital plate rods on sternum VIII are always connected to the ventral wall of the
genital chamber in Trichodectidae, a feature also observed in some Anoplura. The function of
this attachment is not known, although in species with ventrally-positioned parameres and a
median posterior extension to the mesomeral arch, the difference in lengths between the dorsal
wall of the genital chamber and the membrane between the basal apodeme and the subgenital
plate might cause divergence of the apices of the parameres and mesomeral arch extension
during extrusion of the genitalia to allow evertion of the endophallus.
In some Trichodectidae the posterolateral angles of the subgenital plate are greatly extended
into setose lobes termed 'styli' by Eichler (1963) (Figs 46, 68). Abdominal sternal sclerites are
probably derived from a fusion of the sternal plate and the coxal elements of the paired
abdominal appendages (coxopodites) (Matsuda, 1976; Richards & Davies, 1977). In the
subgenital plate of the Acercaria the paired nature of the gonocoxopodite component may be
expressed as a concavity of the posterior margin of the plate and concomitant projection of the
posterolateral angles (Matsuda, 1976). True abdominal styli are derived from the paired
TRICHODECTID MAMMAL LICE 203
abdominal appendages and are serially homologous with either the shaft of the thoracic legs or
the coxal styli (spurs of Matsuda, 1976) of the thoracic legs as present in the Machilidae
(Thysanura) (Matsuda, 1976; Richards & Davies, 1977). Styli are not, therefore, homologous
with the coxopodites, but arise from them, being separated by a clear sulcus. Although true styli
are present in some Psocoptera (Matsuda, 1976), the posterior extensions of the subgenital plate
of Trichodectidae are not demarcated by a sulcus and thus are not homologous with true styli.
The so-called styli of the Trichodectidae are an indication of the gonocoxopodite component of
the subgenital plate, and are here termed 'pseudostyli'. Taxonomic use may be made of the form
of the pseudostyli, which is very variable in the Trichodectidae, but the difficulty of assigning
polarity to transformation series limits the applicability of this character within phylogenetic
studies. The distribution of the pseudostyli through the Trichodectidae, however, is utilised in
the construction of the phylogeny of the family (see below).
Male genitalia
The structure of the male genitalia of lice and the homologies of the major parts are discussed by
Lyal (1983 and in prep., b). The genitalia consist of a more or less sclerotised basal apodeme
supporting caudally a pair of parameres which may be fused ventrally and a pair of mesomeres
which may be fused dorsally; fused to the interior faces of the parameres and mesomeres is the
permanently-everted portion of the eversible endophallus (Lyal, 1983 and in prep., b}. In most
Trichodectidae both parameres and mesomeres are present, the latter usually being fused
apically (Figs 170, 198), the fused portion generally being extended posteriorly (Fig. 224). The
parameres may also be fused, forming a median ventral plate (Fig. 250). The parameres and
mesomeres may meet the basal apodeme together (Fig. 170) or separately (Fig. 250). The full
range of variation of the male genitalia is discussed in the character analysis and taxonomy
sections below.
During copulation the endophallus everts into the female genital chamber. As described
above, the interior of the femlae genital chamber is roughened and lined with scales. The
endophallus is likewise roughened, being covered with small chitinous spicules or larger
sclerites. The probable function of this adaptation is to maintain a firm connection between the
male and female genitalia during copulation. The form of the endophallus and the distribution of
the spicules and sclerites are species-specific and very variable, possibly functioning as pre-
zygotic isolating barriers.
Character survey
As noted above, the characters and character states listed below were selected from a much
larger set of characters and states. In some cases reasons for rejection are discussed below. The
full list of characters and the complete data matrix are presented in Lyal (1983) and are
deposited in the library of the British Museum (Natural History).
1 Posterior of basal apodeme symmetric in vertical plane
lateral struts ('b.a.l.s.') asymmetric in vertical
plane (Figs 193, 195) 1 g
2 Anterior end of basal flat, convex, concave
apodeme (Figs 91, 200, 250)
deeply concave (Fig. 52) 1 g
acuminate (Fig. 173) 1' g
3 Posterior bifurcation of absent
b.a.l.s. present (Fig. 173) 1 g
4 Posterior of b.a.l.s. lacking lateral extension
with lateral extension to
mesomeres (Fig. 239) 1 g
5 Anteposterior spur of absent
b.a.l.s. present (Fig. 83) 1 g
204
6 Shape of basal apodeme
Posterior of b.a.l.s.
8 Posterior of b.a.l.s.
Posterior of b.a.l.s.
10 Basiparameral sclerites
11 Paramere fusion
12 Paramere fusion
13 Paramere fusion
14 Parameral apices
15 Paramere shape
16
Paramere shape
17 Parameral plate shape
18 Parameral orientation
19 Paramere shape
20 Paramere shape
21 Paramere shape
C. H. C. LYAL
not long and 'waisted'
very long, with median
'waist' (Fig. 119)
not modified as below
broad and obtuse in meeting
parameres (Fig. 198)
not modified as below
sharply inturned and convex
(Fig. 200)
not modified as below
incurved to parameres (Figs
144, 145)
absent
present, fused to parameres
present, free
not fused to mesomeral arch
fused to mesomeral arch in
part (Figs 60, 75)
completely fused to
mesomeres (Fig. 74)
not fused to basal apodeme
fused exteriorly to basal
apodeme (Fig. 75)
fused medially or completely
to basal apodeme (Figs 145,
172, 174)
not fused together or closely
associated
fused together
not fused but closely
associated, with line of
division apparent (Figs 223,
224)
unfused
fused
not as described below
fused to shield-shaped plate
(Figs 239, 242)
not as described below
fused as described in
character 15, with antero-
median projection (Fig.
238)
not as described below
produced apically into
incurving points (Fig. 171)
similar
at right-angles to each
other (Fig. 170)
more or less broad, thick
very thin, deflected
asymmetrically (Fig. 174)
more or less broad
narrow rods (Figs 196, 197)
not as described below
r
o
i
r
1
1
TRICHODECTID MAMMAL LICE
205
22 Median internal projection
of parameres
23 Paramere size
24 Paramere shape
25 Paramere shape
26 Paramere shape
27 Base of parameres
28 Base of parameres
29 Paramere and mesomere
shape
30 Reduction of parameres and
mesomeres
31 Mesomeres
32 Mesomeral position
33 Mesomere fusion
34 Lateral desclerotisations
of mesomeral arch
35 Lateral flexions of
mesomeral arch
36 Mesomeral arch
37 Mesomeres basally, between
b.a.l.s.
38 Mesomeral arch mesally
39 Diameter of mesomeral arch
very broad, lanceolate,
scoop-shaped (Fig. 172) 1 g
absent
present 1 g
large or moderate
small discs (Fig. 84) 1 1
not as described below
cylindrical (Figs 169, 205) 1 g
not as described below
basally very narrow, medially
broad (Fig. 198) 1 g
not as described below
with characterically-
differentiated base and
blade (Fig. 51) 1 g
not as described below
broad, club-like (Fig. 192) 1 g
block-like (Fig. 201) 1' g
cuboid (Fig. 52) 1" g
lacking flange
with flange (Fig. 118) 1 g
not as described below
of characteristic asymmetric
form (Fig. 249) 1 g
not reduced as below
characteristically reduced
(Fig. 52) 1 1
characteristically greatly
reduced (Fig. 53) 2 1
present
absent 1 1
reach or fail to reach
b.a.l.s.
extend mesad of b.a.l.s. 1 g
apically fused to form arch
not apically fused 1 1
absent
present (Fig. 118) 1 g
absent
present (Figs 107, 225) 1 g
not as described below
modified into tripartite
arch (Fig. 93) 1 g
not modified as below
sharply directed posteriad
(Fig. 224) 1 g
smooth or with projection
with two nipples (Fig. 221) 1 g
less than half the length of
the permanently-everted
endophallus
more than half the length of
the permanently-everted
endophallus 1 g
206
40
41
42
43
44
45
46
47
48
49
50
51
52
53
54
55
56
57
58
Shape of mesomeral arch
Shape of mesomeral arch
Shape of mesomeral arch
Shape of mesomeral arch
Shape of mesomeral arch
Shape of mesomeral arch
Shape of mesomeres
Median extension of
mesomeral arch
Apex of mesomeral arch
extension
Shape of mesomeral arch
extension
Tongue-like sclerite between
parameres
Everted part of endophallus
Endophallus spiculation
Endophallus spiculation
Endophallus spiculation
Endophallus spiculation
Endophallus spiculation
Gonapophysis setae
Gonapophysis setae
C. H. C. LYAL
as great as the length of the
permanently-everted
endophallus 2
not as described below
with median anterior
deflection (Figs 191, 192) 1
circular or elliptical
rectangular (Fig. 198) 1
not as described below
extension lost, arch 'looped'
(Fig. 247) 1
not as described below
widely circular (Fig. 55) 1
not as described below
pentagonal, convex distally
(Fig. 62) 1
smoothly curved
sharply inturned to
parameres (Fig. 81) 1
extending anteriad to
posterior end of basal apodeme
and sharply recurved (Figs 82,
219)
not as described below
very slender, string-like
(Fig. 52)
present
absent
not bifurcate
bifurcate
not as described below
broadly expanded, lanceolate
absent
present
not sclerotised
sclerotised
not as described below
numerous 'V'-shaped rods
not as described below
comprising large hook-like
spines (Fig. 107) 1
not concentrated about
gonopore
concentrated about gonopore 1
with dense 'V'-shaped patch
about gonopore 2
not as described below
including median row of
hook-like scales 1
not as described below
dense and refrigent in part 1
present
absent 1
lacking sclerotised tubercles
2
g
1
g
1
1
1
g
1
g
1
g
1
g
1
g
TRICHODECTID MAMMAL LICE
59
60
61
62
63
64
65
66
67
68
69
70
71
72
73
Gonapophysis setal
tubercles
Gonapophysis setal
tubercles
Gonapophysis shape
Gonapophysis shape
Gonapophysis shape
Gonapophysis lobe
Apex of gonapophysis
Gonapophysis lobe
Gonapophysis lobe
Gonapophysis lobe
Gonapophysis lobe
Gonapophysis lobe
Gonapophysis junction
with ventral vulval margin
Ventral vulval margin
Ventral vulval margin
some having sclerotised basal
tubercles (Fig. 150)
absent, or present and not
characteristically fused
present and characteristically
fused (Fig. Ill)
absent, or present and in
characteristic pattern (Fig. 153)
present and modified from
characteristic pattern by
loss of apical non-tuberculate
seta (Fig. 146)
not as described below
spoon-shaped (Fig. 156)
not as described below
hook-shaped (Fig. 66)
hook-shaped with spur (Fig.
67)
not explanate ventrally
(other than as discrete lobe)
thinly explanate ventrally
(other than as discrete lobe)
absent
present
acute or obtuse
not projecting beyond lobe
(if present); apparently
absent
not very thick
very thick
absent, or not as described
below
broad, formed of fused
tubercles (Fig. 233)
absent, or not as described
below
narrow, rectangular, formed
from 2 or 3 fused tubercles
(Fig. 177)
negative for character 67
or, if positive, flattened
positive for 67 and folded
anteriorly
positive for 67 and greatly
folded
absent, or not as described
below
produced into spines
posteriorly (Fig. 211)
acute
smoothly continuous
unsclerotised
sclerotised
no shorter than length of
gonapophyses
207
g
r
1
g
1
g
1
1
1
g
1
g
1
g
1
g
2
g
1
g
1
g
1
g
208
74
75
76
77
78
79
80
C. H. C. LYAL
81
82
83
84
85
86
87
88
89
90
91
92
Ventral vulval margin
Ventral vulval margin
Ventral vulval margin
Ventral vulval margin
Subgenital lobe
Subgenital lobe margin
Subgenital lobe surface
Subgenital lobe, basally
Subgenital lobe processes
(ch. 81)
Subgenital lobe processes
(ch. 81)
Subgenital lobe
Subgenital lobe bifurcations
(ch. 84)
Subgenital lobe
Sub-vulval area
Dorsal vulval face
Post-vulval sclerites
Post-vulval sclerites
Common oviduct
Female genital chamber
shorter than length of
gonapophyses (Fig. 102) 1
lacking setal tubercles
with setal tubercles (Fig.
156) 1
not expanded
expanded (Figs 94, 126) 1
not as described below
greatly produced and rounded
(Fig. 113) 1
not as described below
with median narrow projection 1
absent
present 1
not serrate
serrate 1
very serrate 2
smooth dorsally and ventrally
with overlapping pointed
scales (Fig. 210) 1
with small spines (Fig. 208) 2
with many overlapping spines
(Fig. 207) 3
lacking lateral processes
with lateral processes 1
not as described below
thinly sclerotised and
directed posteriad 1
not as described below
membranous and serrate (Figs
155, 157) 1
not bifurcate (Fig. 210)
bifurcate (Fig. 208) 1
not as described below
rectangular, widely separate
(Fig. 176) 1
lacking submarginal setal
patch
with submarginal setal
patch 1
lacking pointed scales
with pointed scales 1
not spinous
spinous 1
present
absent 1
broad, triangular
long, narrow, oriented
parallel to longitudinal
axis 1
not as described below
with folded 'collar' at
branching-point 1
lined with overlapping scales
TRICHODECTID MAMMAL LICE
93 Female genital chamber
94 Female genital chamber
95 Female genital chamber
96 Female genital chamber
97 Female sternum VII
98 Reproduction
99 Male tergite VIII (if
present)
100 Posterior margins of male
tergum IX
101 Male segment IX
102 Pseudostyli
103 Pseudostyli (if present)
104 Pseudostyli (if present)
105 Male segment IX
106 Male sternite VIII
107 Male sternite VIII
(if present)
108 Male perisetal gap
109 Male sternum VI
lined with sclerotised nodules,
some fused
not as described below
with median dorsal longitudinal
fold
not as described below
with median dorsal area clear
of scales
not as described below
with median anterior dorsal
area clear, thinly
sclerotised
not as described below
with spines on dorsal face
lacking processes
with two long spikes on
posterior margin (Fig. 73)
sexual
parthenogenetic
with posterior element absent
or, if present, not fused to
tergite IX
with posterior element fused
to tergite IX
not as described below
greatly expanded (Fig. 79)
without two longitudinal
strengthening sclerites
with two longitudinal
sclerites
absent
present
not as described below
broad, rounded, long (Fig.
68)
rounded apically
angular, pointed apically
(Fig. 78)
lacking single apical
projection
with single apical
projection (?fused pseudo-
styli)
not as described below
characteristically enlarged
(Fig. 159)
not convex posteriorly
characteristically convex
posteriorly (Fig. 218)
present
absent
not as described below
with anterior and posterior
sclerites
209
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
g
1
1
210
110 Female flagellomeres
111 Male flagellomeres
112 Male scape
C. H. C. LYAL
113
115
Male scape setae
114 Male scape setal row
(if present)
Male flagellum
116 Male flagellar 'teeth'
117 Male flagellar 'teeth'
118 Male flagellum
119 Male flagellum
120 Inner face of male
flagellum
121 Male scape
122 Female pedicel
123 Male flagellum
124 Flagellar sensilla placodea
and coeloconica
125 Sitophore sclerite
126 Posterior temple angles
127 Posterior temple margins
unfused (three flagellomeres)
fused to form two flagellomeres 1
fused to form one flagellomere 2
unfused
fused 1
greatly or slightly
expanded
not expanded 1
randomly scattered on
posterior face
in longitudinal row on
posterior face 1
numbering at least three setae
on posterior face
reduced to two setae on
posterior face
lacking apical 'teeth'
with single apical 'tooth'
with two apical 'teeth'
with three apical 'teeth'
with four apical 'teeth'
with eight apical 'teeth'
not on protuberance
on protuberance
articulated basally
fused to flagellum
lacking basal projection
with simple basal projection
with basal projection of
linked 'teeth'
with broad, rough, basal
projection
not as described below
with simple median and basal
projections only
not serrate
serrate
lacking apical projection
with apical projection
lacking projections
with membranous projection
not very long
very long
not in pit
in pit with marginal tongue-
like processes
not as described below
with posterior arms extended
(Fig. 12) 1
lacking projections
with projections (Fig. 129) 1
with very long projections
(Fig. 138)
not as described below
produced and convex 1
1
1
r
1
i
1
i"
g
tint
g
V"
g
1
g
1
g
1
g
r
g
2'
g
1
g
1
g
1
g
1
g
1
g
1
g
TRICHODECTID MAMMAL LICE
128 Osculum
129 Pretarsus
130 Pretarsal claws
131 Post-coxaleoflegUI
132 Sternum II
133 Atrium of thoracic spiracle
134 Abdominal spiracles
135 Abdominal spiracle VIII
136 Abdominal spiracle VII
137 Abdominal spiracle VI
138 Abdominal spiracle III
139 Setae of posterior setal row
of pleurum III
140 Pleurum VII
141 Abdominal tergal setae
142 Male tergum II
143 Male tergum III
144 Male terga II-IV
not deep
deep and with characteristic
anterior marginal convexity
(Fig. 69)
bearing two claws
bearing one claw
lacking ventral spines
with blunt ventral spines
with sharp ventral spines
not fused to abdominal pleurum
II
fused to abdominal pleurum
II, at least in female
not as described below
with sclerotised apophysis
articulated to pleurum II
spherical
tubular
all of similar size
spiracle VIII of male very
small
spiracles VII and VIII very
small in both sexes
present
absent
present
absent
present
absent
present
absent
not stout
stout
lacking tuft of very long
setae
with tuft of very long
setae (Figs 88, 90)
short or of medium length
(Figs 105, 188)
generally very long, obscuring
p.l.s. (if present) (Fig. 158)
lacking specialised setae as
described below
with long stout setae not
found on other terga (Figs
180, 188)
lacking specialised setae as
described below
with isolated pair of long
median setae, longer than
those of tergum IV (Fig. 187)
without specialised setae as
described below
each with single pair of
long median setae (Fig. 182)
1
1
1
r
211
1
g
1
g
1
g
1
1
1'
1
1
1
1
1
1
1
1
1
1
g
1
g
1
g
1
g
1
g
1
g
212
145 Female terga I-VIII
146 Female terga I-IV
147 Male terga II and III
C. H. C. LYAL
148
149
150
151
152
153
154
155
156
157
158
159
160
Female terga I-VIII
Male terga II-III
Male tergum II
Postero-lateral setae
Male terga II-VI
Setae of abdominal sterna
II-IV
Female abdominal setae
Setae of male tergum II
Abdominal setal bases
Abdominal pleurum II
Dorsal projection of
pleurum II
Dorsal projection of
pleurum IV
Ventral projection of
without specialised setae as
described below
each with single pair of
long stout median setae 1
with median setae
lacking median setae 1
without specialised setae as
described below
with group of characteristically
specialised long setae
(Fig. 178) 1
without specialised setae as
described below
each with single pair of
median setae 1
without specialised setae as
described below
with long setae arranged in
straight rows of four or
more (Figs 161, 244) 1
without specialised setae as
described below
with long setae arranged in
curved rows (Fig. 41) 1
absent or possibly present
but not clearly distinguished
from lateral setal group
clearly present on terga II-VI 1
with two median setae the
same size as other setae of
median group
with two median setae
appreciably smaller than
other setae of median group 1
not as described below
short, stout (Fig. 136) 1
not as described below
very long, fine 1
not arising from modified
sclerite
arising from very long,
medially-divided sclerite 1
not enlarged
enlarged 1
not as described below
extending narrowly onto
sternum II 1
extending broadly onto
sternum II 1 '
absent
present, unsclerotised 1
present, sclerotised 2
absent
present 1
absent
TRICHODECTID MAMMAL LICE 213
pleurum IV present 1 g
161 Projections of pleurum not as described below
IV (if present) very long (Fig. 136) 1 g
162 Maleterga VI-VIII (or VI, not as described below
if VII-VIII without with anterior sclerite
sclerites) longitudinally divided
medially 1 g
163 Male terga VI-VII not as described below
with posterior sclerite
longitudinally divided
medially 1 g
164 Abdominal lateral flecks absent
present 1 g
Character analysis
The character analysis is performed in two interconnected parts, the determination of character
polarity and the construction of the cladogram. These are linked through the process of
'reciprocal illumination' (Hennig, 1966), and thus, although they are considered individually in
the two sections below, there is some interaction between the two processes which will be
manifested in discussion. The following two sections are intended to explain the reasons for the
polarity assigned to the characters listed above, and for those groups developed in the cladogram
where characters are apparently arranged non-parsimoniously.
To increase clarity and conciseness the distributions of characters and character states
discussed below are related to taxa and clades developed in the analysis. Clades are referred to
by the names of the taxa (species or genera) on the extreme left (top) and right (bottom) of the
clade as depicted on the cladogram (Figs 23 to 34), reading from left (top) to right (bottom).
Identification of apomorphic states
In many cases an apomorphic character state is identified as such by its distinct complexity and
very limited distribution and, to avoid pointless repetition in the following discussion, such
instances are not examined individually.
The characters are examined under the following headings:
Male genitalia (characters 1-56); Female genitalia (characters 57-97); Reproduction (character
98); Male terminal abdominal segments (characters 99-109); Antennae (characters 110-124);
Head (characters 125-128); Legs (characters 129-130); Postcoxale (character 131); Spiracles
(characters 133-138); Abdominal setae (characters 139-156); Abdominal pleural projections
and modifications (characters 157-161); Abdominal sclerae (characters 132, 162-164).
Male genitalia (characters 1-56)
Most species of Psocodea have symmetric male genitalia and this state is consequently assumed
to be plesiomorphic for the superorder. In a few species of Trichodectidae the genitalia are
asymmetric, but differences in the form of the asymmetry in different species (Figs 170, 174, 249)
suggest that several independent autapomorphies have been developed. In some cases a
characteristic asymmetry is limited to a single species and is therefore of no relevance to phyletic
analysis, but the asymmetries described in characters 1, 18, 19 and 29 are more widely
distributed and are all employed. The distribution of other apomorphies indicates that character
1, the vertical asymmetric deflection of the lateral struts of the basal apodeme ('b.a.l.s.'), has
been developed twice, once in Felicola (S.) bedfordi (Fig. 195), and once in the common
ancestor of F. (F.) cynictis and F. (F.) setosus (Fig. 193).
In most Psocodea the basal apodeme is not fused to the parameres, but in a few Trichodecti-
dae this fusion, considered to be apomorphic, has taken place (character 12). In the Damalinia
(D.) theileri-harrisoni clade the posterior ends of the b.a.l.s. are broad and fused exteriorly to
the parameres (Fig. 75); this fusion (character 12:1) is unlike that found elsewhere in the family
214 C. H. C. LYAL
and is believed on that account to be autapomorphic. In the Trichodectes (S.) retusus-mustelae
clade, the T. (S.)fallax-potus clade, Neolutridia lutrae and Lutridia exilis the parameres are also
fused to the b.a.l.s. (character 12:1'), but there are no features in the fusion pattern to indicate
whether the apomorphy is homologous or convergent in the four clades. There are a number of
possible sequences of gains and reversals. The fusion may have taken place three or four times
with no reversals; once, to be lost at least four times; or twice, to be lost at least twice. The
genitalia of Lutridia spp. differ from those found in the sister-group (the Trichodectes-
Neolutridia clade), being more similar to those of Protelicola. Comparison of L. exilis and L.
matschiei (Figs 144, 145) indicates fusion of the parameres and b.a.l.s. in the former species to be
associated with the virtual detachment of the basal fused portion of the parameres, a unique
feature. For this reason the fusion in L. exilis is deemed to be autapomorphic. The Trichodectes
(S.) retusus-mustelae clade and the T. (S.) fallax-potus clade are placed by character 144 in a
trichotomy with T. (S.) emery i, a species in which the parameres are not fused to the b.a.l.s. If
the fusion is homologous in Trichodectes (Stachiella) and Neolutridia, then it must have been lost
in Werneckodectes, Trichodectes (Trichodectes) and T. (Paratrichodectes), and T. (S.) emeryi.
Parsimony suggests that fusion was developed independently in Neolutridia and the common
ancestor of the two clades in Trichodectes (Stachiella) that possess the character, these latter
being united as sister-groups. Fusion of the parameres to the b.a.l.s. is thus postulated to have
taken place four times in the Trichodectidae, three of those times in the Trichodectes-Lutridia
clade.
In most Psocodea the parameres are not fused together, but such fusion is present, probably
apomorphically, in a number of species of Trichodectidae (character 13:1). In some species
parameral fusion is difficult to observe, as the portion of the permanently-everted endophallus
lying between the parameres is faintly sclerotised, giving the impression that the parameres are
fused together; fusion has probably developed in some cases through sclerotisation of the
endophallus. The distribution of other apomorphies suggests that parameral fusion exhibits
more homoplasy than any other apomorphy in the analysis, being derived 24 times and lost once.
In the Lorisicola (P.) bengalensis-juccii clade the parameres are closely associated with each
other but are not fused (Figs 223, 224), although fusion has sometimes been assumed (e.g.
Werneck, 1948). This proximity is believed to be autapomorphic for the clade (character 13:1').
The form of the fused parameres (parameral plate) may be apomorphic for groups of species
(characters 15, 17).
In a few Trichodectidae the parameres and mesomeres are fused, a probable apomorphy
(character 11). The distribution of other apomorphies and slight differences in the fusion pattern
(Figs 60, 74, 75, 81, 225) indicate some homoplasy in the character.
As described above, the mesomeres are frequently fused apically in the Psocodea, and
consequently this fusion, when found in the Trichodectidae, is deemed to be plesiomorphic.
Loss effusion (i.e. reduction to two unfused mesomeres) is therefore believed to be apomorphic
within the Trichodectidae (character 33), and distribution relative to other apomorphies
indicates that it has occurred several times in the family (see cladogram). In most species of the
Procaviphilus-Eurytrichodectes clade there is a lateral desclerotisation on each side of the
mesomeral arch (Figs 118, 122, 135). This character state is not found elsewhere and this, its
structure, and the distribution of other apomorphies, all indicate its apomorphic status
(character 34). The mesomeral arch may also have lateral flexions (Figs 107, 108, 225), which
can give the arch the appearance of being broken (see Werneck, 1948). This modification is
found in species of the Lorisicola (P.) lenicornis-neoafricanus clade and of Procavicola
( Condylocephalus) , which on the basis of other apomorphies are widely separated on the
cladogram; the lateral flexion of the mesomeral arch (character 35) is consequently believed to
be a convergent apomorphy in the two clades named. In most species of the genus Eutrichophi-
lus the mesomeral arch is divided into three parts by total desclerotisations laterally (Figs 91,
93). This feature is unique within the Phthiraptera and therefore considered apomorphic
(character 36).
In many Trichodectidae a rod-like sclerite terminating posteriorly in a Y-shape or a broad
plate is present longitudinally between the b.a.l.s. (Figs 55, 82, 148). Although this sclerite (the
TRICHODECTID MAMMAL LICE 215
'central sclerotisation') is very clear in some species, it is poorly sclerotised in others, and may be
either absent or obscured by the sclerotisations of the endophallus in mounted specimens. The
uncertainty attached to the observation of this structure has precluded its use in analysis,
although it may be of value taxonomically. The central sclerotisation is probably a derivative of
the basal apodeme, developed for muscle attachment.
A sclerite or pair of sclerites is present anteriorly to the parameres in some Trichodectidae,
and are referred to here as the 'basiparameral sclerites' (Figs 107, 225) (character 10: 1+2). They
may be fused to the parameres (Fig. 225) or to each other (Fig. 93). Their presence is believed to
be apomorphic and their derivation is probably from the basal ventral flanges of the parameres
found apomorphically (character 28) in a number of Trichodectidae (Figs 118, 122). If they are
formed (as suggested here) by detachment of the flange from the main body of the paramere
(Fig. 21), fusion of the sclerites and the parameres is a stage in the transformation series to the
development of free sclerites, but fusion of the basiparameral sclerites to each other is a
'terminal' apomorphic state (although it is not used in the cladistic analysis because of its
extremely limited distribution).
The anterior end of the basal apodeme may be heavily or lightly sclerotised, or apparently not
sclerotised at all. This degree of sclerotisation is very susceptible to modification during
preparation of the specimen, and thus is difficult to assess accurately. The character is not used in
analysis. The most prominent features of the basal apodeme are the lateral struts (b.a.l.s.),
which are generally fairly heavily sclerotised. These struts may approach the anterior of the
basal apodeme in parallel, convergently or divergently; this character is not used in cladistic
analysis because of the difficulty in assigning polarity to the different forms, but is useful
taxonomically, and can help in the determination of the form of the anterior margin of the basal
apodeme. This anterior margin may be straight or broadly convex, shallowly concave, very
deeply concave (Fig. 52) or acuminate (Fig. 173) (character 2). Of these forms the last two are
almost certainly apomorphic within the Trichodectidae; the elongation of the apodeme and the
concomitant parallel-sided concavity is found only in the Bovicola alpinus-tibialis clade
(character 2:1), and the acuminate form is found only in three species of Trichodectes , although
in fact the latter apomorphy imparts little useful information for the construction of the
cladogram. The polarity of the character for the other three states is difficult to assess, and they
are therefore not used in phyletic analysis.
In some species of Damalinia (Tricholipeurus) the b.a.l.s. develop a lateral spur before the
junction with the parameres, probably at the point at which the dorsal and ventral layers of the
basal apodeme separate; this feature (the 'anteposterior spur', Fig. 83) is not found elsewhere in
the Trichodectidae and is considered apomorphic (character 5). The posterior ends of the
b.a.l.s. are most frequently not, or only slightly, expanded laterally, but in some Trichodectidae
they are greatly broadened and scoop-shaped. This very broad form, whilst believed to be
apomorphic, is not used as an apomorphy in phyletic analysis because of difficulties in delimiting
the state. The posterior forking of the b.a.l.s. (Fig. 173) is also considered apomorphic
(character 3), but was probably developed twice in Trichodectes (see cladogram). The basal
apodeme probably extends anteriorly as far as segment VI in the plesiomorphic state, but in a
few Trichodectidae it extends up to segment II; in some cases this lengthening has been
accompanied by a width restriction or 'waisting' medially, and this is believed to be apomorphic
(character 6). Other features of the basal apodeme are, by virtue of their restricted distribution
and concordance with the distributions of other apomorphies, believed to be apomorphic
(characters 4, 7, 8, 9).
Whilst it is not possible to be certain of the plesiomorphic form of the parameres in the
Trichodectidae, it is assumed that this is fairly unspecialised, and that the forms of the parameres
found in groups of species that are also linked by other apomorphies are apomorphic (characters
20-27).
As noted above, apical fusion of the mesomeres is plesiomorphic for the Trichodectidae, so
loss of apical fusion, reduction in size and complete loss of the mesomeres are all considered to
be apomorphic within the family (characters 30, 31, 33). The presence of a median longitudinal
extension to the mesomeral arch is also considered plesiomorphic, as it is present in a number of
216 C. H. C. LYAL
taxa outside the Trichodectidae. Loss of this extension, or modification of its form from 'simple
lanceolate' (Fig. 118), are considered apomorphic within the family (characters 47-49). In the
plesiomorphic state the mesomeres articulate basally with the basal apodeme; articulation of the
basal apodeme with any other part of the mesomeres is considered apomorphic. Mesad
extension of the mesomeres between the b.a.l.s. (character 32) has apparently arisen twice, once
in the Dasyonyginae, and once in the Lorisicola, and in each case providing an autapomorphy
for the clade named. In Lorisicola (P.) bengalensis andphilippinensis the parts of the mesomeres
between the b.a.l.s. are apomorphically deflected posteriad (Fig. 224) (character 37). A similar
recurving of the mesomeres occurs elsewhere in the Trichodectidae (character 45), but in this
case the mesomeres are exterior to the b.a.l.s., and their recurved portions lie ventrally to the
b.a.l.s. This apomorphy is seen as a transformation series of states in Damalinia (Tricho-
lipeurus) (character 45:1-45:2), the most apomorphic of which (45:2) is also exhibited by
Lorisicola mjoebergi, although in this case the recurved parts of the mesomeres are very difficult
to see (Fig. 219). Other modifications to the mesomeres believed to be apomorphic are present
in restricted groups within the Trichodectidae (characters 38-46).
The endophallus may be sclerotised in a number of apomorphic ways within the Trichodecti-
dae (characters 51-56).
Female genitalia (characters 57-97)
The gonapophyses of most Trichodectidae and many other Psocodea bear at least some setae,
which arise directly from the structure and not from tubercles. Absence of setae (character 57)
and development of sclerotised setal tubercles (character 58) are therefore both believed to be
apomorphic within the Trichodectidae. Setal tubercles are found in Protelicola, Procaviphilus
and the Trichodectes-Neolutridia clade (T-N), but as a sister-group relationship between the
latter two is not supported by other apomorphies, and the form of the tubercles differs between
the two clades (Figs 111, 157), the character is probably convergent. The relationship between
Protelicola and the T-N clade is discussed below. The characteristic pattern taken by the
tubercles in each clade is modified by loss (character 60) or fusion (character 59); in both cases
these are believed to be apomorphic modifications because of their concordance with other
apomorphies. Tuberculate setae are also found on the ventral vulval margin of most species in
Trichodectes (Fig. 157); concordance with other apomorphies suggests the apomorphy of this
character (character 74).
The plesiomorphic form of the gonapophyses is not certain, but some forms, because of their
very restricted distribution, are believed to be apomorphic (characters 61, 62, 63); some
convergence in character 62 is indicated by the distribution of other apomorphies. The
development of a lobe on the ventral margin of the gonapophysis is restricted to the Trichodecti-
dae and, for this reason, is believed to be apomorphic within the clade (character 64).
Distribution of other apomorphies suggests that the lobe developed independently in several
different clades, sometimes taking only one form in a clade (character 70), sometimes being
apomorphically modified (characters 66-69). The reduction of the 'spur' - the portion of the
gonapophysis distal to the lobe - is considered apomorphic, as it is confined to two small groups
of taxa within the family (character 65). Where present, the gonapophyses in most Phthiraptera
meet the ventral vulval margin at an angle (Figs 94, 175), but in some Trichodectidae they meet
in a smooth curve (Fig. 154), which may be sclerotised (characters 71, 72). The ventral vulval
margin may extend in a more or less smooth curve between the gonapophyses, as is most
frequently the case in lice with gonapophyses, or it may be produced in some manner (characters
75-78). Each of these projections is considered apomorphic, although the distribution of other
apomorphies suggests that some are homoplastically developed in different clades of the family.
The distinction between the subgenital lobe (character 78) and the expansion of the ventral
vulval margin (character 75, and its apomorphic derivative, character 76), may not be im-
mediately clear, but whilst the former term is applied to structures that arise abruptly from the
margin, the latter is a more extensive posterior production of the whole of the margin. Both of
these apomorphies occur more than once in the Trichodectidae. The form of the subgenital lobe
is variable, though frequently it is marginally serrate, sometimes with the serrations greatly
TRICHODECTID MAMMAL LICE 217
developed (character 79:2). Several other apomorphies, of restricted distribution within the
Trichodectidae, are found in the form of the subgenital lobe (characters 80-83, 85, 86). Two
probable apomorphies, the presence of an internal sclerite in the subgenital lobe and the
presence of lateral setal patch, are not used in the cladistic analysis. The internal sclerite is not
readily observable, but this is probably due to the sclerite being rendered undetectable during
preparation of the specimens (especially the smaller species), and this likelihood precludes its
use. The distribution of the sclerite, where detected, suggests it to be plesiomorphic within the
Trichodectinae, possibly linked to the development of the subgenital lobe in this clade. The
lateral marginal setae appear to share part of the distribution of the lateral processes of the lobe
(character 81), being absent in a few species only, and the two characters are probably closely
associated; for this reason the setal character is not used.
The plesiomorphic form of the genital chamber in the Trichodectidae is not known, but
observations on other Phthiraptera suggest light sclerotisation with a few internal spicules. The
development, in some restricted groups of Trichodectidae, of particular patterns of spicules,
scales, spines and broad sclerotised areas (characters 92-96) is considered apomorphic.
Reproduction (character 98)
Parthenogenetic reproduction (character 98) occurs in a few Trichodectidae, mostly in the
Bovicolinae, but also in the species Geomydoecusscleritus. As all other Phthiraptera reproduce
bisexually, the character is taken as apomorphic. Parthenogenesis appears to have developed at
least four times in the Trichodectidae.
Male terminal abdominal segments (characters 99-109)
The sclerites of the terminal segments of the male trichodectid abdomen are very variable in
presence or absence states, extent, and degree of subdivision. This variability is not, in many
cases, readily associated with transformation series of other characters to which polarity has
been applied, and the plesiomorphic state (and hence apomorphic states) of the characters of
these sclerites is, in most cases, not known. In a few instances the sclerites are distinctly modified
in a restricted group of species, and thus polarity can be assigned (characters 99, 101, 106, 107,
108, 109).
Segment IX and the genital opening are apomorphically positioned more or less dorsally in
many species of Trichodectidae, as discussed above. Distortion due to preparative processes
obscures the position in many of the specimens examined, however, and characters associated
with this positioning cannot be used with any confidence, and are excluded from analysis.
In some species of Trichodectidae the posterior margin of tergum IX is greatly expanded to
produce a double convex lobe (character 100); this development is believed to be apomorphic,
although the distribution of other apomorphies indicates that it developed twice.
The presence of pseudostyli, discussed in detail above, is believed to be apomorphic for the
Trichodectidae (character 102). The plesiomorphic form of the pseudostyli is not known, as the
extant forms are very variable and cannot, in most cases, be resolved into transformation series.
In two cases (characters 103, 104), the pseudostyli are of very distinct form and restricted to
groups of species believed to be holophyletic on other grounds; these character states are
believed to be apomorphic. The presence of a single projection posteriorly on segment IX in
some species of Trichodectidae, believed to be formed of fused pseudostyli, is also considered
apomorphic (character 105). Although found in only three species, the distribution of other
apomorphies suggests two independent developments of this character state.
Antennae (characters 110-124)
In all Trichodectidae the male flagellomeres are fused together (character 111), a state found
elsewhere only in the anopluran families Echinophthiriidae and Hamophthiriidae and therefore
considered apomorphic for the Trichodectidae. Fusion of the flagellomeres has also occurred in
some female Trichodectidae (character 110), but the distribution of other apomorphies suggests
four homoplastic derivations of this apomorphy in the family. The expansion of the scape in the
male to house the enlarged musculature is probably plesiomorphic for the Trichodectidae, as
218 C. H. C. LYAL
similar expansion is found in many Ischnocera and Anoplura. Reduction of this expansion is,
however, apomorphic within the family (character 112), and is believed to have taken place
three times. The flagellum of most male Trichodectidae bears a number of setae modified into
broad pointed 'teeth' (Figs 13, 231), a feature not found in the same form in any other
Phthiraptera, and therefore considered apomorphic for the family. The plesiomorphic number
of 'teeth' is almost certainly two, as this number is the most common in all groups of
Trichodectidae; any variation from this number (to zero, one, three, four or eight) is believed,
therefore, to be apomorphic (character 115). The loss of the basal articulation of the 'teeth'
(character 117) and the development of a supporting protuberance (character 116) are both
apomorphic. In order that the male antennae should clasp the female with maximum efficiency
the 'inner' (posterior) surface of some or all of the antennal segments may be roughened or bear
projections; such developments are considered apomorphic in each form (characters 118-121).
The presence of a membranous projection on the female antenna (character 122) is also believed
to be apomorphic. In most Phthiraptera the setae of the scape are scattered over its surface in no
coherent pattern, and this is true of some Trichodectidae (Fig. 70); in most Trichodectidae,
however, the setae of the dorsoposterior surface of the scape are apomorphically arranged in a
line along the segment (Fig. 13) (character 113). In some cases where a row of setae might be
expected from the construction of the cladogram, the number of the setae involved in the
putative row is only two, and this is believed to represent an apomorphic reduction in number
(character 118).
The sensilla of the antennae in Trichodectidae and other Phthiraptera have been discussed
above; the presence of a fringed pit surrounding the sensilla of the flagellum in two species of
Trichodectidae is unique and believed to be apomorphic (character 124).
Head (characters 125-128)
Although the sitophore sclerite is variable in most Phthiraptera (Haub, 1973), it is comparative-
ly uniform in the Trichodectidae. The form found in most Trichodectidae (Fig. 11) is believed on
this account to be plesiomorphic, and is departed from in Bovicola (Spinibovicola), Dasyonyx
and Eurytrichodectes , where the posterior arms are extended (Fig. 12) (character 125). The
distribution of other apomorphies indicates that the modification is convergent in Bovicola
(Spinibovicola) and Dasyonyx plus Eurytrichodectes.
The posterior margins of the temple are generally broadly rounded in Trichodectidae, but in
species of the genus Eutrichophilus the convexity is much greater than in the rest of the family
(Fig. 87); this development is believed to be apomorphic (character 127). In the three species of
the genus Eurytrichodectes (only two of which are described) and the four of Procavicola
(Condylocephalus) the posterior temple angles are developed into pointed projections (charac-
ter 126), these being very long in the former genus (character 126: 2). This modification is not
found elsewhere in the Trichodectidae, although small rounded projections are found in some
Dasyonyx spp. and some Damalinia spp. The presence of pointed projections is believed, on the
basis of the distribution of other characters, to be homoplastic in the two genera mentioned.
The form of the osculum has been largely excluded from consideration in the cladistic analysis
because of the direct influence of the hair of the host (see above). However, in the Damalinia
(D.) theileri-baxi clade it is quite different from other species of Trichodectidae (Fig. 69), and is
here suggested to be apomorphic (character 128).
Legs (characters 129, 130)
The loss of one tarsal claw on each leg (character 129) is an apomorphy associated with
ectoparasitism on mammals. This character is proposed as an autapomorphy of the Trichodecti-
dae, although it may be autapomorphic for a postulated holophyletic group comprising the
Anoplura, Rhyncophthirina and Trichodectidae (Lyal, 1985).
A number of Psocodea have teeth on the 'inside' face of the tarsal claws, and many
Trichodectidae have what appears to be a single basal tooth (Fig. 15). The occurrence of teeth all
along the 'inside' face of the tarsal claws (Figs 14, 15) is restricted in the Trichodectidae to
Dasyonyx, .and is believed to be autapomorphic for the genus (character 130). The two
TRICHODECTID MAMMAL LICE 219
subgenera of Dasyonyx have tarsal claw teeth of different forms: D. (Dasyonyx) have sharp
slender teeth (Fig. 14), whilst D. (Neodasyonyx) have blunter, broader teeth (Fig. 15). These
two forms may be co-apomorphies, indicating that the subgenera are sister-groups, or one may
be the plesiomorphic state. No other characters have been found within Dasyonyx that indicate
reliable sister-group relationships within the genus. In this study the two tooth forms are
accepted as co-apomorphies and the subgenera retained, but further work on the genus may
cause this hypothesis to be challenged.
Postcoxale (character 131)
In most species of Trichodectidae the metathoracic postcoxale is either not sclerotised or
sclerotised and small, but the polarity of the transformation series with the extreme states
'sclerotised' and 'not sclerotised' is not known. In Procaviphilus (Meganarionoides) and some
species of Dasyonyx (Dasyonyx) the postcoxale is greatly enlarged and heavily sclerotised, an
apomorphic state not included in the analysis for reasons given below. A further apomorphic
condition, the fusion of the postcoxales, is also found in some members of the same subgenera.
The presence of the sclerotised postcoxale is difficult to determine in some of the smaller species
of Dasyonyx, but in any case the distribution of other apomorphies indicates convergence of the
postcoxale characters in the two subgenera. Neither apomorphy is used in the analysis.
In P. (Meganarionoides), uniquely, the postcoxale is fused to abdominal pleurum II (charac-
ter 131), and this apomorphy is used in the analysis.
It is interesting that the great development of the postcoxales, restricted to lice parasitic on
hyraxes, is morphologically convergent on the development of the apophysis of abdominal
sternum II (character 132), which is found in another group of hyrax lice.
Spiracles (characters 133-138)
For the Phthiraptera (and the Trichodectidae) the plesiomorphic number of spiracles is a single
thoracic pair and six abdominal pairs; further reduction in the number of abdominal spiracles is
apomorphic. The numbers of abdominal spiracles in the different species of Trichodectidae are
summarised in Table 1. Each reduction is considered to be an apomorphy (characters 135-138),
though some homoplasy has occurred. Inspection of the distribution of other apomorphies
indicates that reduction to five, four and one pair of spiracles has occurred once, reduction to
three and two pairs twice, and reduction to none eight times. Because of the sequential pattern
of spiracle loss, apomorphy 136 is always associated on the cladogram with apomorphy 135, 137
with 136 and 135, and 138 with 137, 136 and 135.
In most species of lice all of the abdominal spiracles have atria of roughly the same size; as has
been pointed out in earlier discussion, however, some Trichodectidae have posterior spiracles
with atria much smaller than those more anterior on the abdomen (character 134). This
difference in size is believed to be an apomorphic reduction.
In most species of lice the atrium of the thoracic spiracle is as broad or broader than long;
species of the genus Cebidicola, however, have a tubular atrium associated with the thoracic
spiracle (character 133). This modification of form is believed to be apomorphic.
Abdominal setae (characters 140-156)
Study of abdominal setal patterns throughout the Psocodea suggests that the plesiomorphic
pattern is a row of setae running around the abdomen on each of segments I to VIII. On
trichodectid abdominal pleura II-VII this row of setae (referred to here as the posterior setal
row or 'p.s.r.') is generally clear, and is absent in only a few species. The distribution of other
apomorphies indicates this absence to be apomorphic, although limited to a very few, distantly-
related, species. In some Trichodectidae, the p.s.r. of pleura II, III and IV comprises setae that
are much stouter than those of other pleura, and the distribution of other apomorphies suggests
the apomorphic status of each of these, although each exhibits some homoplasy. Preliminary
analysis indicated that specialisation of the p.s.r. on pleura II and IV conveys little phylogenetic
information, and only character 139 (specialisation of the p.s.r. on pleurum III) is used in the
final cladistic analysis. The setae on pleura VIII and IX are frequently longer than the setae of
220 C. H. C. LYAL
the p.s.r. on anterior pleura; the extreme length of these setae in Eutrichophilus is, however,
recognised as apomorphic (character 140).
The setal row on sterna II and III is usually similar to the row on other sterna, but in the two
species of Eurytrichodectes the setae of these two rows are short, stout and conical (Fig. 136).
This unique feature is believed to be apomorphic (character 153).
The tergal setal row of many Trichodectidae, especially males, is clearly composed of four
discrete groups - two lateral and two median - each separated by a gap (Fig. 22) . The positioning
of the groups and the number of setae in them are useful taxonomic characters, and may be
utilised as landmarks for the identification of particular setae. The groups are, however, difficult
to use in phyletic analysis because of the difficulty of assigning polarity to any transformation
series.
In some Trichodectidae (and in no other Phthiraptera) a seta- termed here the 'posterolateral
seta' or 'p.l.s.' - is present posterolaterally on each side of terga II-VI (Fig. 22). The restriction
of distribution of this seta within the Phthiraptera suggests that its presence is apomorphic
(character 151). In some cases there is more than one p.l.s. on each side of the tergum (Fig. 201);
this is believed to be an apomorphy but its sporadic occurrence (in terms of clades indicated by
other apomorphies) has led to its omission from the cladistic analysis. The presence of the p.l.s.
is difficult to assess in some species, either because the lateral group may be reduced in number
or because the lateral group is composed of very long setae. In the former case, a single seta in
the position of the p.l.s. may be this seta (the lateral group being absent), or it may be the sole
remaining seta of the lateral group (the p.l.s. being absent) (Fig. 159). In the latter case (most
Trichodectes species and the Neotrichodectinae), the most lateral seta of the lateral group
frequently lies slightly posterior to the rest of the row (Fig. 158) and a more differentiated p.l.s.
is absent. In both these cases the p.l.s. is postulated to be present, though modified. The
anterior setae
posterior setal
row
postero-lateral
seta
lateral
gap
median
gap
median setal
group
lateral setal
group
Figs 21, 22 Trichodectid morphology. 21, postulated evolution of basi-parameral sclerite by detachment
of basal flange of paramere. 22, Abdominal setal arrangement, illustrated by anterior terga and pleura of
male.
TRICHODECTID MAMMAL LICE 221
distribution of other characters suggests, however, that secondary loss of the p.l.s. has occurred
within some taxa.
Some setae of the median tergal group, particularly in males, may be specialised. In males of
Neotrichodectes spp. the two central setae of the united median groups, or two setae very near
the centre (perhaps separated by one or two unmodified setae) are very much smaller than the
other setae of the row (Fig. 229). These 'tergocentral microsetae' (character 152) are found
nowhere else, and this and the concordance of their distribution with that of apomorphy 118:
l'+2' suggests the apomorphy of the character state. In Felicola (S.) pygidialis and F. (S.)
macrurus the median group of tergum III is modified in a distinctive manner in the males (Fig.
178), and this modification is assumed to be apomorphic for the two species (character 147).
Some or all of the setae of the median group on terga II and III of male Trichodectidae may be
enlarged relative to the other tergal setae. This enlargement occurs sporadically both within and
outside the family, and each case is believed to be autapomorphic. Within the Trichodectidae, in
males of Geomydoecus (Thomomydoecus), G. (G.) copei, Trichodectes (Paratrichodectes)
ovalis and ugandensis, and the undescribed sister-subspecies of Trichodectes (T.) galictidis, the
setae of the median row on both terga II and III are enlarged but remain in a straight row (Figs
161 , 244) ; this arrangement is believed to be apomorphic, but probably convergent in each of the
four groups (character 149). In Bovicola (Spinibovicola) hemitragi and multispinosa a similar
enlargement is confined to some of the setae of tergum II, and the lines are curved (Fig. 41)
(character 150). In males of Felicola, the holophyly of which is supported by several apomor-
phies, the median setal group is reduced to a single, greatly enlarged seta (Fig. 188). The
apomorphic status of this character (character 142) is indicated by its restricted distribution and
correlation with other apomorphies. It is notable that the setae are single, but of normal size
(very small) in the Felicola (F.) rahmi-viverriculae clade (Fig. 183), and secondarily increased in
number to six in Felicola (S.) bedfordi and F. (F.) setosus (Figs 180, 189). On the basis of other
apomorphies, the former is believed to be a single autapomorphic reversal, whilst the latter is
believed to be a convergent gain.
The sclerite from which the pair of setae arises may be long and of characteristic shape (Fig.
181); this feature is found only in conjunction with the enlarged setae (character 142) and is
postulated to be apomorphic (character 155).
In many males of Felicola the median setal group on terga III-VII is also reduced to a single
seta, although in most cases this does not approach the size of the seta of tergum II. In the
Felicola (S.) cooleyi-quadraticeps clade this reduction has taken place on terga III-VII, but the
setae are similar in length to those of tergum II, the latter being reduced relative to those of other
species of Felicola and the former enlarged (Fig. 186) (character 143). This apomorphy is
convergent on the apomorphic setal pattern of the males of Trichodectes (Stachiella) (character
144), although in this case the setae are all generally long and stout. The median setal group of
the female tergum may also be reduced to a single seta (character 145) or lost (character 146).
The distribution of these female apomorphies is as follows. The reduction of the median group
to a single seta is found only in the Trichodectes (S.) fallax-octomaculatus clade; the sister-
species, T. (S.) potus, and the sister-group to this clade, the T. (S.) retusus-mustelae clade, lack
the female median group entirely. The sister-species to the whole T. (S.) retusus-potus clade, T.
(S.) emeryi, has the median group unreduced, numbering three setae, on terga I and II, reduced
to one seta or absent on tergum III, and absent on terga IV- VIII. It is not certain whether setal
loss in the female has taken place only once, the setae being regained in the fallax-octomaculatus
clade, or has taken place independently three times (in emeryi, the retusus-mustelae clade, and
potus).
The length of the abdominal setae is difficult to employ in phyletic analysis because of the
problem of establishing the polarity of the transformation series 'very short - medium - very
long'. The restricted distributions of the two extremes of the series (concordance with other
apomorphies) indicate their apomorphic status, however. The very short, sparse setae (Fig. 183)
are found in no other Phthiraptera but the Felicola-Lorisicola clade and some Trichodectes spp. ,
and are probably apomorphic but convergent in the two groups. The very long refringent setae
of some Trichodectidae (Fig. 158) are considered apomorphic for a similar reason (character
222 C. H. C. LYAL
141), though in this case similar setae are found in some Philopteridae. These long setae
probably evolved twice in the Trichodectidae: once in Trichodectes and once in the Neo-
trichodectinae. The fine, long setae of the females of the Felicola (S.) cooleyi-quadraticeps clade
are found in no other group and are considered apomorphic (character 154).
The setal bases - the circular 'pits' of the setal articulations - are of fairly constant size relative
to the setae in most Trichodectidae. However, in the Bovicola (B.) alpinus-tibialis clade the
bases are noticeably large in relation to the setae, and seem to have a double margin. This
feature needs to be examined using the scanning electron microscope to elucidate its true
structure, but examination using the light microscope is sufficient to detect its presence. This
feature is here considered as apomorphic (character 156).
Abdominal pleural projections and modifications (characters 157-161)
In many Trichodectidae the dorsoposterior and/or the ventroposterior pleural angles project on
pleura II, III or IV (see discussion above, and Table 2). Projections on these segments of the
type found here do not occur elsewhere in the Phthiraptera, and are therefore considered as
apomorphic (characters 158-160). Preliminary analysis reveals that the projections on pleurum
III contribute no useful phyletic information, so the apomorphy is omitted from the final
cladistic analysis. Variation in the degree of development of the dorsal and ventral lobes of the
projection on pleurum IV is omitted for the same reason, except for the extreme development in
Eurytrichodectes (character 161). Both the presence of a dorsal and a ventral projection on
pleurum IV (characters 159 and 160 respectively) are included in the analysis, though the latter
apomorphy is reversed in some clades. The presence of a projection on pleurum II (character
158) provides a synapomorphy for Geomydoecus spp., which are also united as a holophyletic
group on other grounds; the projection on this pleurum is found convergently in Trichodectes
(Paratrichodectes) zorillae. The sclerotisation of the dorsal projection on pleurum II (character
158:2) is an autapomorphy of Geomydoecus (Thomomydoecus) (and some species in Geomy-
doecus s.str. - see discussion below -) and T. zorillae, but sclerotisation of the projections on the
other pleura is very variable, and is not used in cladistic analysis.
In Damalinia (Damalinia) pleurum II extends on to sternum II, and the pleurite is expanded
at the expense of the sternite (character 157). This extension may be broad (character 157:1') or
narrow (character 157:1), but the more plesiomorphic state of these two (should they not be
co-apomorphies) is not known. The species with a broad ventral extension of pleurite II also
possess a more or less extensive dorsal extension, but this is not found in species with a narrow
ventral extension. In this treatment the two forms of the ventral extension are used to
characterise each of two sister-groups, but this hypothesis is open to challenge, as the group
indicated by character 157:1 has no other supporting apomorphy.
Abdominal sclerae (characters 132, 162-164)
As noted above, in species of the trichodectid genus Procavicola sternite II is greatly developed
as a heavily sclerotised internal apophysis, articulated to pleurum II (character 132). The
presence of this unique structure is considered apomorphic.
The presence of the lateral flecks and their associated small sclerite is considered apomorphic,
as the structure occurs in no Phthiraptera other than the Trichodectidae (character 164).
The pleura, sterna and terga of the trichodectid abdomen may be sclerotised or not; although
some groups (identified on the basis of other apomorphies) may, in general, be more or less
sclerotised, the polarity of the transformation series 'sclerotised - not sclerotised' for each
segment cannot be determined, and these characters are not used in the cladistic analysis. The
male abdomen may have a characteristic sclerotisation dorsally, in that the terga may have
anterior and posterior sclerites; this feature is present in some Anoplura, but is probably
convergent in this suborder. The presence of doubled tergal sclerites in male Trichodectidae is
believed to be apomorphic for the family, but is not used in cladistic analysis because of the large
number of reversals. The tergal sclerites are not further modified in most male Trichodectidae,
but in some there is longitudinal division of the anterior or posterior sclerites (characters 162 and
163 respectively), this division being accepted as apomorphic.
TRICHODECTID MAMMAL LICE
223
D. (Trlchollpeurus)
Eutrichophllus
00) Dasyonyginae
i o
CD) Trichodectini
Felicola
_ - -* -C* -* CT) CD -r^
j^wojoj t/i oj <_n **
Neotrichodectinae
Fig. 23 Cladogram of Trichodectidae. Clades numbered 1-12 are resolved in Figs 24-34. The Eu-
trichophilinae (Genus Eutrichophllus only) is not resolved further. For explanation of numbered
apomorphies see text.
Cladistic analysis
The holophyly of both monotypic and polytypic species is accepted without the need for
justification, so species-level autapomorphies have not been indicated unless they are homoplas-
tic with character states elsewhere on the cladogram. Omission of the autapomorphies of species
saves both space in the data matrix and time taken for analysis, and for the same reasons many
sister-species pairs are justified on the cladogram with fewer autapomorphies than are available.
Of the 187 apomorphic character states used in the analysis, 86 are postulated to have been
developed more than once or to have been secondarily lost, 363 such homoplasies being
proposed. When, in the analysis, a choice is available between postulating one reversal or a pair
of homoplastic gains (i.e. three clades in a holophyletic group are involved and the topography
of the tree is not affected whichever the choice), the latter is chosen (e.g. character 13:1 in the
Damalinia theileri-appendiculata clade). This choice is made so that the distribution of
apomorphic character states can more easily be discerned on the cladogram. The number of
224
C. H. C. LYAL
Blsonlcolg s. seaeclmdecembrll
Fig. 24 Cladogram of Bovicolinae (part) (clades 1-4 of Fig. 23: genera Bovicola, Bisonicola, Tragulicola
and Werneckiella). For explanation of numbered apomorphies see text.
TRICHODECTID MAMMAL LICE
225
D. (D.) oppendlculata
Fig. 25 Cladogram of Damalinia (Damalinia) (clade 5 of Fig. 23). For explanation of numbered
apomorphies see text.
homoplasies could be slightly reduced without affecting the topology of the tree because, as
explained below, the less parsimonious presentation is sometimes chosen to make the clado-
gram more informative and less potentially misleading. In a number of places in the discussion of
apomorphic state identification above, reference is made to single apomorphic character states
arising more than once. This is superficially contradictory, especially if the terms apomorphy
and homology are equated. Such situations have been detected during one 'round' of reciprocal
illumination. Should all such convergent apomorphic states be receded as separate apomorphies
(which they are believed to be, even if they cannot be differentiated morphologically) the
cladogram would appear more parsimonious.
The loss of the median extension of the mesomeral arch (character 47) is placed on the
cladogram 17 times, frequently in combination with the loss of apical fusion of the mesomeres
(character 33). These apomorphies are not arranged in the most parsimonious manner on the
cladogram, as can be seen by inspection of the Bovicolinae. As presented, the cladogram depicts
the loss of the extension 11 times in this subfamily. A more parsimonious arrangement of the
apomorphies is achieved by postulating characters 33 and 47 as synapomorphic for Damalinia
(Damalinia) and Damalinia (Cervicola), and character 47 as synapomorphic for two clades:
Werneckiella plus Tragulicola and Bisonicola, and Bovicola (Bovicola) plus B. (Lepikentron)
and B. (Spinibovicola) . This arrangement reduces the number of proposed homoplasies of
character 47 to six within the Bovicolinae, and reduces the number of polychotomies on the
cladogram. Alternatively, the loss of the extension might be postulated to have occurred only
once, in the common ancestor of the Bovicolinae, and regained six times (B. crassipes, D.
226
C. H. C. LYAL
-* O)
' ro
3 9
~* ~*-
+
PO
00
CO
r r\> (J *
o
Q
o CTI
ro co co -*
OJ -* -^ O
Fig. 26 Cladogram of Damalinia (Cervicola) (clade 6 of Fig. 23). For explanation of numbered
apomorphies see text.
r-CH
-+9-
)la extrorius
C, semlarmatus
Procovicola (P.) 16 species
P. (Condylocephalus) 1 indf leldi
P. (C.) hopkinsl
Fig. 28 Cladogram of Dasyonyginae (clade 8 of Fig. 23, part: genera Cebidicola and Procavicold) . Clade
13 is resolved in Fig. 29. For explanation of numbered apomorphies see text.
TRICHODECTID MAMMAL LICE
227
co en - co
' o-9-m-
mallnla (Trlcholipeurus) aepycer
Fig. 27 Cladogram of Damalinia (Tricholipeurus) (clade 7 of Fig. 23). For explanation of numbered
apomorphies see text.
elongata, D. moschatus, D. clayi and the D. albimarginata-indica clade). The most parsimo-
nious hypothesis is that the structure was lost in the ancestor of Bovicolinae as suggested above,
regained twice (B. crassipes and the D. (T.) albimarginata-elongata clade), and secondarily lost
twice (D. (T.) lineata-victoriae and D. (T.) pakenhami-bedfordi). This last hypothesis, although
more parsimonious than the distribution on the tree presented, does not change the topology of
the tree. The distribution of character 47 is not as apparent from inspection of the tree in its most
parsimonious distribution as it is in the tree presented, as the more scattered distribution of the
losses and reversals obscures the alternative possible distributions and implies a spurious
confidence in the tree as supported by them.
The distribution of character 33 (the loss of mesomeral fusion) in Werneckiella is not
presented in the most parsimonious manner. There is great difficulty in the observation of this
character state in Werneckiella, and the morphological difference between 'loss of fusion' and
'fusion' is very slight. A detailed examination of the species of this genus for other characters to
228
C. H. C. LYAL
, 00
00 CO -^ O
Fig. 29 Cladogram of Dasyonyginae (part; clade 13 of Fig. 28: genera Procaviphilus , Dasyonyx and
Eurytrichodectes). For explanation of numbered apomorphies see text.
TRICHODECTID MAMMAL LICE
229
--m-
Trlchodectes
CO O) -
01 co y;
Fig. 30 Cladogram of Trichodectini (clade 9 of Fig. 23, part: genera Protelicola, Lutridia, Neolutridia and
Werneckodectes). Clade 14 is resolved in Fig. 31. For explanation of numbered apomorphies see text.
complete a full analysis was not made, character 33 only being noted because it occurs elsewhere
on the cladogram. It is possible but not likely that the distribution of character 33 as observed is
supported by other apomorphies, but the proposal of holophyletic groups within the genus on
the basis of the observations made of this single character would be unwise. It is notable that
Werneckiella fulva and W. neglecta, which differ in the state of character 33, are otherwise very
similar, the females apparently being indistinguishable (Emerson & Price, 1979), and it is very
likely that they are sister-species.
The arrangement of Protelicola, Lutridia and the Trichodectes-Neolutridia clade (T-N) on
the cladogram (Figs 30, 31) does not accord with the most parsimonious distribution of the
apomorphies. The cladogram contains four convergences for 'gain' apomorphies: 12:1' (fusion
of parameres and b.a.l.s.) is postulated as homoplastic in Lutridia, Neolutridia and Trichodectes
(Stachiella); 13:1 (fusion of parameres to each other) is postulated as homoplastic in Protelicola
and Lutridia; 20 (development of rod-shaped parameres) is postulated as homoplastic in
Protelicola and Lutridia; and 58 (development of tubercles for the gonapophysis setae) is
postulated as homoplastic in Protelicola and T-N. Apomorphy 12: 1 ' has been discussed in detail
above, and the distribution suggested in the cladogram is believed consistent with the morpholo-
gical evidence. Apomorphy 58 could be considered in two ways other than that presented: as an
autapomorphy supporting the sister-group relationship of Protelicola and T-N, or as an
autapomorphy of the Trichodectini (the Trichodectes-Protelicola clade), reversed in Lutridia.
The first alternative is not supported by the distribution of any other apomorphies, whereas the
two alternative arrangements are each indicated by more than one apomorphy (see below); the
sister-group relationship of Protelicola and T-N is therefore rejected. The plesiomorphic
arrangement of the gonapophysis tubercles in T-N is clearly distinct from the arrangement in
Protelicola. If the tubercles are postulated to be homologous in the two clades two further
apomorphies (the form of the tubercles in each clade) would have to be proposed, as neither
form appears to be plesiomorphic with respect to the other. This manipulation does not affect
the topology of the cladogram (whatever the position of Lutridia), and does not clarify the
relationships of the clades involved, so the hypothesis of convergence of character 58 in
Protelicola and T-N is retained. The other two apomorphies may now be considered together as
they both suggest the sister-group relationship of Protelicola and Lutridia. The alternative
hypothesis (of the cladogram as presented) is supported by apomorphies 72 (development of a
sclerotisation along the ventral vulval margin) and 47 (loss of the median extension of the
mesomeral arch). 'Loss' characters are given much less weight than 'gain' characters in this
analysis, so character 47 should be left out of consideration. The sister-group relationship of
Protelicola and Lutridia is therefore supported by two apomorphies and the relationship
proposed on the cladogram supported by one. As noted in the generic descriptions below,
however, an undescribed species of Protelicola has been seen which does not share apomorphy
20. The cladistic position of this species with respect to the other two species in the genus has not
230
C. H. C. LYAL
I O-OD+B
CH -si CO CO L
oo .0. r\j .0. -
Co r\> en -
-0-0
1 OO-O-B-B-H-
no -*-*
CJ en ui
j. oo to
oo A -I to en
** O> *
to f
CTJ u.
- A CO <0 - -
O CO CO CO - t CO
Fig. 31 Cladogram of Trichodectes (clade 14 of Fig. 30). For explanation of numbered apomorphies see
text.
TRICHODECTID MAMMAL LICE
231
Fig. 32 Cladogram of Felicola (clade 10 of Fig. 23) . For explanation of numbered apomorphies see text.
232
Ul ~%l CO 00 O3 GO
L. (L.) spenceri
L. (L.) sudamerlcanus
L. (L.) hercvnlanus
L. (L.) Siem
L. (Pnradoxuroecus) bengalensls
L. (P.) philipplnensls
L. (P.) luce 11
L. (P.) nspldorhvnchus
Fig. 33 Cladogram of Lorisicola (clade 11 of Fig. 23). For explanation of numbered apomorphies see
text.
TRICHODECTID MAMMAL LICE 233
a Geomydoecus (G.) unresolved species
G. (G.) copel
G. (G.) thonoiwus
G. (G.) duchesnensls
G. (G_. ) dakotensls
- G. (Thoniomydoecus) wardl
G. (I. ) a - z clade
Neotrlchodectes (N.) thoraclcus
N. (N.) mlnutus
N. (N.) osbornl
N. (N.) mephltldls
N. (N.) wolffhuegell
r o
N. (Trigonodectes) bgrbarge
N. (Nasulcola) pallldus
N. (Lakshmlnarayanella) gastrodes
N. (L. ) cunmilngsl
N. (ConeeatliQlg) chilensls
N. (C.) interruptofosclatus
N . ( C . ) semlstrlotus
N. (C.) arlzonae
Fig. 34 Cladogram of Neotrichodectinae (clade 12 of Fig. 23: genera Neotrichodectes and Geomydoecus).
For explanation of numbered apomorphies see text.
been determined because of the poor state of preservation of the specimens, but its existence
raises the possibility that character 20 is an apomorphy not of Protelicola but of only two species
within the genus (the alternative being a reversal in the undescribed species). Character 20 is
also homoplastically developed in Felicola and Bovicola (Lepikentron) . If this character is
disregarded, apomorphies 72 and 13:1 must be compared for their comparative likelihood of
homoplasy. Apomorphy 13:1 is homoplastically developed at 22 other points on the cladogram
whilst 72 is found elsewhere only in Bovicola (Lepikentron) . Apomorphy 72 should clearly be
given much more weight than 13:1 in construction of the cladogram, and 20 is considered of
uncertain value in view of the undescribed species of Protelicola. For these reasons the
cladogram is retained as proposed, even though it is not maximally parsimonious.
The dorsal projection of pleurum IV (character 159) is lost in the Trichodectini, but
postulated as secondarily regained in Werneckodectes and Trichodectes (Paratrichodectes)
zorillae. The form of the projection is different in the two species, however, which indicates the
independent development of the structure.
Whilst the genus Geomydoecus s.l. is almost certainly holophyletic, this probably does not
234 C. H. C. LYAL
apply to either of the two included subgenera (Fig. 34). The question of holophyly should be
addressed first in the smaller subgenus Geomydoecus (Thomomydoecus). All but one of the
included species (G. (T.) war.di) have characteristically asymmetric male genitalia (character
29), and are proposed on this basis to be a holophyletic group (the asymmetricus-zacatecae clade
or 'a-z clade'). G. (T.) wardi and the a-z clade share the following apomorphies: posterolateral
temple margin with single stout seta and associated shorter, finer setae (a character not included
in the data matrix); male parameral plate apically pointed (character 14); gonapophysis
smoothly continuous with ventral vulval margin (character 71); male abdominal terga II and HI
with median setal group comprising exceptionally long, stout setae (character 149); and pleural
projections sclerotised, especially in females (character 158:2). The possession of a single stout
temple seta is unique to these species, but may be a reduction from the two stout setae found in
this position in some Geomydoecus (Geomydoecus). Apomorphies 14, 71 and 149 are also
shared by G. (G.) copei, and this species has the mesomeral arch and parameral plate very
slender, approaching the shape of the genitalia of the a-z clade more closely than does G. (T.)
wardi; the posterolateral temple margin lacks any specially-modified setae, but this may be due
to secondary loss. Apomorphies 14, 71 and 158:2 are shared by the G. (G.) thomomyus-
dakotensis clade, but the male genitalia are considerably broader than those of G. (T.) wardi,
and the mesomeral arch lacks a median extension (an autapomorphy of the clade). This clade
has a further autapomorphy in the form of the posterolateral setae of the temple margin, which
comprise a single long fine seta and associated shorter fine setae. As with G. (G.) copei, the
plesiomorphic form of the temple setae is unknown, and could have been the form found in G.
(Thomomydoecus) . Other species of G. (Geomydoecus) have a single apex to the parameral
plate (character 14), but do not share any of the other apomorphies mentioned. G. (T.) wardi,
the a-z clade, and the G. (G.) thomomyus-dakotensis clade are all parasitic on Thomomys spp. ,
whilst G. (G.) copei is a parasite of Orthogeomys hispidus; both host genera are parasitised by
other members of Geomydoecus (Geomydoecus) .
The apomorphies listed above plainly do not support unequivocally any of the three possible
sister-group relationships of the a-z clade without invoking homoplasy to an unjustifiable
extent. It is apparent, however, that G. (Geomydoecus) is paraphyletic with respect to G.
(Thomomydoecus) and that the latter subgenus is possibly polyphyletic. A full phylogenetic
analysis of the 102 species and subspecies of Geomydoecus, which would have been necessary to
resolve the problem, was not attempted. The hosts of the genus are all geomyid rodents, the
systematic and taxonomic understanding of which is of questionable accuracy (Price, pers.
comm.). For the purposes of this study the subgeneric concepts proposed by Price & Emerson
(1972) are retained.
Taxonomic history of Trichodectidae
Burmeister (1838) divided the Mallophaga into two families, Liotheidae and Philopteridae, the
latter comprising the two genera Philopterus and Trichodectes . Kellogg (1896) proposed the
suborders Amblycera and Ischnocera for Liotheidae and Philopteridae (sensu Burmeister)
respectively, and erected the family Trichodectidae for the genus Trichodectes.
Mjoberg (1910) described Damalinia and Eutrichophilus , the second and third genera of
Trichodectidae, and Stobbe (1913a) described a fourth genus, Eurytrichodectes. Stobbe (1913b)
revised the family for the first time. Ewing (1929) described four further genera and provided a
key to all eight, although Ferris (1929) regarded E wing's new genera as of 'most dubious value'.
Bedford (1929, 19320, 19326, 1936) described a further 10 genera, two of which were junior
synonyms of genera proposed by Ewing (1929), thus bringing the total to 16; Ewing (1936)
provided a key to 14 of these. Keler (1938a) recognised 24 genera, 10 of them new (although one
of these had been published previously by Keler, 1934 as a nomen nudum). The two genera
omitted by Keler (1938a) were the same two previously omitted by Ewing (Cebidicola Bedford,
1936 and Lorisicola Bedford, 1936); the three species included in these genera were placed by
Keler with two others, also from primates, in his new genus Meganarion (although with the
proviso 'without, of course, intending to establish the congeneric status of these species'). Keler
TRICHODECTID MAMMAL LICE 235
(1938fl) provided a key to most of the genera described in his paper (with the sole exception of
Meganarion) and to many of the species. Werneck (1941) introduced the subgenus concept to
the taxonomy of Trichodectidae , describing three new subgenera in two of the four genera of the
family parasitic on hyraxes. During the decade following Keler's (1938a) review of the family a
number of new genera were described, bringing the total number of available names in the
genus-group to 43 by the end of 1948. Werneck (1948, 1950) reviewed all the Trichodectidae,
and recognised 20 genera (though one of these doubtfully) and three subgenera. No genera and
only one subgenus have been described since 1948. The most recent name to be added
(Lakshminarayanella Eichler, 1982) was published as a replacement name for a junior
homonym, and brings the total of available names to 45. There have been no revisionary works
of the family since those of Werneck (1948, 1950), although Hopkins & Clay (1952), when
cataloguing the 'Mallophaga', accepted 13 genera, some of these doubtfully, and Eichler (1963)
recognised 38 genera (with no subgenera).
The problem of establishing criteria by which taxa can be distinguished at the generic level
received early attention. An attempt to identify morphological characters for this purpose was
initiated by Bedford (1929). Bedford (1932) provided a more thorough discussion, and
concluded that whilst the shape of the head, the presence or absence of abdominal sclerites and
the form of the female gonapophyses were of value, the form of the male genitalia and the
number of abdominal spiracles provided no useful guide. Ewing (1936) came to a quite different
conclusion, regarding abdominal spiracle number as 'the most important generic character'.
Bedford (1939) realised the unworkability of any system involving a priori assessment of
morphological characters for generic discrimination, although he still felt that abdominal
spiracle number was not of value at the generic level, and noted that (morphological) generic
characters 'may not be very striking'. To supplement or replace morphological characters
Bedford (1939) made use of host data, the possibility of which was first discussed by Kellogg
(1913, 1914) and Harrison (19166). Bedford (1939) wrote: 'Before placing a species in a new
genus one should ask oneself: would it be possible to say from what kind of host the parasite was
taken off had it not been recorded? If it is impossible to answer the question, then one should be
justified in placing it in a new genus.' Hopkins (1941) used this principle to a certain extent in his
discussion of Felicola. He also discussed the morphological characters used for the discrimina-
tion of the genus from others, and pointed out that 'the singling out of one character [on which to
base genera] . . . only tends to obscure natural relationships'. Werneck (1936) perceived and
treated the problem of generic discrimination in a rather different way from those described
above. He noted that whilst the type-species (and sometimes a few species similar to the type) of
each of the described genera were quite distinct, other species showed intermediate characters.
The existence of these 'transitional forms' convinced Werneck that there was no validity in the
separate genera, so he synonymised them all (with the exception of some genera not found in
South America, which were outside the scope of the paper). Bedford (1939) regarded this action
as 'unwarrantable' and reinstated all the genera. Hopkins (1942, 1943) reviewed the characters
used to separate the genera of Trichodectidae parasitic on carnivores and antelopes respect-
ively. In each case he found annectent species as described by Werneck (1936), and took similar
action, though modified by the belief that louse genera should somehow reflect host taxa.
Hopkins (1942) therefore accepted three genera of Trichodectidae parasitic on carnivores and
later (1943) accepted one genus parasitic on antelopes. Werneck (1948, 1950), although less
influenced by the host data, recognised more genera than had Hopkins. He accepted the
morphologically 'distinct' species and species groups as genera, and placed annectent species in
the genus which they most closely resembled. Hopkins (1949) 'conceded subgeneric status to
many groups which seem likely to be accepted by systematists whose views . . . differ from mine'
and recognised 14 genera and 20 subgenera. Hopkins & Clay (1952) synonymised some of the
subgenera accepted by Hopkins (1949), but raised others to generic status. Ledger (1980) held
views similar to Hopkins & Clay (1952), although in some cases followed the views of Hopkins
(1949); the resultant generic arrangement still involved fewer genera than accepted by Werneck
(1948, 1950), and many more subgenera. Emerson & Price (1981), however, could 'find no basis
for rejecting the classification of Trichodectidae given by Werneck (1948, 1950)', and suggested
236
C. H. C. LYAL
that the 'question of genera vs subgenera will perhaps continue until Mallophaga have been
described from all likely hosts'. A number of other workers also follow these views. A third
group of taxonomists (e.g. Eichler, 1963; Zlotorzycka, 1972) have accepted not only all the
genera recognised by Werneck (1948, 1950), but also a number of genera that Werneck
considered as junior synonyms; the subgenus category is not used, however. The present generic
placement of most of the Trichodectidae is thus a matter of some contention and a review of the
variations in status of some genera and subgenera is presented in Tables 3-5.
a
1
1
i
I
o
5
|
o
Ou
3
c
'3
01
M
f.
u
M
C
1
c
3
a.
(t:
h
o
M
C
e-
1. Damalinia MjSberg, 1910
1
1
1
1
1
1
1
1
1
1
1
2. Bovicola Ewing, 1929
2
2
2
= 1
1 (2)
2
1 (2)
2
1 (2)
2
2
3. Trieholipeurus Bedford, 1929
3
3
3
= 1
1 (J)
3
1 (3)
3
1 (3)
3
1 (3)
4. Cervicola Keler, 1938a
4
4
1 (4)
= 1
= 1 6 =2
4
-1 s.l.
= 2
1 (4)
5. Holakartikos Kler, 1938a
5
5
-
5
= 2
= 2
5
=1 s.l.
= 2
2 (5)
6. Lepikentron Kgler, 1938a
6
6
-
6
= 2
1 (6)
6
=1 s.l.
= 2
2 (6)
7. Rhabdopedilon Keler, 1938a
7
7
-
1 (7)
= 2
= 6
7
=1 s.l.
- 2
= 2
8. Uerneckiella Eichler, 1940
8
-
1 (8)
= 2
= 2
8
=1 s.l.
= 2
8
9. B. (Spinibovicola) subgen. n.
2 (9) [from 2 1
10. Bisonicola gen. n.
10 [from 2|
11. Tragulicola gen. n.
11 [from 1|
Table 3 Generic concepts in the Bovicolinae. The genera included in the table are represented by
numbers 1-11; '2' indicates that the genus (Bovicola) is given full generic status, '1(2)' indicates that the
genus (Bovicola) is considered a subgenus (of Damalinia) and ' = 1' indicates that the genus is considered
as a junior synonym (of Damalinia). The generic name Bovidoecus Bedford, 1929 is omitted, as it was
synonymised with Bovicola by Bedford (1932a) and has not since been used. There is no reference in the
table to Werneck (1936), who treated all genera of Trichodectidae as synonyms of Trichodectes .
i
i
i
1
1
X
r>
10
o
a.
*O ~-f
I
^
M
C
M
c
e ^
*~*
C cr
-.
1
41
M
C
jg
jf ^
on
w '
09
a.
ft
a
a.
c
o
o
_o
!
M
3
X
M
J
*
1. Tfichodectee Nitzsch, 1818
1
2
1
2
1
= 1
1
2
1
1 (2)
1
1 (2)
1
2
1
1 (2)
1
2
1
3. Galiatobius Keler, 1938fc
3
3
- 1
= 1
1 (3)
1
3
= 1
= i
= 1
4. iMtridia KSler, 1938a
4
4
= 1
4
1 (4)
4
4
1 (4)
4
4
5. Staohiella Keler, 1938a
5
5
- 1
5
1 (5)
1 (5)
5
1 (5)
5
1 (5)
6. Vrsodectes Keler, 1938a
6
6
= 1
= 1
1 (6)
= 1
6
= 1
- 1
- 1
7. Potusdia Conci, 1942
= 1
1 (7)
- 1
7
- 1
- 1
= 1
8. Tr-igonodeates Keler, 1944
1
1 (8)
= 1
8
- 1
- 1
2 (8) [to Neotrichodectinae)
9. Uerneakodeates Conci, 1946
= 1
1 (9)
1
9
- 1
- 1
9
10. Seolutridia gen. n.
10 [from 4]
11. T. (Pamtriahodeatee) subgen. n.
1 (11) [from 5]
Table 4 Generic concepts in the Trichodectini (plus Neotrichodectes and Trigonodectes) . Coding as for
Table 3. Protelicola is included in Table 5. The generic name Grisonia Keler, 1938a is omitted and its
replacement name, Galictobius Keler, 1938ft, used throughout.
TRICHODECTID MAMMAL LICE
237
s
1
-a-
<r
^t
<r
>>
o
r-i
sC
c
Q.
*Q >
f
~
S
c
&
s
c S
S
T
2
g
*
s
jtf
o.
c
.u
J*
g
to
T3
b
0)
It
%
=
3
.
*
u
E
1. Feliaola Ewing, 1929
1
1
1
1
1
1
1
i
1
1
1
2. Protelioola Bedford, 1932
-
2
2
2
= 1
1 (2)
= 1
2
1 (2)
= 3
2 [to Trichodectinil
3. Suricatoecus Bedford, 1932
- 1
3
3
- 1
3
1 (3)
1 (3)
3
1 (3)
3
1 (3)
4. Lorisicola Bedford, 1936
-
4
-
4
4
4
4
4
4
5. Fastigatosoulum Kler, 1939
5
= 1
= 3
-
= 3
5
= 1 s.l.
= 3
= 1
6. Eiahlerella Conci, 1942
= 3
-
= 3
6
- 1 s.l.
= 3
= 3
7. Paradoxwr-oeaus Conci, 1942
= 1
-
= 1
7
= 1 s.l.
- 1
4 (7)
8. Neofeliaola Werneck, 1948
8
1 (8)
1 (8)
8
1 (8)
8
= 7
9. Parafelicola Werneck, 1948
9
1 (9)
1 (9)
9
1 (9)
9
= 7
Table 5 Generic concepts in the Felicolini (plus Protelicold). Coding as for Table 3. The generic names
Bedfordia and Felicinia (a junior homonym and an absolute synonym respectively) have not been
included as their status has not varied; they are discussed in the comments following the description of
Felicola s. str. below.
Hopkins (1941) presented an explanation for the presence of so many annectent species in the
Trichodectidae and a justification for synonymising many of the genera, writing: 'I believe the
explanation to be that the Trichodectidae are in the process of dividing up into genera; in some
cases the divergence has proceeded far enough for us to recognise the segregates as generically
distinct, but in a much greater number of cases extreme members of a group may have become
strikingly distinct whilst the others remain as connecting-links which entirely undo our attempts
to find characters peculiar to the group.' Hopkins' statement implies that the species in a genus
are somehow evolving as a unit, and is linked to the typological approach to taxonomy.
Relationships between three or more taxa, if assessed by a simple count of character states (i.e.
not distinguishing between plesiomorphies, apomorphies and homoplasies) are frequently
reticulate in aspect (Simpson, 1961; Hennig, 1966; Mayr, 1969). The greater the number of
homoplasies the more complex the reticulum is likely to be, and the more difficult it is to
combine the taxa into groups. If genera are constructed on this principle some morphologically
distinct species and species-groups will be distinguished, 'linked' by annectent species, with the
concomitant absence of 'gaps' between genera - precisely the problem with traditional group-
ings of the Trichodectidae. The 'problem' of annectent species is therefore engendered by the
typological approach; the difficulties in distinguishing supra-specific groups are also a result of
this, but combined in the Trichodectidae with a high degree of homoplasy.
The difficulties discussed above have discouraged authors from attempting to produce keys to
genera of the Trichodectidae. Since the key to genera published by Keler (1938a), very few have
been published, and none included all the genera. Keler (1944) produced a key to some genera,
slightly emended from Keler (1938), but a promised second half to the paper containing the rest
of the key was never published. Werneck (1948, 1950), despite describing all of the genera, did
not attempt to produce a key. A few keys have since been published in faunistic works, for
example Toulechkoff (1955) produced a key (in Bulgarian) to the genera found in Bulgaria, and
Zlotorzycka (1972) published a rather inaccurate key (in Polish) to the genera found in Poland.
Although most authors follow Kellogg (1896) in their conception of the Trichodectidae, and
retain familial rank for the group (e.g. Hopkins & Clay, 1952; Hopkins, 1960; Ledger, 1980;
Emerson & Price, 1983), the rank of the group has been raised by others. Keler (19380) raised
the Trichodectidae to superfamily level and included three families: Trichodectidae, Bovicoli-
dae and Dasyonygidae, the latter two being described as new. In the Trichodectidae he placed
four subfamilies, all of which he indicated to be new: Trichodectinae, Felicolinae, Eury-
trichodectinae and, dubiously placed in this family, Eutrichophilinae (Fig. 35). Eichler (1940)
238
C. H. C. LYAL
MSIOHIX
EURYTRICHODECTES
EUTRICHOPHILVS
PROCAVICOLA
PROCAVIPHILUS
BEDFORDIA
FELICOLA
PROTELICOLA
SURICATOECUS
GEOMYDOECUS
GRISONIA
LUTRIDIA
NEOTRICHODECTES
STACHIELLA
TRICHODECTES
URSODECTES
BOVICOLA
CERVICOLA
DAWLINIA
HOLAKARTIKOS
LEPIKEKTRON
RHABDOPEDILON
TRICHOLIPEURUS
DASYONYGIDAE
EURYTRICHODF.CTINAF
EUTRICHOPHILINAE
FELICOLINAE
TRICHODECTIDAE
TRICHODECTINAE
BOVICOLIDAE
Incertae sedis
MEGANARION __
Fig. 35 Classification of Trichodectoidea' according to Keler (1938a).
described two further subfamilies in the Trichodectidae (sensu Keler, 19380): Lymeoninae and
Cebidicolinae. Eichler (1941) described the new subfamily Damaliniinae in the Bovicolidae,
attributing the nominate subfamily to himself. He also transferred the Eurytrichodectinae and
the Eutrichophilinae to the Dasyonygidae, again attributing the newly-defined nominate
subfamily to himself. Eichler (1941) considered the rank of the whole group to be not
superfamily but 'family group' and termed it the Trichodectiformia (attributed to Keler, 19380);
the subfamily Trichodectinae is also attributed to Keler (19380) but the nominate family is
attributed to Burmeister (1838). The classification proposed by Eichler (1941) is depicted in Fig.
36. Keler (1944) retained the Eurytrichodectinae in Trichodectidae and moved the Eutrichophi-
linae to Bovicolidae; the subfamilies described by Eichler (1940, 1941) were not mentioned (Fig.
37). The Trichodectidae were attributed by Keler (1944) to Kellogg (1896) but the nominate
subfamily to Keler (19380). The superfamily rank was retained and the 'family group' not
mentioned. Hopkins (1949) regarded the families proposed by Keler (19380) as subfamilies and
the subfamilies as at most tribes. Hopkins also, following his synonymy of Bovicola with
Damalinia, considered that 'Bovicolinae must be known as Damaliniinae'. Eichler (1963)
retained the higher ranks and included 'interfamilia Trichodectiformia' within the superfamily
Trichodectoidea. He moved the Lymeoninae to Dasyonygidae (wherein he retained Euryt-
richodectinae), but accepted the move of Eutrichophilinae to Bovicolidae proposed by Keler
(1944). Eichler (1963) also proposed the division of the Trichodect(oidea) into tribes, indicating
their existence and composition by variations in typography in the list of genera presented
(Eichler, 1963: 159, lines 31-37). Eichler (1963) did not publish any of the tribal names, but
Lakshminarayana (1976) listed all of them. Neither Eichler (1963) nor Lakshminarayana
(1976), however, gave any statement that purported 'to give characters differentiating the
tax(a); or ... a definite bibliographic reference to such a statement' as is required by the
International Code of Zoological Nomenclature for any name published after 1930 (Article 13).
TRICHODECTID MAMMAL LICE
239
DASYOtlYX
EURYTRICHODECTES
EUTRICHOPHILUS
PROC AVI COLA
PROCAVIPHILUS
BEDFORDIA
FELICOLA
PROTELICOLA
SURICATOECUS
GALICTOBIUS
GEOMYDOECUS
LUTRIDIA
NEOTRICHODECTES
STACHIELLA
TRICHODECTES
URSODECTES
LYMEOH
CEBIDICOLA
LORISICOLA
MEGANARIOHOIDES
CERVICOLA
DAMALINIA
TRICHOLIPEURVS
BOVICOLA
HOLAKARTIKOS
LEPlKENTROtI
RHABDOPEDILOtl
UERNECKIELLA
TRICHOPHI LOPTEPUS
DASYONYGINAE
EURYTRICHODECTINAE
EUTRICHOPHILIHAE
FELICOLINAE
TRICHODECTINAE
LYMEONINAE
CEBIDICOLINAE
DAMALINIINAE
BOVICOUNAE
DASYONYGIDAE
TRICHODECTIDAF,
BOVTCOLTDAE
TRIC1IOP11ILOPTERIDAE
Fig. 36 Classification of Trichodectiformia' according to Eichler (1941).
None of the (11) names, therefore, are available for taxonomic use. Eichler (1963) attributed
Trichodectoidea, Trichodectiformia, Trichodectidae and Trichodectinae to Burmeister (1838)
but, whilst (correctly) attributing Bovicolidae and Dasyonygidae to Keler (19380) he attributed
the nominate subfamilies of both to Eichler (1941). The classification proposed by Eichler
(1963) is depicted in Fig. 38. Keler (1969) proposed a classification similar to that proposed by
Keler (1944), but omitting a number of genera (Fig. 39).
Article 36 of the International Code of Zoological Nomenclature (1984) states that all
categories in the family-group (tribe, subfamily, family, superfamily and any supplementary
categories, according to Article 35a) are co-ordinate, and a name established for any category
within the group is available with its original date and author for a taxon with the same type
genus in each of the categories. The Trichodectinae, Trichodectidae, Trichodectiformia and
Trichodectoidea should therefore all have the same date and author. The first use of a
family-group name based on the type genus Trichodectes was by Kellogg (1896), who ranked
'the Nitzschian families as suborders, the Nitzschian genera as families, and the Nitzschian
subgenera, the genera of present-day writers, as genera.' Kellogg (1908) attributed the
Trichodectinae (the only subfamily included in the Trichodectidae, which bore no attribution)
to 'Burmeister (?)'. As explained above, subsequent attribution has frequently been to
Burmeister (1838) and, in the case of some co-ordinate names, to Keler (19380). Burmeister
(1838) did not mention any taxon in the family-group with the type genus Trichodectes. All
240
EURWRICHODECTES
FASTIGATOSCULUM 1
SURICATOECUS
MEGANARIONOIDES
GALICTOBIUS
GEOMXDOECUS
LUTRIDIA
NEOTRICHODECTES
STACHIELLA
TRICHODECTES
TRIGONODECTES
URSODECTES
EUTRICHOPHILUS
PROCAVICOLA
1 ,2
1,2
PROCAVIPHILUS
BOVICOLA 1
CERVICOLA 1
DAMALINIA 1
HOLAKARTIKOS 1
LEPIKENTRON 1
RHABDOPEDILON 1
TRICHOLIPEURUS 1
TRICHOPHILOPTERUS
C. H. C. LYAL
EURYTRICHODECTINAE
FELICOLINAE
DASYONY
TRICHODECTINAE
EUTRICHOPIIILINAE
BOVICOLINAE
TRICIIODECTIDAF.
BOVICOLinAE
TRICHOPMILOPTERIDAE
Fig. 37 Classification of Trichodectoidea' according to Keler (1944). Position of genus inferred from
Keler (1938a). Position of genus inferred from key in Keler (1944).
family-group names with the type genus Trichodectes should therefore be attributed to Kellogg
(1896). Trichodectidae Kellogg, 1896 has been placed on the Official List of Family-group
Names in Zoology (Opinion 627, Bull. zool. Nom. 19: 91-96, (1962)). The names Bovicolinae,
Bovicolidae, Dasyonyginae and Dasyonygidae should all be attributed to Keler (1938a), not
Eichler (1941). The action of Hopkins (1949) is synonymising the senior family-group (Bovicoli-
nae) with the junior (Damaliniinae) was taken because he believed Bovicola and Damalinia (the
type-genera) to be synonyms, and Damalinia is senior to Bovicola. This action is incorrect under
Article 40 of the Code, however, which states that, at least after 1961, in the case of type-genus
synonymy the senior family-group name is to be used for the family-group taxon that contains
both senior and junior synonyms. This Rule can be set aside for such an action if taken before
1961 , if the name has 'won general acceptance' (Article 40b). The subfamily Damaliniinae sensu
Hopkins (1949) has rarely if ever been used since, whilst the name Bovicolidae (= Damaliniinae
sensu Hopkins) has been employed by Eichler (1963) and Keler (1969). The action of Hopkins
(1949) is therefore rejected.
Keler (1944) included the Trichophilopteridae - a family containing a single genus, parasitic
on Lemurs - within the Trichodectoidea, although Keler (1969) referred this family to the
Philopteroidea. Eichler (1963) retained the Trichophilopteridae in the Trichodectoidea, but
distinguished it as 'interfamily Trichophilopteriformia' as opposed to 'interfamily Trichodecti-
formia'. Stobbe (19130), Ferris (1933) and Werneck (1948) all considered the affinities of
TRICHODECTID MAMMAL LICE
"Tribes"
241
DASYONYX
NEODASYOKYX
PROCAVIPHILUS
EVRYTRICHODECTES
PROC AVI COLA
CONDYLOCEPHALUS
MEGAHARIOHOIDES
LYHEON
FELICOLA
SURICATOECVS
EICHLERELLA
FASTIGATOSCULUM
FELICOMORPHA
PROTELICOLA
NEOFELICOLA
PARAFELICOLA
PARADOXUROECUS
TRICHODECTES
OALICTOBIUS
POTVSDIA
TRIGONODECTES
URSODECTES
WERtiECKODECTES
STACHIELLA
NEOTRICHODECTES
LUTRIDIA
GEOMYDOECVS
CEBIDICOLA
LORISICOLA
BOVICOLA
HOLAKARTIKOS
LEPIKENTRON
RHABDOPEDILOH
WERKECKIELLA
DAMALINIA
CERVICOLA
TRICHOLIPEURUS
EUTRICHOPHILUS
Zl
Zl
=1
Zl
DASYONYGINAE
F.URYTRICHODECTINAE
ID LYMEONINAE
FELICOUNAE
Zl
ZJ
Zl
Zl
TRICHODECTTNAE
Zl
n
CEBIDICOLTNAF
BOVICOLINAE
DAMALINIINAE
EUTRICHOPHILTNAE
DASYONYGIDAE
TRICHODECTIDAE
Fig. 38 Classification of Trichodectiformia' according to Eichler (1963). The 'tribes' are indicated by
square brackets in the appropriate column, the first genus in each 'tribe' being intended by Eichler (1963)
as the type-genus. See text.
Trichophilopterus to lie with the Thilopteridae' rather than with the Trichodectidae. In this
study no apomorphies were found to indicate a sister-group relationship between Trichophilop-
terus and all or part of the Trichodectidae.
Proposed classification
The proposed classification is derived from the results of a cladistic analysis of the Trichodecti-
dae (Trichodectiformia sensu Eichler, 1963) at the species level (Figs 23-34). The species are
grouped on the four criteria discussed below, and ranked according to the principles of phyletic
sequencing.
Holophyly. The classification includes, as far as possible, only holophyletic groups. Some genera
and subgenera, however, may be found not to follow this criterion (see discussion of Damalinia
s. str., Dasyonyx and Geomydoecus below).
Utility. Genera are ideally of 'moderate' size and relative morphological uniformity. If a genus is
large and diverse, recognition is difficult and useful discussion on many aspects of biology or
distribution prohibited; if genera are too small, identification is time-consuming and discussion
242
C. H. C. LYAL
DASYONYX
NEODASmim
PROCAVIPHILVS
EURtTRICHODECTES
PROCAVICOLA
PKOTELICOLA
FELICOLA
SVRICATOECUS
NEOFELICOLA
CEBIDICOLA
LORISICOLA
GEOMYDOECUS
STACHIELLA
TRICHODECTES
TRIGONODECTES
URSODECTES
EVTRICHOPHILUS
BOVICOLA
CERVICOLA
DAHALINIA
LEPIKENTRON
TRICHOLIPEURUS
DASYONYGINAE
EURYTRICHODECTIHAE
FELICOLINAE
TRICHODECTINAE
EUTRICHOPHILINAE
BOVICOLINAE
DASYONYGIDAE
TRICHODECTIDAF.
BOVICOLIDAE
1-/ERNECK1ELLA
Fig. 39 Classification of Trichodectoidea' according to Keler (1969).
again impeded. No 'absolute' size can be recommended, however, as the most satisfactory size
will depend on a number of properties of the species, and must (in this study) conform to such
limitations as are imposed by the criterion of holophyly. Subfamilies are chosen in this study to
aid discussion by providing names for holophyletic groups of genera, and to fulfil the logic of the
phyletic sequencing convention.
Stability. To ensure that the classification has maximal stability the generic concepts accepted in
this study conflict as little as possible with established usage.
Distinctness. To facilitate identification, taxa in the genus-group should be as distinct from one
another as possible. The requirements of Mayr (1969) that genera must be separated by a
decided gap, and that the 'size' of the gap should be inversely proportional to the size of the
taxon, are not necessarily compatible with the criterion of holophyly followed here, however,
and the problem of annectent species (that 'fill' any such gap) has been discussed above. Despite
the apparent drawback of adherence to holophyletic groups at the expense of inter-generic
'gaps', it has been possible in this study to produce a key to the genera of Trichodectidae (see
discussion of keys to genera of Trichodectidae above).
The formation of the genera of Trichodectidae is discussed below, to give an indication of the
rationale behind each decision. The genera are discussed by subfamily, and the division into
subfamilies is discussed last.
Bovicolinae (Figs 24-27, 40)
This clade was not resolved fully in the analysis, and a primary pentachotomy was obtained. The
clade has been treated as a single genus (see Table 4), but the diversity of morphology and of
hosts indicates that this concept is too broad to be of great value. Subdivision of the clade into
smaller holophyletic groups increases the value of the classification for information retrieval,
and leads to the acceptance of genera that approach the concepts of Werneck (1950). To obtain a
TRICHODECTID MAMMAL LICE 243
measure of conformity each of the five branches of the clade has been accorded generic status.
The monobasic genera Bisonicola and Tragulicola require no comment, and the genus
Werneckiella was revised by Moreby (1978). Of the two remaining clades, one corresponds
approximately to a restricted concept of Bovicola, the other to Damalinia plus Tricholipeurus
(sensu Werneck, 1950). The Bovicola clade (genus Bovicola) has a primary tetrachotomy, with
most of the species belonging to only one of the four resultant clades (Fig. 24). The species in this
large clade are morphologically more similar to one another than they are to any of the species in
the other three clades. To recognise this morphological divergence (and thus facilitate identifica-
tion), and to demonstrate in the classification the extent of the phylogenetic knowledge, the four
branches are each accorded subgeneric status. The Damalinia plus Tricholipeurus clade (genus
Damalinia} has a primary trichotomy (Fig. 23), and it is clear that discussion of the genus will be
facilitated by the recognition of each of these branches as a subgenus. Damalinia s. str.
comprises two major clades, each characterised by the form of an apomorphic development of
abdominal pleurum II onto the sternum. As discussed above, these two forms may be
co-apomorphies, or may represent two states in a transformation series. If the latter interpreta-
tion is correct, one of the clades is probably paraphyletic with respect to the other.
Eutrichophilinae
Only the single genus Eutrichophilus is included, with no change in generic concept.
Dasyonyginae (Figs 28, 29)
The previously-accepted generic concepts in this subfamily remain essentially unchanged at the
subgenus level. The only change is the transfer of the subgenus Meganarionoides from
Procavicola to Procaviphilus , and the inclusion of Procaviphilus sclerotis and P. serraticus in P.
(Meganarionoides). Subgenera are used (as in Werneck, 1941, 1950; Ledger, 1980) as no
advantage accrues from regarding each of the clades so recognised as a full genus, and
application of the principles of phyletic sequencing allows retention of all the currently-used
generic and subgeneric names with no higher taxa required, whereas recognition of all these as
genera would require the description of a number of intercalating family-group taxa.
It is notable that one of the two subgenera of Dasyonyx may be paraphyletic with respect to
the other, as the subgenera are characterised by apomorphic developments of the teeth of the
tarsal claws. These may be co-apomorphies or two states in a transformation series (see above).
If the latter interpretation is correct, one of the subgenera is probably paraphyletic with respect
to the other.
Trichodectinae (Figs 23, 30-33, 40)
The first dichotomy in this clade splits it roughly into Felicola (sensu Ledger, 1980, but without
Protelicola and with Lorisicola) on one side and Trichodectes (sensu Ledger, 1980, but without
Neotrichodectes and Trigonodectes , and with Protelicola) on the other. The diversity of
morphology of the lice, and the variety of hosts infested, indicates that the very broad generic
concepts endorsed by Ledger (1980) are too inclusive to be of great value in data-retrieval and
discussion. For this reason the genera proposed here are smaller than those of Ledger (1980)
and, in some cases, approach the concepts held by Werneck (1948).
Most of the species in the Trichodectes side of the initial dichotomy arise from the three
branches of an apical trichotomy (Figs 30, 31). The branch of this trichotomy comprising the
pinguis-galictidis clade corresponds roughly to the concept of Trichodectes held by Werneck
(1948), whilst the other two branches (the ovalis-zorillae clade and the emeryi-potus clade)
correspond roughly to Stachiella sensu Werneck (1948) (though fallax, octomaculatus andpotus
were placed in Trichodectes by Werneck, 1948). However, placing two of the three clades of the
trichotomy in a taxon Stachiella and excluding the third results in a group that is not
holophyletic. Recognising each of the three branches of the trichotomy as a separate genus is
undesirable, as the three intergrade phenetically. The course followed here is to recognise the
244 C. H. C. LYAL
genus Trichodectes comprising all three branches, each of these being considered a subgenus
(Fig. 40). Using the principle of phyletic sequencing the sister-group of Trichodectes is also
considered a genus, for which the name Werneckodectes is available. Likewise the next three
branches of this clade are also considered genera. This process necessitates dividing Werneck's
genus Lutridia into two genera, but retention of the genus as it stood calls for recognition of a
paraphyletic group in the classification, and, although the species in the two clades comprising
Lutridia (sensu Werneck, 1948) are superficially similar, some of these similarities may be
homoplastic.
The other branch of the initial dichotomy of the Trichodectinae clade comprises, as noted
above, most of the species consigned to Felicola by Ledger (1980) plus the single species of the
genus Lorisicola (sensu Werneck, 1950). The two branches of this clade (Fig. 23) are each
considered as genera which, taking the most senior available names, are known as Felicola and
Lorisicola. For reasons of utility, each genus is divided into two holophyletic subgenera. None of
the genera or subgenera coincides with any previous generic concept, as such concepts relied
heavily on head shape and abdominal spiracle number, both of which characters have proved to
be subject to a considerable degree of homoplasy.
In order to maintain the logic of phyletic sequencing, if the Felicola-Lorisicola clade is to be
considered as comprising two genera, the rank of this clade and of the Trichodectes-Protelicola
clade must be equal and formally recognised. Use of the tribal category permits this, and the
family-group names Trichodectini and Felicolini are available (see full classification below). It
must be stressed that these tribes are inserted to maintain the formal structure of the
classification, and are not intended (or believed) to have any other significance.
Neotrichodectinae (Fig. 34)
The first dichotomy in this subfamily divides the clade into those species previously assigned to
the genus Geomydoecus on one side, and species from Neotrichodectes, Lakshminarayanella
and Trichodectes (sensu Werneck, 1948) on the other. The two branches will be discussed
separately.
The genus Geomydoecus as previously recognised is fairly uniform in morphology, distribu-
tion and host species, and may be identified readily. To divide this genus into others would
inhibit rather than encourage discussion, and the genus is retained in its present form. The two
subgenera as proposed by Price & Emerson (1972) are also retained though, as indicated above,
neither are holophyletic groups.
The other branch of the primary dichotomy comprises the 10 species previously assigned to
the genus Neotrichodectes (considered a subgenus of Trichodectes by Hopkins, 1949 and Ledger,
1980), the two species previously assigned to the genus Lakshminarayanella (formerly Ly-
meon), and a single species formerly placed in Trichodectes by most authors (T. barbarae}. The
clade is plainly close to the established concept of Neotrichodectes, and it is preferable that this
name is applied to as much of the group as possible. The 10 species of Neotrichodectes auctt. do
not form a holophyletic group, however, though morphologically they are quite uniform.
Inclusion of T. barbarae is unlikely to create problems, but Lakshminarayanella (as Lymeon)
has been placed by some authors in a subfamily of its own (Eichler, 1940, 1963), and considered
close to the hyrax lice (Keler, 1944; Hopkins, 1949; Eichler, 1963). If Lakshminarayanella is
synonymised with Neotrichodectes and given no formal recognition it is likely to be raised from
synonomy by future workers because of its distinctive morphology, leaving Neotrichodectes
paraphyletic. The course taken here is to recognise Lakshminarayanella as a subgenus of
Neotrichodectes, which necessitates recognition of four other (holophyletic) subgenera, names
already being available for two of these. Application of the principles of phyletic sequencing
permits equal ranking of the subgenera within the genus.
Subfamilies (Fig. 40)
To divide the family into 'manageable' holophyletic groups for the purposes of discussion and to
maintain the logic of phyletic sequencing, supra-generic groupings had to be employed. Use of
.o
"
TRICHODECTID MAMMAL LICE
Bovlcola (Bovlcola)
B. (Holakartlkos)
B. (Leplkentron)
B. (spinlpovlcola)
Blsonlcola
Tragullcola
Hernecklella
iianiallnlo (DnmallDJa)
- D. (Cervlcola)
D. (Trlchollpeurus)
Eutrlchophllus
Procavlcola (Procavlcola)
P. (Condylocephalus)
Procavlphllus (Procavlphllus)
P. (MegQngrj.QQQides)
Dasyonyx (Dasyonyx)
D. (Neodasyonyx)
F-un'trlchodectes
- Lutrldla
Neolutrldla
Werneckodectes
Trlchodectes (Trlchodectes)
I. (Paratrichodectes)
I. (Stachlella)
Felicolc (Felicola)
F. (Surlcatoecus^
Lorlsicold (LorisjCQlg)
- L. (Paradoxuroecus)
Neotrlchodectes (Neotrlchodectes)
N. (Trigonodectes)
ft. (Nasulcola)
N. (Lakshmlnarcyonella)
N. (Conepotlcola)
Geomydoecus (Geomydoecus)
G. (Thomomydoecus)
245
> BOVICOLINAE
EUTRICHOPHILINAE
> DASYONYGINAE
*\
> TRICHODECTINAF.
> NEOTRICHODECTINAE
Fig. 40 Cladogram of the genera and subgenera of Trichodectidae, with subfamily assignments.
the principles of phyletic sequencing permitted the use of the subfamily category throughout
(with the addition of the tribes mentioned above). The limits of the subfamilies were chosen for
maximum utility, modified by the dictates 01 the sequencing convention. It would be surprising,
given the high degree of homoplasy of structures in the Trichodectidae, 4f the subfamilies
fulfilled the criterion 'distinctness' described above and were readily distinguishable. A key to
246 C. H. C. LYAL
subfamilies is provided, however, largely to satisfy the requirements of the International Code of
Zoological Nomenclature (Article 13) for a description to accompany any new name for,
although names were available for most of the subfamilies, a single new name is required.
A complete classification of the Trichodectidae to generic level is set out below in phyletic
sequence (as recommended by Wiley, 1979, 1981).
Sequenced classification of the Trichodectidae
Family TRICHODECTIDAE Kellogg, 1896
Subfamily BOVICOLINAE Keler, 1938 (all genera sedis mutabilis)
Genus BOVICOLA Ewing, 1929 (all subgenera sedis mutabilis)
Subgenus BOVICOLA Ewing, 1929
Subgenus HOLAKARTIKOS Keler, 1938
Subgenus LEPIKENTRONKeler, 1938
Subgenus SPINIBOVICOLA subgen. n.
Genus BISONICOLA gen. n.
Genus WERNECKIELLA Eichler, 1940
Genus TRAGULICOLA gen. n.
Genus DAMALINIA Mjoberg, 1910 (all subgenera sedis mutabilis)
Subgenus DAMALINIA Mjoberg, 1910
Subgenus CERVICOLA Keler, 1938
Subgenus TRICHOLIPEURUS Bedford, 1929
Subfamily EUTRICHOPHILINAE Keler, 1938
Genus EUTRICHOPHILUS Mjoberg, 1910
Subfamily DASYONYGINAE Keler, 1938
Genus CEBIDICOLA Bedford, 1936
Genus PROCAVICOLA Bedford, 1932
Subgenus PROCAVICOLA Bedford, 1932
Subgenus CONDYLOCEPHALUSWemeck, 1941
Genus PROVCAVIPHILUS Bedford, 1932
Subgenus PROCAVIPHILUS Bedford, 1932
Subgenus MEGANARIONOIDES Eichler, 1940
Genus DASYONYX Bedford, 1932
Subgenus DASYONYX Bedford, 1932
Subgenus NEODASYONYXWerneck, 1941
Genus EURYTRICHODECTESStobbe, 1913
Subfamily TRICHODECTINAE Kellogg, 1896
Tribe TRICHODECTINI Kellogg, 1896
Genus PROTELICOLA Bedford, 1932
Genus LUTRIDIA Keler, 1938
Genus NEOLUTRIDIA gen. n.
Genus WERNECKODECTESCond, 1946
Genus TRICHODECTESNitzsch, 1818 (all subgenera sedis mutabilis)
Subgenus TRICHODECTESNitzsch, 1818
Subgenus PA RATRICHODECTES subgen. n.
Subgenus STACHIELLA Keler, 1938
Tribe FELICOLINI Keler, 1938
Genus FELICOLA Ewing, 1929
Subgenus FELICOLA Ewing, 1929
Subgenus SURICATOECUS Bedford, 1932
Genus LORISICOLA Bedford, 1936
Subgenus LORISICOLA Bedford, 1936
Subgenus PARADOXUROECUSConci, 1942
Subfamily NEOTRICHODECTINAE subfam. n.
Genus NEOTRICHODECTES Ewing, 1929
Subgenus NEOTRICHODECTES Ewing, 1929
Subgenus TRIGONODECTESKeler, 1944
Subgenus NASUICOLA subgen. n.
Subgenus LAKSHMINARAYANELLA Eichler, 1982
TRICHODECTID MAMMAL LICE 247
Subgenus CONEPATICOLA subgen. n.
Genus GEOMYDOECUSEwing, 1929
Subgenus GEOMYDOECUSEv/ing, 1929 (paraphyletic)
Subgenus THOMOMYDOECUS Price & Emerson, 1972 (polyphyletic?)
Descriptions of genera and subgenera
The generic and subgeneric descriptions below are arranged by subfamily in the order of the
sequenced classification of the Trichodectidae (see above).
Descriptions are set out in the following order: paragraph one - head, both sexes, with details
of sexually-dimorphic features of antennae, if present; paragraph two - thorax, both sexes,
omitting mention of the anterior setae (on the post-temporal margin) which are present in all
species; paragraph three - abdomen, both sexes, with details of sexually-dimorphic features of
the setae, sclerites or shape, if present; fourth paragraph - female terminalia and genitalia; fifth
paragraph - male subgenital plate, terminalia and genitalia. Descriptions are given of each
genus as a whole, even where subgenera are present. The descriptions of subgenera (if any are
present) follow that of the genus in which they are placed, and give only subgeneric characters,
so that some of the paragraphs listed above may be omitted. Characters that vary between
subgenera, if mentioned in the generic description, are indicated by an asterisk (**).
Each description is followed by an indication of the host group or groups parasitised, and by
any pertinent comments on the taxonomy, morphology or biology of some or all of the included
species. A check-list of all species included in each genus or subgenus is also given, the names
being placed in alphabetical order. Following each species name in the check-lists is an
indication of the number of specimens of each sex examined in the study.
Two species have not been placed, and are considered incertae sedis. Trichodectes baculus
Schommer, 1913; type-host: Caprahircus Linnaeus. Trichodectes tigris Ponton, 1870; type-host:
Felis tigris Linnaeus. These species are discussed by Werneck (1950).
The subfamilies, genera and subgenera are keyed (p. 335).
BOVICOLINAE Keler
Genus BOVICOLA Ewing
The genus Bovicola comprises four subgenera.
DESCRIPTION. Anterior of head with osculum absent or, if present, broad and shallow* ; pulvinus of normal
length or short and not attaining anterior margin of head; dorsal preantennal sulcus present or absent*;
clypeal marginal carina not or only slightly broadened medially, or broadened to variable degree into bar
with posterior and anterior margins roughly parallel, bar either straight and at right angles to long axis of
head or curved and parallel to anterior margin of head* ; anterolateral margin of head smoothly rounded;
preantennal portion of head short, outline broadly rounded or trapezoid*. Temple margin smoothly
convex or with posterior projection*, sometimes convexly produced posteriad*. Male scape expanded or
not expanded* , with setal row apparently present or setae randomly scattered; flagellomeres fused in males
and females; male flagellum with two basally-articulated 'teeth' and interior face not 'roughened'. Dorsum
of head with more or less abundant setae , short , long or of moderate length * . Sitophore sclerite unmodified
or with posterior arms extended*.
Thorax with more or less abundant setae, short, long, or of moderate length, frequently longest on
postero-lateral margin of pterothorax*.
Abdomen oval or elongate, frequently tapering posteriorly more in male than in female*. Abdominal
spiracles present on segments HI- VIII. Abdominal setae variable*; anterior setae always present on
pleura, sometimes on sterna and terga; postero-lateral setae absent. Abdominal pleural projections
absent. Sclerites present at least on sterna III-VII (males) and III- VIII (females), terga II-VII (males) and
III-IX (females) and pleura II- VIII; male terga with posterior sclerites present or absent*.
Gonapophyses with marginal setae; ventral lobe present, though sometimes not pronounced*. Gona-
pophyses meet ventral vulval margin acutely, not linked by sclerotised band. Ventral vulval margin not
sclerotised, or sclerotised only medially; subgenital lobe absent, though small median membranous
projection may be present (Fig. 42)*. Genital chamber sometimes with median antero-dorsal area lacking
scales or spicules*.
248
C. H. C. LYAL
41
45
43
Figs 41-45 Bovicola species. 41, fi. (Spinibovicola) hemitragi, cf abdominal terga I and II. 42, fi. (5.
jellisoni, $ terminalia, ventral. 43, B. (Lepikentron) breviceps, 9 terminalia, ventral. 44, B. (Holakarti
kos) crassipes, $ gonapophysis, ventral. 45, B. (B.) caprae, 9 gonapophysis, ventral.
TRICHODECTID MAMMAL LICE
249
250
57
Figs 51-57 Bovicola species, cf genitalia. 51, B. (B.) bovis. 52, B. (B.) caprae. 53, B. (B.) concavifrons.
54, B. (B.) concavifrons, detail of right paramere (p) and mesomere (m). 55, B. (Lepikentron) breviceps.
56, B. (Spinibovicola) hemitragi. 57, B. (Holakartikos) crassipes.
Male subgenital plate variable*. Pseudostyli present or absent*. Male genital opening dorsal or
postero-dorsal. Male genitalia variable*.
HOSTS. Bovidae, Cervidae and Camelidae (Artiodactyla).
COMMENTS. Some species of Bovicola are parthenogenetic, males being rare or unknown.
A summary of the varying taxonomic treatments of Bovicola, its subgenera and synonyms, is presentee
in Table 3.
TRICHODECTID MAMMAL LICE 251
Subgenus BOVICOLA Ewing
(Figs 3, 42, 45, 46, 48, 49, 51-54)
Bovicola Ewing, 1929: 193. Type-species: Trichodectes caprae Gurlt, by original designation.
Bo vidoecus Bedford, 1929: 518. Type-species: Pediculus bovis Linnaeus, by original designation. [Synony-
my by Bedford, 1932a: 356.]
Rhabdopedilon Keler, 1938a: 453. Type-species: Trichodectes longicornis Nitzsch, by original designation.
[Synonymy by Werneck, 1950: 59.]
DESCRIPTION. Clypeal marginal carina not broadened medially, or more or less broadened into bar with
posterior margin straight or matching curvature of osculum. Temple margin smoothly convex, lacking
projection on postero-lateral angle, not convexly produced posteriad to great extent. Male scape not
expanded or only slightly expanded. Dorsum of head with setae short or of moderate length, of greater
abundance anteriorly than posteriorly. Sitophore sclerite unmodified.
Thorax with lateral and dorsal setae long and of moderate length, sometimes abundant and numerous on
disc of prothorax and pterothorax, otherwise less abundant and sparsely scattered on disc of pterothorax
with only two setae present on disc of prothorax; setae present along lateral margins and posteriorly
(dorsally) on prothorax and pterothorax; posterior setal row of prothorax marginal, with median gap
present or absent; posterior setal row of pterothorax submarginal, with median gap absent, row
incorporating two very long setae between postero-lateral and postero-median angles or, if setae generally
abundant on thorax, postero-lateral setae of pterothorax longer than others.
Abdominal setae short, long or of medium length; setal bases, at least of setae of posterior setal row on
sterna and terga, enlarged, clearly with doubled margins. Pregenital sclerites present on sterna and terga
(where present) of all segments, except sometimes tergum I and (independently) tergum VIII of males;
terga of males, at least of segments IV- VI, with both anterior and posterior sclerites.
Gonapophyses with lobe rectangular, acute, rounded or not pronounced; marginal setae confined to
lobe, long. Ventral vulval margin not sclerotised; convex, biconvex with median indentation, or convex
with small median membranous projection (Fig. 42); margin smooth or spinose. Postgenital area lacking
spinose patch. Genital chamber with antero-median dorsal area lacking spicules, scales or other decora-
tion , either very narrow and strongly-defined or wide and ill-defined , or with very narrow longitudinal fold .
Male subgenital plate variable; stqrnites VII and IX present, fused to s.g.p.r., sternite VIII absent or, if
present, fused or not fused to s.g.p.r. (Figs 46, 48). Pseudostyli absent (Fig. 49) or, if present, setose and
lobulate (Figs 46, 48). Basal apodeme very concave anteriorly, the sides of the concavity frequently being
parallel, though sometimes obscure. Parameres with broad basal flange or block; sometimes very reduced.
Basiparameral sclerites present and fused, or absent. Mesomeres, if fused apically, forming very narrow
arch lacking median extension; otherwise mesomeres not fused, sometimes very reduced and obscure.
Male genitalia depicted in Figs 51-54.
HOSTS. Bovidae and Cervidae (Artiodactyla).
COMMENTS. Some of the species in this subgenus are parthenogenetic.
SPECIES INCLUDED
alpinus Keler, 1942 (5 cf , 3 $)
bovis (Linnaeus, 1758) (7 cf, 137 9)
caprae (Gurlt, 1843) (c.50 Cf , c.50 $)
concavifrons (Hopkins, 1960) [Recalled from synonymy with longicornis (Nitzsch).] (2 cf , 98 9)
jellisoni Emerson, 1962 (10 cf , 10 $)
7/mbafus(Gervais, 1844) (c.SOcT, c.65 9)
longicornis (Nitzsch, 1818) (44 $)
oreamnidis (Hopkins, 1960) (holotype cf)
ov/s(Schrank, 1781) (59 cf, 64 9)
tarandi (Mjoberg, 1910) (2 $ , 7 nymphs)
t/Wa/is(Piaget, 1880) (c.100 9)
Subgenus HOLAKARTIKOS Keler gen. rev., stat. n.
(Figs 44, 47, 57)
Holakartikos Keler, 1938a: 461. Type-species: Trichodectes pilosus Piaget (nee Giebel) [= Trichodectes
crassipes Rudow], by original designation.
DESCRIPTION. Anterior of head with osculum absent; pulvinus very short, not attaining anterior margin of
252 C. H. C. LYAL
head; dorsal preantennal sulcus absent, though ventral preantennal sulcus sometimes present; clypeal
marginal carina not always pronounced and not, or only slightly, broadened medially; preantennal portion
of head very short, outline smoothly and shallowly rounded. Temple margin smoothly convex, lacking
projection on postero-lateral angle, convexly produced posteriad. Male scape very slightly expanded, with
setae randomly scattered. Dorsum of head with abundant setae of moderate length; temple with long
postero-lateral marginal setae. Sitophore sclerite unmodified.
Thorax with abundant setae, long and of moderate length, present on margins and disc of prothorax and
pterothorax; setae longest on the rounded postero-lateral angles of prothorax and ptero thorax.
Abdomen with long setae of posterior setal row, and shorter anterior setae, present on sterna, terga and
pleura (where present) of all segments (Fig. 47). Pre-genital sclerites sometimes very faint, present on
sterna and terga (where present) of all segments except tergum I and sternum II; male terga lacking
posterior sclerites.
Gonapophyses with broadly rounded lobe smoothly continuous with ventral margin; marginal setae
long, present all along ventral margin, including lobe. Ventral vulval margin not sclerotised; produced into
three weakly-developed lobes. Postgenital pleural area with patch of short, spine-like setae. Genital
chamber lacking dorsal non-ornamented area or fold.
Male subgenital plate with s.g.p.r. not joining sternites VII and VIII, and sometimes failing to contact
either or both; sternites sometimes very faintly sclerotised, obscure; s.g.p.r. with broad lateral flange on
VIII and IX (Fig. 47). Pseudostyli absent. Male genital opening dorsal. Basal apodeme long, not concave
anteriorly. Parameres long, slender, with basal block and flange. Basiparameral sclerites absent. Meso-
meres fused apically, with median extension present (see comments below). Male genitalia depicted in Fig.
57.
HOSTS. Bovidae (Artiodactyla).
COMMENTS. The only included species is not known to be parthenogenetic.
Werneck (1950) failed to recognise the median extension of the mesomeral arch, and considered it
absent.
Holakartikos was considered a synonym of Bovicola by Werneck (1950) and Emerson & Price (1981); a
more extensive history of the variations in status of this subgenus is presented in Table 3.
SPECIES INCLUDED
crassipes(Rudow, 1866) (24 cf, 31 $)
Subgenus LEPIKENTRONKeler gen. rev., stat. n.
(Figs 43, 55)
Lepikentron Keler, 1938a: 452. Type-species: Trichodectes breviceps Rudow, by original designation.
DESCRIPTION. Anterior of head with osculum absent; pulvinus very short, not attaining anterior of margin
of head; dorsal preantennal sulcus absent; clypeal marginal carina not pronounced, not broadened
medially; preantennal portion of head shorter in male than female, outline broadly and smoothly rounded.
Temple margin smoothly convex, lacking projection on postero-lateral angle, not convexly produced
posteriad to great extent. Male scape expanded, with setal row apparently present, though setae may be
scattered randomly. Dorsum of head with setae of moderate length, slender; setae sparsely distributed,
more abundant anteriorly than posteriorly. Sitophore sclerite unmodified.
Thorax with lateral and dorsal setae slender, long and of moderate length; setae present postero-
laterally and posteriorly on prothorax and along lateral margins and posteriorly (dorsally) on pterothorax;
posterior setal row of prothorax submarginal, sparse, with large median gap; posterior setal row of
pterothorax submarginal, with small median gap, setae shorter medially than laterally, with two long setae
laterally; pair of setae, widely spaced, present on disc of prothorax dorsally; setae not present on disc of
pterothorax.
Abdominal setae of moderate length, slender; anterior setae never present on sterna and terga.
Pre-genital sclerites present on terga II-VII (males) and terga III-IX (females) and sterna III-VII (males)
and III- VIII (females); male terga III-VII with both anterior and posterior sclerites, though the posterior
elements may be very faintly sclerotised and difficult to see.
Gonapophyses with small, pronounced lobe and broad tapering spur (Fig. 43); marginal setae confined
to lobe. Ventral vulval margin sclerotised medially; shallowly convex or biconvex (Fig. 43). Postgenital
pleural area lacking spinous patch. Genital chamber lacking median non-ornamented area or fold.
Male subgenital plate with sternite VII fused to s.g.p.r. and sternites VIII and IX absent; s.g.p.r. broad.
Pseudostyli absent. Male genital opening poste/o-dorsal. Basal apodeme slightly longer than parameres,
TRICHODECTID MAMMAL LICE 253
convex anteriorly. Parameres long, slender. Basiparameral sclerites absent. Mesomeres not apically fused;
each with median desclerotisation, and appearing as two rods (Fig. 55).
HOSTS. Camelidae (Artiodactyla).
COMMENTS. Only one male of the single included species is known and the species may be parthenogenetic.
The subgenus was treated as a synonym of Bovicola by Werneck (1950) and Emerson & Price (1981); a
more extensive history of the variations in status of this subgenus is presented in Table 3.
SPECIES INCLUDED
6rev/ceps(Rudow, 1866) (1 cf , 24 $)
Subgenus SPINIBOVICOLA subgen. n.
(Figs 12, 41, 50, 56)
Type-species: Trichodectes hemitragi Cummings.
DESCRIPTION. Anterior of head with osculum absent; pulvinus very short, but attaining anterior margin of
head; dorsal preantennal sulcus absent; clypeal marginal carina broadened medially into straight bar with
posterior margin slightly irregular; anterolateral margin of head smoothly rounded; preantennal portion
short, with outline rounded, almost straight anteriorly. Temple margin convex laterally, straight posterior-
ly, with short posteriorly-directed projection on postero-lateral angle bearing two setae. Male scape
slightly expanded, with setae randomly scattered. Dorsum of head with abundant setae of moderate
length. Sitophore sclerite with posterior arms extended (Fig. 12).
Thorax with dorsal and lateral setae abundant, long or of moderate length, present marginally and on
disc of prothorax and pterothorax; longest setae present postero-laterally on pterothorax.
Abdomen tapering more acutely in male than in female. Abdomen with posterior setal row comprising
long setae on sterna, terga and pleura, anterior setae shorter; anterior setae present on sterna, terga and
pleura (where present) of all abdominal segments except sometimes tergum I; male tergum II with curved
row of 3-4 long stout setae on each side, modified from posterior setal row, these setae being linked by a
curved sclerite (modified tergite) (Fig. 41). Pre-genital sclerites present on terga II-VII or VIII (males) and
II-IX (females) and sterna II-VII (males) and II- VIII (females); male terga lacking posterior sclerites.
Gonapophyses with broadly rounded lobe smoothly continuous with ventral margin; marginal setae
long, confined to lobe. Ventral vulval margin not sclerotised; convex. Postgenital pleural area lacking
spinose patch. Genital chamber lacking dorsal non-ornamented area or fold.
Male segment IX produced posteriorly into narrow, sclerotised extension; subgenital plate tapering
characteristically, comprising sternites VI and VII linked by s.g.p.r. (Fig. 50). Pseudostyli absent. Male
genital opening dorsal. Basal apodeme not as long as parameres, not concave anteriorly. Parameres fused
basally, long and tapering to pointed apices. Basiparameral sclerites absent. Mesomeres absent, or
represented by very short sclerites, not apically fused. Male genitalia depicted in Fig. 56.
HOSTS. Bovidae (Artiodactyla).
COMMENTS. Neither of the two included species is known to be parthenogenetic.
SPECIES INCLUDED
hemitragi (Cummings, 1916) (20 cf , 26 $)
multispinosus Emerson & Price, 1979 (8 C?, 11 $)
Genus BISONICOLA gen. n.
(Figs 58-60)
Type-species: Bovicola sededmdecembrii Eichler.
DESCRIPTION. Anterior of head with osculum absent, though pulvinus attaining margin; margin anteriorly
to pulvinus membranous, hyaline; dorsal preantennal sulcus present; clypeal marginal carina slightly
broadened medially; anterolateral margin of head smoothly convex; preantennal portion of head short or
longer, but not as long as postantennal portion, outline broadly rounded, though slightly truncate
anteromedially. Temple margin broadly and smoothly convex. Male scape expanded, with setae randomly
scattered; flagellomeres fused in males and females; male flagellum with two or three basally-articulated
'teeth'; male flagellum not 'roughened' on interior face. Dorsum of head with numerous setae of medium
length. Sitophore sclerite unmodified.
Thorax with lateral and dorsal setae long and of moderate length; setae present along lateral margins and
254
C. H. C. LYAL
58
60
Figs 58-60 Bisonicola sedecimdecembrii. 58, $ terminalia, ventral. 59, O" subgenital plate, setae omitted.
60, cf genitalia.
posteriorly (dorsally) on prothorax and pterothorax; posterior setal row on prothorax marginal, with
median gap; posterior setal row of pterothorax submarginal, with median gap absent; pterothorax with
posterior setal row incorporating two very long setae between postero-lateral and postero-median angles;
pair of setae, widely spaced, present on disc of prothorax dorsally; scattered setae sometimes present
posteriorly on disc of pterothorax dorsally. Atria of thoracic spiracles very large.
Abdomen oval. Abdominal spiracles present on segments III- VIII. Abdomen with setae short and of
medium length; anterior setae present on sterna, terga and pleura; postero-lateral setae absent. Abdomin-
al pleural projections absent. Sclerites present on sterna, terga and pleura (where present) of all pre-genital
abdominal segments except tergum I; male terga V, VI and VII with anterior and posterior sclerites.
Gonapophyses broad, truncate; setae present along postero-median margin; ventral lobe absent (Fig.
58). Gonapophyses meet ventral vulval margin acutely, not linked by sclerotised band. Ventral vulval
margin not sclerotised; very short, more or less straight; subgenital lobe absent.
Male subgenital plate with sternites VII and IX fused to s.g.p.r., sternite VIII present but not fused to
s.g.p.r.; s.g.p.r. heavily sclerotised, widest on sternum VIII (Fig. 59). Pseudostyli present, large, broadly
triangular (Fig. 59). Male genital opening dorsal. Parameres separate, rod-like, fused to mesomeral arch.
Basiparameral sclerites absent. Mesomeres fused apically; median extension absent. Male genitalia
depicted in Fig. 60.
HOSTS. Bovidae (Artiodactyla).
SPECIES INCLUDED
sedecimdecembrii sedecimdecembrii (Eichler , 1946) comb. n. from Bovicola (5 d", 4 9)
sedecimdecembrii bison (Blagoveshtchenski, 1967) comb. n. from Bovicola (2 C?, 1 $)
TRICHODECTID MAMMAL LICE 255
Genus WERNECKIELLA Eichler gen. rev.
(Figs 61, 62)
Werneckiella Eichler, 1940: 160. Type-species: Trichodectes equi Denny, by original designation.
DESCRIPTION. Anterior of head with osculum absent; pulvinus not attaining margin; dorsal preantennal
sulcus present; clypeal marginal carina slightly broader medially than laterally, or not broadened;
anterolateral margin of head smoothly convex; preantennal portion of head not long, outline broadly
rounded, sometimes slightly flattened anteriorly. Temple margin convex or rectangular. Male scape
expanded, with setae randomly scattered; flagellomeres fused in males and females; male flagellum with
two basally-articulated 'teeth'; male flagellum not 'roughened' on interior face. Dorsum of head with
abundant short setae. Sitophore sclerite unmodified.
Thorax with lateral and dorsal setae short and of medium length; setae present along lateral margins and
posteriorly (dorsally) on prothorax and pterothorax; posterior setal row on prothorax marginal, with
median gap; posterior setal row on pterothorax marginal or submarginal, with no median gap; posterior
setal row of pterothorax incorporating two very long setae with intervening shorter setae between
postero-lateral and postero-median angles; prothorax with seta or setae on disc laterally (dorsally).
Abdomen elongate-oval. Abdominal spiracles present on segments III-VIII. Abdomen with setae short
and of moderate length; anterior setae frequently present on sterna and terga, always present on pleura;
postero-lateral setae absent. Abdominal pleural projections absent. Sclerites present on abdominal pleura
II- VII and sterna and terga of at least abdominal segments II- VII; male terga with single sclerites only.
Gonapophyses broad, truncate, with median faces almost parallel to one another; marginal setae of
moderate length; ventral lobe absent. Gonapophyses meet ventral vulval margin acutely, not linked by
sclerotised band. Ventral vulval margin not sclerotised, very short, straight; subgenital lobe absent.
Male subgenital plate with sternite VII present and fused to s.g.p.r., sternites VIII and IX absent;
s.g.p.r. sinuate and broadest on sternum VIII (Fig. 61). Pseudostyli present, small, simple setose lobes.
Male genital opening postero-dorsal. Parameres long, straight or flared and twisted medially, sometimes
fused basally. Basiparameral sclerites absent. Mesomeres fused or almost fused into pentagonal mesomer-
al arch with median extension absent; mesomeres broadest basally (external to b.a.l.s.) and more or less
broad distally; mesomeres projecting basally between b.a.l.s. to contact parameres. Male genitalia
depicted in Fig. 62.
HOSTS. Equidae (Perissodactyla) and Bovidae (Artiodactyla).
COMMENTS. Some species are parthenogenetic, the males being unknown.
Werneckiella was considered a subgenus of Damalinia by Hopkins (1949) and a synonym of Bovicola by
Werneck (1950); it is here raised from synonomy with Bovicola. A more extensive history of the variations
in status of this genus is presented in Table 3. The genus was revised by Moreby (1978).
SPECIES INCLUDED
aspilopyga (Werneck, 1956) comb. n. from Bovicola (9 cf , 11 $)
equi (Denny, 1842) comb. n. from Bovicola (3 cT, c. 100 <j>)
fulva (Emerson & Price, 1979) comb. n. from Bovicola (4 cf , ?1 $)
neglecta (Keler, 1942) comb. n. from Bovicola (5 cf , 6 $)
ocellata (Piaget, 1880) comb. n. from Bovicola (17 $)
zebrae Moreby, 1978 comb. rev. from Bovicola (1 cf , 2 <j>)
zuluensis (Werneck, 1950) comb. n. from Bovicola (13 (?, 13 <j>)
Genus TRAGULICOLA gen. n.
(Figs 63, 64)
Type-species: Damalinia traguli Werneck.
DESCRIPTION. Anterior of head with osculum present, broad; dorsal preantennal sulcus present; clypeal
marginal carina broadened medially into less heavily sclerotised dorsal sclerite, which is broad, posteriorly
convex and with median posterior projection (Fig. 63); anterolateral margin of head straight or slightly
sinuate; preantennal portion of head as long as its maximum width, outline trapezoid. Temple margin
convex or rectangular. Male scape expanded, with setal row present and comprising at least four setae;
flagellomeres fused in males and females; male flagellum with two basally-articulated 'teeth'. Dorsum of
head with setae short posteriorly and of moderate length anteriorly. Sitophore sclerite unmodified.
Thorax with dorsal setae short or of moderate length; prothorax and pterothorax with marginal or
256
C. H. C. LYAL
I
64
62
Figs 61-64 Bovicolinae species. 61, 62, Werneckiella equi C?, (61) terminalia; (62) genitalia. 63, 64,
Tragulicola traguli, (63) $ head, dorsal; (64) cf genitalia.
submarginal posterior setal row, the longest setae being posterolaterally except in the male, which has a
pair of long setae medially on the posterior row of the pterothorax; male with setal patch centrally on disc of
pronotum, but no other setae present on disc of either sex.
Abdomen elongate, with male tapering to more acute posterior angle than female. Abdominal spiracles
present on segments III- VIII; atria oblate-spheroids, very large. Abdomen with setae of moderate length,
the longest being those comprising the pleural posterior setal row, particularly of the posterior pleura;
anterior setae present on all pleura, but not sterna or terga; postero-lateral setae absent. Abdominal
pleural projections absent. Sclerites present on sterna, terga and pleura (where present) of all abdominal
segments except tergum I, which is reduced and obscure; male terga HI-VI with anterior and posterior
sclerites.
Gonapophyses broad medially, tapering smoothly distally; ventral margin with long, abundant setae;
ventral lobe absent. Gonapophyses meet ventral vulval margin acutely, not linked by sclerotised band.
Ventral vulval margin not sclerotised: convex; subgenital lobe absent; marginal spines present, though
difficult to see. Female genital chamber with dorsal wall lacking spicules over narrowly triangular area
anteromedially.
Male subgenital plate with sternites IX and VIII linked by s.g.p.r., but sternite VII not attache<
TRICHODECTID MAMMAL LICE
257
Pseudostyli present, short, conical; median ventro-posterior projection also present, longer than pseudo-
styli. Male genital opening postero-dorsal. Basal apodeme acuminate apically. Parameres broad, triangu-
lar, poorly-sclerotised, asymmetrically deflected (may be artifact of preparation, though deflected the
same way in all specimens seen). Basiparameral sclerites absent. Mesomeres fused apically, median
extension absent; mesomeral arch fused to b.a.l.s. about one-third length of basal apodeme anteriad from
posterior end. Male genitalia depicted in Fig. 64.
HOSTS. Tragulidae (Artiodactyla).
SPECIES INCLUDED
fragu/j (Werneck, 1950) comb. n. from Damalinia (70 cf , 75 $)
Genus DAMALINIA Mjoberg
The genus Damalinia comprises three subgenera.
DESCRIPTION. Anterior of head with osculum present, narrow or broad, deep or shallow, or osculum
absent, in which case head as described below for D. (T.) conectens*; dorsal preantennal sulcus present or
absent*; clypeal marginal carina more or less broadened medially and of variable form*; anterolateral
margin of head straight, slightly concave, slightly sinuate or convex; preantennal portion of head of
variable length, outline triangular, trapezoid, rectangular or rounded*. Temple margin smoothly convex,
sometimes convexly produced posteriad* , with posterolateral angle sometimes developed laterally or with
small posterior rounded projection*. Male scape expanded, with setal row present or setae randomly
scattered*; flagellomeres fused in males and females; male flagellum with two or three basally-articulated
'teeth'* and interior face serrate or 'roughened'. Dorsum of head with setae sparse or more or less
abundant, short or of moderate length, frequently longer along the anterolateral margins and across the
clypeus than elsewhere. Sitophore sclerite unmodified.
Thorax with dorsal and marginal setae short, long or of moderate length, frequently longest on
posterolateral margin of pterothorax. Prothorax with setae sparse or absent on anterolateral margin;
posterior setal row marginal, though directed onto disc medially and median setal pair sometimes isolated,
row more or less sparse, with median gap between setae (other than isolated median pair) present,
sometimes wide*; single seta frequently present on dorsal disc anterolaterally. Pterothorax with posterior
setal row marginal or submarginal, sometimes irregular or 'doubled'*, median gap present or absent,
posterior setal row incorporating two very long setae with intervening shorter setae between postero-
lateral and postero-median angles; setae absent from disc.
Abdomen oval, elongate, or very elongate and narrow*. Abdominal spiracles present on segments
III- VIII. Abdominal setae short or of moderate length, frequently longer on pleura than on sterna and
65
Figs 65-67
meyeri,
Damalinia species. 65, D. (Tricholipeurus) elongata, $ terminalia, ventral. 66, D. (Cervicola)
gonapophysis, ventral. 67, D. (C.) hendrickxi, $ gonapophysis, ventral.
258
C. H. C. LYAL
68
Figs 68-70 Damalinia species. 68, D. (D.) crenelata, C? abdomen. 69, D. (D.) baxi, $ head, dorsal. 70,
D. (Cervicola) martinaglia, cf scape.
terga; anterior setae present on all pleura except, occasionally, pleurum II, rarely on sterna and terga;
postero-lateral setae absent. Abdominal pleural projections absent. Pregenital sclerites present on sterna,
terga and pleura (where present) of all segments except tergum I , sometimes absent or very small on pleura
(Fig. 80)*; male terga with or without posterior sclerites*.
Gonapophyses variable; ventral margin without rounded lobe, but sometimes with hook-shaped
projection*; marginal setae present. Gonapophyses meet ventral vulval margin acutely, not joined by
sclerotised band. Ventral vulval margin sclerotised or not sclerotised; subgenital lobe present or absent*.
Dorsal margin of vulva and post-vulval area with or without pointed scales*. Common oviduct, at
branching point, with or without collar (see subgenus Cervicola)*.
Male subgenital plate variable, sternites VII and VIII always being present and fused to s.g.p.r. , sternite
IX sometimes modified. Pseudostyli absent or, if present, of variable form*. Posterior margins of tergum
TRICHODECTID MAMMAL LICE
259
76
Figs 71-76 Damalinia (D.) species. 71, D. theileri, cf terminalia. 72, D. appendiculata, cf terminalia. 73,
D. theileri, $ terminalia. 74-76, cf genitalia of (74) D. orientalis; (75) D. neotheileri; (76) D. crenelata.
260
C. H. C. LYAL
77
78
79
80
Figs 77-80 Damalinia species, d" terminalia. 77, D. (Cervicola) natalensis, ventral, setae omitted. 78, D.
(C.) martinaglia, ventral, setae omitted. 79, D. (Tricholipeurus) indica, ventral, setae omitted. 80, D.
(T.) aepycerus.
IX sometimes greatly expanded * . Male genital opening postero-dorsal or dorsal . Male genitalia variable * .
HOSTS. Bovidae and Cervidae (Artiodactyla).
Subgenus DAMALINIA Mjoberg
(Figs 11, 68, 71-76)
Damalinia Mjoberg, 1910: 69. Type-species: Trichodectes crenelatus Piaget, by monotypy.
DESCRIPTION. Anterior of head with osculum present, narrow or broad, deep or shallow; dorsal preanten-
nal sulcus present; clypeal marginal carina broadened medially either into more or less developed simple
TRICHODECTID MAMMAL LICE
261
81
Figs 81-86 Damalinia species, cf genitalia. 81, D. (Tricholipeurus) victoriae. 82, D. (T.) indica. 83, D.
(T.) aepycerus. 84, D. (Cervicola) hopkinsi. 85, D. (C.) meyeri. 86, D. (C.) reduncae.
bar with posterior margin straight or concave, or into more or less broad U-shaped sclerite, or into broad,
heavily-sclerotised margin of deep osculum; anterolateral margin of head convex or slightly sinuate, in the
latter case slightly concave at junction of margin and clypeof rental sulcus and convex anteriorly;
preantennal portion longer or shorter than posterior portion, outline triangular, trapezoid or rounded,
sometimes with slight protuberances on either side of osculum. Temple margin smoothly convex, slightly
produced posteriad, sometimes with postero-lateral angle developed laterally, or with small posterior
rounded projection. Male scape with setae randomly scattered; male flagellum with two basally-articulated
'teeth'. Dorsum of head with setae more or less abundant.
262 C. H. C. LYAL
Pterothorax with posterior setal row sometimes irregular or 'doubled'.
Abdomen oval or elongate, sometimes very narrowly elongate. Abdominal setae present anteriorly on
all pleura, occasionally on terga and sterna, but only laterally and as irregularity or 'doubling' of posterior
setal row. Pleurum II with sclerite extending broadly or narrowly onto sternum II and sometimes tergum
II, frequently 'crowding' sternite or tergite II (Fig. 68); pleurites not reduced in size or absent; tergum I
lacking sclerite; male terga V and VI (at least) with both anterior and posterior sclerites.
Gonapophyses variable, sometimes hook-shaped, though lacking distal spur, more frequently obtuse,
sometimes with ventral (median) margin concave or convex; ventral margin with setae long or of moderate
length, abundant, setae sometimes also present on anterior margin. Ventral vulval margin not sclerotised,
sometimes short, straight or convex; subgenital lobe absent, though posterior margin of sternum VII
sometimes developed into two spikes (Fig. 73). Dorsal margin of vulva and postvulval area usually without
pointed scales. Common oviduct without 'collar'.
Male subgenital plate with sternites VII, VIII and IX fused to s.g.p.r. , sternite IX and postgenital sclerite
sometimes fused; s.g.p.r. more heavily sclerotised than sternites (Fig. 68). Pseudostyli absent or, if
present, long and broad, parallel-sided or with basal constriction (Figs 68, 71), or long and narrow (Fig.
72). Posterior margin of tergum IX not greatly expanded. Parameres more or less broad, sometimes fused
together. Basiparameral sclerites absent. Mesomeres unfused apically, sometimes fused to b.a.l.s. in
characteristic manner (Fig. 75) or to parameres, basally or in entirety, in the latter case apparently absent.
Endophallus lacking spicular patch (cf. subgenus Tricholipeurus). Male genitalia depicted in Figs 7476.
HOSTS. Bovidae (Artiodactyla).
COMMENTS. Emerson & Price (1982) distinguish their new species orientalis (described in Bovicold) from
the very similar species thompsoni Bedford on the following grounds: The female of B. orientalis is smaller
than that of B. thompsoni and the lateral margins of the forehead are even [sic] so slightly indented for B.
orientalis and always even for B. thompsoni; the median plates on tergites II- VIII are of different shapes
for the two species; the chaetotaxy of terminal abdominal segments is different, with each gonapophysis
having at least 20 median and anterior setae for B. thompsoni [orientalis having, according to the preceding
description, 12-17 setae]; and the posterior margin of the temple of B. thompsoni has small projections that
are not present for B. orientalis. The male of B. thompsoni is unknown.' Five male and seven female
paratypes of B. orientalis were examined in this study, together with a further eight males and eight females
from the same host (Capricornis crispus swinhoei) not examined by Emerson & Price when they prepared
their description of orientalis, and three females of thompsoni, including the holotype. Treating the
supposed distinguishing characters in order, the three specimens of thompsoni are larger than any of the
females from C. c. swinhoei; whilst none of the specimens of thompsoni has an indentation on the forehead
(at the junction of the clypeo-frontal sulcus with the margin), not all of the specimens from Capricornis c.
swinhoei have either; the shapes of abdominal tergites II- VIII are not significantly different in the two
species; the chaetotaxy of the terminal abdominal segments is not different, and no specimen of thompsoni
has more than 17 setae on the median and anterior margins of the gonapophyses; the small projections of
the posterior temple margins are present in all specimens of orientalis. It seems, therefore, that the
differences between the two species are twofold: host (orientalis being described from Capricornis crispus
swinhoei and thompsoni being known only from Capricornis sumatrensis sumartrensis) , and size. The
biological significance of the latter character is not clear, and the two species may be found to differ in other
characters not so far discovered. Until a larger sample can be examined, collected from more localities, no
taxonomic action is taken to reduce the rank or synonymise orientalis, though on the basis of the
information so far available the species probably should not stand.
SPECIES INCLUDED
adenota (Bedford, 1936) comb. n. from Bovicola (39 cf , 35 9)
appendiculata (Piaget, 1880) (19 cf , 25 9)
fcaxiHopkins, 1947 (16 cf , 24 9)
chorleyi (Hopkins, 1941) (21 cf , 19 9)
crenelata (Piaget, 1880) (27 cf , 20 9)
dimorpha (Bedford, 1939) comb. n. from Bovicola (syntype 9)
/a/ireii/io/zi'(Eichler, 1949) comb. n. from Tricholipeurus
harrisoni (Cummings, 1916) (3 cf , 3 9)
hiltt (Bedford, 1934) comb. n. from Bovicola (42 cf , 27 9)
neotheileri Emerson & Price, 1971 (1 cf , 6 9)
orientalis (Emerson & Price, 1982) comb. n. from Bovicola (108 cf , 135 9)
ornate Werneck, 1957 (Holotype cf )
pelea (Bedford, 1934) comb. n. from Bovicola
TRICHODECTID MAMMAL LICE 263
semitheileri Emerson & Price, 1971 (holotype cf , allotype $)
theileri Bedford, 1928 (2 cf . 4 $)
thompsoni (Bedford, 1936) comb. n. from Bovicola (3 $)
Subgenus CERVICOLA Keler gen. rev., stat. n.
(Figs 66, 67, 70, 77, 78, 84-86)
Cervicola Keler, 1934: 263. [Nomen nudum.]
Cervicola Keler, 1938a: 460. Type-species: Trichodectes tibialis Keler (nee Piaget) [= Trichodectes meyeri
Taschenberg], by original designation.
DESCRIPTION. Anterior of head with osculum present, narrow or broad, deep or shallow; dorsal preanten-
nal sulcus present; clypeal marginal carina broadened medially, either into simple bar with posterior
margin straight or convex but occasionally with median posterior projection, or into longer posteriorly-
developed sclerite with posterolateral angles more or less convex and more or less pronounced median
posterior projection; anterolateral margin of head straight or slightly sinuate, in the latter case slightly
concave at junction of margin and clypeof rental sulcus and convex anteriorly; preantennal portion of head
as long as or shorter than postantennal portion, outline triangular, trapezoid or rounded. Temple margin
smoothly convex, slightly produced posteriad. Male scape with setae randomly scattered; male flagellum
with two basally-articulated 'teeth'. Dorsum of head with setae more or less abundant, sometimes less so
posteriorly than anteriorly.
Pterothorax with posterior setal row single.
Abdomen oval or elongate. Abdominal setae present anteriorly on all pleura except, occasionally,
pleurum II, but may be very short, fine and difficult to see; anterior setae never on sterna and terga.
Pleurum II never with sclerite extending onto sternum II; pleurites not reduced in size or absent; male terga
V and VI (at least) with both anterior and posterior sclerites.
Gonapophyses hook-shaped, apex of curved portion acute or rounded, sometimes with distal (dorsal)
spur (Figs 66, 67). Gonapophyses with setae long or of moderate length on posterior margin and sometimes
on apex of 'hook', smaller setae sometimes present on anterior margin of 'hook'. Ventral vulval margin not
sclerotised. Subgenital lobe absent. Dorsal margin of vulva and post-vulval area with pointed scales.
Common oviduct at branching-point with folded and more or less apparent 'collar', sometimes partially
sclerotised and refracting light when viewed in phase-contrast or bright field transmitted light.
Male subgenital plate with sternites VII, VIII and IX fused to s.g.p.r. , though sternite IX sometimes not
complete, s.g.p.r. sometimes not attaining posterior margin of segment IX, perisetal gaps sometimes
absent; s.g.p.r. more heavily sclerotised than sternites. Pseudostyli absent or, if present, apically pointed
and more or less broad (Fig. 78) or apically rounded and very narrow (Fig. 77). Posterior margins of male
tergum IX not greatly expanded. Parameres broad or narrow, larger or smaller than mesomeres, may be
reduced to small discs, in which case mesomeres absent; parameres fused or unfused; apices sometimes
widely divergent. Basiparameral sclerites present or absent. Mesomeres absent or, if present, not fused
apically, nor fused to parameres or b.a.l.s. Endophallus lacking spicular patch (cf. subgenus Tricho-
lipeurus).^ Male genitalia depicted in Figs 84-86.
HOSTS. Bovidae and Cervidae (Artiodactyla).
COMMENTS. Cervicola was treated as a synonym of Damalinia by Werneck (1950), and of both Damalinia
and Bovicola by Hopkins & Clay (1952, pp. 102 and 67 respectively). The history of the variations in status
of Cervicola is presented in Table 3.
SPECIES INCLUDED
annectens Hopkins, 1943 comb. rev. from Tricholipeurus (21 cf , 25 $)
forficula (Piaget, 1880) (5 cf , 8 $)
hendrickxi Hopkins, 1947 (4 cf , 6 $)
hopkinsi Bedford, 1936 (15 cf , 28 $)
lerouxi (Bedford, 1930) comb. n. from Tricholipeurus (9 $)
maa/ Emerson & Price, 1973 (holotype cf , allotype $)
martinaglia (Bedford, 1936) (34 cf , 27 $)
meyeri meyeri (Taschenberg, 1882) (5 cf , 54 $)
meyeri hydropotis (Dobroruka, 1975) comb. n. from Cervicola
meyeri sika (Dobroruka, 1975) comb. n. from Cervicola
muntiacus (Seguy, 1948) (12 cf , 12 $)
natalensis Emerson, 1964 (2 cf , 2 $)
264 C. H. C. LYAL
reduncae (Bedford, 1929) stat. n., comb. n. from Tricholipeurus [Raised from subspecies of D.
trabeculae.} (2% <3, 29%)
trabeculae (Bedford, 1929) comb. n. from Tricholipeurus (10 cf , 10 $)
Uganda? (Werneck, 1950) stat. n., comb. n. from Tricholipeurus [Raised from subspecies of D.
trabeculae.] (14 cf , 13 $)
Unless otherwise stated, all species were previously considered as placed in Damalinia s. str.
Subgenus TRICHOLIPEURUS Bedford stat. n.
(Figs 4, 7, 8, 13,65,79-83)
Tricholipeurus Bedford, 1929: 514. Type-species: Tricholipeurus aepycerus Bedford, by original designa-
tion.
DESCRIPTION. Anterior of head variable, one of two types.
(a) Osculum absent; pulvinus short, not attaining anterior margin of head; dorsal preantennal sulcus
absent; clypeal marginal carina insignificant, not, or only slightly, broadened medially; anterolateral
margin of head straight posteriorly, convex anteriorly; preantennal portion of head as long as postantennal
portion, outline rounded anteriorly (D. (T.) conectens only).
(b) Osculum present, rarely deep; dorsal preantennal sulcus present; clypeal marginal carina broadened
medially either into simple bar with posterior margin straight or convex, or into longer posteriorly-
developed U- or W-shaped sclerite, with posterolateral angles more or less acutely convex, and frequently
with more or less pronounced median posterior projection; anterolateral margin straight, slightly sinuate,
or slightly concave; preantennal portion of head longer or shorter than postantennal portion but not short,
outline trapezoid or rectangular. Temple margin smoothly convex, more or less convexly produced
posteriorly. Male scape with setal row frequently present, comprising four or five setae, though setae
sometimes more or less randomly scattered; male flagellum with two or three basally- articulated 'teeth'.
Dorsum of head with setae sometimes sparse.
Pro thorax with median gap of posterior setal row sometimes almost the width of the posterior margin.
Pterothorax with posterior setal row single.
Abdomen usually very elongate, narrow. Abdominal setae present anteriorly on all pleura except,
occasionally, pleurum II, but may be very short, fine and difficult to see; anterior setae never on sterna and
rarely on terga. Pleurum II never with sclerite extending onto sternum II; pleural sclerites frequently
reduced to small anterior plate or absent; tergum I sometimes lacking sclerite; male terga lacking posterior
sclerites, or with both anterior and posterior sclerites present on at least segments V and VI (though may be
present on any segments up to II- VIII).
Gonapophyses variable, sometimes hook-shaped with distal spur and marginal setae on posterior
(dorsal) margin of 'hook', or long with ventral margin convex, straight or sinuate (though not with lobe),
apically pointed or rounded; marginal setae of moderate length. Ventral vulval margin sometimes
sclerotised; subgenital lobe present, variable, not marginally serrate, ventrally smooth or scaled. Dorsal
margin of vulva and post-vulval area without pointed scales. Common oviduct without 'collar'.
Male subgenital plate with sternites VII, VIII and IX fused to s.g.p.r.; perisetal gaps sometimes large,
occasionally absent; s.g.p.r. heavily sclerotised and sternites sometimes very lightly scerlotised; s.g.p.r.
sometimes curved or sinuate (Fig. 80). Pseudostyli absent or, if present, variable, short and rounded or
apically angular, posteriorly or medially directed, narrow or broad, sometimes fused to form single caudal
projection. Posterior margins of male tergum IX frequently greatly expanded (Fig. 79). Lateral struts of
basal apodeme sometimes with anteposterior spur (Fig. 83). Parameres unfused or, if fused, plate apically
pointed or bifurcate. Basiparameral sclerites present or absent. Mesomeres unfused apically or, if fused,
symmetric or asymmetric, median extension absent or present; mesomeral arch frequently recurved
abruptly at base to contact parameres, sometimes extended between b.a.l.s.; mesomeres not fused to
parameres or b.a.l.s. Endophallus with patch of regularly-arranged and numerous spicules sometimes very
apparent. Male genitalia depicted in Figs 81-83.
HOSTS. Bovidae and Cervidae (Artiodactyla).
COMMENTS. Damalinia (T.) longiceps is included following the statement of Clay & Hopkins (1955) that it
resembles D. (T.) spinifer Hopkins 'most closely among known species'.
Tricholipeurus has been treated as a genus, synonym (Hopkins, 1943) and subgenus (Hopkins, 1949) of
Damalinia; the history of the variations in status of Tricholipeurus is presented in Table 3.
SPECIES INCLUDED
aepycerus (Bedford, 1929) comb. n. from Tricholipeurus (1 cf , 1 $)
TRICHODECTID MAMMAL LICE 265
albimarginata (Werneck, 1936) comb. n. from Tricholipeurus (9 cf , 13 $)
antidorcus (Bedford, 1931) comb. n. from Tricholipeurus (11 cf, 18 9)
bvdfordi (Hill , 1922) comb. n. from Tricholipeurus (2 cf , 2 $ )
clayi( Werneck, 1938) comb. n. from Tricholipeurus (14 cf , 12 $)
conectens Hopkins, 1943 comb. rev. from Tricholipeurus (18 cf , 16 $)
cornuta cornuta (Gervais, 1844) comb. n. from Tricholipeurus (20 cf , 27 $)
cornuta ourebiae Hopkins, 1943 comb. rev. from Tricholipeurus (18 cf , 17 $)
dorcephali( Werneck, 1936) comb. n. from Tricholipeurus (2 cf , 2 $)
elongata (Bedford, 1934) comb. n. from Tricholipeurus (10 cf , 10 $)
indica (Werneck, 1950) comb. n. from Tricholipeurus (65 cf , 60 $)
//jieafa (Bedford, 1920) comb. n. from Tricholipeurus (75 cf , 83 $)
//peuro/des(Megnin, 1884) comb. n. from Tricholipeurus (106 cf , 91 $)
70iigiceps(Rudbw, 1866)
moschatus (Emerson & Price, 1971) comb. n. from Tricholipeurus (holotype cf , disassociated cf head)
pakenhami (Werneck, 1947) comb. n. from Tricholipeurus (21 cf , 20 $)
parallels (Osborn, 1896) comb. n. from Tricholipeurus (50 cf , 106 $)
parkeri (Hopkins, 1941) comb. n. from Tricholipeurus (10 cf , 10 $)
spinifer Hopkins, 1943 comb. rev. from Tricholipeurus (17 cf , 16 $)
victoriae Hopkins, 1943 comb. rev. from Tricholipeurus (33 Cf , 35 $)
EUTRICHOPHILINAE Keler
Genus EUTRICHOPHILUS Mjoberg
(Figs 87-93)
Eutrichophilus Mjoberg, 1910: 71. Type-species: Eutrichophilus cercolabes Mjoberg, by subsequent
designation (Harrison, 1916a: 21).
DESCRIPTION. Anterior of head with osculum present or absent, but pulvinus always attaining margin;
dorsal preantennal sulcus absent; clypeal marginal carina with median expansion absent or slight, or
present as broad or narrow parallel-sided bar with transverse margins convex, straight, or concave;
anterolateral margin of head straight or convex; preantennal portion of head long or short; outline
triangular, rounded or broadly trapezoid. Temple margin convex or with posterolateral angle apparent;
temples greatly expanded posteriad (Fig. 87). Male scape expanded, with longitudinal setal row present
and comprising two setae; male flagellomeres fused; female flagellomeres fused or flagellum comprising
two flagellomeres; male flagellum very long, with two basally-articulated 'teeth'. Dorsum of head with
setae short or long, sometimes longer anteriorly than posteriorly. Sitophore sclerite unmodified.
Thorax with dorsal setae short or of moderate length marginally or submarginally on posterior of
prothorax and pterothorax, absent from disc of both; one or two long setae on posterolateral margins of
pterothorax.
Abdomen oval and elongate. Abdominal spiracles present on segments III- VIII. Abdominal setae short
or of moderate length, with tufts of long setae on at least pleurum VIII, sometimes also pleurum VII
(males) or IX (females) (Fig. 88); anterior setae present on all pleura but absent from sterna and terga;
postero-lateral setae present. Abdominal pleura lacking projections dorsally or ventrally. Sclerites present
on sterna, terga and pleura (where present) of all pre-genital abdominal segments except, sometimes,
tergum I, which may be very small; male terga, at least of abdominal segments V and VI, with anterior and
posterior sclerites.
Gonapophyses frequently large, broadly triangular or rounded, ventral margin lacking lobe but with
more or less dense marginal setae which are long or of moderate length. Gonapophyses meet ventral vulval
margin acutely, not linked by sclerotised band. Ventral vulval margin not sclerotised; smoothly convex,
with or without median indentation or setose projection; subgenital lobe absent. Female terminalia
depicted in Fig. 90.
Male subgenital plate with sternites VII and VIII present and fused to s.g.p.r. , IX absent or, if present,
fused to s.g.p.r. (Fig. 88). Pseudostyli absent. Male genital opening dorsal, male segment IX posterior.
Parameres long or short, narrow or broad; with basiparameral sclerite or flange sometimes present and
fused medially, thus linking parameres, but otherwise unfused. Mesomeres present, fused apically to form
arch with no median extension; arch smoothly rounded, or with lateral desclerotisations, in which case
median portion is straight and at right angles to lateral portions, very poorly sclerotised and thin, or absent.
Male genitalia depicted in Figs 89, 91, 92.
266
C. H. C. LYAL
Figs 87-89 Eutrichophilus species. 87, E. minor, $ head, dorsal. 88, E. setosus, cf abdomen. 89,
setosus, cf genitalia.
HOSTS. Erethizontidae (Rodentia).
SPECIES INCLUDED
cerco/aftesMjoberg, 1910 (29 cf , 25 9)
comitans Werneck, 1950 (6 cf , 2 $)
cordicepsMjoberg, 1910 (23 cf , 27 $)
exiguus Werneck, 1950 (holotype cf , allotype $)
guyanensis Werneck, 1950 (8 Cf , 7 $)
/o&afusEwing, 1936 (5 cf , 8 $)
maximus Bedford, 1939 (11 cf , 11 $)
mexicamis(Rudow, 1866) (50 cf , 50 $)
minor Mjoberg, 1910 (34 cf , 27 $)
moojeni Werneck, 1945 (3 cf , 3 })
setosus (Giebel, 1874) (102 cf , 102 $)
TRICHODECTID MAMMAL LICE
267
91
Figs 90-93 Eutrichophilus species. 90, E. maximus, 9 terminalia. 91, E. guyanensis, cf genitalia. 92, E.
guyanensis, cf genitalia, detail. 93, E. moojeni, cf genitalia.
DASYONYGINAE Keler
Genus CEBIDICOLA Bedford
(Figs 94-99)
Cebidicola Bedford, 1936: 52. Type-species: Trichodectes armatus Neumann, by original designation.
Meganarion Keler, 1938a: 465. Type-species: Trichodectes armatus Neumann, by original designation.
[Synonymy by Eichler, 1941.]
268
C. H. C. LYAL
Fig. 94 Cebidicola armatus, 9 terminalia.
,/
Fig. 95 Cebidicola armatus, cf abdomen.
TRICHODECTID MAMMAL LICE
98
Figs 96-99 Cebidicola species. 96, C. semistriatus , $ head, dorsal. 97, C. armatus, 9 head, dorsal. 98, C.
armatus, cf genitalia. 99, C. extrarius, C? genitalia.
DESCRIPTION. Anterior of head with osculum present, deep; dorsal preantennal sulcus present or absent;
clypeal marginal carina broadened medially into dorsal, posteriorly convex, sclerite; anterolateral margin
straight, slightly convex, concave or sinuate anteriorly, more or less abruptly concave at junction with
clypeof rental sulcus, with or without anterior sclerotised projection on either side of osculum (Figs 96, 97);
preantennal outline broadly triangular. Temple margin convex or slightly acute and angular posterolateral-
ly, with eyes more or less prominent (Figs 96, 97). Male scape expanded, with setal row present and
comprising two or more setae; male flagellomeres fused, with two basally-articulated 'teeth'; female
flagellomeres fused or unfused. Dorsum of head with setae short or of moderate length, sparse. Sitophore
sclerite unmodified.
Tarsal claws lacking ventral spines or teeth. Postcoxale absent or present, not greatly developed.
Thoracic setae present dorsally only along posterior and posterolateral margins of pterothorax; setae short
medially, longer laterally. Atrium of thoracic spiracle tubular or conical.
Abdomen oval, sometimes tapering posteriorly more in male than female. Abdominal spiracles present
on segments III- VIII. Abdominal setae short or of moderate length, longest on pleura VI- VIII; anterior
setae sometimes present on pleura and laterally on sterna and terga; postero-lateral setae present,
sometimes numbering more than one per site. Pleural projection present ventrally on abdominal pleurum
IV, large, sclerotised. Sclerites present on all abdominal pleura, on at least abdominal terga II to VIII and
at least abdominal sterna V to VII; male terga, at least on segments V-VII, with anterior and posterior
sclerites (Fig. 95).
Gonapophyses broad, especially medially, though lobe absent; marginal setae long, densely crowded.
Gonapophyses meet ventral vulval margin acutely, not linked by sclerotised band. Ventral vulval margin
not sclerotised; sometimes expanded, otherwise smoothly convex (Fig. 94); bilobed median spinose
projection may be present, but subgenital lobe absent. Median longitudinal sclerite sometimes present on
female sternum VIII (Fig. 94).
Male subgenital plate with sternite VII fused to s.g.p.r. , VIII fused or not fused to s.g.p.r. , and IX absent
270 C. H. C. LYAL
or, if present, not fused to s.g.p.r. Pseudostyli absent or, if present, small, slender, incurved (Fig. 95). Male
genital opening postero-dorsal or dorsal. Parameres fused or unfused; basiparameral sclerites present,
fused. Mesomeres present, fused or unfused and, if fused, median extension absent; mesomeral arch
produced basally between b.a.l.s. to contact parameres, which do not meet b.a.l.s. Male genitalia depicted
in Figs 98, 99.
HOSTS. Cebidae (Primates).
SPECIES INCLUDED
armatus (Neumann, 1913) (3 cf , 4 $)
exfrariusWerneck, 1950 (21 cf , 13 $)
semiarmatus (Neumann, 1913) (12 (J, 12 <J>)
Genus PROCA VICOLA Bedford
The genus Procavicola comprises two subgenera.
DESCRIPTION. Anterior of head with osculum present, semicircular; dorsal preantennal sulcus present
clypeal marginal carina broadened medially into dorsal, posteriorly convex, sclerite (Fig. 106); conus
large, as long as female scape; anterolateral margin of head straight or convex anteriorly, more or less
abruptly concave at junction with clypeofrontal sulcus; preantennal outline broadly triangular. Temple
margin smoothly convex, sometimes produced posteriad, or with postero-lateral angle developed into
posteriorly-projecting triangular or rounded process*. Male scape expanded, with setal row represente(
by two setae only; male flagellomeres fused, with two basally-articulated 'teeth'; female flagellomeres
unfused and closely associated, fused to two closely associated annulations, or completely fused*. Dorsum
of head with setae short, sparse; anterior margin of head with setae longer than on disc. Sitophore sclerite
unmodified.
Tarsal claws lacking ventral spines or teeth. Postcoxale of metathoracic leg absent or present; if present
may be well developed, but not to the same degree as described for Procaviphilus (Meganarionoides) am
not fused to abdominal pleurite II. Thoracic setae present dorsally only along posterior and latero
posterior margins of prothorax and pterothorax; setae short, except for laterally on pterothorax, where o
moderate length.
Abdomen oval, more or less elongate (Figs 103, 105). Abdominal spiracles present on segments
III- VIII, all approximately the same size. Abdominal setae short or of moderate length; anterior setae
present on pleura only; postero-lateral setae present, sometimes numbering more than one per site*
Pleural projections present dorsally and ventrally on abdominal pleurum IV, sclerotised. Sclerites presen
on sterna, terga and pleura of all abdominal segments except, occasionally, tergum I* ; male terga, at leas
of segments IV- VI, with anterior and posterior sclerites; second abdominal sternum with broad
heavily-sclerotised apophysis underlying sternite, articulating with median extensions of abdomina
pleurite II (Figs 103, 105).
Gonapophyses broad, lacking lobe; marginal setae lacking tubercles, occasionally on small conica
protuberances (Figs 101, 102). Gonapophyses meet ventral vulval margin acutely, not linked by sclerotisec
band. Ventral vulval margin not sclerotised; expanded, sometimes W-shaped medially (Fig. 100),
sometimes broadened posteriorly (Fig. 101), sometimes contracted, shorter than length of gonapophyses
(Fig. 102)*.
Male subgenital plate with sternites VII and VIII present and fused to s.g.p.r., sternite IX absent or, if
present, fused to s.g.p.r. and perisetal gaps small*; s.g.p.r. not always attaining posterior margin of
segment IX*. Pseudostyli absent (Figs 103-105). Male genital opening postero-dorsal. Male genitalia very
variable* ; parameres not fused, basiparameral sclerites present or absent* , mesomeres fused or unfused*.
HOSTS. Procaviidae (Hyracoidea).
Subgenus PROCA VICOLA Bedford
(Figs 100, 104-106, 109)
Procavicola Bedford, 1932: 711. Type-species: Trichodectes sternatus Bedford, by original designation.
DESCRIPTION. Temple margin smoothly convex, sometimes projecting posteriorly, but never with posterior
membranous or lightly-sclerotised process.
Postero-lateral setae present, single at each site. Sclerites present on sterna, terga and pleura of all
abdominal segments except segment I, where tergal sclerite absent.
TRICHODECTID MAMMAL LICE
271
100
Figs 100-102 Procavicola species, $ terminalia, ventral. 100, P. (P.) natalensis. 101, P. (Condylocepha-
lus) lindfieldi. 102, P. (C.) dissimilis.
Ventral vulval margin expanded as described, much broader than length of gonapophyses (Fig. 100).
Male subgenital plate with sternites VII and VIII present and fused to s.g.p.r., sternite IX absent;
s.g.p.r. may be very slender, may not attain posterior margin of segment IX (Figs 104, 105). Parameres
unfused, more or less narrow, rod-like, sometimes asymmetrically curved. Basiparameral sclerites
frequently present, fused or separate. Mesomeres not fused, short. Endophallus lacking large, hook-like
sclerites. Male genitalia depicted in Fig. 109.
HOSTS. Procaviidae (Hyracoidea).
SPECIES INCLUDED
affijMsWerneck, 1941 (10 C?, 13 $)
272
C. H. C. LYAL
106
Figs 103-106 Procavicola species. 103, P. (Condylocephalus) dissimilis , C? abdomen. 104, P. (P.)
vicinus, C? terminalia. 105, P. (P.) eichleri, cf abdomen. 106, P. (P.) natalensis, $ head, dorsal.
TRICHODECTID MAMMAL LICE
273
107
108
109
Figs 107-109 Procavicola species, cf genitalia. 107, P. (Condylocephalus) dissimilis. 108, -P. (C.)
dissimilis, detail of right paramere and mesomere at junction with basal apodeme. 109, P. (P.)
pretoriensis.
brucei Werneck, 1941 (43 cf , 48 $)
eichleri Werneck, 1941 (45 cf , 56 $)
emarginatus (Bedford, 1928) (16 5, 19 $>)
furca Bedford, 1939 (1 cf)
heterohyracis Bedford, 1932 (3 cf , 1 $)
lopesi lopesi Bedford, 1939 (32 cf , 40 $)
lopesi vicinus Werneck, 1941 (8 cf , 10 9)
mokeetsi Bedford, 1939 (20 cf , 15 $)
natalensis Bedford, 1932 (23 cf , 12 $)
parvus Bedford, 1932 (8 Cf , 6 $)
pretoriensis Bedford, 1932 (23 Cf , 21 $)
sAoanusMaltbaek, 1937
sternal us (Bedford, 1928) (9 cf, 8 ?)
subparvus Bedford, 1932 (6 cf , 9 $)
thorntoni Hopkins, 1942 (18 cf , 22 $)
ugandensis Werneck, 1941 (17 cf , 18 $)
274 C. H. C. LYAL
Subgenus CONDYLOCEPHALUS Werneck
(Figs 101-103, 107, 108)
Condylocephalus Werneck, 1941: 497 [as subgenus of Procavicola Bedford]. Type-species: Procavicola
(Condylocephalus) bedfordi Werneck, by original designation.
DESCRIPTION. Temple margin convex, with posteriorly-projecting membranous or lightly-sclerotised
process, more apparent in male than female, triangular or as small rounded bump (linfieldi females).
Female flagellomeres unfused, but closely associated.
Abdomen with postero-lateral setae present, frequently doubled, trebled or numerous at each site.
Sclerites present on sterna, terga and pleura of all abdominal segments.
Ventral vulval margin expanded, sometimes broadened posteriorly and broader than length of gona-
pophyses (Fig; 101), otherwise narrower, width less than length of gonapophyses (Fig. 102).
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r., with perisetal gaps
small (Fig. 103). Parameres unfused, curved, with anterolateral projections, not asymmetric. Basipar-
ameral sclerites present, fused or unfused. Mesomeres fused apically; mesomeral arch with median
extension and lateral double flexion (Figs 107, 108). Endophallus ornamented with large, hook-like
sclerites (Fig. 107).
HOSTS. Procaviidae (Hyracoidea).
COMMENTS. Though Condylocephalus has been treated by most authors as a subgenus of Procavicola,
Eichler (1963) considered it to have full generic status.
SPECIES INCLUDED
bedfordi Werneck, 1941 (2 cf , 1 $)
dissimilis Werneck, 1941 (64 cf , 58 $)
hopkinsi Werneck, 1941 (24 cT, 29 $)
JbMlfieW(Hill, 1922) (77 cT, 65 $)
univirgatus (Neumann, 1913) (33 cf , 32 $)
Genus PROCAVIPHILUS Bedford
The genus Procaviphilus comprises two subgenera.
DESCRIPTION. Surface of head, thorax and abdomen frequently covered with clearly- visible scales or
sclerotised nodules.
Anterior of head variable, one of two types: 'procaviphilus' or 'procavicola'*.
(a) 'procaviphilus' type. Osculum absent or, if present, slightly concave only; dorsal preantennal sulcus
absent; clypeal marginal carina broadened medially into straight or slightly curved bar (Fig. 112); conus
small, not as long as female scape; anterolateral margin of head straight or convex anteriorly, no abrupt
concavity at junction with clypeofrontal sulcus; preantennal outline trapezoid (Fig. 112).
(b) 'procavicola' type. Osculum present, semicircular; dorsal preantennal sulcus present or absent; clypeal
marginal carina broadened medially into dorsal, posteriorly convex, sclerite (Fig. 106); conus large, as long
as female scape; anterolateral margin of head convex anteriorly, more or less abruptly concave at junction
with clypeofrontal sulcus; preantennal outline broadly triangular.
Temple margin smoothly convex, more or less projecting posteriorly. Male scape expanded, with setal
row represented by two setae only; male flagellomeres fused, with two basally-articulated 'teeth'; female
flagellomeres unfused , though sometimes very closely associated . Dorsum of head with setae short , sparse .
Sitophore sclerite unmodified.
Tarsal claws lacking ventral spines or teeth. Postcoxale of leg III absent or, if present, frequently
enlarged, heavily sclerotised, displaced posteriad to occupy abdominal sternum II, and fused to sclerite of
abdominal pleurum II, in which case gap between postcoxales sometimes obscured by sternite II*.
Thoracic setae present dorsally only along posterior and lateroposterior margins of prothorax and
pterothorax; setae short, except for laterally on pterothorax, where of moderate length.
Abdomen oval-elongate. Abdominal spiracles present on segments III to VIII, though sometimes very
small and possibly non-functional on VIII*. Abdominal setae short or of moderate length; anterior setae
present on pleura only, sparse; postero-lateral seta present. Pleural projections present ventrally and
dorsally on abdominal pleurum IV, sclerotised. Sclerites, frequently faint, present on sterna, terga and
pleura of all abdominal segments except I; male terga, especially tergum VI, frequently with anterior and
posterior sclerites.
Gonapophyses with setae non-tuberculate and, frequently, tuberculate; setal tubercles, if present,
TRICHODECTID MAMMAL LICE
275
1 10
112
1 13
Figs 110-113 Procaviphilus species. 110, P. (Meganarionoides) n. baculatus, $ terminalia, ventral. Ill,
P. (M.) n. baculatus, $ gonapophysis, ventral. 112, P. (P.)f. granuloides, $ head, dorsal. 113, P. (M.)
scutifer, $ terminalia, ventral.
sometimes fused characteristically*; lobe absent or, if apparently present, formed of fused tubercles and
thick, with submarginal setae (Fig. 111). Gonapophyses meet ventral vulval margin acutely, not linked by
sclerotised band. Ventral vulval margin not sclerotised; expanded, frequently with posterior broadening
(Fig. 110), sometimes lengthened (Fig. 113) or with median lobulate process (Fig. 110)*.
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r., perisetal gaps very
small or absent, rarely large (Figs 114, 115). Pseudostyli absent. Male genital opening posterodorsal. Male
genitalia very variable: basal apodeme short or long with median constriction*; parameres unfused,
frequently with basal flange, or fused, faintly or clearly, and parameral plate apically bifurcate. Basipar-
ameral sclerites absent. Mesomeres fused, basally extending between b.a.l.s. to contact parameres, which
276
C. H. C. LYAL
TRICHODECTID MAMMAL LICE
277
116
120
Figs 116-122 Procaviphilus species, cf genitalia. 116, P. (Meganarionoides) angolensis. 117, P. (P.) f.
granuloides. 118, P. (P.) dubius. 119, P. (M.) n. neumanni. 120, P. (M.) jordani. 121, P. (M.)jordani,
detail of junction of mesomere, paramere and basal apodeme. 122, P. (M.) serraticus.
278 C. H. C. LYAL
may also contact b.a.l.s. (Figs 116, 117, 119-121); mesomeral arch with median extension more or less
broad, lateral desclerotisations sometimes apparent (Figs 118, 122)*.
HOSTS. Procaviidae (Hyracoidea) and Cercopithecidae (Primates).
Subgenus PROCAVIPHILUS Bedford
(Figs 112, 117, 118)
Procaviphilus Bedford, 1932: 725. Type-species: Procaviphilus ferrisi Bedford, by original designation.
DESCRIPTION. Anterior of head of 'procaviphilus' type.
Postcoxale absent or, if present, not greatly developed and not fused to abdominal pleurite II.
Abdominal spiracles all the same size.
Gonapophyses with setal tubercles present but not fused; gonapophyses not apically truncate. Ventral
vulval margin expanded, sometimes W-shaped medially, not broadened posteriorly.
Male genitalia with basal apodeme short, attaining abdominal segment VII or VI, lacking median
constriction; mesomeral arch with lateral desclerotisations.
HOSTS. Procaviidae (Hyracoidea).
SPECIES INCLUDED
duWus Werneck, 1941 (10 cf , 10 $)
ferrisi ferrisi Bedford, 1932 (12 cf , 5 $)
ferrisi granuloides Bedford, 1939 (26 C? , 19 9)
ferrisi Wndei Werneck, 1946 (3 cf , 4 $)
granulates (Ferris, 1930) (13 cf , 17 $)
Aarrisi Werneck, 1946 (15 cf , 19 $)
robertsi (Bedford, 1928) (16 cf , 18 $)
Subgenus MEGANARIONOIDESEichler
(Figs 110, 111, 113-116, 119-122)
Meganarionoides Eichler, 1940: 159. Type-species: Trichodectes colobi Kellogg, by original designation.
Acondylocephalus Werneck, 1941: 478 [as subgenus of Procavicola Bedford]. Type-species: Trichodectes
congoensis Ferris, by original designation. [Synonymy by Werneck, 1946: 85.]
DESCRIPTION. Anterior of head of 'procaviphilus' type or, more frequently, of 'procavicola' type.
Postcoxale of leg III enlarged as described in description of Procaviphilus s.l., and fused to abdominal
pleurite II, at least in female.
Gonapophyses with setal tubercles absent (in which case gonapophyses characteristically broad and
vulval margin produced posteriad as in Fig. 113), or present basally, marginally and submarginally, and
fused characteristically to form basal process (Fig. Ill); gonapophyses more or less truncate. Ventral
vulval margin expanded, as described for Procaviphilus s. str. or, more frequently, broadened posteriorly,
sometimes produced posteriad (Fig. 113) or with median lobulate process (Fig. 110).
Male genitalia with basal apodeme attaining abdominal segment VII or VI or, more frequently, long,
attaining segment III or II, with median constriction (Figs 116, 119); mesomeral arch with or without
lateral desclerotisation.
HOSTS. Procaviidae (Hyracoidea) and Cercopithecidae (Primates).
COMMENTS. There has been some disagreement in the literature over the correct host of one species in this
subgenus. Most species included in P. (Meganarionoides) are parasites of Procaviidae, as are all other
species in the subfamily Dasyonyginae (other than the three species of Cebidicola, which are included in
the subfamily for the first time in this study). One species, however, P. (M.) colobi (Kellogg, 1910), was
described from the monkey Colobus guereza caudatus Thomas. Keler (1938a) included this species with
the others described from Primates in his genus Meganarion (an objective synonym of Cebidicola},
although realising that the species were not truly congeneric. Eichler (1940) described the new genus
Meganarionoides for colobi, and placed it with Cebidicola and Lorisicola in the new subfamily Cebidicoli-
nae. Werneck (1946) recognised the identity of colobi with the hyrax lice, and synonymised Meganar-
ionoides with Procavicola (Acondylocephalus) Werneck, 1941, the subgenus thus taking the name
Procavicola (Meganarionoides). Werneck (1946) also suggested that Colobus was not the true host of P.
colobi, but that the louse was probably a parasite of Dendrohydrax validus subsp. He suggested that the
host record of the type-specimens was erroneous and due to mislabelling (the collection having included
TRICHODECTID MAMMAL LICE 279
both Colobus and Dendrohyrax), and that a second record was due to contamination (other hyrax-lice
having been associated with the specimens of colobi). Hopkins (1949) reported having examined 25 skins
of Colobus polykomos, which he identified as the 'supposed host', without having found any Trichodecti-
dae, and agreed with Werneck (1946) that Dendrohyrax validus subsp. was the correct host. Hopkins &
Clay (1952) also identified D. validus subsp. as the host, the record from 'Colobus caudatus' being termed
an 'error'. Eichler (1963) agreed, and removed Meganarionoides from the Cebidicolinae and placed it in
the Dasyonygidae with Procavicola (Fig. 38). Emerson & Price (1981) include P. colobi, without
comment, as a parasite of Dendrohyrax validus validus, although the association with the nominate
subspecies of this animal has not appeared elsewhere in the literature. Kuhn & Ludwig (1964), however,
reported a specimen of Colobus guereza with 'hundreds of eggs and adult and larval Procavicola on it, all
clasping the hairs tightly; most of them on the back and on the throat', and were able to state that the
monkey had not been in contact with a Dendrohyrax or any other Procaviidae after its death. They
concluded There is no doubt . . . that Colobus guereza is a natural host of Procavicola (Meganarionoides)
colobi.' In view of the fact that there are now three records of the species from Colobus guereza and none
from any member of the Procaviidae, this conclusion seems fully justified.
Meganarionoides was, as described above, treated as a subgenus of Procavicola by Werneck (1946). In
this he has been followed by most authors, although Eichler (1963) considered it to be a full genus. Before
the present study Meganarionoides had not been placed as a subgenus of Procaviphilus.
SPECIES INCLUDED
a/rica/ius( Werneck, 1941) comb. n. from Procavicola (5 cf , 5 $)
angolensis (Bedford, 1936) comb. n. from Procavicola (8 cf , 11 9)
colobi (Kellogg, 1910) comb. n. from Procavicola (1 d", 1 $)
congoensis (Ferris, 1930) comb. n. from Procavicola (22 cf , 25 $)
jordani (Bedford, 1936) comb. n. from Procavicola (2 cf , 2 $)
muesebecki (Emerson & Price, 1969) comb. n. from Procavicola (10 cf , 10 $)
/it'i/niann/neuma/in/(Stobbe, 1913) comb. n. from Procavicola (2 cf , 1 $)
neumanni baculatus (Ferris, 1930) comb. n. from Procavicola (13 cf , 14 $)
sclerotis sclerotis Bedford, 1932 [treated as Procaviphilus s. str. by previous authors] (10 cf , 17 $)
sclerotis major Maltbaek, 1937 [treated as Procaviphilus s. str. by previous authors]
scutifer (Werneck, 1941) comb. n. from Procavicola (14 cf , 19 $)
serraticus (Hill, 1922) [treated as Procaviphilus s. str. by previous authors] (50 cf , 70 $)
tendeiroi (Emerson, 1965) comb. n. from Procavicola (3 Cf , 2 $)
Genus DASYONYX Bedford
The genus Dasyonyx. comprises two subgenera.
DESCRIPTION. Anterior of head with osculum present, variable in degree of excavation* ; dorsal preantennal
sulcus absent; clypeal marginal carina broadened medially, with posterior curvature of broadened portion
similar to curvature of osculum* (Fig. 123); conus not large; anterolateral margin of head straight, convex
or concave, though not very concave at junction with clypeofrontal suture; preantennal portion of head
short or long, outline broadly triangular, trapezoidal or rounded*. Temple margin shallowly convex,
sometimes with small rounded projection postero-laterally*. Male scape expanded, with setal row present;
male flagellomeres fused, with two basally-articulated 'teeth'; female flagellomeres unfused. Dorsum of
head with setae of moderate length. Sitophore sclerite with posterior arms extended (cf. Fig. 12), though
sclerite difficult to see.
Tarsal claws with ventral spines or teeth (Figs 14, 15)*. Post-coxale of metathoracic leg absent or, if
present* enlarged, though not as described for Procaviphilus (Meganarionoides) and not fused to
abdominal pleurite II. Thorax with dorsal setae present only posteriorly on prothoracic margin and
posteriorly and posterolaterally on pterothoracic margin; setae short anteriorly, longer posteriorly with the
longest setae on the posterolateral margins of the pterothorax; setae generally sparse.
Abdomen broadly oval, with male segment IX not projecting greatly (Fig. 129). Abdominal spiracles
present on segments III to VIII, all approximately the same size, frequently inconspicuous. Abdominal
setae of moderate length; anterior setae present on pleura only; postero-lateral setae present. Pleural
projections present dorsally and ventrally on abdominal pleurum IV, sclerotised. Sclerites present on
sterna, terga and pleura (where present) of all abdominal segments except I; male terga, at least of
abdominal segment VI, with anterior and posterior sclerites.
Gonapophyses with sparse marginal setae and variably-developed lobe ventrally, the lobe bearing two
apical or subapical setae and frequently being serrate along dorsal (posterior) margin (Fig. 126).
280
C. H. C. LYAL
TRICHODECTID MAMMAL LICE
281
282
C. H. C. LYAL
Fig. 129 Dasyonyx (D.) validus, C? abdomen.
Gonapophyses meet ventral vulval margin acutely, not linked by sclerotised band. Ventral vulval margin
not sclerotised; greatly expanded, sometimes with postero-lateral angular projections (Fig. 126); subgenit-
al lobe absent.
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r. , but variably modified
(Figs 124, 125, 127-129), frequently lacking perisetal gap. Pseudostyli absent. Male genital opening
postero-dorsal or dorsal. Parameres fused or, if unfused, then with basal flanges (Figs 130, 132, 134, 135).
Basiparameral sclerites absent. Mesomeres fused; mesomeral arch with median extension and lateral
desclerotisations; mesomeres more or less produced basally between b.a.l.s. to meet parameres, which
sometimes do not contact b.a.l.s. (Figs 130, 132-135).
HOSTS. Procaviidae (Hyracoidea).
Subgenus DASYONYX Bedford
(Figs 14, 123-127, 129-132, 134, 135)
Dasyonyx Bedford, 19326: 720. Type-species: Dasyonyx validus Bedford, by original designation.
TRICHODECTID MAMMAL LICE
283
Figs 130-135 Dasyonyx species, C? genitalia. 130, D. (D.) validus. 131, D. (D.) validus, endophallus,
sclerites shown by dashed outlines. 132, D. (D.) ovalis, 133, D. (Neodasyonyx) ruficeps. 134, D. (D.)
guineensis. 135, D. (D.) minor.
DESCRIPTION. Osculum deeply concave; preantennal outline of male head subtriangular or subtrapezoidal.
Temple margin frequently with small rounded projection postero-laterally (Fig. 123).
Tarsal claws with ventral spines slender and sharp (Fig. 14). Postcoxale of metathoracic leg generally
present and enlarged, though not as described for Procaviphilus (Meganarionoides).
HOSTS. Procaviidae (Hyracoidea).
284 C. H. C. LYAL
SPECIES INCLUDED
bedfordi Werneck, 1945 (11 cT, 16 $)
dendrohyracis (Ferris, 1930) (12 cf , 15 $)
guineensis Werneck, 1941 (5 cf , 3 $)
hopkinsi Werneck, 1941 (33 cf , 32 $)
minor Bedford, 1939 (3 cf , 1 $)
oculatus Bedford, 1928
ovalis Bedford, 1932 (36 cf , 35 $)
smallwoodae Emerson & Price, 1969 (7 cf , 3 $)
validus validus Bedford, 1932 (22 cf , 22 $)
validus ugandensis Werneck, 1941 (29 cf , 26 <j>)
Subgenus NEODASYONYX Werneck
(Figs 15, 128, 133)
Neodosyonyx Werneck, 1941: 543 [as subgenus of Dosyonyx Bedford]. Type-species: Dosyonyx trans-
vaalensis Bedford, by original designation.
DESCRIPTION. Osculum shallowly concave; preantennal region of male head short, outline rounded.
Temple margin lacking postero-lateral projection.
Tarsal claws with broad ventral teeth (Fig. 15). Postcoxale of metathoracic leg absent.
HOSTS. Procaviidae (Hyracoidea).
SPECIES INCLUDED
capensis Emerson, 1965 (holotype cf , allotype $)
d/acantfii/s(Ehrenberg, 1828) (9 cf, 8 $)
nairobiensis Bedford, 1936 (89 cf , 92 $)
ruficeps Emerson, 1964 (15 cf , 13 $)
transvaalensis Bedford, 1932 (24 cf , 31 $)
waterburgensis Bedford, 1932 (5 cf , 6 $)
Genus EURYTRICHODECTES Stobbe
(Figs 9, 16, 136-138)
Eurytrichodectes Stobbe, 1913o: 111. Type-species: Eurytrichodectes paradoxus Stobbe, by monotypy.
DESCRIPTION. Anterior of head with osculum absent or, if present, very shallowly concave; dorsal
preantennal sulcus absent; clypeal marginal carina slender, not greatly developed medially or, if de-
veloped, in the form of a median posteriorly-directed narrow-based triangle; anterolateral margin of head
slightly sinuate; preantennal portion of head very short, outline broadly triangular or trapezoid. Temple
margin produced posteriorly into broad triangular spike, almost as long as prothorax or, if shorter,
attaining front of pronotum (Fig. 138). Male scape expanded, with longitudinal setal row comprising only
two setae; male flagellomeres fused, though semicircular sclerite at apex may be vestige of terminal'
flagellomere; two basally-fused 'teeth' present on male flagellum; female flagellomeres unfused; membra-
nous projection present on female pedicel (Fig. 138). Dorsum of head with setae short, sparse. Sitophore
sclerite with posterior arms extended (cf. Fig. 12), though sclerite difficult to see.
Tarsal claws ridged ventrally, lacking teeth or spines (Fig. 16). Postcoxale of metathoracic leg absent.
Thorax with dorsal setae present on posterior margin of prothorax and on posterior margin of pterothorax;
setae short, sparse.
Abdomen broad and oval, sometimes with male terminal segments tapering and projecting slightly.
Abdominal spiracles present on segments III to VIII, all approximately of the same size. Abdominal setae
short on sterna IV to IX, terga and pleura, some tergal setae very short; sterna II and III with stout, conical
setae (Fig. 136); anterior setae present, sparse on pleura; setal row on terga and sterna may be irregular;
postero-lateral setae present. Pleural projections present dorsally and ventrally on abdominal pleurum IV,
sclerotised, very long, reaching or almost reaching posterior margin of pleurum V. Sclerites present on
sterna, terga and pleura (where present) of all abdominal segments except, sometimes, tergum I; male
terga, at least of abdominal segments II to VI, with anterior and posterior sclerites.
Gonapophyses with ventral marginal setae present, each with a small, conical tubercle; ventral lobe
absent. Gonapophyses meet ventral vulval margin smoothly, not linked by sclerotised band. Ventral vulval
margin not sclerotised; medially expanded and trapezoid; subgenital lobe absent.
TRICHODECTID MAMMAL LICE
285
ig. 136 Eurytrichodectes paradoxus , cf abdomen.
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r. , with setal gaps very
small or absent (Fig. 136). Pseudostyli absent. Male genital opening dorsal. Parameres not fused,
sometimes apically bifurcate. Basiparameral sclerites absent. Mesomeres fused; mesomeral arch with
median extension (see comments below); mesomeral arch lacking lateral desclerotisations; mesomeres
sometimes produced basally between b.a.l.s. to contact parameres. Male genitalia depicted in Fig. 137.
HOSTS. Procaviidae (Hyracoidea).
COMMENTS. The illustration of the the mesomeral arch of E. paradoxus is reproduced upside-down in
Werneck (1941: 452).
SPECIES INCLUDED
machadoi Werneck, 1958 (holotype cf , allotype $)
paradoxus Stobbe, 1913 (29 cf , 28 $)
286
C. H. C. LYAL
138
137
Figs 137, 138 Eurytrichodectesparadoxus. 137, cf genitalia. 138, $ head, dorsal.
TRICHODECTINAE Kellogg
Genus PROTELICOLA Bedford gen. rev.
(Figs 139, 140)
Protelicola Bedford, 1932a: 355. Type-species: Protelicola intermedius Bedford, by monotypy.
DESCRIPTION. Anterior of head with osculum present; clypeal marginal carina broadened just laterally to
osculum, tapering medially and interrupted by dorsal preantennal sulcus; anterolateral margin of head
straight or convex; preantennal outline of head triangular or rounded. Temple margin convex. Male scape
slightly expanded, with longitudinal setal row present and comprising three or four setae; flagellomeres
fused in males and females; male flagellum with two basally-articulated 'teeth'. Dorsal setae of head short
or of moderate length, longest and most abundant anteriorly. Sitophore sclerite unmodified.
Thorax with setae long or of moderate length laterally and dorsally, with setae on postero-lateral angles
of pterothorax shorter, more spine-like. Prothorax with setae present sparsely on lateral and posterior
margins; median gap present, wide; single seta present antero-laterally on disc. Pterothorax with setae
present postero-laterally and submarginally posteriorly; median gap present, wide; no setae present on
disc.
Abdomen rounded, similar in shape in males and females. Abdominal spiracles present on segments
III- VIII. Abdominal setae of moderate length; anterior setae present on pleura only; postero-lateral setae
absent. Abdominal pleura lacking projections. Abdominal segments with tergal sclerites absent except
tergite IX in female; pleural sclerites present on pleura II and III; sternites IV- VII present in male and
V-VII present in female, very slender and difficult to see in both sexes.
Gonapophyses with basal setae and rectangular lobe on ventral margin, lobe formed from more or less
fused setal tubercles. Gonapophyses meet ventral vulval margin smoothly, not linked by sclerotised band.
Ventral vulval margin not sclerotised. Subgenital lobe present, not marginally serrate, but sometimes with
apical papillae. Female genital chamber with small lapped scales on walls, dorsal wall with median
longitudinal anterior 'slit' where scales are absent.
Male subgenital plate comprising very slender sternite VII and s.g.p.r. only, s.g.p.r. not reaching
posterior of sternum IX. Male genital opening dorsal or postero-dorsal. Pseudostyli absent. Basal
apodeme with b.a.l.s. widely divergent anteriorly. Parameres long, slender, fused basally, projecting
anteriorly between b.a.l.s. (but see second paragraph of 'comments' below). Basiparameral sclerites
TRICHODECTID MAMMAL LICE
287
139
Figs 139, 140 Protelicola species, C? genitalia. 139, P. hyaenae. 140, undescribed species.
absent. Mesomeres fused apically to form arch, with median projection present, broad (see second and
third paragraphs of 'comments' below). Male genitalia depicted in Fig. 139.
Alimentary canal with numerous small spines in crop.
HOSTS. Hyaenidae and Protelidae (Carnivora).
COMMENTS. Protelicola was treated as a subgenus and a junior synonym of Felicola by Hopkins (1949) and
Werneck (1948) respectively; its most recent placement (Emerson & Price, 1981) was as a synonym of
Felicola. A more detailed history of the variations in status of Protelicola is presented in Table 5.
In the British Museum (Natural History) collection there is a slide bearing 1 nymphal, 3 female and 2
male (one of which is teneral) lice from Proteles cristatus termes. Hopkins has identified the lice as
Protelicola intermedius. The females are indistinguishable from females of P. intermedius s. str., but the
male genitalia are very distinct, with the parameres completely fused to form a broad plate, the mesomeral
arch wide, and the b.a.l.s. with a small postero-lateral projection contacting the mesomeres (Fig. 140). In
all other respects the males resemble P. intermedius s. str. It seems that the males represent a new species,
differing from Protelicola intermedius by the structure of the genitalia, but the identity of the females is
doubtful. In view of the limited number of specimens available, the species is not formally described.
Bedford (19320) described P. intermedius from Proteles cristatus, the first louse known from a hyaena.
Hopkins (1960) described P. intermedius hyaenae (in Felicola) from Hyaena brunnea, distinguishing it
from the nominate subspecies by the size, the outline of the preantennal portion of the head, and the male
288 C. H. C. LYAL
genitalia. Ledger (1980) raised P. i. hyaenae to specific rank although Emerson & Price (1981) retained its
subspecific status. Hopkins (1960) indicated three features of the male genitalia in which the two taxa
differ: the greater anterior divergence of the b.a.l.s. in P. i. intermedius , the shape of the parameres (which
he presumed to be completely fused in P. i. intermedius, and are only basally fused in P. i. hyaenae), and
the absence of the mesomeral arch extension in P. i. intermedius. The type-series of P. i. intermedius has
not been seen in the present study, but a series of specimens from Proteles cristatus has been examined;
these specimens agree with Bedford's and Hopkins' descriptions of P. i. intermedius in all but details of the
male genitalia. The male genitalia of this series were found to be very similar to those of P. hyaenae, with
the b.a.l.s. variably divergent anteriorly, the parameres fused only basally and the mesomeral arch with a
median broad extension (although this is very thinly sclerotised and difficult to see in both taxa). The male
genitalia of P. hyaenae are depicted in Fig. 139. The two species may be distinguished by the smaller size
and shorter preantennal region of P. hyaenae (see photographs in Hopkins, 1960).
SPECIES INCLUDED
hyaenae (Hopkins, 1960) stat. rev., comb. n. from Felicola (holotype cf , allotype $)
intermedius Bedford, 1932 comb. rev. from Felicola (19 cf, 48 9)
Genus LUTRIDIA Keler
(Figs 141-145)
Lutridia Keler, 1938a: 433. Type-species: Trichodectes exilis Giebel, by original designation.
DESCRIPTION. Anterior of head with osculum absent; dorsal preantennal sulcus absent; clypeal marginal
carina broadened medially to form dorsal sclerite with three posteriorly-directed projections (Fig. 142);
preantennal portion of head short, outline smoothly rounded. Temple margin convex or rectangular. Male
scape not expanded, longitudinal setal row comprising two setae positioned distally on segment; flagello-
meres fused in males and females; male flagellum lacking 'teeth'. Dorsum of head with setae short or of
moderate length, sparse. Sitophore sclerite unmodified.
Thorax with dorsal setae long or of moderate length, limited to posterior and postero-lateral margins of
prothorax and pterothorax.
Abdomen oval or slightly elongate-oval, with male segment IX projecting slightly posteriorly (Fig. 141).
Abdominal spiracles present on segments III- VIII. Abdominal setae as follows: pleurum II with setae
sparse, stout and short, anterior setae and p.s.r. present; pleurum III with setae short and of moderate
length, stout, very sparse (exilis) or with p.s.r. present (matschiei); pleura IV- VII or VIII lacking setae,
those on VIII, if present, very small and posteriorly positioned; sternal setae short and stout or longer
(about two-thirds length of segment); sternum II with median gap of posterior setal row small or absent;
sterna III-IV and VII- VIII (matschiei) or III- VIII (exilis) with setae very sparse, setae absent from sterna
V-VI of L. matschiei; terga I-IV (males) or I-III (females) with median setal group including one or two
setae as long as segment, other terga with median group absent or comprising shorter setae; terga with
lateral seta or setae generally present, of moderate length on terga II-III, shorter on more posterior
segments (these setae may represent either lateral setal group or postero-lateral seta); anterior setae
present only on pleurum II. Abdominal pleura lacking projections. Sclerites absent from abdominal
pleura, present on terga IV- VIII (males) or terga IV-IX (females) and sterna III- VIII (males) or sterna
IV- VIII (females); sclerites frequently very faint, may not be seen; male terga lacking posterior sclerites.
Gonapophyses with non-tuberculate setae on ventral margin; ventral lobe not present. Gonapophyses
meet ventral vulval margin smoothly, linked by sclerotised band. Ventral vulval margin sclerotised, with
chord at 90 degrees to long axis of abdomen, submarginal non-tuberculate setae present. Subgenital lobe
present, small, rectangular, sometimes serrate along posterior margin (Fig. 143).
Male subgenital plate comprising s.g.p.r. only or with s.g.p.r. linked by broad sternite VIII (and possibly
VII); in either case s.g.p.r. not attaining segment IX. Pseudostyli absent. Male genital opening dorsal.
Basal apodeme slender, long, attaining at least abdominal segment III. Parameres long, slender, basally
fused; basal fused portion may be partially detached from rest of parameres; parameres sometimes fused to
b.a.l.s. Mesomeres absent. Male genitalia depicted in Figs 144, 145.
HOSTS. Lutrinae (Carnivora: Mustelidae).
COMMENTS. Lutridia has been treated as a synonym and a subgenus of Trichodectes (Hopkins, 1942 and
Hopkins, 1949 respectively), though the most recent treatment (Emerson & Price, 1981) considered
Lutridia as, a full genus. The history of the variations in status of Lutridia is presented in Table 4.
TRICHODECTID MAMMAL LICE
289
141
143
142
144
145
Figs 141-145 Lutridia species. 141, L. exilis, cf abdomen. 142, L. matschiei, $ head, dorsal. 143, L.
matschiei, $ terminalia. 144, L. matschiei, cf genitalia. 145, L. exilis, cf genitalia.
SPECIES INCLUDED
exi//s(Nitzsch, 1861) (3 cf , 10 $)
matec/iJd(Stobbe, 1913) (36 cf , 38 $)
Genus NEOLUTRIDIA gen. n.
(Figs 146, 147)
Type-species: Trichodectes lutrae Werneck.
DESCRIPTION. Anterior of head with osculum absent, though pulvinus attaining margin; dorsal preantennal
sulcus absent; clypeal marginal carina slightly broadened medially at junction with pulvinus; preantennal
outline broadly and smoothly rounded. Temple margin rectangular. Male scape not expanded; longitudin-
290 C. H. C. LYAL
al setal row present, comprising four setae; flagellomeres fused in males and females; male flagellum
lacking 'teeth'. Dorsum of head with setae short or of moderate length, sparse; temple margin with two or
three longer setae. Sitophore sclerite unmodified.
Prothorax with two setae of medium length on posterior margin; pterothorax with one or two short,
spine-like setae anterolaterally and six to ten long setae dorsally on posterior margin.
Abdomen oval, with male segment IX projecting posteriorly. Abdominal spiracles present on segments
III-VIII. Abdominal setae as follows: pleurum III with short, stout setae anteriorly and posteriorly; pleura
III-VIII lacking setae; terga I- VI (males) or I-IV (females) with central seta of median groups as long as
segment, setae otherwise short; terga VII-VIII (males) or V-IX (females) with short setae, sparse; sterna
with stout, short setae, sparse; anterior setae present only on pleurum II; postero-lateral setae absent.
Abdominal pleura lacking projections. Sclerites absent from abdominal pleura and sterna, but present,
slender, on at least terga III- VII (males) or V-IX (females), though may be very faint and not seen; male
terga lacking posterior sclerites.
Gonapophyses with small ventral lobes formed from fused setal tubercles. Gonapophyses meet ventral
vulval margin smoothly, linked by sclerotised band. Ventral vulval margin sclerotised, with chord at 90
degrees to long axis of abdomen; submarginal non-tuberculate setae present. Subgenital lobe present,
large, with lateral spine-like projections and associated setae present basally (Fig. 146).
Male subgenital plate represented by sternite VIII with lateral arms extending anteriad. Pseudostyli
absent. Male genital opening dorsal. Basal apodeme attaining abdominal segment III, not slender.
Parameres broad, scoop-shaped, not fused together, but fused to b.a.l.s. Mesomeres absent. Male
genitalia depicted in Fig. 147.
HOSTS. Lutrinae (Carnivora: Mustelidae).
SPECIES INCLUDED
lutrae (Werneck, 1937) comb. n. from Lutridia (1 cf , 1 $)
Genus WERNECKODECTES Conci gen. rev.
(Fig. 148)
Werneckodectes Conci, 1946: 59. Type-species: TrichodectesferrisfWerneck, by original designation.
DESCRIPTION. Osculum absent, though pulvinus attains anterior margin of head; dorsal preantennal sulcus
present; clypeal marginal carina broadened slightly medially; preantennal portion of head short, outline
146
147
148
Figs 146-148 Trichodectini species. 146, Neolutridia lutrae, $ terminalia, ventral. 147, N. lutrae, cf
genitalia (after Werneck). 148, Werneckodectes ferrisi, cf genitalia (after Werneck).
TRICHODECTID MAMMAL LICE 291
smoothly and broadly rounded. Temple margin convex. Male scape expanded; flagellomeres fused in
males and females; male flagellum with two basally-articulated 'teeth' and basal projection. Dorsum of
head with setae of moderate length.
Thorax with dorsal setae longest postero-laterally on pterothorax and along posterior margin on
pterothorax; shorter setae present submarginally along posterior of pterothorax and on postero-lateral
angles of prothorax, where there is a small setal patch; disc and posterior margin of pronotum each with
pair of small setae.
Abdomen oval, tapering posteriorly rather more in male than female. Abdominal spiracles present on
segments III- VIII. Abdominal setae numerous, anterior setae being present on sterna, terga and possibly
pleura of all segments; anterior setae smaller than setae of posterior setal rows on each segment;
postero-lateral setae, if present, obscured by large number of other setae. Abdominal pleurum IV with
ventral projection in male, possibly with dorsal projection in female; pleurum III possibly with dorsal
projection in female. Sclerites present on all abdominal pleura, on terga V-IX (males) or terga VII-IX
(females) and on sterna III-VI (males only - sclerites absent on female sterna); male terga without
posterior tergites, but anterior tergites, where present, with median longitudinal division.
Gonapophyses with non-tuberculate setae on ventral margin; ventral lobe present, small; setae on lobe
stout, short, whilst setae distal to lobe longer, more slender. Gonapophyses meet ventral vulval margin
smoothly, linked by broad sclerotised band. Ventral vulval margin sclerotised, with chord at 90 degrees to
long axis of abdomen; marginal non-tuberculate setae present, stout, short. Subgenital lobe present,
broad, with lateral rounded projections and associated setae present basally.
Male subgenital plate with sternite VII fused to s.g.p.r., sternites VIII and IX absent; s.g.p.r. with
sinuate margins. Pseudostyli absent. Male genital opening dorsal. Parameres not as long as basal apodeme,
not fused together. Basiparameral sclerites absent. Mesomeres not apically fused, abutting parameres and
b.a.l.s. basally. Male genitalia depicted in Fig. 148.
HOSTS. Ursidae (Carnivora).
COMMENTS. Werneckodectes has been treated as a synonym and a subgenus of Trichodectes (by Hopkins,
1942 and Hopkins, 1949 respectively) ; its most recent placement was as a synonym of Trichodectes. A more
comprehensive history of the variations in status of Werneckodectes is given in Table 4.
SPECIES INCLUDED
ferrisi(Werneck, 1944) comb. rev. from Trichodectes.
Genus TRICHODECTES Nitzsch
The genus Trichodectes comprises three subgenera.
DESCRIPTION. Anterior of head with osculum present or absent* , but always with pulvinus attaining margin;
dorsal preantennal sulcus present or absent; clypeal marginal carina broadened to variable extent medially
to form simple bar with posterior margin straight or concave, or carina broadened into dorsal sclerite which
is heavily-sclerotised laterally (dorsal to margin of clypeus and pulvinus) and lightly or very lightly-
sclerotised medially (posterior to the osculum) , more or less convex posteriorly or U-shaped with median
posterior process* ; antero-lateral margin of head straight, convex or sinuate* ; preantennal portion of head
long or short*, outline broadly rounded, broadly triangular, trapezoid or only slightly produced anteriad
between coni*. Temple margin convex, rectangular or produced laterally*. Male scape expanded or not
expanded* ; longitudinal setal row present, comprising at least four setae; flagellomeres fused in males and
females; male flagellum with one, two or four basally-articulated 'teeth' or 'teeth' absent*. Dorsum of head
with setae short, of moderate length or long, longest setae generally present along posterior temple
margin; setae sometimes sparse. Sitophore sclerite unmodified.
Thorax with prothoracic dorsal setae sparse, short or of moderate length* posteriorly and postero-
laterally, absent from disc; pterothorax with setae on postero-lateral angles short and spine-like or of
moderate length, dorsal setae otherwise present on posterior margin only, long or short, numerous, sparse
or absent*.
Abdomen oval, male segment IX sometimes slightly projecting posteriad, but usually positioned
dorsally on the abdomen (Figs 156, 158, 160, 164, 165). Abdominal spiracles present on segments HI-IV,
III-V, III- VII or III- VIII*; spiracle on segment VIII, if present, sometimes much smaller than those on
segments III- VII*. Abdomen with at least some tergal and sternal setae as long as segment, or setae very
short, sparse and absent from pleura V and VI*; terga with lateral and median groups of setae frequently
distinct; tergal setae numerous, or median group reduced to a single seta or absent*; male terga II and III
sometimes with median group comprising exceptionally long, stout setae (Fig. 161)*; anterior setae
292
C. H. C. LYAL
151
Figs 149-152 Trichodectes (T.) species, $ terminalia. 149, T. (T.) emersoni. 150, T. (T.) canis. 151, T.
(T.)galictidis. 152, T. (T.) p. pinguis, ventral.
TRICHODECTID MAMMAL LICE
293
I
I
CH-
o
s
Tf
IT)
294
C. H. C. LYAL
157
Figs 155-157 Trichodectes (Stachiella) species, $ terminalia. 155, T. (S.) octomaculatus . 156, T. (S.)
erminiae. 157, T. (S.) potus.
present only on pleurum II; postero-lateral setae presumed absent, or presence obscured by numerous
long setae or reduction (or absence) of lateral setal group. Abdominal pleura lacking projections, or
projections present dorsally on pleura II, III and IV* (Fig. 160). Abdominal sclerites variable, present or
absent*; male terga with or without anterior and posterior sclerites*.
Gonapophyses with separate tuberculate setae and single apical non-tuberculate seta on ventral margin
(Figs 149, 151, 153, 156, 159), or tubercles more closely associated (Fig. 155) or all setae non-tuberculate
(Figs 152, 154)*; ventral lobe absent. Gonapophyses meet ventral vulval margin smoothly, linked by
sclerotised band, or band absent. Ventral vulval margin sclerotised or, rarely, not sclerotised; with chord at
TRICHODECTID MAMMAL LICE
295
00
in
296
C. H. C. LYAL
TRICHODECTID MAMMAL LICE
297
B
^H
I
^2
K'
3
h
^
^ s
il
298
C. H. C. LYAL
167
166
170
Figs 166-171 Trichodectes species, C? genitalia. 166, T. (Paratrichodectes) ovalis. 167, T. (P.) zorillae.
168, T. (P.) ugandensis. 169, T. (T.) canis. 170, T. (T.) galictidis. 171, T. (T.) emersoni.
90 degrees to long axis of abdomen; marginal setae present, tuberculate or non-tuberculate. Subgenital
lobe present, usually with marginal serrations and lateral basal projections, the latter sometimes with
associated setae.
Male subgenital plate absent (Figs 158, 160-162, 164), represented only by s.g.p.r. (Fig. 163) or enlarged
sternite VIII (Fig. 159), with sternites VII, VIII and IX present and fused to s.g.p.r. (Fig. 165), or of the
latter form but with sternite VIII divided medially. Pseudostyli absent. Male genital opening dorsal or
postero-dorsal. Parameres fused or separate, fused to b.a.l.s. or free*. Basiparameral sclerites absent.
Mesomeres absent or present, fused or unfused; if mesomeres fused apically, median extension absent.
TRICHODECTID MAMMAL LICE
299
174
Figs 172-174 Trichodectes (Stachiella) species, c? genitalia. 172, T. (S.) octomaculatus. 173, T. (S.)
emeryi. 174, T. (S.) erminiae.
HOSTS. Canidae, Mustelidae, Procyonidae, Ursidae and Viverridae (Carnivora).
COMMENTS. The different concepts of the extent of the genus Trichodectes held by various workers are
summarised in Table 4.
Subgenus TRICHODECTES Nitzsch
(Figs 10, 149-152, 158, 159, 162, 163, 169-171)
Trichodectes Nitzsch, 1818: 294. Type-species: 'Trichodectes canis DeGeer (syn. T. latus N.)', by
subsequent designation [Johnston & Harrison, 1911: 326].
Ursodectes Keler, 1938a: 435. Type-species: Trichodectes pinguis Burmeister, by original designation.
[Synonymy by Hopkins, 1942: 444.]
Grisonia Keler, 1938a: 464. [No type-species designated] [Homonym of Grisonia Gray, 1843: 68].
Galictobius Keler, 1938fc: 228. [Replacement name for Grisonia Keler.] Type-species: Trichodectes
galictidis Werneck, by original designation. [Synonymy by Hopkins, 1942: 444.]
DESCRIPTION. Male scape expanded; male flagellum with two or four basally-articulated 'teeth'.
Abdominal spiracles present on segments III- VIII, spiracle on segment VIII not smaller than those on
anterior segments. Abdomen with at least some tergal and sternal setae as long as segment, or setae very
short, sparse and absent from pleura V and VI (kuntzi and emersoni); male terga II and III sometimes with
300 C. H. C. LYAL
median group comprising exceptionally long, stout setae (undescribed sister-species to T. galictidis), tergal
setae never with median group reduced to single seta or absent (except sometimes on tergum I).
Abdominal pleura lacking projections. Abdominal sternal sclerites absent, or present on sterna V-VIII
only; abdominal tergal sclerites present or, more usually, absent; if tergites present on male, then not with
anterior and posterior sclerites on each segment.
Female genital chamber with ventral wall frequently obscure, dorsal wall bearing sclerotised nodules,
sometimes fused together.
Male subgenital plate absent, represented by s.g.p.r. only, or by enlarged sternite VIII only. Parameres
fused to form plate or unfused; symmetric or asymmetric; not fused to b.a.l.s. Faintly-sclerotised
tongue-like sclerite of uncertain homology sometimes present dorsally between parameres if mesomeres
absent (Fig. 169). Mesomeres absent or, if present, fused or unfused. Male genitalia depicted in Figs
169-171.
HOSTS. Canidae, Mustelidae, Ursidae and Viverridae (Carnivora).
COMMENTS. Trichodectes Nitzsch, 1818 was placed on the Official List of Generic Names in Zoology, with
the type-species Trichodectes canis DeGeer, by Opinion 627 of the International Commission of Zoologic-
al Nomenclature (1962, Bulletin of Zoological Nomenclature 19: 91).
SPECIES INCLUDED
canis (DeGeer, 1778) (c.100 cf, c.100 $)
emersoni Hopkins, 1960 (15 cf , 12 $)
ga/icfMsWerneck, 1934 (15 cf , 16 $; also 2 cf, 1 $ of an undescribed sister-species)
kuntzi Emerson, 1964 (15 cf , 14 $)
me/is (Fabricius, 1805) (c.60 cf , c.60 $)
pinguispinguisBurmeister, 1838 (1 cf , 5 9)
pinguis euarctidos Hopkins, 1954 (20 cf , 20 $)
vosse/er/Stobbe, 1913 (2 cf , 7 $)
Subgenus PARATRICHODECTESsubgen. n.
(Figs 6, 154, 160, 161, 166-168)
Type-species: Trichodectes ovalis Bedford.
DESCRIPTION. Anterior of head with osculum present; clypeal marginal carina broadened medially into
dorsal sclerite which is heavily-sclerotised laterally and lightly-sclerotised medially, more or less convex
posteriorly or U-shaped with median posterior process; antero-lateral margin of head convex or sinuate;
preantennal portion of head not long, outline rounded or subtriangular. Temple margin convex or
rectangular. Male scape not greatly expanded; male flagellum with two basally-articulated 'teeth'.
Thorax with prothoracic dorsal setae sparse, of moderate length posteriorly and postero-laterally,
absent from disc; pterothorax with setae on postero-lateral angle long dorsally, short and spine-like
ventrally; dorsal posterior pterothoracic setae submarginal, long, comprising two pairs with wide median
gap.
Abdominal spiracles present on segments III- VII. Abdomen with at least some tergal and sternal setae
as long as segment; setae present on all pleura; tergal setae numerous, median group not reduced to a single
seta or absent except sometimes on tergum I or on posterior terga of males only , if male terga II and III with
median setal group comprising exceptionally long, stout setae; postero-lateral setae presumed absent,
though may be present as the most lateral seta of lateral group, which is frequently situated more
posteriorly than other setae. Abdominal sterna and terga with or without sclerites; male terga, if sclerites
present, with anterior sclerites only.
Female genital chamber with dorsal wall not bearing sclerotised nodules.
Male subgenital plate unsclerotised. Parameres separate or thinly fused to each other; symmetric or
asymmetric; not fused to b.a.l.s. Tongue-like sclerite not present. Mesomeres absent. Male genitalia
depicted in Figs 166-168.
HOSTS. Mustelinae (Carnivora: Mustelidae).
SPECIES INCLUDED
ovalis Bedford, 1928 comb. rev. from Stachiella (15 cf , 12 $)
ugandensis Bedford, 1936 comb. rev. from Stachiella (44 cf , 47 $)
zor/tfaeStobbe, 1913, comb. rev. from Stachiella (17 cf , 26 $)
TRICHODECTID MAMMAL LICE 301
Subgenus STACHIELLA Keler stat. n.
(Figs 153, 155-157, 164, 165, 172-174)
Stachiella Keler, 1938a: 428. Type-species: Trichodectes pusillus Nitzsch [= Pediculus mustelae Schrank],
by original designation.
Potusdia Conci, 1942: 141. Type-species: Trichodectes potus Werneck, by original designation. [Synony-
mised with Trichodectes by Werneck, 1948: 110; syn. n. of Stachiella.]
DESCRIPTION. Clypeal marginal carina broadened medially into dorsal sclerite which is heavily-sclerotised
laterally and lightly or very lightly-sclerotised medially, more or less convex posteriorly or U-shaped with
median posterior process; antero-lateral margin of head smoothly convex; preantennal portion of head
long or short, outline broadly rounded, sometimes only slightly produced anteriad between coni. Male
scape not, or only slightly, expanded; male flagellum with one or two basally-articulated 'teeth' or 'teeth'
absent.
Thorax with prothoracic dorsal setae sparse, of moderate length posteriorly and postero-laterally,
absent from disc; ptero thorax with setae on postero-lateral angles short and spine-like or of moderate
length, dorsal posterior setae submarginal, long, comprising one or more pairs with wide median gap, or
absent.
Abdominal spiracles present on segments III-IV, III-V, or III-VIII; spiracle on segment VIII, if
present, sometimes much smaller than those on segments III-VII. Abdomen with at least some tergal and
sternal setae as long as segment; setae present on all pleura; terga with median setal group of male reduced
to a single seta on most segments, lateral group small; female tergal setae of similar arrangement or with
median group absent (see discussion on p. 221 above); male terga II and III never with median group
comprising exceptionally long, stout setae; postero-lateral setae presumed absent, though perhaps present
as the most lateral seta of lateral setal group, which is frequently situated more posteriorly than other setae
of the group. Abdominal pleura lacking projections. Abdominal sterna variably sclerotised, with sternites,
if present, most commonly on posterior segments; abdominal terga with sclerites present on segments
III-VIII or III-IX, sometimes on I and II; male terga frequently with both anterior and posterior sclerites,
though posterior sclerites may be faintly-sclerotised or absent; abdominal pleurum II frequently sclero-
tised, otherwise pleura unsclerotised.
Gonapophyses with separate or closely-associated tuberculate setae and single apical non-tuberculate
seta on ventral margin. Gonapophyses meet ventral vulval margin smoothly, linked by sclerotised band.
Ventral vulval margin sclerotised. Female genital chamber with dorsal wall not bearing sclerotised
nodules.
Male subgenital plate absent, or with sternites VII, VIII and IX present and fused to s.g.p.r., or of this
form but with sternite VIII divided medially. Parameres separate or thinly fused to one another;
symmetric, asymmetric or asymmetrically deflected; fused or not fused to b.a.l.s. Tongue-like sclerite not
present. Mesomeres absent, present, small and unfused, or present and fused. Male genitalia depicted in
Figs 172-174.
HOSTS. Mustelidae: Mustelinae and Procyonidae (Carnivora).
COMMENTS. Stachiella has been considered a synonym and a subgenus of Trichodectes (Hopkins, 1942 and
Hopkins, 1949 respectively), although its most recent treatment (Emerson & Price, 1981) was as a valid
genus. Potusdia has also been considered as a valid genus, subgenus of Trichodectes and synonym of
Trichodectes but has not, before this study, been considered a synonym of Stachiella. A more comprehen-
sive history of the variations in status of Stachiella and Potusdia is given in Table 4.
SPECIES INCLUDED
divaricat us Harrison, 1915 comb. rev. from Stachiella
emeryi Emerson & Price, 1974 [treated by Emerson & Price (1974, 1981) as Trichodectes s. str.] (4 cT, 17
?)
erminiae (Hopkins, 1941) comb. n. from Stachiella (92 d" , 100 $)
fallax Werneck, 1948 [treated by Werneck (1948) and all subsequent authors as Trichodectes s. str.]
(2cf,2$)
jacobi (Eichler, 1941) comb. n. from Stachiella
king! McGregor, 1917 comb. rev. from Stachiella (14 cf , 25 <j>)
larseni (Emerson, 1962) comb. n. from Stachiella (58 cf , 57 $)
mustelae (Schrank, 1903) comb. n. from Stachiella (18 cf , 23 $)
octomaculat us Paine, 1912 [treated by all authors as Trichodectes s. str.] (48 cf , 58 $)
302 C. H. C. LYAL
potus Werneck, 1934 [treated by Werneck (1948) and Emerson & Price (1981) as Trichodectes s. str.]
(19cf,25$)
refiisusrefususBurmeister, 1838 comb. rev. from Stachiella (1 cf , 1 $)
retusus martis ( Werneck, 1948) comb. n. from Stachiella (1 cf)
retusussalfii (Conci, 1940) comb. n. from Stachiella
Genus FELICOLA Ewing
The genus Felicola comprises two subgenera.
DESCRIPTION. Anterior of head with osculum present or absent; dorsal preantennal sulcus present or
absent; clypeal marginal carina, if osculum absent, with very slight median broadening, or, if osculum
present," carina broadened medially into dorsal sclerite which is heavily-sclerotised laterally (dorsal to
margin of clypeus and pulvinus) and very lightly-sclerotised medially (posterior to osculum); anterolateral
margin of head straight, slightly sinuate or convex; preantennal portion of head long or short, outline
triangular or broadly rounded*. Temple margin rectangular or convex. Male scape variably expanded or
not expanded* , with longitudinal setal row present and comprising at least three setae; flagellomeres fused
in males and females; male flagellum with one or three basally-articulated 'teeth', or variable number of
'teeth' fused to flagellum, or 'teeth' absent*. Dorsum of head with setae short, of moderate length or long,
but in any case frequently longer than abdominal tergal setae; setae sparse, though most numerous
anteriorly and along lateral margins. Sitophore sclerite unmodified.
Thorax with dorsal setae long or of moderate length though frequently short and spine-like on
postero-lateral angles of pterothorax, not present on disc of prothorax or pterothorax.
Abdomen oval or elongate, frequently terminating in more or less acute projection of segment IX in the
male (Figs 179, 181, 183-189). Abdominal spiracles absent, or present on segment III, segments III-IV, or
segments III-V*. Abdominal setae short, of moderate length, or occasionally long and fine* (Figs 182, 186,
188); male tergum II frequently with 2-6 very long setae medially (Figs 179, 180, 187, 189); abdominal
pleurum III frequently with posterior setal row comprising stout, conical setae (Figs 181, 182, 188);
anterior setae, if present, only on pleurum II; postero-lateral setae present or absent. Abdominal pleurum
III with projections absent, or, if present, dorsal or ventral, sclerotised or unsclerotised; abdominal
pleurum IV with projections absent or, if present, dorsal and occasionally ventral, sclerotised or
unsclerotised. Abdominal sclerotisation variable; sternal, tergal and pleural sclerites, if present, generally
on anterior segments, becoming less clear on posterior segments; male terga sometimes with anterior and
posterior sclerites, at least on tergum VI*.
Gonapophyses with non-tuberculate setae and rounded or rectangular lobe present on ventral margin;
apical spur present or absent* . Gonapophyses meet ventral vulval margin acutely, not linked by sclerotised
band. Ventral vulval margin not sclerotised; straight or concave, with chord less than 90 degrees to long
axis of abdomen; subgenital lobe present, apically single or bifurcate, with margins serrate, at least
posteriorly (Fig. 175) (see comment below).
Male subgenital plate not present, though sternite VIII sometimes with postero-lateral projections
probably homologous with lateral rods of subgenital plate (Fig. 181). Pseudostyli absent. Male genital
opening postero-dorsal or dorsal; segment IX frequently produced posteriad. Parameres generally long
and slender (Figs 191, 192, 197, 201), occasionally broader (Fig. 198); frequently fused completely or,
more usually, basally only. Basiparameral sclerites absent. Mesomeres present or absent; if present,
unfused (Fig. 200) or, if fused, median extension absent*. Male genitalia depicted in Figs 190-206.
HOSTS. Herpestidae, Viverridae, Felidae and Canidae (Carnivora).
COMMENTS. Emerson & Price (1980) distinguish the females of their new species Suricatoecus occidentals
(transferred to Felicola in this study) from other species in the 'helogale Group' (equivalent to the
congoensis-occidentalis clade) by the presumed absence of the subgenital lobe in occidentalis . Examina-
tion of the type-series of this species, however, reveals that the subgenital lobe, although very fine, is
present in all the females. A second species of this clade, close to F. helogaloidis, has been taken from skins
of Crossarchus obscurus (the host of F. occidentalis) and specimens are in the collection of the British
Museum (Natural History).
Subgenus FELICOLA Ewing
(Figs 5, 175, 179-185, 190-193, 196-198, 202-204)
FelicolaEwing, l.vi.1929: 121, 122, 192. Type-species: Trichodectes subrostratus Burmeister [attributed to
Nitzsch], by original designation.
TRICHODECTID MAMMAL LICE
303
304
C. H. C. LYAL
S
>*;
TRICHODECTID MAMMAL LICE
305
CO
a
306
C. H. C. LYAL
183
Figs 183-185 Felicola (F.) species. 183, F. (F.) calogaleus, cf abdomen. 184, F. (F.) helogale, C?
terminalia. 185, F. (F.) hopkinsi, cf terminalia.
Felicinia Bedford, -.x.1929: 519. Type-species: Trichodectes subrostratus Burmeister [attributed to
Nitzsch], by original designation. [Synonymised by Bedford, 1932a: 536.]
Bedfordia Keler, 1938a: 463. Type-species: Felicola helogale Bedford, by original designation. [Homonym
oi Bedfordia Fahrenholz, 1936: 55.] [Synonymised with Felicola by Hopkins, 1941: 36.]
Fastigatosculum Keler, 1939: 11. [Replacement name for Bedfordia Keler.] [Synonymised with Surica-
toecus by Werneck, 1948.]
DESCRIPTION. Preantennal portion of head with outline narrowly or broadly triangular or rounded. Male
flagellum with 'teeth' absent or, if present, numbering one, two, three or four and fused to flagellum, not
basally articulated.
Abdominal spiracles absent, or present on segments III-IV or III-V. Abdominal setae very short or of
moderate length.
Gonapophyses with rounded lobe on ventral margin; spur present or absent.
TRICHODECTID MAMMAL LICE
307
CTJ
b
308
C. H. C. LYAL
O)
CO
o
oo
S
V.
8
TRICHODECTID MAMMAL LICE
309
193
Figs 190-194 Felicola species, C? genitalia. 190, F. (F.) calogaleus. 191, F. (F.) inaequalis, endophallus
partially everted. 192, F. (F.) inaequalis, endophallus not everted. 193, F. (F.) setosus; 194, F.
(Suricatoecus) fahrenholzi.
Everted portion of endophallus frequently thinly sclerotised (Figs 190, 196, 204). Mesomeres present
and fused or absent.
HOSTS. Felidae, Herpestidae and Viverridae (Carnivora).
COMMENTS. F. genettae (Fresca) is included on the basis of the figures and description of Fresca (1924)
which, although poor, suggest an affinity with the calogaleus-viverriculae clade. If this is a correct
placement, the host record of Genetta genetta rhodanica is anomalous. The various treatments of the junior
310
C. H. C. LYAL
195
196
199
Figs 195-199 Felicola species, O" genitalia. 195, F. (Suricatoecus) bedfordi. 196, F. (F.) robertsi. 197, F.
(F.) hopkinsi. 198, F. (F.) congoensis. 199, F. (F.) helogale.
synonyms of Felicola are summarised in Table 5; omitted from the table is Conci (1946) who, like Eichler
(1941, 1963) regarded Fastigatosculum as a valid genus.
SPECIES INCLUDED
calogaleus (Bedford, 1928) (14 cf , 24 $)
congoensis (Emerson & Price, 1967) comb. n. from Suricatoecus (23 cf , 31 $)
cynictis (Bedford, 1938) (13 C?, 19 9)
genettae (Fresca, 1924)
TRICHODECTID MAMMAL LICE
311
200
205
Figs 200-206 Felicola species, cf genitalia. 200, F. (Suricatoecus) acutirostris. 201, F. (S.) vulpis. 202, F.
(F.) liberiae. 203, F. (F.) minimus, parameres displaced slightly apart. 204, F. (F.) subrostratus. 205, F.
(S.) decipiens. 206, F. (S.) decipiens, detail of paramere.
312 C. H. C. LYAL
helogale Bedford, 1932, comb. rev. from Suricatoecus (3 cf , 7 $)
Ae/oga7oid/s( Werneck, 1948) comb. n. from Suricatoecus
hopkinsi Bedford, 1936 comb. rev. from Suricatoecus (2 cf , 1 $)
inaegua/is(Piaget, 1880) (10 cf , 19 $)
/ifreriae Emerson & Price, 1972 (7 cf , 3 $)
minimus Werneck, 1948 (15 cf , 28 $)
occidental!* (Emerson & Price, 1981) comb. n. from Suricatoecus (4 cf , 5 9)
ra/i/ni Emerson & Stojanovitch, 1966 (9 cf , 16 $)
robertsi Hopkins, 1944 (8 cf , 15 $)
rohani Werneck, 1956 (68 cf , 62 $)
setosus Bedford, 1932 (14 Cf, 18 ?)
suferosfrafus(Burmeister, 1838) (103 cf, 120 $)
viverriculae (Stobbe, 1913) comb. n. from Parafelicola (21 cf, 26 9; 6 cf, 6 $ of undescribed
sister-species)
zeylonicus Bedford, 1936 (6 cf , 7 $)
Subgenus SURICATOECUS Bedford stat. n.
(Figs 176-178, 186-189, 194, 195, 200, 201, 205, 206)
Suricatoecus Bedford, 1932a: 354. Type-species: Trichodectes cooleyi Bedford, by monotypy.
EichlerellaCond, 1942: 140. Type-species: Trichodectes vulpis Denny, by original designation. [Synonymy
by Werneck, 1948: 172.]
DESCRIPTION. Preantennal portion of head with outline narrowly triangular or broadly rounded. Male
scape not, or very slightly, expanded; male flagellum with 'teeth' absent or, if present, numbering one or
three and basally articulated.
Abdominal spiracles absent, or present on segment III or segments III-V. Abdominal setae very short,
of moderate length, or long and fine. Tergal and sternal sclerites generally present on abdomen, though
less clear on posterior segments; male terga never with posterior sclerites.
Gonapophyses with rounded or rectangular lobe on ventral margin; spur present.
Everted portion of endophallus never sclerotised.
HOSTS. Canidae and Herpestidae (Carnivora).
COMMENTS. Suricatoecus has been treated not only as a genus, but also as a synonym and a subgenus of
Felicola (Bedford, 1936 and Hopkins, 1949 respectively); a more complete history of the variations in
status of Suricatoecus and of Eichlerella is provided in Table 5. Eichler (1963) included the manuscript
name Felicomorpha in his catalogue, without providing any further details, the name therefore being a
nomen nudum. In an earlier, unpublished work, Eichler had attributed this name to Keler m.s. , and noted
the type-species, which is a junior synonym of T. vulpis Denny. Felicomorpha is, however, not an available
name.
SPECIES INCLUDED
acutirostris (Stobbe, 1913) [treated as Felicola s. str. by previous authors] (2 cf , 2 $)
bedfordi Hopkins, 1942 [treated as Felicola s. str. by previous authors] (4 cf , 11 9)
cooleyi (Bedford, 1929) comb. n. from Suricatoecus (30 cf , 28 $)
decipiens Hopkins, 1941 comb. rev. from Suricatoecus (9 cf , 9 $)
fahrenholzi( Werneck, 1948) comb. n. from Suricatoecus (16 cf , 16 $)
fennecus (Emerson & Price, 1981) comb. n. from Suricatoecus
guinlei( Werneck, 1948) comb. n. from Suricatoecus (4 cf , 10 $)
macrurus Werneck, 1948 [treated as Felicola s. str. by previous authors] (23 cf , 19 $)
pygidialis Werneck, 1948 [treated as Felicola s. str. by previous authors] (36 cf , 41 $)
quadraticeps (Chapman, 1897) comb. n. from Suricatoecus (5 cf , 9 $)
vulpis (Denny, 1842) comb. n. from Suricatoecus (18 Cf , 25 $)
Genus LORISICOLA Bedford
The genus Lorisicola comprises two subgenera.
DESCRIPTION. Anterior of head with osculum present or absent*; dorsal preantennal sulcus present or
absent; clypeal marginal carina broadened medially, median sclerite variable*; anterolateral margin of
TRICHODECTID MAMMAL LICE
313
210
211
Figs 207-211 Lorisicola species. 207, L. (L.) mjoebergi, 9 abdomen. 208, L. (Paradoxuroecus)
bengalensis, 9 subgenital lobe, ventral. 209, L. (P.) africanus, $ head, dorsal. 210, L. (L.) spenceri, $
terminalia, ventral, setae omitted apart from on gonapophyses. 211, L. (L.) felis, $ gonapophysis,
ventral.
314
C. H. C. LYAL
212
213
Figs 212, 213 Lorisicola (L.) species, C? terminalia. 212, L. (L.) similis. 213, L. (L.) spenceri.
head sinuate or convex; preantennal portion of head of variable length, outline more or less broadly
triangular or rounded, sometimes, if osculum absent, convexly produced anteriorly*. Temple margin
convex, rectangular, or slightly produced laterally*. Male scape expanded or only slightly expanded, with
longitudinal setal row present and comprising at least two setae*; male and female flagellomeres fused;
male flagellum with one or two basally- articulated teeth* . Dorsum of head with setae short or of moderate
length, sparse. Sitophore sclerite unmodified.
Thorax with dorsal setae short or of moderate length* though frequently short and stout on postero-
lateral angles of pterothorax, not present on disc or medially posteriorly on prothorax or pterothorax.
Abdomen oval or elongate, male segment IX not produced greatly. Abdominal spiracles absent, or
present on segments III- VI or III- VIII ; posterior two pairs of spiracles , if six pairs present , sometimes very
small and possibly non-functional*. Abdominal setae short or very short, frequently sparse dorsally;
abdominal pleurum III frequently with posterior setal row comprising stout, conical setae (Fig. 218);
anterior setae absent except on pleurum II; postero-lateral setae present or absent, sometimes numbering
more than one per site* (Fig. 207). Pleural projections present dorsally and sometimes ventrally on
pleurum IV, sclerotised or unsclerotised. Abdominal sclerotisation variable; sterna with sclerites absent
TRICHODECTID MAMMAL LICE
315
}
214
215
Figs 214, 215 Lorisicola species, cf terminalia. 214, L. (L.) mjoebergi. 215, L. (Paradoxuroecus) laticeps.
except for subgenital plate (in male), or present on more posterior segments (VII, VI + VII, V-VII,
IV-VII or III-VII); terga with sclerites on I-VIII, II- VIII, III-VIII or IV-VIII; pleura with sclerites on at
least II, sometimes also on III and IV; male terga with anterior and posterior sclerites present on at least
terga IV-VII, or posterior sclerites not present*.
Gonapophyses with ventral marginal non-tuberculate setae; rounded or rectangular lobe present or
absent on ventral margin*; apical spur present or absent*. Gonapophyses meet ventral vulval margin
acutely, not linked by sclerotised band. Ventral vulval margin not sclerotised; straight or concave, with
chord less than 90 degrees to long axis of abdomen; subgenital lobe present, ventral surface more or less
covered in overlapping scales*.
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r. , VII and VIII fused to
s.g.p.r. and IX absent, or VII fused to s.g.p.r. and VIII and IX absent or present but not fused to s.g.p.r. *.
Pseudostyli absent. Male genital opening postero-dorsal or dorsal; male segment IX lying dorsally on
abdomen. Parameres short, broad, sometimes fused. Basiparameral sclerites present or absent*. Meso-
meres present, fused; median extension present or absent* ; mesomeres extending basally between b.a.l.s.
to contact parameres, or terminating exteriorly to b.a.l.s.*. Male genitalia depicted in Figs 219-226.
HOSTS. Felidae, Herpestidae and Viverridae (Carnivora) and Lorisidae (Primates).
316
C. H. C. LYAL
Fig. 216 Lorisicola (Paradoxuroecus) acuticeps, C? abdomen.
Subgenus LORISICOLA Bedford
(Figs 207, 210-214, 219, 220)
Lorisicola Bedford, 1936: 51. Type-species: Trichodectes mjoebergi Stobbe, by original designation.
DESCRIPTION. Anterior of head with osculum present; clypeal marginal carina broadened medially to form
rectangular or W-shaped sclerite, or broadened slightly to either side of osculum, very lightly sclerotised
posterior to osculum. Temple margin rectangular or slightly produced laterally. Male scape expanded or
only slightly expanded, with setal row comprising two setae; male flagellum with basally-articulated 'teeth'
on projection.
Abdominal spiracles absent, or present on segments III-VIII; posterior two pairs of spiracles, if six pairs
present, sometimes very small and possibly non-functional. Males with posterior tergal sclerites absent.
Gonapophyses with rounded lobe, or lobe absent (Fig. 207); apical spur present or absent (Fig. 211).
Subgenital lobe covered in overlapping pointed scales (Fig. 210) or spines (Fig. 207).
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r. (Fig. 214), or with
sternite VII fused to s.g.p.r. and VIII and IX present but not fused to s.g.p.r. (Fig. 212), or of the latter
form but lacking sternite IX (Fig. 213), or lacking VIII and IX but VII very broad. Basiparameral sclerites
absent. Mesomeres fused, with median extension. Mesomeres extending basally between b.a.l.s. to
TRICHODECTID MAMMAL LICE
317
Fig. 217 Lorisicola (Paradoxuroecus) bengalensis, cf abdomen.
contact parameres (Fig. 222), terminating exteriorly to b.a.l.s. (Fig. 220), or extending anteriad to
posterior end of b.a.l.s. and abruptly reversing, lying ventrally to b.a.l.s., though reversed portion is
difficult to see (Fig. 219).
HOSTS. Felidae and Viverridae (Carnivora) and Lorisidae (Primates).
COMMENTS. The original spelling mjobergi is emended here to mjoebergi in accordance with Article 32(d)(i)
of the International Code of Zoological Nomenclature (1984).
SPECIES INCLUDED
americanus (Emerson & Price, 1983) comb. n. from Felicola
braziliensis (Emerson & Price, 1983) comb. n. from Felicola
caffra (Bedford, 1919) comb. n. from Felicola (1 cf , 1 $)
/e7is(Werneck, 1934) comb. n. from Felicola (1 cf , 1 ?)
hercynianus (Keler, 1957) comb. n. from Felicola (6 cf , 6 $)
ma/a^s/aiius(Werneck, 1948) comb. n. from Trichodectes (4 cf , 5 9)
nyoeftergi(Stobbe, 1913) (c.100 cf , c.100 $)
neofelis (Emerson & Price, 1983) comb. n. from Felicola
siamensis (Emerson, 1964) comb. n. from Felicola (5 cf , 2 <j>)
similis (Emerson & Price, 1983) comb. n. from Felicola (1 (5", 1 9)
spenceri (Hopkins, 1960) comb. n. from Felicola (8 Cf , 5 $)
sudamericanus (Emerson & Price, 1983) comb. n. from Felicola
318
C. H. C. LYAL
Fig. 218 Lorisicola (Paradoxuroecus) juccii, C? abdomen.
Subgenus PAlMDOXl/tfOECI/S Conci gen. rev., stat. n.
(Figs 208, 209, 215-218, 221-226)
Paradoxuroecus Conci, 1942: 141. Type-species: Paradoxuroecus juccii Conci, by original designation.
Parafelicola Werneck, 1948: 226. Type-species: Trichodectes acuticeps Neumann, by original designation.
Syn. n.
Neofelicola Werneck, 1948: 235. Type-species: Neofelicola aspidorhynchus Werneck, by original designa-
tion. Syn. n.
DESCRIPTION. Anterior of head with osculum present, in which case clypeal marginal carina broadened
slightly to either side of osculum and very lightly sclerotised posterior to osculum, or osculum absent, in
which case clypeal marginal carina broadened medially to form posteriorly-convex or straight heavily
sclerotised bar (Fig. 209); outline of preantennal portion of head more or less broadly rounded or
triangular, sometimes, if osculum absent, convexly produced anteriorly (Fig. 209). Temple margin convex
or rectangular. Male scape expanded or only slightly expanded, with longitudinal setal row comprising at
least three setae; male flagellum with two basally-articulated 'teeth', only rarely on projection.
Thorax with dorsal setae short.
Abdominal spiracles absent, or present on segments III- VI. Postero-lateral setae present or absent, but
never numbering more than one per site if present. Males with posterior tergal sclerites absent, or anterior
and posterior sclerites present on at least terga IV- VII.
Gonapophyses with lobe present, rounded or rectangular; apical spur present. Subgenital lobe bilobate,
with scales modified into short spines in some cases, though spines may be sparse (Fig. 208).
TRICHODECTID MAMMAL LICE
319
320
C. H. C. LYAL
224
223
225
226
Figs 223-226 Lorisicola (Paradoxuroecus) species, C? genitalia. 223, L. (P.) juccii. 224, L. (P.)
bengalensis. 225, L. (P.) acuticeps. 226, L. (P.) acuticeps, endophallus everted.
Male subgenital plate with sternites VII, VIII and IX present and fused to s.g.p.r. (Fig. 216), VII and IX
present and fused to s.g.p.r. but VIII absent (Fig. 215), or VII and VIII present and fused to s.g.p.r. but IX
absent (Figs 217, 218). Parameres short, broad, not fused together (see second paragraph under
'comments' below), but may be very closely associated (Figs 223, 224). Basiparameral sclerites present
(Fig. 225) or absent. Mesomeres fused apically with median extension present or, if absent, arch with two
apical nipples (Fig. 221); mesomeral arch sometimes with lateral double flexion (Fig. 225); mesomeres
extending basally between b.a.l.s. to contact parameres, sometimes sharply recurved posteriorly between
b.a.l.s. (Fig. 224). Endophallus frequently with spicular collar, sometimes V-shaped, around gonopore
(Fig. 225).
HOSTS. Herpestidae and Viverridae (Carnivora).
COMMENTS. Paradoxuroecus has been considered by most authors, following Werneck (1948), to be a
synonym of Felicola; it is here recalled from synonymy and placed as a subgenus of Lorisicola. Neofelicola
and Parafelicola were both considered by Hopkins (1949) to be subgenera of Felicola. A more complete
history of the variations in the status of Paradoxuroecus, Neofelicola and Parafelicola is presented in Table
5.
Werneck (1948) figured the parameres of aspidorhynchus, sumatrensis and juccii as fused together;
examination of the type-material of the first two species and of numerous specimens of the last has revealed
that this is not the case, although in L. juccii the parameres are very closely associated with each other.
TRICHODECTID MAMMAL LICE 321
Lorisicola (Paradoxuroecus) bengalensis (Werneck, 1948) was described in Neofelicola from three
females, the male being unknown. These females were taken from a museum skin of Paradoxurus
hermaphrodytus canus Miller, which was itself collected in Thailand. Female lice subsequently collected
from P. hermaphrodytus subspp. in Thailand agree with Werneck's (1948) description of N. bengalensis.
Emerson (1965) describes the male of bengalensis, collected from the type host in Thailand, and
distinguishes a new species, N. philippinensis , collected from Paradoxurus philippinensis.
Numerous specimens of lice from a number of subspecies of Paradoxurus hermaphrodytus have been
examined during the course of this study, and it has become apparent that lice of two clades are present: the
Felicola (F.) zeylonicus-viverriculae clade and the Lorisicola (P.) philippinensis-juccii clade.
These clades may be distinguished by the following characters.
F. (F.) zeylonicus-viverriculae clade - Three pairs of abdominal spiracles; male flagellum with 'teeth' not
basally articulated; parameres long, slender, extending anteriorly between b.a.l.s.; mesomeres not
basally extending between b.a.l.s. ; mesomeral arch lacking median extension; female subgenital lobe
broad, smooth ventrally, with long, flattened marginal spines.
L. (P.) philippinensis-juccii clade - Four pairs of abdominal spiracles; male flagellum with 'teeth' basally
articulated; parameres short, broad, closely-associated with one another but not fused, not extending
between b.a.l.s., if reaching them; mesomeres basally extending between b.a.l.s. to contact para-
meres; mesomeral arch with median extension; female subgenital lobe narrow, apically bilobate,
covered ventrally with small pointed scales, lacking marginal spines.
The female described as N. bengalensis by Werneck (1948) is, by the structure of the subgenital lobe and
the number of abdominal spiracles, a member of the L. (P.) philippinensis-juccii clade, as are both male
and female of N. philippinensis as described by Emerson (1965). The male described by Emerson (1965) as
N. bengalensis is, however, a member of the F. (F.) zeylonicus-viverriculae clade and therefore not the true
male of N. bengalensis. Female lice of the latter clade are now known from Paradoxurus hermaphrodytus
subspp. , as are males of the former.
The male of L. bengalensis has genitalia indistinguishable from those of L. philippinensis. Emerson
(1965) distinguishes L. philippinensis from L. bengalensis by the male genitalia, the greater number of
sternal and tergal setae in both sexes of the former, and the greater number of vulval setae in the former.
The characteristics of the male genitalia, as stated above, are the same in the two species. Study of the large
sample of specimens now available indicates that the vulval setal number of L. philippinensis is not outside
the range of L. bengalensis. The tergal and sternal setae in the males are the same, as are the sternal setae in
the females. The tergal setae of the female paratype of L. philippinensis in the British Museum (Natural
History) collection are the same as those of L. bengalensis, but do not agree with the figure in Emerson
(1965), where far more setae are depicted. The host species Paradoxurus philippinensis is at best a
subspecies of P. hermaphrodytus, and all other subspecies appear to harbour L. bengalensis. L. philip-
pinensis (Emerson) is consequently provisionally synonymised with L. bengalensis (Werneck), subject to
examination of the female allotype of philippinensis.
SPECIES INCLUDED
acuticeps (Neumann, 1902) comb. n. from Parafelicola (28 cf , 34 <J?)
africanus (Emerson & Price, 1966) comb. n. from Parafelicola (15 cf , 7 $)
aspidorhynchus ( Werneck, 1948) comb. n. from Neofelicola (6 cf , 7 9)
bengalensis ( Werneck, 1948) comb. n. from Neofelicola (26 cf , 59 $)
juccU Conci, 1942 comb. n. from Felicola (82 cf , 81 ?)
laticeps( Werneck, 1942) comb. n. from Suricatoecus (10 cf , 13 $)
/en/cora/s ( Werneck, 1948) comb. n. from Parafelicola (9 cf , 22 <J>)
mungos (Stobbe, 1913) comb. n. from Suricatoecus (1 cf , 3 $)
neoafricanus (Emerson & Price, 1968) comb. n. from Parafelicola (holotype cf , 4 $)
paralaticeps ( Werneck, 1948) comb. n. from Suricatoecus (1 cf , 4 $)
philippinensis (Emerson, 1965) comb. n. from Neofelicola (5 cf , 5 $)
sumatrensis ( Werneck, 1948) comb. n. from Neofelicola (4 cf , 4 $)
wernecW (Hopkins, 1941) comb. n. from Parafelicola (7 cf , 11 ?)
NEOTRICHODECTINAE subfam. n.
Type-genus: Neotrichodectes Ewing
Genus NEOTRICHODECTES Ewing
The genus Neotrichodectes comprises five subgenera.
322
227
228
229
231
Figs 227-231 Neotrichodectes species. 227, N. barbarae, cf anterior abdominal segments, dorsal. 228, N.
mephitidis, 9 abdominal pleura II and III, dorsal. 229, N. mephitidis, cf terminalia. 230, N. barbarae, cf
temple margin. 231, N. semistriatus, cf flagellum.
DESCRIPTION. Anterior of head with osculum present, sometimes very shallow* ; dorsal preantennal sulcus
present, not always clear; clypeal marginal carina broadened medially into dorsal sclerite of variable form*
which is always more heavily sclerotised laterally (dorsal to margin of clypeus and pulvinus) than medially;
antero-lateral margin of head convex, straight or sinuate*; preantennal portion of head short or longer,
sometimes as long as postantennal portion*, outline broadly rounded or triangular*. Temple margin
broadly convex. Male scape expanded, with longitudinal setal row present and comprising at least four
setae; flagellomeres fused in males and females; male flagellum with two or more basally-articulated 'teeth'
TRICHODECTID MAMMAL LICE
323
233
235
Figs 232-235 Neotrichodectes species, $ terminalia. 232, N. mephitidis, ventral. 233, N. chilensis,
showing setal distribution. 234, N. chilensis, showing sclerites. 235, N. barbarae.
distally, and with basal toothed or rough projection sometimes present* (Fig. 231); female pedicel lacking
membranous projection. Dorsum of head with most setae short or of moderate length, more or less sparse,
sometimes with long seta on posterior margin of temple* (Fig. 230). Sitophore sclerite unmodified.
Thorax with dorsal setae long or of moderate length* though frequently short and spine-like on
postero-lateral angles of prothorax and pterothorax; posterior margin of pronotum with four setae and
wide median gap, posterior of pterothorax dorsally with setae more abundant, marginal or submarginal,
median gap present or absent.
Abdomen broadly rounded, not greatly projecting posteriad in male (Fig. 229). Abdominal spiracles
absent. Abdominal setae grenerally abundant, as long as segment or sparse, shorter, with long setae
present only on posterior pleura* ; terga, especially of males, with median and lateral setal groups distinct,
though median groups generally united; male terga II- VI with median two setae much smaller than other
setae in the row, sometimes separated by one or more longer setae (Fig. 229); anterior setae present on
324
236
237
Figs 236, 237 Neotrichodectes species, $ terminalia. 236, N. gastrodes. 237, N. pallidus.
pleura II and HI only; postero-lateral setae presumed absent, though possibly present as the most lateral
seta of lateral group, which is frequently situated more posteriorly than other setae. Abdominal pleura
lacking projections, except for small sclerotised or unsclerotised projection dorsally on pleurum III of
female N. mephltidis. Abdominal terga and sterna lacking sclerites, except for lateral rods of subgenital
plate in male and, sometimes, tergite IX in female; abdominal pleura usually unsclerotised, sometimes
with sclerites on pleura II, III, IV and, in females, VIII*.
Gonapophyses with or without lobe on ventral margin, very variable*. Gonapophyses meet ventral
vulval margin smoothly or acutely, but not linked by sclerotised band. Ventral vulval margin not
sclerotised; generally convex medially*, subgenital lobe present or absent*. Common oviduct not notably
striate (cf. Geomydoecus}.
Male subgenital plate with only s.g.p.r. present (Fig. 229). Pseudostyli absent. Male genital opening
TRICHODECTID MAMMAL LICE
325
240
Figs 238-242 Neotrichodectes species, C? genitalia. 238, N. (Neotrichodectes) mephitidis. 239, N.
(Nasuicola) pallidus. 240, N. (Trigonodectes) barbarae. 241, N. (Lakshminarayanella) gastrodes. 242,
N. (Conepaticola) chilensis.
dorsal. Parameres fused to form single plate with apex pointed or bifid*. Basiparameral sclerites absent.
Mesomeres fused apically; mesomeral arch with median extension*; mesomeres basally abut postero-
lateral projections of b.a.l.s. but do not contact parameral plate (Fig. 242), or contact neither b.a.l.s. nor
parameral plate (Fig. 240).
HOSTS. Mustelidae and Procyonidae (Carnivora) and Bradypodidae (Edentata).
COMMENTS. Neotrichodectes has been considered a synonym and a subgenus of Trichodectes (by Hopkins,
1942 and Hopkins, 1949 respectively); a more complete history of the variations in status of Neo-
trichodectes is presented in Table 4.
326 C. H. C. LYAL
Subgenus NEOTRICHODECTES Ewing
(Figs 228, 229, 232, 238)
Neotrichodectes Ewing, 1929: 194. Type-species: Goniodes mephitidis Packard, by original designation.
DESCRIPTION. Anterior of head with clypeal marginal carina broadened into dorsal sclerite which is more or
less convex posteriorly, sometimes almost circular; preantennal portion of head sometimes as long as
postantennal portion. Male flagellum with two basally-articulated 'teeth' distally, and with toothed
projection sometimes present basally (Fig. 231). Temple with long seta present on posterior margin.
Thoracic and abdominal setae long, abundant. Abdominal pleura lacking projections, except for small
sclerotised or unsclerotised projection dorsally on pleurum III of female N. mephitidis (Fig. 228).
Abdomen lacking sternal, tergal and pleural sclerites, except for lateral rods of male subgenital plate.
Gonapophyses broad, membranous, with ventral marginal setae absent or, if present, basal only (Fig.
232); ventral lobe absent. Ventral vulval margin with lobe present though difficult to see; lobe serrate, at
least along posterior margin (Fig. 232). Female genital chamber with dorsal wall bearing slanting scales
laterally, lightly sclerotised and lacking scales or other decoration medially (anteriorly).
Parameral plate slender, apically bifid, with median basal extension reaching anteriorly between b.a.l.s.
Mesomeral arch with median extension pointed; mesomeres basally abut postero-lateral extensions of
b.a.l.s. Male genitalia depicted in Fig. 238.
HOSTS. Mustelidae and Procyonidae (Carnivora).
COMMENTS. Neotrichodectes s. str. is most readily distinguished from other subgenera of Neotrichodectes by
characters of the female terminalia: the retention of the plesiomorphic form of the gonapophyses (found
also in Geomydoecus) and the apomorphic development of a membranous subgenital lobe (cf . the very
different structure in N. (Trigonodectes)). Assignment of male insects to the subgenus relies on absence of
the basal flagellar projection (a character-state reversal not undergone by the type-species of the
subgenus), the presence of an anterior development of the parameral plate between the b.a.l.s., and the
slenderness of that parameral plate in relation to its length. Neotrichodectes wolffhuegeli (Werneck) is
known only from the male, although Werneck (1948) predicted that the female would be very similar to
that of N. chilensis (placed in N. (Conepaticola) in this study), and so must be assigned to subgenus on the
basis of male characters. N. wolffhuegeli does have a toothed projection on the base of the male flagellum,
although, as indicated above, cannot be eliminated from Neotrichodectes s. str. on that basis. The form of
the parameral plate of N. wolffhuegeli is much the same as members of Neotrichodectes s. str. and on this
evidence the species is placed in the subgenus. Emerson (pers. comm.), however, suggests that the male
genitalia of N. wolffhuegeli lie within the limits of permissible variation of N. chilensis, which he therefore
considers as a junior synonym; N. wolffhuegeli and N. chilensis are found on the same host, although N.
chilensis is also found on several other species of the host genus. All specimens identified as N. wolffhuegeli
(by Werneck and in the present study) have a much narrower parameral plate than those identified as N.
chilensis. However, the parameral plate, whilst not extending anteriad between the b.a.l.s. in most
specimens ofN. chilensis, does do so in some. In those specimens of TV. chilensis where there is no anterior
extension of the parameral plate the endophallus lacks large, heavily-sclerotised teeth, or such teeth are
few in number; in those where the extension is present the teeth are correspondingly more developed. The
degree of development of the endophallus teeth seems to be proportional to the degree of development of
the anterior margin of the parameral plate in N. chilensis and such teeth are present and well developed in
N. wolffhuegeli. The sample of specimens of both species was too small to permit any correlations of
genitalia type with host species or geographical distribution, though within N. chilensis specimens
exhibiting both extremes were found from the same host in the same area. For the purposes of this study,
the two species are treated as separate, N. wolffhuegeli is assigned to Neotrichodectes s. str., and N.
chilensis is assigned to Neotrichodectes (Conepaticola). This conclusion is regarded as the most satisfactory
for the data presently available, but further collecting from species of the host genus (Conepatus) is needed
to clarify the situation.
The original spelling of the species name wolffhugeli is emended here to wolffhuegeli in accordance with
Article 32 (d)(i) of the International Code of Zoological Nomenclature (1984).
SPECIES INCLUDED
mephitidis (Packard, 1873) (c.50 cf , c.100 $)
minutus (Paine, 1912) (64 cf , 76 $)
osftorn/Keler, 1944 (16 cf , 21 $)
f/ioraciciis(Osborn, 1902) (11 cf, 9 $)
wolffhuegeli ( Werneck, 1936) (1 cf)
TRICHODECTID MAMMAL LICE 327
Subgenus TRIGONODECTES Keler gen. rev., stat. n.
(Figs 227, 230, 235, 240)
Trigonodectes Keler, 1944: 179, 185. Type-species: Trichodectes barbarae Neumann, by original designa-
tion.
DESCRIPTION. Anterior of head with osculum very shallow; clypeal marginal carina broadened medially
into dorsal U-shaped sclerite with median posterior process; anterolateral margin of head convex;
preantennal portion of head with outline rounded. Male flagellum with two basally-articulated 'teeth'
distally, and with toothed projection present basally. Temple with long seta present on posterior margin.
Thoracic and abdominal setae long, abundant. Abdominal pleura lacking projections. Abdomen lacking
sternal, tergal and pleural sclerites, except for lateral rods of male subgenital plate.
Gonapophyses slender, apically acute, sclerotised, with strong setae present along ventral margin;
ventral lobe absent. Gonapophyses meet ventral vulval margin acutely. Ventral vulval margin with median
lobulate projection, with margin not serrate, and submarginal setal row present (Fig. 235). Female genital
chamber lacking lateral slanting scales and anterior sclerotised area on dorsal wall, but both ventral and
dorsal walls bearing numerous scales with posterior spinules.
Parameral plate triangular or shield-shaped, pointed apically. Mesomeral arch with median extension
rounded, covered in small tubercles; mesomeres extend basally anterior to ends of b.a.l.s., and do not
contact b.a.l.s. or parameral plate. Basal apodeme lacking postero-lateral projections on b.a.l.s. Male
genitalia depicted in Fig. 240.
HOSTS. Mustelinae (Carnivora: Mustelidae).
COMMENTS. Trigonodectes has been treated as a synonym and a subgenus of Trichodectes (by Werneck,
1948 and Hopkins, 1949 respectively); in this study it is raised from synonymy with Trichodectes and
barbarae is placed for the first time in Neotrichodectes. A more complete history of the varying status
accorded to Trigonodectes is presented in Table 4.
SPECIES INCLUDED
barbarae (Neumann, 1913) comb. n. from Trichodectes (13 C?, 13 $)
Subgenus NASUICOLA subgen. n.
(Figs 237, 239)
Type-species: Trichodectes pallidus Piaget.
DESCRIPTION. Anterior of head with osculum shallow; clypeal marginal carina broadened into dorsal
rectangular sclerite; antero-lateral margin of head convex; preantennal portion of head not as long as
postantennal portion, outline broadly rounded. Male flagellum with two basally-articulated 'teeth' distally,
and with toothed projection present basally. Temple with long seta present on posterior margin.
Thoracic and abdominal setae long, abundant. Abdominal pleura lacking projections. Abdomen lacking
sternal, tergal and pleural sclerites, except for lateral rods of male subgenital plate.
Gonapophyses with rounded ventral lobe with submarginal setae; spur distal to lobe very short.
Gonapophyses meet ventral vulval margin acutely. Ventral vulval margin convex, but subgenital lobe or
lobulate process not present. Female terminalia depicted in Fig. 237. Female genital chamber with dorsal
wall bearing slanting scales laterally, spines medially (spines most apparent anteriorly, though may be
obscured).
Parameral plate apically bifid, lacking median basal extension reaching anteriorly between b.a.l.s.
Mesomeral arch with median extension pointed; mesomeres basally abut postero-lateral extensions of
b.a.l.s. Male genitalia depicted in Fig. 239.
HOSTS. Procyonidae (Carnivora).
SPECIES INCLUDED
pallidus (Piaget, 1880) [treated by previous authors as Neotrichodectes s. str.] (65 cf , 75 $)
Subgenus LAKSHMINARAYANELLA Eichler stat. n.
(Figs 236, 241)
Lymeon Eichler, 1940: 158. Type-species: Trichodectes gostrodesCummings, by monotypy. [Homonym of
Lymeon Foerster, 1868: 176.]
Lakshminarayanella Eichler, 1982: 83. [Replacement name for Lymeon Eichler.]
328 C. H. C. LYAL
DESCRIPTION. Anterior of head with osculum present, deep; clypeal marginal carina broadened medially
into posteriorly convex bar, parallel to curvature of osculum; antero-lateral margin of head straight or
slightly sinuate; preantennal portion of head short in male, longer in female; outline broadly triangular.
Male flagellum with seven basally-articulated 'teeth' distally, and with roughened projection present
basally. Temple with no long setae present on posterior margin.
Thorax with dorsal setae of moderate length, longest on postero-lateral angle of pterothorax; ptero-
thorax with setae sparse along posterior dorsal margin.
Abdominal setae of moderate length, not as long as segment except on posterior pleura; terga with
lateral and median setal groups not clearly distinct, median gap sometimes pronounced. Abdominal pleura
lacking projections. Abdomen lacking tergal and sternal sclerites except for lateral rods of male subgenital
plate and tergite IX of female; abdominal pleura II, III, IV and, in female, VIII, with sclerites, though that
of IV sometimes very small.
Gonapophyses broad, very thick; ventral lobe present, thick, with setae along posterior margin; spur
distal to lobe not short, rounded apically (Fig. 236). Gonapophyses meet ventral vulval margin acutely.
Ventral vulval margin convex, but subgenital lobe or lobulate process not present. Female genital chamber
with dorsal wall bearing slanting scales laterally, spines medially (spines most apparent anteriorly).
Parameral plate apically bifid, sometimes projecting slightly anteriad between b.a.l.s. Mesomeral arch
with median extension pointed, broad basally; mesomeres basally abut postero-lateral extensions ol
b.a.l.s. Male genitalia depicted in Fig. 241.
HOSTS. Bradypodidae (Edentata).
SPECIES INCLUDED
cummingsi (Eichler, 1943) comb. n. from Lakshminarayanella
gastrodes (Cummings, 1916) comb. n. from Lakshminarayanella (4 cf, 6 $ , 4 nymphs)
Subgenus CONEPATICOLA subgen. n.
(Figs 231, 233, 234, 242)
Type-species: Neotrichodectes semistriatus Emerson & Price.
DESCRIPTION. Anterior of head with clypeal marginal carina broadened medially into dorsal U-shapec
sclerite with median posterior process; antero-lateral margin of head convex or sinuate; preantenna
portion of head not as long as postantennal portion, outline broadly rounded. Male flagellum with two
basally-articulated 'teeth' distally, and with toothed projection basally. Temple with long seta present on
posterior margin.
Thoracic and abdominal setae long, abundant. Abdominal pleura lacking projections. Abdomen lacking
sternal and tergal sclerites, except for sternal rods of male subgenital plate and tergite IX of female;
abdominal pleura II, III and VIII in female sometimes with sclerites, though that of III may be very small.
Gonapophyses with ventral lobe present, large, apparently comprising fused setal tubercles, with setae
along posterior margin and anterior margin, the latter frequently directed posteriad; spur distal to lobe
present, not short, frequently obtuse apically. Gonapophyses meet ventral vulval margin acutely. Ventral
vulval margin convex, but subgenital lobe or lobulate process not present. Female terminalia depicted in
Figs 233, 234. Female genital chamber with dorsal wall bearing slanting scales laterally, spines medially,
sometimes lightly sclerotised and lacking scales, spines or other decoration antero-medially.
Parameral plate apically bifid, sometimes projecting slightly anteriad between b.a.l.s. Mesomeral arch
with median extension pointed; mesomeres basally abut postero-lateral extension of b.a.l.s. Male genitalia
depicted in Fig. 242.
HOSTS. Mustelinae and Mephitinae (Carnivora: Mustelidae).
SPECIES INCLUDED
arizonaeWerneck, 1948 (18 C?, 39 $)
cMens/sWerneck, 1948 (c.50 C?, c.50 $)
interruptofasciatus (Kellogg & Ferris, 1915) (27 cf, 46 $)
semjsfriatus Emerson & Price, 1976 (5 cf , 5 $)
All the above species have been treated previously as Neotrichodectes s. str.
Genus GEOMYDOECUSEwing
The genus Geomydoecus comprises two subgenera.
TRICHODECTID MAMMAL LICE
329
DESCRIPTION. Anterior of head with osculum present; dorsal preantennal sulcus present; clypeal marginal
carina broadened medially into posteriorly convex bar; antero-lateral margin of head convex or sinuate;
preantennal portion of head not long, outline broadly triangular or rounded. Temple margin broadly
convex. Male scape expanded; longitudinal setal row present and comprising at least three setae; male
scape sometimes with median posterior projection; flagellomeres fused in males and females; male
flagellum with two basally-articulated 'teeth'; female pedicel with membranous postero-ventral projection
(Fig. 245), sometimes obscure. Dorsum of head with setae short or of moderate length, more or less sparse;
temple margin sometimes with specialised long, fine or short and stout latero-posterior setae*. Sitophore
sclerite unmodified.
244
245
Figs 243-245 Geomydoecus species. 243, G. (G. ) calif ornicus , $ terminalia. 244, G. (Thomomydoecus)
asymmetricus , cf anterior of abdomen, dorsal. 245, G. (T.)pattoni, $ antenna.
330
C. H. C. LYAL
Fig. 246 Geornydoecus (Thomomydoecus) minor, 9 abdomen. Hatched areas indicate damage to
specimen.
Thorax with dorsal setae long or of moderate length; posterior margin of pronotum with four setae and
wide median gap, posterior margin of pterothorax dorsally with varying number of marginal or submargin-
al setae.
Abdomen broadly rounded or more elongate and tapered, particularly in male. Abdominal spiracles
absent. Abdomen with at least some setae as long as segment; setae generally abundant; terga, especially
of males, with median and lateral setal groups distinct, though median groups generally united; male terga
II- VI without median setae shorter than others; male terga II and III sometimes with median group
comprising exceptionally long, stout setae* (Fig. 244); anterior setae present on pleura II and III only;
postero-lateral setae sometimes clearly present (Fig. 244), otherwise obscured, though may be present as
most lateral seta of lateral group, which is frequently situated more posteriorly than other setae.
Abdominal pleura with projections dorsally on pleura II, III, IV and, at least in female, ventrally on IV,
sclerotised (Fig. 246) or unsclerotised*; projections generally more apparent in females than males.
Abdominal terga and sterna lacking sclerites, except for lateral rods of male subgenital plate and,
sometimes, terga II-IV of male*; abdominal pleura II and sometimes III and IV sclerotised, at least in
female; other pleura unsclerotised.
TRICHODECTID MAMMAL LICE
331
249
248
Figs 247-250 Geomydoecus species, cf genitalia. 247, G. (G. ) thomomyus. 248, G. (G.) actuosi. 249, G.
(Thomomydoecus) minor. 250, G. (T.) wardi.
332 C. H. C. LYAL
Gonapophyses broad, membranous, with ventral marginal setae, if present, generally basal only; ventral
lobe absent (Figs 243, 246). Gonapophyses meet ventral vulval margin smoothly or acutely, but not linked
by sclerotised band. Ventral vulval margin not sclerotised; generally convex or very convex medially;
subgenital lobe not present. Female genital chamber with dorsal wall bearing slanting scales laterally,
lightly sclerotised and lacking scales or other decoration antero-medially. Common oviduct generally with
distinct striae.
Male subgenital plate with only s.g.p.r. present. Pseudostyli absent. Male genital opening dorsal.
Parameres fused to form single plate with apex pointed or bifid*. Basiparameral sclerites absent.
Mesomeres fused apically; mesomeral arch with or without median extension; mesomeres basally abut
b.a.l.s., which sometimes have postero-lateral extensions; mesomeres do not contact parameral plate.
Male genitalia depicted in Figs 247-250.
HOSTS. Geomyidae (Rodentia).
COMMENTS. A few of the species are parthenogenetic.
No more than 25 specimens of most species were examined during the course of this study, though in
many cases large numbers were available.
Detailed descriptions of all species of Geomydoecus s.l. and a phenetic treatment of the genus may be
found in Hellenthal & Price (1976, 1980), Price (1974, 1975), Price & Emerson (1971, 1972), Price &
Hellenthal (1975a, 19756, 1976, 1979, 1980a, 19806, 1980c, 1981a, 19816), Price & Timm (1979), Timm &
Price (1979, 1980).
Subgenus GEOMYDOECUS Ewing
(Figs 243, 247, 248)
Geomydoecus Ewing, 1929: 193. Type-species: Trichodectes geomydis Osborn, by original designation.
DESCRIPTION. Temple margin sometimes with two short, stout setae latero-posteriorly, or single long, fine
seta latero-posteriorly.
Male abdominal terga II and III only rarely with median setal group comprising exceptionally long, stout
setae (G. copei). Pleural projections rarely sclerotised. Male terga II-IV not sclerotised.
Male genitalia not asymmetric. Parameral plate apically pointed or bifid. Male genitalia depicted in Figs
247, 248.
HOSTS. Geomyidae (Rodentia).
SPECIES INCLUDED
actuosi Price & Hellenthal, 1981 (25 cf , 25 9)
albati Price & Hellenthal, 1981 (25 cf , 25 9)
alcorni Price & Emerson, 1971 (6 cf , 6 $)
aUeni Price & Emerson, 1971 (2 cf , 6 $)
angularis Price & Hellenthal, 1981 (25 cf , 25 9)
aureiaurei Price & Hellenthal, 1981 (25 cf , 25 $)
aurei grahamensis Price & Hellenthal, 1981 (25 (?, 25 9)
bajaiensis Price & Hellenthal, 1981 (25 cf , 25 9)
bulleri Price & Emerson, 1971 (25 cf , 25 $)
californicus (Chapman, 1897) (25 cf , 25 $)
centra/is Price & Hellenthal, 1981 (25 cf , 25 9)
chapini Werneck, 1945 (12 cf , 16 $)
cherriei Price, 1974
chiapensis Price & Emerson, 1971 (18 cf , 20 $)
chihuahuae chihuahuae Price & Hellenthal, 1979 (25 cf , 25 9)
chihuahuae emersoni Price & Hellenthal, 1979 (25 cf , 17 9)
clausonae Price & Hellenthal, 1981 (25 cf , 25 9)
dittooi Price & Hellenthal, 1981 (25 cf , 25 9)
copei Werneck, 1945 (25 cf , 25 9)
corojiado/ Barrera, 1961 (25 cf , 25 9)
costaricensis Price & Emerson, 1971 (7 cf, 7 9)
crovelloi Price & Hellenthal, 1981 (25 cf , 25 9)
dakotensis Price & Emerson, 1971 (25 cf , 25 9)
dalgleishiTimm & Price, 1979 (2 cf , 2 9)
TRICHODECTID MAMMAL LICE 333
dariensis Price & Emerson, 1971 (10 cf , 14 9)
duchesnensis Price & Emerson, 1971 (5 cf , 8 9)
ewingi Price & Emerson, 1971 (25 cf , 25 9)
expansus(Duges, 1902) (25 cf , 25 $)
art/in/ Price & Hellenthal, 1981 (26 cf , 25 9)
fulvescens Price & Emerson, 1971 (25 cf , 25 <j>)
fu/vi Price & Hellenthal, 1979 (25 cf , 25 9)
geomydis(Osborn, 1891) (25 cf, 25 $)
guadalupensis Hellenthal & Price, 1980 (25 cf , 25 9)
AeanejiTimm & Price, 1980 (25 cf , 25 ?)
hoffmanni Price & Hellenthal, 1976 (25 cf , 25 9)
/Hieji Price & Hellenthal, 1980 (25 cf , 25 9)
idahoensis Price & Emerson, 1971 (25 cf , 25 9)
illinoensis Price & Emerson, 1971 (25 cf , 25 9)
jaliscoensis Price & Hellenthal, 1981 (25 cf , 25 9)
joaesi Price & Emerson, 1971 (4 cf , 5 9)
Hmitaris limitaris Price & Hellenthal, 1981 (25 cf , 25 9)
limitaris bakeri Price & Hellenthal, 1981 (25 cf , 25 9)
limitaris halli Price & Hellenthal, 1981 (25 cf , 25 9)
limitaris tolteci Price & Hellenthal, 1981 (25 cf , 25 9)
mart/ni Price & Hellenthal, 1975 (25 cf , 25 9)
mcgregori Price & Emerson, 1971 (25 cf , 25 9)
merriami Price & Emerson, 1971 (25 cf , 25 9)
mexicanus Price & Emerson, 1971 (25 cf , 25 $)
mobilensisPrice, 1975 (25 9)
jmiscufi Price & Hellenthal, 1981 (25 cf , 25 9)
nayaritensis Price & Hellenthal, 1981 (16 cf , 25 9)
nebrathkensisTimm & Price, 1980 (25 cf , 25 9)
oklahomensis Price & Emerson, 1971 (25 cf , 25 9)
oregonus Price & Emerson, 1971 (25 cf , 25 9)
panamensis Price & Emerson, 1971 (23 cf , 22 9)
pattoni Price & Hellenthal, 1979 (12 cf , 8 9)
perotensis perotensis Price & Emerson, 1971 (25 cf , 25 9)
perotensis irolonis Price & Emerson, 1971 (25 cf , 25 9)
polydentatus Price & Emerson, 1971 (25 cf , 25 9)
quadridentatus Price & Emerson, 1971 (25 cf , 25 9)
scleritus (McGregor, 1917) (3 cf , 25 9)
setzeri Price, 1974 (6 cf , 8 9)
shastensis Price & Hellenthal, 1980 (25 cf , 25 9)
sinaloae Price & Hellenthal, 1981 (25 cf , 25 9)
sp/dta/Timm & Price, 1980 (25 cf , 25 9)
subcalifornicus Price & Emerson, 1971 (25 cf , 25 9)
subgeomydis Price & Emerson, 1971 (25 cf , 25 9)
subnuhili Price & Hellenthal, 1975 (25 cf , 25 9)
tamaulipensis Price & Hellenthal, 1975 (3 cf , 25 9)
texanustexanusEv/ing, 1936 (25 cf , 25 9)
texanus tropicalis Price & Hellenthal, 1975 (25 cf , 25 9)
thomomyus (McGregor, 1917) (25 cf , 25 9)
tolucae Price & Emerson, 1971 (25 cf , 25 9)
traubi Price & Emerson, 1971 (25 cf , 25 9)
trichopi Price & Emerson, 1971 (25 cf , 25 9)
truncatus Werneck, 1950 (25 cf , 25 9)
umbrini Price & Emerson, 1971 (25 cf , 25 9)
ustulati ustulatiPrice & Hellenthal, 1975 (25 cf , 25 9)
ustulati clarkii Price & Hellenthal, 1975 (25 cf , 25 9)
veracruzensis Price & Emerson, 1971 (25 cf , 25 9)
warmanae Price & Hellenthal, 1981 (25 cf , 25 $)
welleri w?77eri Price & Hellenthal, 1981 (25 cf , 25 9)
wellerimultilineat us Price & Hellenthal, 1981 (25 cf , 25 9)
334 C. H. C. LYAL
werneckl Price & Emerson, 1971 (25 cf , 25 $)
yucatanensis Price & Emerson, 1971 (25 cf , 25 $)
Subgenus THOMOMYDOECUS Price & Emerson
(Figs 244-246, 249, 250)
Thomomydoecus Price & Emerson, 1972: 464 [as subgenus of Geomydoecus Ewing]. Type-species:
Geomydoecus (Thomomydoecus) jamesbeeri Price & Emerson, by original designation.
DESCRIPTION. Temple margin with single stout seta and finer, shorter adjacent setae latero-posteriorly.
Male abdominal terga II and III with median setal group comprising exceptionally long, stout setae.
Pleural projections sclerotised, at least in female (Fig. 246). Male terga II-IV sometimes with sclerites.
Gonapophyses meet vulval margin smoothly.
Male genitalia symmetric (Fig. 250) or asymmetric (Fig. 249). Parameral plate apically pointed.
HOSTS. Thomomyusspp. (Rodentia: Geomyidae).
COMMENTS. As discussed on p. 232, the subgenus Geomydoecus is probably paraphyletic and Thomomy-
doecus paraphyletic or even polyphyletic. Geomydoecus (Thomomydoecus) was raised to generic status by
Hellenthal & Price (1984), on the basis that there are 'sufficient' morphological differences between it and
Geomydoecus s. str. This action, reversed here, is consistent with the purely phenetic approach employed
by the authors, but cannot be reconciled with the cladistic methods used in this study. The division of
Geomydoecus into subgenera is retained, even though neither of the two subgenera is 'natural' (i.e.
holophyletic), because insufficient work has been done to resolve the relationships properly, and the
independent unit of the classification, the genus Geomydoecus, is holophyletic. Raising Thomomydoecus
to generic status, however, would produce two independent units in the classification (Geomydoecus and
Thomomydoecus) that differ from all the others in not being holophyletic.
Geomydoecus (T.) byersi (Hellenthal & Price) was described after the completion of this manuscript and
has not been included in the cladistic analysis.
SPECIES INCLUDED
asymmetricus Price & Hellenthal, 1980 (25 cf , 25 $)
birneyi Price & Hellenthal, 1980 (25 cf , 25 $)
byersi (Hellenthal & Price, 1984) comb. n. from Thomomydoecus
dickermani Price & Emerson, 1972 (25 cf , 25 $)
genowaysi Price & Emerson, 1972 (25 cf , 25 $)
green Price & Hellenthal, 1980 (2 cf )
jamesbeeri Price & Emerson, 1972 (8 cT, 10 $)
johnhafneri Price & Hellenthal, 1980 (25 cf , 25 $)
markhafneri Price & Hellenthal, 1980 (25 cf , 25 $)
minor Werneck, 1950 (25 cf , 25 ?)
neocopei Price & Emerson, 1971 (2 cf , 1 9)
orizabae Price & Hellenthal, 1980 (10 cf , 26 $)
peregrin! Price & Hellenthal, 1980 (4 cf , 4 $)
potteri Price & Hellenthal, 1980 (16 cf , 25 $)
timmi Price & Hellenthal, 1980 (25 cf , 25 $)
wardi Price & Emerson, 1971 (25 cf , 25 $)
williamsiPrice & Hellenthal, 1980 (8 cf , 16 $)
zacatecae Price & Hellenthal, 1980 (25 cf , 25 $)
Keys to Trichodectidae
Two keys are provided: a key to subfamilies and a key to genera and subgenera. The latter key
contains all genera and subgenera, and it is not necessary to use the subfamily key as an
introduction to it.
The subfamily key is included because of the formal requirement that any newly described
taxon must be accompanied by some form of description in order to make the name available.
The key to the five subfamilies of Trichodectidae therefore serves to distinguish Neotrichodecti-
nae subfam. n. from all others.
TRICHODECTID MAMMAL LICE 335
Key to subfamilies
1 No abdominal spiracles present; majority of tergal and sternal setae at least two-thirds length of
segment or, if not, median setal group on tergum II comprising at least three setae (and,
frequently, median groups running together). New World
NEOTRICHODECTINAE subfam. n.(p. 321)
- At least one pair of abdominal spiracles present or, if not, majority of abdominal sternal and
tergal setae less than two-thirds length of segment or median setal group on tergum II
comprising only one seta. Old and New World 2
2 Female subgenital lobe present, frequently with serrate margin, at least posteriorly; if margin
of subgenital lobe smooth, gonapophyses meet vulval margin smoothly (Fig. 153); female
flagellomeres fused; abdominal spiracles numbering six or fewer pairs. Parasitic on Carni-
vora and Primates. Old and New World TRICHODECTINAE Kellogg, 1896 (p. 286)
- Female subgenital lobe absent or, if present, not marginally serrate and gonapophyses meet
vulval margin acutely; female flagellomeres fused or unfused; abdominal spiracles numbering
six pairs, though spiracles on segment VIII may be very small and inconspicuous, possibly
non-functional (some species of Procaviphilus (Meganarionoides)). Not parasitic on Carni-
vora. Old and New World 3
3 Dorsal or ventral projection present on abdominal pleurum IV; mesomeral arch generally
produced basally between b.a.l.s.; female antennal flagellomeres generally not fused, or
only partially fused; parasitic on hyraxes and primates. Old and New World
DASYONYGINAE Keler, 1938 (p. 267)
- Pleurum IV lacking any projection; mesomeral arch rarely produced basally between b.a.l.s. ;
female flagellomeres generally fused or, if not, then female with long setal tufts on
abdominal pleura VIII and IX; not parasitic on hyraxes or primates 4
4 Posterior margin of temple generally produced, very convex; (Fig. 87) ; very long setae present
on at least pleurum VIII, sometimes also on pleura VII (male) or IX (female); basiparameral
sclerites present; mesomeral arch lacking extension if complete, otherwise tripartite, median
part sometimes obscure (Figs 91, 93); if mesomeral arch entire, male genitalia as in Fig. 89,
temple margins not greatly produced, and female with two flagellomeres; parasitic on New
World porcupines (Erethizontidae) EUTRICHOPHILINAE Keler, 1938 (p. 265)
- Temples not so developed; setae on pleurum VIII not exceptionally long; basiparameral
sclerite present or absent; mesomeral arch, if present, with or without extension, but arch
never tripartite; female flagellomeres fused; pseudostyli frequently present; parasitic on
Artiodactyla and Perissodactyla. Old and New World BOVICOLINAE Keler, 1938 (p. 247)
Key to genera and subgenera
1 No abdominal spiracles present ; maj ority of tergal and sternal setae at least two-thirds length of
segment or, if not, median setal group on tergum II comprising at least three setae (and,
frequently, median groups running together) . New World 2
- At least one pair of abdominal spiracles present or, if not, majority of abdominal sternal and
tergal setae less than two-thirds length of segment or median setal group on tergum II
comprising only one seta. Old and New World 8
2 Abdominal pleura II-IV with dorsal projections (Fig. 246), though most apparent in females
and sometimes very inconspicuous; male lacking tergocentral microsetae; latero-posterior
corner of temple margin frequently with single long fine seta or one or two shorter, stout
setae; female pedicel with dorsal membranous projection (Fig. 245) (sometimes obscure).
[Geomvidae](GOAfyDQEO75s.l.,p.328) 3
- Abdominal pleura lacking dorsal projections, or single membranous projection present on
pleurum IV only (Fig. 228); male with tergocentral microsetae on abdominal terga II-VI
(Fig. 227): long seta frequently present on temple margin but shorter stout seta not
developed; female pedicel lacking any projection. (NEOTRICHODECTESs.l. , p. 321) 4
3 Pleural projections on pleurum II sclerotised; temple margin with single stout seta and
associated smaller finer setae; male abdominal terga II and III with rows of enlarged setae
(Fig. 244); parameral plate with single apical point; male genitalia symmetric or asymmetric
GEOMYDOECUS (THOMOMYDOECUS)(p. 334)
Pleural projection on pleurum II unsclerotised or, if sclerotised, posterolateral temple margin
336 C. H. C. LYAL
with single long fine seta and associated smaller setae; temple margin with or without
specialised setae but not with single stout seta; male abdominal terga II and III rarely with
rows of specialised setae (G. copei only); parameral plate with single apical point or apically
bifid; male genitalia symmetric GEOMYDOECUS (GEOMYDOECUS)(p. 332)
4 Female subgenital lobe present, with serrate margins; female genital chamber with clear, flat
dorsal region but lacking single scattered spines; gonapophyses broad, membranous, lacking
lobe; parameral plate slender, with basal projection between b.a.l.s. (Fig. 238). [Mustelidae
and Procyonidae] NEOTRICHODECTES (NEOTRICHODECTES) (p. 326)
Female subgenital lobe absent or, if present, lobe with smooth margins and longitudinal setal
rows (Fig. 235); female genital chamber, if with clear flat dorsal area, then with single spines
scattered over it; gonapophyses not broad and membranous, frequently with lobe; paramer-
al plate broad, with very limited projection between b.a.l.s 5
5 Ventral vulval margin with lobulate process with smooth margins and longitudinal rows of
setae (Fig. 235) ; gonapophyses slender, sclerotised, lacking lobe ; parameral plate with single
apical point; mesomeral arch extension broad, clubbed (Fig. 240). [Mustelidae]
NEOTRICHODECTES (TRIGONODECTES)(p. 327)
- Ventral vulval margin convex, but not produced; gonapophyses not slender, lobe present;
parameral plate with apex bifid ; mesomeral arch with pointed extension 6
6 Large species, over 2-75 mm long; male flagellum with 7 articulated 'teeth'; female gonapo-
physes thick, with lobe and spur (Fig. 236); female pleurum VIII sclerotised; abdominal
setae relatively small, not attaining following setal row. [Bradypodidae]
NEOTRICHODECTES (LAKSHMINARAYANELLA)(p. 327)
Smaller species, under 2-25 mm long; male flagellum with 2 articulated 'teeth'; female
gonapophyses otherwise; female pleurum VIII not sclerotised; abdominal setae long,
attaining or nearly attaining setal bases of following setal row. [Carnivora] 7
7 Gonapophyses with flat lobe and small spur (Fig. 237); male mesomeral arch extension
attaining end of parameral plate (Fig. 239) [Procyonidae]
NEOTRICHODECTES (NASUICOLA)(p. 327)
- Gonapophyses with lobe comprising fused setal tubercles, long spur present (Fig. 233); male
mesomeral arch extension reaching beyond apex of parameral plate (Fig. 242). [Mustelidae]
NEOTRICHODECTES (CONEPATICOLA)(p. 328)
8 Five pairs of abdominal spiracles present; vulval margin sclerotised, with or without setal
tubercles, and meeting gonapophyses smoothly; subgenital lobe present; parameres not
fused to b.a.l.s.; mesomeres absent; postcoxale absent; abdominal segments II-V with
median setal group present, comprising at least three setae. [Mustelidae]
TRICHODECTES (PARATRICHODECTES)(p. 300)
Other than five pairs of abdominal spiracles present, though spiracles on segment VIII may be
very small, inconspicuous and possibly non-functional (some species of Procaviphilus
(Meganarionoides) as described in key couplet 26, and some species of Trichodectes
(Stachiella) , as described in key couplet 14) 9
9 Abdominal pleura V-VI (at least) lacking setae 10
Abdominal pleura III-VIII (at least) with posterior setal row and , sometimes , anterior setae ... 12
10 Abdominal tergal setae on segments I-VI less than half length of segment, shorter than
postero-lateral setae; pleura V-VI lacking setae (Fig. 159); male flagellum with two
basally-articulated 'teeth'; mesomeres present, unfused; parameres fused, with distinct
inturned apices arising from plate (Fig. 171); subgenital lobe bifurcate, with long basal
lateral processes (Fig. 149). [Mustelidae]
TRICHODECTES (TRICHODECTES) (in part) (p. 299)
At least some setae on abdominal terga I-III as long or longer than segment and postero-lateral
seta, and postero-lateral setae sometimes absent; pleura IV- VII (at least) lacking setae (Fig.
141); male flagellum lacking 'teeth'; mesomeres absent; parameres unfused or united at base
only; subgenital lobe not bifurcate or only slightly so; basal processes of subgenital lobe
absent or, if present, not long (Fig. 146). [Lutrinae] 11
11 Posterior setal row present on pleurum III; parameres slender, rod-like, fused basally (Figs
144, 145); subgenital lobe lacking basal lateral processes; gonapophyses lacking setal
tubercles (Fig. 143) LUTRIDIA(p. 288)
Posterior setal row not present on pleurum III; parameres broad, not fused to each other (Fig.
TRICHODECTID MAMMAL LICE 337
147); subgenital lobe with basal lateral processes; gonapophyses with setal tubercles
(Fig. 146) NEOLUTRIDIA(p. 289)
Ventral vulval margin meets gonapophyses smoothly, joined by sclerotised band; subgenital
lobe present, frequently with basal lateral processes; sternal setae on at least segments
III-VI attaining or nearly attaining base of following setal row; dorsum of head with setae
sparse; male scape expanded or, if not, parameres fused to b.a.l.s 13
Ventral vulval margin meets gonapophyses acutely or, if meeting smoothly, not joined by
sclerotised band; subgenital lobe present or absent, but, if present, never with basal lateral
processes (except Damalinia (Tricholipeurus) elongata; see Fig. 65); sternal setae on
segments III-VI not attaining base of following setal row, usually less than three-quarters
length of segment or, if longer, either female genitalia not as described and dorsum of head
with dense setal covering (Bovicola (Holakartikos) and B. (Spinibovicola)) or male scape
not expanded and parameres not fused to b.a.l.s 15
Pleurum IV with dorsal projection; anterior setae present on abdominal terga and sterna.
[Ursidae] WERNECKODECTES(p.29Q)
Pleurum IV without dorsal projection; anterior setae not present on abdominal terga and
sterna 14
Male abdominal terga II-IV (at least) with median setal group reduced to one seta (Fig. 164);
parameres fused to b.a.l.s. (Figs 172, 174) or characteristically asymmetric (Fig. 173); female
abdominal terga III- VII (at least) with median setal group reduced to one seta or absent.
[Mustelidae and Procyonidae] TRICHODECTES (STACHIELLA) (p. 301)
Tergal setae of both sexes more abundant, with at least two setae in median setal group;
parameres not fused to-b.a.l.s. [Canidae, Viverridae, Ursidae and Mustelidae]
TRICHODECTES (TRICHODECTES) (in part) (p. 299)
Posterior setal row of pleurum III with setae stouter than those of p.s.r. of pleurum V (Figs 182,
188) or , if not , species with four pairs of abdominal spiracles ; otherwise species with 0,1,2,3,
4 or 6 pairs of abdominal spiracles ; subgenital lobe present ; gonapophyses with lobe present 1 6
Posterior setal row of pleurum III with setae not stouter than those of p.s.r. of pleurum V; six
pairs of abdominal spiracles present , if gonapophyses with lobe , then subgenital lobe absent 19
Abdominal spiracles numbering 6, 4 or pairs; if no abdominal spiracles present, then female
with gonopore surrounded by spicular refringent patch, or gonapophysis lobe comprising
two fused tubercles, or antennal sensilla in pit with peripheral tongue-like projections;
female subgenital lobe frequently with overlapping scales or spines; male mesomeral arch
always present, with median extension or two apical nipples; mesomeres produced basally
between b.a.l.s. or, if not, antennal sensilla as described above; parameres usually broad,
contacting mesomeres only, not b.a.l.s.; male abdominal tergum II lacking specialised setae
of median group. (LORISICOLAs.L, p. 312) 17
Abdominal spiracles numbering 3, 2, 1 or pairs; if no abdominal spiracles present, then
female gonopore not surrounded by spicular refringent patch, gonapophysis lobe not
comprising two fused setal tubercles; antennal sensilla of male and female never in pit with
peripheral tongue-like projections; female subgenital lobe never with overlapping scales or
spines; male mesomeres fused, unfused or absent; if mesomeres fused, mesomeral arch
never with median extension or apical nipples; parameres frequently narrow, rod-like,
contacting mesomeres, b.a.l.s. or both; male abdominal tergum II frequently with long,
specialised setae (Figs 179, 180, 187-189). (FELICOLAs.L, p. 302) 18
Male antennal flagellum with 'teeth' on projection; six pairs of abdominal spiracles present or
abdominal spiracles absent, in which case antennal sensilla in pit with peripheral tongue-like
projections. [Felidae, Viverridae and Lorisidae] LORISICOLA (LORISICOLA) (p. 316)
Male antennal flagellum with 'teeth' not on projection, or, if projection present, mesomeral
arch lacking extension; four pairs of abdominal spiracles present or abdominal spiracles
absent, in which case male gonopore surrounded by spicular patch (Fig. 225). [Viverridae
andHerpestidae] LORISICOLA (PARADOXUROECUS)(p. 318)
Male antennal flagellum with one or three basally-articulated 'teeth', or 'teeth' absent, in which
case male abdominal tergum III with median setal group reduced to one seta of similar size to
those on tergum II, which are not greatly enlarged, and parameres not fused; female
gonapophysis with lobe and spur present, lobe rounded or rectangular and formed of fused
tubercles ; subgenital lobe bifid or not ; if subgenital lobe bifid , lobes pointed , rounded or with
338 C. H. C. LYAL
rectangularly obtuse posterior margins (Fig. 176); everted portion of male endophallus
never sclerotised; abdominal spiracles numbering 0, 1 or 3 pairs. [Herpestidae and Canidae]
FELICOLA (SURICATOECUS)(p. 312)
Male antennal flagellum with one, two, three or four nonarticulated 'teeth', or 'teeth' absent, in
which case male abdominal tergum III with median setal group reduced to one seta much
smaller than those on tergum II, which are greatly enlarged, and parameres fused, at least
basally; female with gonapophysis lobe rounded, with or without spur, but lobe never
rectangular; subgenital lobe bifid or not; if subgenital lobe bifid, lobes of various shapes, but
never with rectangularly obtuse posterior margins; everted portion of male endophallus
frequently thinly sclerotised (Figs 190, 196); abdominal spiracles numbering 0, 2 or 3 pairs.
[Herpestidae, Viverridae and Felidae] FELICOLA (FELICOLA)(p. 302)
19 Dorsal x>r ventral projection present on abdominal pleurum IV (Fig. 105); mesomeral arch
generally produced basally between b.a.l.s.; female antennal flagellomeres generally not
fused, or only partially fused. [Procaviidae and Primates] 20
Pleurum IV lacking any projection; mesomeral arch rarely produced basally between b.a.l.s. ;
female flagellomeres generally fused or, if not, then female with long setal tufts on
abdominal pleura VIII and IX (see couplet 27) ; not parasitic on hyraxes or primates 27
20 Sitophore sclerite modified , with posterior arms extended (Fig. 12) (sclerite difficult to see) 21
- Sitophore sclerite unmodified (Fig. 1 1) (not, generally, difficult to see) 23
21 Tarsal claws with ventral teeth or spines; temple margin with or without small rounded
projection; pleural projection on abdominal pleurum IV not elongate. [Procaviidae]
(DASYONYXs.L,p.279) 22
Tarsal claws lacking ventral teeth or spines; temple margin with long, broad, triangular
projection (Fig. 138); pleural projection on abdominal pleurum IV long (Fig. 136). [Pro-
caviidae] EURYTRICHODECTES(p.284)
22 Tarsal claws with sharp, fine spines (Fig. 14) DASYONYX (DASYONYX) (p. 282)
Tarsal claws with broad, saw-like teeth (Fig. 15) DASYONYX (NEODASYONYX)(p. 284)
23 Abdominal sternum II with broad, heavily-sclerotised band articulated with abdominal
pleurum II (Fig. 105); setal row of male scape comprising only two setae; basiparameral
sclerites present. [Procaviidae] (PROCAVICOLAs.L, p. 270) 24
Abdominal sternum II lacking sclerotised band or, if sclerotised band present, this is fused to
abdominal pleurum II or medially broken; setal row of male scape numbering more than two
setae; basiparameral sclerites absent or, if present, thoracic spiracle with tubular atrium and
female flagellomeres fused 25
24 Posterior angle of temple with small projection ; mesomeral arch with lateral double flexion and
median extension; endophallus with large hook-like spines (Fig. 107)
PROCAVICOLA (CONDYLOCEPHALUS)(p. 274)
Posterior angle of temple lacking projection; mesomeres unfused and lacking lateral double
flexion and median extension; endophallus lacking large hook-like spines (Fig. 109)
PROCAVICOLA (PROCAVICOLA)(p. 270)
25 Atrium of thoracic spiracle spherical; mesomeral arch with median extension and lateral
desclerotisations; gonapophyses with setal tubercles or, if not, postcoxale greatly developed
and fused to abdominal pleurum II. [Procaviidae and Cercopithecidae] (PROCAVIPHILUS
s.l.,p.274) 26
Atrium of thoracic spiracle tubular; mesomeral arch lacking median extension and not
desclerotised laterally; gonapophyses lacking setal tubercles; postcoxale not greatly de-
veloped and fused to abdominal pleurum II. [Cebidae] CEBIDICOLA (p. 267)
26 Parameres with basal flange, sometimes fused faintly; perisetal gap of male subgenital plate
absent; postcoxale not fused to abdominal pleurum II; setal tubercles of gonapophyses not
fused characteristically. [Procaviidae] PROCA VIPHILUS (PROCA VIPHILUS)(p. 278)
Parameres usually lacking basal flange; perisetal gap of male subgenital plate present or, if
absent, parameres fused together and articulated with mesomeral arch as in Fig. 120, and
mesomeral arch produced basally along b.a.l.s. (Fig. 121); postcoxale fused to abdominal
pleurum II, at least in females; setal tubercles of gonapophyses fused characteristically (Fig.
Ill) or, if not, ventral vulval margin as in Fig. 113. [Procaviidae and Cercopithecidae]
PROCAVIPHILUS (MEGANARIONOIDES)(p. 278)
TRICHODECTID MAMMAL LICE 339
27 Posterior margins of temple generally produced, very convex (Fig. 87) ; very long setae present
on at least pleurum VIII (Figs 88, 90), sometimes also on pleurum VII (males) or IX
(female); basiparameral sclerites present; mesomeral arch lacking extension if complete,
otherwise tripartite, median part sometimes obscure (Figs 101, 102); if mesomeral arch
entire, male genitalia as in Fig. 100, temples not greatly produced, and female with two
flagellomeres, otherwise female flagellomeres fused. [Erethizontidae]
EUTRICHOPHILUS(p. 265)
- Temples not so developed; setae on pleurum VIII not exceptionally long; basiparameral
sclerites present or absent; mesomeral arch, if present, with or without extension, but never
tripartite; female flagellomeres fused 28
28 Parameres narrow , rod-like and fused basally ; mesomeral arch with broad lobe-like extension ;
b.a.l.s. widely divergent anteriorly (Fig. 139); gonapophyses with setal tubercles; pseudo-
styli absent. [Protelidae and Hyaenidae] PROTELICOLA(p. 286)
- Parameres not narrow and fused basally or, if so, then b.a.l.s. not widely divergent anteriorly;
mesomeral arch without broad lobulate extension; gonapophyses lacking setal tubercles;
pseudostyli frequently present 29
29 Subgenital lobe present ; endophallus with dense patch of regularly-arranged spicules or , if not ,
parameral plate with single apex (Fig. 81); mesomeral arch entire, with abrupt bend to
enable bases to meet parameres (Figs 81, 82), or mesomeres unfused and b.a.l.s. with
anteposterior spur (Fig. 83); interior face of male flagellum serrate (Fig. 13); abdominal
sterna never with anterior setae; long, slender species. [Bovidae and Cervidae]
DAMALINIA (TRICHOLIPEURUS)(p. 264)
- Subgenital lobe absent or, if present, as small flap (Fig. 42) and species with anterior setae on
abdominal sterna (Bovicola jellisoni); sternum VII sometimes developed posteriorly into
two projecting spikes (Damalinia theileri, neotheileri and semitheileri, Fig. 73); endophallus
lacking spicular patch; parameres with apices free; mesomeres apically fused, unfused or
absent, but abrupt bend not present; b.a.l.s. lacking anteposterior spur; interior face of male
flagellum without serrations ; broader species 30
30 Dorsal face of vulva with pointed scales; gonapophyses hook-shaped (Fig. 66); common
oviduct, at branching point, with folded and more or less apparent collar, sometimes
partially sclerotised and refracting transmitted light; mesomeres unfused; abdominal
pleurum never extending ventrally onto abdominal sternum II; interior face of male
flagellum serrate. [Bovidae and Cervidae] DAMALINIA (CERVICOLA)(p. 263)
Dorsal face of vulva lacking pointed scales; gonapophyses not hook-shaped or, if they are, then
abdominal pleurum II extending onto sternum II (Fig. 68); common oviduct lacking 'collar'
as described above; mesomeres fused, unfused or absent; interior face of male flagellum with
or without serrations 31
31 Abdominal pleurum II with sclerite extending onto sternum II and occasionally tergum II,
sometimes at the expense of sternite or tergite (Fig. 68); mesomeres unfused, may be fused
to parameres and apparently absent; pseudostyli absent or, if present, broad or narrow (Figs
68, 71, 72); interior face of male flagellum serrate. [Bovidae]
DAMALINIA (DAMALINIA) (p. 260)
- Abdominal pleurum II not extending onto sternum II; mesomeres fused, unfused or absent;
pseudostyli , if present , not as figured above ; interior face of male flagellum lacking serrations 32
32 Atria of abdominal spiracles large, clear; mesomeral arch fused to b.a.l.s. ; parameres broad,
asymmetrically deflected (Fig. 64) ; thorax with setae sparse dorsally in female, but male with
median patch of setae on pro thorax; head elongate, trapezoid, with deep osculum present
(Fig. 63) [Tragulidae] TRAGULICOLA(p. 255)
- Atria of abdominal spiracles not large; mesomeral arch not fused to b.a.l.s. or, if it is,
parameres and mesomeres also fused (Fig. 60); parameres not broad or asymmetrically
deflected; thoracic setae less sparse but male thorax lacking central setal patch; head not
elongate but rounded , osculum absent or, if present, not deep (Fig. 3) 33
33 Parameres fused to mesomeres (Fig. 60); pseudostyli present, apically angular (Fig. 59);
gonapophyses broad, truncate (Fig. 58); osculum absent, but anterior margin of head slightly
flattened or concave medially, with hyaline border where pulvinus attains margin. [Bovidae]
BISONICOLA(p. 253)
- Parameres not fused to mesomeres; pseudostyli, if present, apically rounded; gonapophyses
340 C. H. C. LYAL
not broad, and with lobe variably apparent, or, if gonapophyses broad and truncate
(Werneckiella), then pulvinus not attaining anterior margin of head, which is smoothly
rounded and lacks a median hyaline border 34
34 Gonapophyses broad, truncate; mesomeres of characteristic pentagonal form (Fig. 62).
[EquidaeandBovidae] WERNECKIELLA (p. 255)
Gonapophyses with more or less discrete lobe (Figs 42, 43); mesomeres, if present, not
pentagonal; base of parameres frequently heavily block-like. [Bovidae, Cervidae and
Camelidae](JJOV7COLAs.l.,p.247) 35
35 Species with more or less dense covering of long setae; anterior setae present on abdominal
terga, sterna and pleura, slightly shorter than setae of posterior setal rows on these elements
(Fig. 47) ; gonapophyses with very limited lobe formation (Fig. 44) 36
- Species with shorter setae or, if setae long, then sparsely distributed and anterior setae not
present on abdominal terga and sterna; gonapophyses generally with more developed lobe
(Figs43,45) 37
36 Sitophore sclerite with posterior arms extended (Fig. 12); male with specialised setae on
abdominal tergum II (Fig. 41); mesomeres absent (Fig. 56); female lacks spinose patch on
postgenitalpleuralarea BOVICOLA (SPINIBOVICOLA)(p. 253)
- Sitophore sclerite with posterior arms not extended; male lacking specialised setae on
abdominal tergum II; mesomeres present, fused apically, with median extension (Fig. 57);
female with spinose setal patch on postgenital pleural area
BOVICOLA (HOLAKARTIKOS)(p. 251)
37 Setae on head long, fine; osculum absent; preantennal sulcus absent; gonapophyses with very
distinct lobe of characteristic form (Fig. 43); male genitalia with mesomeres not fused,
bipartite (Fig. 55) BOVICOLA (LEPIKENTRON)(p. 252)
- Setae on head not long and fine; osculum present or absent; preantennal sulcus present or
absent ; gonapophyses with less distinct lobe , of different form (Figs 42 , 45) ; male genitalia of
different form, mesomeres never bipartite BOVICOLA (BOVICOLA) (p. 251)
References
Bedford, G. A. H. 1929. Anoplura (Siphunculata and Mallophaga) from South African Hosts. Annual
Report of the Director of Veterinary Services, Union of South Africa 15: 501-549.
1932o. Trichodectidae (Mallophaga) found on African Carnivora. Parasitology 24: 350-364.
19326. Trichodectidae (Mallophaga) parasitic on Procaviidae. Proceedings of the Zoological Society
of London 3: 709-730.
1936. Notes on species of Trichodectidae with descriptions of new genera and species. Onderstepoort
Journal of Veterinary Science and Animal Industry 7: 33-58.
Blagoveshtchenski, D. I. 1956. Structure and taxonomic significance of the genitalia of Mallophaga. [In
Russian] Parazitologjcheskiy Sbornik Zoologicheskogo Instituta Akademiya Nauk. S.S.S.R. 16: 1-88.
Bouvier, G. 1945. De 1'hemophagie de quelques Mallophages des animaux domestiques. Schweizer Archiv
fur Tierheilkunde 87: 429-434.
Buckup, L. 1959. Der Kopf von Myrsidea cornicis (De Geer) (Mallophaga-Amblycera). Zoologische
Jahrbiicher. (Anatomic) 77: 241-288.
Burmeister, H. 1838. Handbuch der entomologie. 2 (2), Gymnognatha: 418-443.
Chapman, R. F. 1982. The insects. Structure and function, xiv + 919 pp. London, Sydney, Auckland,
Toronto.
Clay, T. 1946. Mallophaga miscellany. - No. 3. The Trabecula. Annals and Magazine of Natural History
(11) 13: 355-359.
1951. An introduction to a classification of the avian Ischnocera (Mallophaga): Part I. Transactions of
the Royal Entomological Society of London 102: 171-194.
1969. A key to the genera of Menoponidae (Amblycera: Mallophaga: Insecta). Bulletin of the British
Museum (Natural History) (Entomology) 24: 1-76.
1970. The Amblycera (Phthiraptera: Insecta). Bulletin of the British Museum (Natural History)
(Entomology) 25: 75-98.
Clay, T. & Hopkins, G. H. E. 1955. Notes on the Rudow Collection of Mallophaga at Hamburg.
Mitteilungen aus den Hamburgischen Zoologischen Museum und Institut 53: 49-73.
Conci, C. 1942. Diagnosi preliminari di tre nuovi generi e di una nuova specie di Trichodectinae. Bolletino
della Societa Entomologica Italiana 74: 140-142.
TRICHODECTID MAMMAL LICE 341
1946. Un nuovo genere di Trichodectidae Sudamericano. Bolletino delta Societa Entomologica
Italiana 76: 59.
Cracraft, J. 1974. Phylogenetic models and classification. Systematic Zoology 23: 71-90.
Cummings, B. 1913. On some points of the anatomy of the mouth-parts of the Mallophaga. Proceedings of
the Zoological Society of London 1913: 128-141.
Eichler, W. 1940. Notulae Mallophagologicae. I. Neue Gattungen und Subfamilien von Haarlingen.
Zoologischer Anzeiger 129: 158-162.
1941. Zur Klassifikation der Lauskerfe (Phthiraptera Haeckel: Rhynchophthirina, Mallophaga und
Anoplura). Archivfilr Naturgeschichte (N.F.) 10: 345-398.
1963. Phthiraptera. 1. Mallophaga. Bronns, Klassen und Ordnungen des Tierreichs 5 (3) 7: vii + 291
pp. Leipzig.
1982. Notulae Mallophagologicae. XIII. Goliathipon nov. gen. und weitere neue Taxa der Gattungs-
stufe. Deutsche Entomologische Zeitschrift 29: 81-88.
Emerson, K. C. 1965. A new species of Mallophaga from the Philippine Islands. Journal of the Kansas
Entomological Society 38: 68-69.
Emerson, K. C. & Price, R. D. 1979. Two new species ofBovicola (Mallophaga: Trichodectidae). Journal
of the Kansas Entomological Society 52: 747-750.
1980. A new species of Suricatoecus (Mallophaga: Trichodectidae) from the western cusimanse,
Crossarchus obscurus (Carnivora: Viverridae). Florida Entomologist 63: 505-508.
1981. A host-parasite list of the Mallophaga on mammals. Miscellaneous Publications of the
Entomological Society of America 12: 1-72.
1982. A new species of Bovicola (Mallophaga: Trichodectidae) from the Formosan Serow, Capricor-
nis crispus swinhoei (Artiodactyla: Bovidae). Pacific Insects 24: 186-188.
1983. A review of the Felicola felis complex (Mallophaga: Trichodectidae) found on New World cats
(Carnivora: Felidae). Proceedings of the Entomological Society of Washington 85: 1-9.
Ewing, H. E. 1929. A manual of external parasites, xiv + 225 pp. Springfield & Baltimore.
1936. The taxonomy of the mallophagan family Trichodectidae, with special reference to the New
World fauna. Journal of Par asitology 22: 233-246.
Fahrenholz, H. 1936. Zur Systematik der Anopluren. Zeitschrift fur Parasitenkunde 9: 50-57.
Ferris, G. F. 1929. Review: a manual of external parasites, by Henry Ellsworth Ewing. Entomologist's
News 40: 337-341.
1931. The louse of elephants. Haematomyzus elephantis Piaget (Mallophaga: Haematomyzidae).
Parasitology 23: 112-127.
1933. The mallophagan genus Trichophilopterus. Parasitology 25: 468-471.
1951. The sucking lice. Memoirs of the Pacific Coast Entomological Society 1: 320 pp.
Foerster, A. 1868. Synopsis der Familien und Gattungen der Ichneumonen. Verhandlungen des Naturhis-
torischen Vereins der Preussischen Rheinlande und Westfalens 25: 135-221.
Fresca, A. G. 1924. Mal6fagos del Museo de Madrid. III. Un Trichodectido nuevo. Boletin de la Real
Sociedad Espanola de Historia Natural 24: 75-78.
Gray, J. E. 1843. List of specimens of Mammalia in the collection of the British Museum, xxviii + 216 pp.
London.
Harrison, L. 1915. The respiratory system of Mallophaga. Parasitology 8: 101-127.
1916a. The genera and species of Mallophaga. Parasitology 9: 1-156.
19166. The relation of the phylogeny of the parasite to that of the host. Report of the British
Association for the Advancement of Science 85: 476-477.
Haub, F. 1967. Der Kopf von Pseudomenopon pilosum (Scopoli) (Mallophaga-Amblycera). Zoologische
Jahrbucher. (Anatomic) 84: 493-558.
1972. Das Cibarialsklerit der Mallophaga-Amblycera und der Mallophaga-Ischnocera (Insecta).
Zeitschrift filr Morphologic und Okologie der Tiere 73: 249-261.
1973. Das Cibarium der Mallophagen/Untersuchungen zur morphologischen Differenzierung.
Zoologische Jahrbucher. (Anatomic) 90: 483-525.
1977. Das Cibarialsklerit von Arten der Familie Colpocephalidae (Burmeister 1838) Mallophaga/
Amblycera. Untersuchungen zur Grossenund Formkonstanz funktionsbezogener Strukturen in der
Uberordnung Psocodea. Zoologische Jahrbucher. (Anatomic) 97: 294-322.
Hellenthal, R. A. & Price, R. D. 1976. Louse-host associations of Geomydoecus (Mallophaga: Trichodecti-
dae) with the yellow-faced pocket gopher, Pappageomys castanops (Rodentia: Geomyidae). Journal of
Medical Entomology 13: 331-336.
1980. A review of the Geomydoecus subcalifornicus complex (Mallophaga: Trichodectidae) from
Thomomys pocket gophers (Rodentia: Geomyidae), with a discussion of qualitative techniques and
342 C. H. C. LYAL
automated taxonomic procedures. Annals of the Entomological Society of America 73: 495-503.
1984. A new species of Thomomydoecus (Mallophaga: Trichodectidae) from Thomomys bottae
pocket gophers (Rodentia: Geomyidae). Journal of the Kansas Entomological Society 57: 231-236.
Hennig, W. 1966. Phylogenetic systematics. Translated by Davis, D. D. & Zangerl, R. 263 pp. Urbana,
Chicago and London.
Hopkins, G. H. E. 1941. New African Mallophaga. Journal of the Entomological Society of South Africa 4:
32-47.
1942. Notes on Trichodectidae. Revista Brasileira de Biologia 2: 439453.
1943. Notes on Trichodectidae (Mallophaga) 1. Revista Brasileira de Biologia 3: 11-28.
1949. The host-associations of the lice of mammals. Proceedings of the Zoological Society of London
119: 387-604.
1960. Notes on some Mallophaga from mammals. Bulletin of the British Museum (Natural History)
(Entomology) 10: 75-95.
Hopkins, G. H. E. & Clay, T. 1952. A check list of the genera and species of Mallophaga. 362 pp. London.
Imms, A. D. 1939. On the antennal musculature in insects and other Arthropods. Quarterly Journal of
Microscopical Science 81: 273-320.
Johnston, T. H. & Harrison, L. 1911. Notes on some mallophagan generic names. Proceedings of the
Linnean Society of New South Wales 36: 321-328.
Keler, S. von 1934. Mallophaga von Polen. Die Familie Trichodectidae. Bulletin del' Academic Polonaise
des Sciences et des Lettres (B) 2: 259-267.
1938a. Baustoffe zu einer Monographic der Mallophagen. I Teil: Uberfamilie der Trichodectoidea.
Nova Acta Leopoldina 5: 395-467.
19386. Uber einige Mallophagen aus Paraguay und Kamerun. Arbeiten iiber Morphologische u.
Taxonomische Entomologie 5: 228-241.
1939. Baustoffe zu einer Monographic der Mallophagen. II Teil: Uberfam. Nirmoidea (1). Nova Acta
Leopoldina (N.F.) 8: 1-254.
1944. Bestimmungstabelle der Uberfamilie Trichodectoidea (Mallophaga). Stettiner Entomolog-
ischer Zeitung 105: 167-191.
1957. Der Haarling der Wildkatze (Felicola hercynianus n. sp.). Mallophaga: Trichodectidae.
Deutsche Entomologische Zeitschrift 4: 172-178.
1966. Zur Mechanik der Nahrunsaufname bei Corrodentien. Zeitschrift fur Parasitenkunde 27: 64-79.
1969. Mallophaga (Federlinge und Haarlinge). Handbuch derZoologie. Berlin. 4 (2) 11/17: 1-72.
Kellogg, V. L. 1896. New Mallophaga, I, - with special reference to a collection made from maritime birds
of the bay of Monterey, California. Proceedings of the California Academy of (Natural) Sciences (2) 6:
31-168.
1908. Mallophaga. In Wytsman, P., Genera Insectorum 66: 1-87.
1910. Corrodentia: Mallophaga. Wissenschaftliche Ergebnisse der Schwedischen Zoologischen Ex-
pedition dem Kilimandjaro, dem Meru und dem Umgebenden Massaisteppen Deutsch-Ostafrikas 1905-
1906 15(4): 43-56.
1913. Distribution and species-forming of ecto-parasites. American Naturalist 47: 129-158.
1914. Ectoparasites of mammals. American Naturalist 48: 257-279.
Kim, K. C. & Ludwig, H. W. 1978. Phylogenetic relationships of parasitic Psocodea and taxonomic
position of the Anoplura. Annals of the Entomological Society of America 71: 910-922.
1982. Parallel evolution, cladistics, and classification of parasitic Psocodea. Annals of the Entomol-
ogical Society of America 75: 537-548.
Kuhn, H.-J. & Ludwig, H. W. 1964. Mallophaga on catarrhine monkeys: Colobus guereza, a natural host
of Procavicola colobi. Nature 203: 424425.
Lakshminarayana, K. V. 1976. Nomenclatural changes in Phthiraptera - some suggestions. Angewandte
Parasitologie 17: 160-167.
Ledger, J. A. 1980. The arthropod parasites of vertebrates in Africa south of the Sahara. Volume IV.
Phthiraptera (Insecta). Publications of the South African Institute for Medical Research no. 56: 1-327.
Lyal, C. H. C. 1983. Taxonomy, phylogeny and host relationships of the Trichodectidae (Phthiraptera:
Ischnocera). Ph.D. Thesis, University of London. 602 pp.
1985. Phylogeny and classification of the Psocodea, with particular reference to the lice (Insecta:
Phthiraptera). Systematic Entomology 10: 145-165.
in prep. [a]. Host relationships of the Trichodectidae (Phthiraptera: Ischnocera).
in prep. [b]. The external genitalia of Psocodea, with particular reference to the lice.
Matsuda, R. 1965. Morphology and evolution of the insect head. Memoirs. American Entomological
Institute 4: 1-334.
TRICHODECTID MAMMAL LICE 343
1970. Morphology and evolution of the insect thorax. Memoirs of the Entomological Society of
Canada 76: 1-431.
1976. Morphology and evolution of the insect abdomen - with special reference to developmental
patterns and their bearings upon systematics. viii + 532 pp. Oxford & New York.
Mayer, C. 1954. Vergleich Untersuchungen am Skeletl-Muskelsystem des Thorax der Mallophagen unter
Berucksichtigung des Nervensy stems. Zoologische Jahrbiicher. (Anatomic) 74: 1-188.
Mayr, E. 1969. Principles of systematic zoology, x + 428 pp. Bombay.
Mjoberg, E. 1910. Studien iiber Mallophagen und Anopluren. Arkivfor Zoologi 6 (13): 1-296.
Moreby, C. 1978. The biting louse genus Werneckiella (Phthiraptera: Trichodectidae) ectoparasitic on the
horse family Equidae (Mammalia: Perissodactyla). Journal of Natural History 12: 395-412.
Murray, M. D. 1957a. The distribution of the eggs of mammalian lice on their hosts. I. Description of the
oviposition behaviour. Australian Journal of Zoology 5: 13-18.
1957 b. The distribution of the eggs of mammalian lice on their hosts. II. Analysis of the oviposition
behaviour of Damalinia ovis (L.). Australian Journal of Zoology 5: 19-29.
Nelson, B. C. 1972. A revision of the New World species of Ricinus (Mallophaga) occurring on
Passeriformes (Aves). University of California Publications in Entomology 68: 1-175.
Nelson, G. 1972. Phylogenetic relationship and classification. Systematic Zoology 21: 227-231.
1974. Classification as an expression of phylogenetic relationship. Systematic Zoology 22: 344-359.
Nitzsch, C. L. 1818. Die Familien und Gattungen der Thierinsecten (Insecta Epizoica). Magazin der
Entomologie (Germar) 3: 261-316.
Price, R. D. 1974. Two new species of Geomydoecus from Costa Rica pocket gophers (Mallophaga:
Trichodectidae). Proceedings of the Entomological Society of Washington 76: 41-44.
1975. The Geomydoecus (Mallophaga: Trichodectidae) of the southeastern U.S.A. pocket gophers
(Rodentia: Geomyidae). Proceedings of the Entomological Society of Washington 77: 61-65.
Price, R. D. & Emerson, K. C. 1971. A revision of the genus Geomydoecus (Mallophaga: Trichodectidae)
of the New World pocket gophers (Rodentia: Geomyidae). Journal of Medical Entomology 8: 228-
257.
1972. A new subgenus and three new species of Geomydoecus (Mallophaga: Trichodectidae) from
Thomomys (Rodentia: Geomyidae). Journal of Medical Entomology 9: 463-467.
Price, R. D. & Hellenthal, R. A. 1975o. A reconsideration of Geomydoecus expansus (Duges) (Mallo-
phaga: Trichodectidae) from the yellow-faced pocket gopher (Rodentia: Geomyidae). Journal of the
Kansas Entomological Society 48: 3342.
19756. A review of the Geomydoecus texanus complex (Mallophaga: Trichodectidae) from Geomys
and Pappageomys (Rodentia: Geomyidae). Journal of Medical Entomology 12: 401^08.
1976. The Geomydoecus (Mallophaga: Trichodectidae) from the hispid pocket gopher (Rodentia:
Geomyidae). Journal of Medical Entomology 12: 695-700.
1979. A review of the Geomydoecus tolucae complex (Mallophaga: Trichodectidae) from Thomomys
(Rodentia: Geomyidae) based on qualitative and quantitative characters. Journal of Medical Ento-
mology 16: 265-274.
1980a. The Geomydoecus oregonus complex (Mallophaga: Trichodectidae) of the western United
States pocket gophers (Rodentia: Geomyidae). Proceedings of the Entomological Society ofWashington
82: 25-38.
19806. The Geomydoecus neocopei complex (Mallophaga: Trichodectidae) of the Thomomys
umbrinus pocket gophers (Rodentia: Geomyidae) of Mexico. Journal of the Kansas Entomological
Society 53: 567-580.
1980c. A review of the Geomydoecus minor complex (Mallophaga: Trichodectidae) from Thomomys
(Rodentia: Geomyidae). Journal of Medical Entomology 17: 298-313.
1981o. Taxonomy of the Geomydoecus umbrini complex (Mallophaga: Trichodectidae) from
Thomomys umbrinus (Rodentia: Geomyidae) in Mexico. Annals of the Entomological Society of
America 74: 37-47.
19816. A review of the Geomydoecus californicus complex (Mallophaga: Trichodectidae) from
Thomomys (Rodentia: Geomyidae). Journal of Medical Entomology 18: 1-23.
Price, R. D. & Timm, R. M. 1979. Description of the male of Geomydoecus scleritus (Mallophaga:
Trichodectidae) from the southeastern pocket gopher. Journal of the Georgia Entomological Society 14:
162-165.
Richards, O. W. & Davies, R. G. 1977. Imms' general textbook of entomology 10th Edition. 2 volumes, vii
+ 1354 pp. London.
Rudolph, D. 1982a. Occurrence, properties and biological implications of the active uptake of water
vapour from the atmosphere in Psocoptera. Journal of Insect Physiology 28: 111-121.
344 C. H. C. LYAL
1982ft. Site, process and mechanism of active uptake of water vapour from the atmosphere in the
Psocoptera. Journal of Insect Physiology 28: 205-212.
1983. The water-vapour uptake system of the Phthiraptera. Journal of Insect Physiology 29: 15-25.
Sikora, H. & Kichler, W. 1941. Uber Kopulationseigentumlichkeiten der Mallophagen. Zeitschrift fur
Morphologic und Okologie der Tiere 38: 80-84.
Simpson, G. G. 1961. Principles of animal taxonomy . 247 pp. New York.
Sneath, P. H. A. & Sokal, R. R. 1973. Numerical taxonomy - The principles and practice of numerical
classification, xc + 573 pp. San Francisco.
Snodgrass, R. E. 1935. Principles of insect morphology . 667 pp. New York.
Stobbe, R. 1913a. Mallophagen. 1. Beitrag: Neue Formen von Saugetieren (Trichophilopterus und
Eurytrichodectes nn. gg.). Entomologische Rundschau 30: 105-106, 111-112.
1913ft. Mallophagen. 3. Beitrag: Die Trichodectiden des Berliner Museums fur Naturkunde.
Sitzungsberichten der Gesellschaft naturforschen der Freunde, Berlin 8: 365-383.
Symmons, S. 1952. Comparative anatomy of the mallophagan head. Transactions of the Zoological Society
of London 27: 349-436.
Timm, R. M. & Price, R. D. 1979. A new species of Geomydoecus (Mallophaga: Trichodectidae) from the
Texas pocket gopher, Geomys personatus (Rodentia: Geomyidae) . Journal of the Kansas Entomological
Society 52: 264-268.
1980. The taxonomy of Geomydoecus (Mallophaga: Trichodectidae) from the Geomys bursarius
complex (Rodentia: Geomyidae). Journal of Medical Entomology 17: 126-145.
Toulechkoff, K. 1955. The Trichodectidae (Trichodectoidea, Mallophaga) ectoparasitic on mammals of
Bulgaria. [In Bulgarian.] Izvestiya na Zoologicheskiya Institut. Bulgarska Akademiya na Naukite. Sofiya
4-5: 423-434.
Weber, H. 1936. Copeognatha. In Schulz, P. (Ed.), Biologic der Tiere Deutschlands 39: 1-50.
Werneck, F. L. 1936. Contribuigao ao conhecimento dos Mallophagos encontrados nos mammifero
sul-americanos. Memorias do Instituto Oswaldo Cruz 31: 391-589.
1941. Os Mal6fagos dos Procaviideos. Memorias do Instituto Oswaldo Cruz 36: 445-576.
1946. Sobre alguns ma!6fagos de procaviideos. Revista Brasileira de Biologia 6: 85-97.
1948. Os Maldfagos de Mamiferos. Parte I: Amblycera e Ischnocera (Philopteridae e parte de
Trichodectidae) Edicao da Revista Brasileira de Biologia, 243 pp. Rio de Janeiro.
1950. Os Maldfagos de Mamiferos. Parte II: Ischnocera (continuacao de Trichodectidae) e Rhyn-
cophthirina. Edi^ao de Memorias do Instituto Oswaldo Cruz. 207 pp.
Wiley, E. O. 1979. An annotated Linnaean hierarchy, with comments on natural taxa and competing
systems. Systematic Zoology 28: 308-337.
1981. Phylogenetics. The theory and practice ofphylogeneticsystematics. xv + 439 pp. New York.
Zlotorzycka, J. 1972. Mallophaga. Goniodoidea i Trichodectoidea. Klucze do Oznaczania Owadow Polski
15 (3): 1-48.
Index
Synonyms are in italics; original generic combinations are indicated for identical specific epithets.
Acondylocephalus 278 angularis 332 baxi 262
actuosi 332 annectens 263 bedfordi, Dasyonyx 284
acuticeps321 antidorcus 265 bedfordi, Felicola 312
acutirostris 312 appendiculata 262 bedfordi, Procavicola 274
adenota 262 arizonae 328 bedfordi, Trichodectes 265
aepycerus 264 armatus 270 Bedfordia 306
affinis 271 aspidorhynchus 321 bengalensis 321
africanus, Parafelicola 321 aspilopyga 255 birneyi 334
africanus, Procavicola 279 asymmetricus 334 bison 254
albati 332 aurei 332 Bisonicola 253
albimarginata 265 Bovicola 247
alcorni332 baculatus279 Bovidoecus 251
alleni 332 baculus 247 bovis 251
alpinus 251 bajaiensis 332 braziliensis 317
americanus 317 bakeri 333 breviceps 253
angolensis 279 barbarae 327 brucei 273
bulleri 332
byersi 334
caffra 317
californicus 332
calogaleus 310
canis 300
capensis 284
caprae 251
Cebidicola 267
centralis 332
cercolabes 266
Cervicola 263
chapini 332
cherriei 332
chiapensis 332
chihuahuae 332
chilensis 328
chorleyi 262
clarkii 333
clausonae 332
clayi 265
cliftoni 332
colobi 279
comitans 266
concavifrons 251
Condylocephalus 274
conectens 265
Conepaticola 328
congoensis, Suricatoecus 310
congoensis, Trichodectes 279
cooleyi 312
copei 332
cordiceps 266
cornuta 265
coronadoi 332
costaricensis 332
crassipes 252
crenelata 262
crovelloi 332
cummingsi 328
cynictis 310
dakotensis 332
dalgleishi 332
Damalinia 257
dariensis 333
Dasyonyx 279
decipiens 312
dendrohyracis 284
diacanthus 284
dickermani 334
dimorpha 262
dissimilis 274
divaricatus 301
dorcephali 265
dubius 278
duchesnensis 333
Eichlerella 312
eichleri 273
elongata 265
emarginatus 273
emersoni, Geomydoecus 332
emersoni, Trichodectes 300
emeryi 301
INDEX
equi 255
cr mi niac 301
euarctidos 300
Eurytrichodectes 284
Eutrichophilus 265
ewingi 333
exiguus 266
exilis 289
expansus 333
extimi 333
extrarius 270
fahrenholzi, Suricatoecus 312
fahrehholzi, Tricholipeurus 262
fallax 301
Fastigatosculwn 306
Felicinia 306
Felicola 302
felis 317
fennecus 312
ferrisi, Procaviphilus 278
ferrisi, Trichodectes 291
forficula 263
fulva255
fulvescens 333
fulvi 333
furca 273
galictidis 300
Gatictobius 299
gastrodes 328
genettae 310
genowaysi 334
geomydis 333
Geomydoecus 328
grahamensis 332
granulatus 278
granuloides 278
greeri 334
Grisonia 299
guadalupensis 333
guineensis 284
guinlei 312
guyanensis 266
halli 333
harrisi 278
harrisoni 262
heaneyi 333
helogale 312
helogaloidis 312
hemitragi 253
hendrickxi 263
hercynianus 317
heterohyracis 273
hilli 262
hindei 278
hoffmanni 333
Holakartikos 251
hopkinsi, Damalinia 263
hopkinsi, Dasyonyx 284
hopkinsi, Felicola 312
hopkinsi, Procavicola 274
hueyi 333
hyaenae 288
hydropotis 263
345
idahoensis 333
illinoensis 333
inaequalis 312
indica 265
intermedius 288
interruptofasciatus 328
irolonis 333
jacobi 301
jaliscoensis 333
jamesbeeri 334
jellisoni251
johnhafneri 334
jonesi 333
jordani 279
juccii 321
kingi 301
kuntzi 300
Lakshminarayanella 327
larseni 301
laticeps 321
lenicornis 321
Lepikentron 252
lerouxi 263
liberiae 312
Iimbatus251
limitaris 333
lindfieldi 274
lineata 265
lipeuroides 265
lobatus 266
longiceps 265
Iongicornis251
lopesi 273
Lorisicola 312
lutrae 290
Lutridia 288
Lymeon 327
maai 263
machadoi 285
macrurus 312
major 279
malaysianus 317
markhafneri 334
martinaglia 263
martini 333
martis 302
matschiei 289
maximus 266
mcgregori 333
Meganarion 267
Meganarionoides 278
melis 300
mephitidis 326
merriami 333
mexicanus, Eutrichophilus 266
mexicanus, Geomydoecus 333
meyeri 263
minimus 3 12
minor, Dasyonyx 284
minor, Eutrichophilus 266
minor, Geomydoecus 334
minutus 326
346
C. H. C. LYAL
mjoebergi 317
mobilensis 333
mokeetsi 273
moojeni 266
moschatus 265
muesebecki 279
multilineatus 333
multispinosus 253
mungos 321
muntiacus 263
musculi 333
mustelae 301
nairobiensis 284
Nasuicola 327
natalensis, Damalinia 263
natalensis, Procavicola 273
nayaritensis 333
nebrathkensis 333
neglecta 255
neoafricanus 321
neocopei 334
Neodasyonyx 284
Neofelicola 318
neofelis 317
Neolutridia 289
neotheileri 262
Neotrichodectes 321
neumanni 279
occidentalis 312
ocellata, 255
octomaculatus 301
oculatus 284
oklahomensis 333
oreamnidis251
oregonus 333
orientalis 262
orizabae 334
ornata 262
osborni 326
ourebiae 265
ovalis, Dasyonyx 284
ovalis, Trichodectes 300
ovis 251
pakenhami 265
pallidus 327
panamensis 333
Paradoxuroecus 318
paradoxus 285
Parafelicola 318
paralaticeps 321
parallela 265
Paratrichodectes 300
parkeri 265
parvus 273
pattoni 333
pelea 262
peregrini 334
perotensis 333
philippinensis 321
pinguis 300
polydentatus 333
potteri 334
potus 302
Potusdia 301
pretoriensis 273
Procavicola 270
Procaviphilus 274
Protelicola 286
pygidialis 312
quadra ticeps 312
quadridentatus 333
rahmi 312
reduncae 264
retusus 302
Rhabdopedilon 251
robertsi, Felicola 312
robertsi, Trichodectes 278
rohani 312
ruficeps 284
salfii 302
scleritus 333
sclerotis 279
scutifer 279
sedecimdecembrii 254
semiarmatus 270
semistriatus 328
semitheileri 263
serraticus 279
setosus, Felicola 312
setosus, Trichodectes 266
setzeri 333
shastensis 333
shoanus 273
siamensis 317
sika 263
similis 317
sinaloae 333
smallwoodae 284
spenceri 317
spickai 333
Spinibovicola 253
spinifer 265
Stachiella 301
sternatus 273
subcalifornicus 333
subgeomydis 333
subnubili 333
subparvus 273
subrostratus 312
sudamericanus 317
sumatrensis 321
Suricatoecus 312
tamaulipensis 333
tarandi 251
tendeiroi 279
texanus 333
theileri 263
Thomomydoecus 334
thomomyus 333
thompsoni 263
thoracicus 326
thorntoni 273
tibialis251
tigris 247
timmi 334
tolteci 333
tolucae 333
trabeculae 264
traguli 257
Tragulicola 255
transvaalensis 284
traubi 333
Trichodectes 291
Tricholipeurus 264
trichopi 333
Trigonodectes 327
tropicalis 333
truncatus 333
ugandae 264
ugandensis, Dasyonyx 284
ugandensis, Procavicola 273
ugandensis, Trichodectes 300
umbrini 333
univirgatus 274
Ursodectes 299
ustulati 333
validus 284
veracruzensis 333
vicinus 273
victoriae 265
viverriculae 312
vosseleri 300
vulpis 312
wardi 334
warmanae 333
waterburgensis 284
welleri 333
wernecki, Felicola 321
wernecki, Geomydoecus 334
Werneckiella 255
Werneckodectes 290
williamsi 334
wolffhuegeli 326
yucatanensis 334
zacatecae 334
zebrae 255
zeylonicus 312
zorillae 300
zuluensis 255
British Museum (Natural History)
An introduction to the Ichneumonidae of Australia
/. D. Gauld
In the important field of biological and integrated control of pests the parasitic Hymenoptera
are of particular significance, and this work considers one of the largest families of Parasitica,
the Ichneumonidae. The group has received little attention in Australia - though it has already
been utilized successfully in curtailing the ravages caused by accidentally introduced pests. For
selective control programmes to be effective, however, a sound knowledge of the biology of
both the pest and its parasites is essential - and a sound taxonomic base is vital for the
development of such knowledge.
Ironically, considering the group's economic importance, the parasitic Hymenoptera is
amongst the least studied of any group of living organisms, and taxonomic difficulties have
presented major problems to many entomologists working with the Parasitica. An
Introduction to the Ichneumonidae of Australia will go a long way towards rectifying this
situation, being a taxonomic treatment, by genus, of the Australian ichneumonids, a
comprehensive illustrated identification guide, and a summary of all available information on
the group. It will also serve as an introduction to the biology and distribution of Australian
ichneumonids, and provide a check-list of the described species and an index to their known
hosts. It provides an important revision of ichneumonid nomenclature in order to bring the
group into line with the generally accepted principles of zoological nomenclature.
1984, 413pp, 3 maps, 580 figs. Paperback. 565 00896 X 40.00
Titles to be published in Volume 51
The ichneumon-fly genus Banchus (Hymenoptera) in the Old World
By M. G. Fitton
The phytogeny, classification and evolution of parasitic wasps of the subfamily Ophioninae
(Ichneumonidae)
By I. D.Gauld
A cladistic analysis and classification of trichodectid mammal lice (Phthiraptera: Ischnocera)
By C. H. C. Lyal
The British and some other European Eriococcidae (Homoptera: Coccoidea)
By D. J. Williams
Photoset by Rowland Phototypesetting Ltd, Bury St Edmunds, Suffolk
Printed in Great Britain by Henry Ling Ltd, Dorchester
Bulletin of the
British Museum (Natural History)
x %/ s
PRESENTED
The British and some other European
Eriococcidae (Homoptera: Coccoidea)
D. J. Williams
Entomology series
Vol51 No 4
28 November 1985
The Bulletin of the British Museum (Natural History) , instituted in 1949, is issued in four
scientific series, Botany, Entomology, Geology (incorporating Mineralogy) and Zoology,
and an Historical series.
Papers in the Bulletin are primarily the results of research carried out on the unique and
ever-growing collections of the Museum, both by the scientific staff of the Museum and by
specialists from elsewhere who make use of the Museum's resources. Many of the papers are
works of reference that will remain indispensable for years to come.
Parts are published at irregular intervals as they become ready, each is complete in itself,
available separately, and individually priced. Volumes contain about 300 pages and several
volumes may appear within a calendar year. Subscriptions may be placed for one or more of
the series on either an Annual or Per Volume basis. Prices vary according to the contents of
the individual parts. Orders and enquiries should be sent to:
Publications Sales,
British Museum (Natural History),
Cromwell Road,
London SW75BD,
England.
World List abbreviation: Bull. Br. Mus. nat. Hist. (Ent.)
Trustees of the British Museum (Natural History), 1985
The Entomology series is produced under the general editorship of the
Keeper of Entomology: Laurence A. Mound
Assistant Editor: W. Gerald Tremewan
ISBN 565 06014 4
ISSN 0524-6431 Entomology series
Vol51No4pp347-393
British Museum (Natural History)
Cromwell Road
London SW7 5BD Issued 28 November 1985
^v-,-", ,r, ...6^ (. !* -v^V
'
2 a MOV h
The British and some other European Eriococcidae
(Homoptera: Coccoidea)
D. J. Williams
Commonwealth Institute of Entomology, c/o British Museum (Natural History), Cromwell
Road, London SW7 5BD
Contents
Synopsis 347
Introduction 347
Acknowledgements and depositories 348
Morphology 348
Biology and economic importance 350
Eriococcidae Cockerell 351
Key to genera of British Eriococcidae 352
Cryptococcus Douglas 352
Eriococcus Targioni Tozzetti 356
Kuwanina Cockerell 384
Noteococcus Hoy 384
Ovaticoccus Kloet 384
Pseudochermes Nitsche 385
References 389
Index 393
Synopsis
Eleven of the 12 indigenous British Eriococcidae, Cryptococcus fagisuga Lindinger, Eriococcus devonien-
sis (Green), E. glyceriae Green, E. greeni Newstead, E. inermis Green, E. insignis Newstead, E. munroi
(Boratynski), E. placidus Green, E. pseudinsignis Green, E. spurius (Modeer) and Pseudochermes fraxini
(Kaltenbach), are redescribed; the twelfth species, E. cantium, is described as new. Four introduced
species, E. lagerstroemiae Kuwana, Kuwanina parva (Maskell), Noteococcus hoheriae (Maskell) and
Ovaticoccus agavium (Douglas), are discussed. Coccus buxi Fonscolombe, the type-species of Eriococcus
Targioni Tozzetti, is redescribed, as are Coccus spurius Modeer, Acanthococcus aceris Signoret and
Rhizococcus gnidii Signoret, the type-species of Gossyparia Signoret, Acanthococcus Signoret and
Rhizococcus Signoret respectively.
In agreement with some earlier works ? these three genera remain synonyms of Eriococcus but Greenisca
Borchsenius and Kaweckia Koteja & Zak-Ogaza are newly synonymised with Eriococcus. R. gnidii is
newly synonymised with E. thy mi (Schrank). An unnamed species of Eriococcus near E. buxi, occurring in
the U.S.S.R. and Turkey, is also discussed and lectotypes of the British species are designated, where
appropriate. There are sections also on morphology, biology and economic importance.
Introduction
The family Eriococcidae has been formerly associated with the Pseudococcidae, or mealybugs,
but the two are now considered to be distinct, and Afifi (1968) has given good characters to
separate the adult males of both families. Ferris (1957a, 1957ft) discussed numerous genera,
based on the female, that might be included in the Eriococcidae, and defined the family, mainly
on negative characters or without certain characters normally found in other families of scale
insects. Many genera that have been included in the family have never been studied critically and
their affinities may lie elsewhere.
In order to establish the true relationships and possible phylogeny of the Eriococcidae, Dr J.
M. Cox, British Museum (Natural History), and the present writer are currently studying the
group on a world basis, and it is hoped to publish on these aspects later. In the meantime this
Bull. Br. Mus. Nat. Hist. (Ent . ) 51 (4) : 347-393 Issued 28 November 1985
348 D. J. WILLIAMS
paper has been written to facilitate identification of the British species currently assigned to the
family, and to discuss some European genera, the type-species of which are not British although
such genera have been recorded from Britain.
Williams (1984) discussed briefly the distribution of the family and commented on its possible
origins in Gondwanaland. It is poorly represented in the tropics and much speciation has taken
place in the more temperate areas. The North American fauna was described by Ferris (1955)
(under the name Dactylopiidae) and by Miller & McKenzie (1967). Much of the Palaearctic
fauna has been discussed in recent years by Borchsenius (1949), Danzig (1962, 1980), Dziedicka
& Koteja (1971) and Tereznikova (1981). Some interesting species from South America have
been described by Miller & Gonzalez (1975) and a thorough study of the 75 New Zealand species
was made by Hoy (1962). The British species were listed by Boratynski & Williams (1964). Of
the 481 species included by Hoy (1963) in his catalogue of world species, almost 150 are known
from Australia. It is doubtful, however, if more than one or two Australian species can be
identified from the present literature and, because the Australian Region probably includes
some of the most interesting and bizarre species, it is clear that a definition of the family and its
limits can only be made after the Australian species have been studied adequately. Species from
the Oriental Region, although at present rather few, are virtually unrecognisable from the
literature and need revising.
Acknowledgements and depositories
The writer is much indebted to Dr A. Kaltenbach, Naturhistorisches Museum, Vienna, who has
kindly supplied for study specimens from V. Signoret's collection mentioned in the text. Mrs D.
Matile-Ferrero, Museum National d'Histoire Naturelle, Paris, has sent for study valuable
material of Rhizococcus gnidii and Eriococcus thy mi collected in France, for which the writer
extends his sincere thanks. Dr J. M. Cox, British Museum (Natural History), has given much
help in discussions during the preparation of the manuscript.
Most of the material studied is deposited in the British Museum (Natural History) (BMNH)
but other depositories of original material mentioned are the Imperial Agricultural Experiment
Station, Tokyo (IAES) and the New Zealand Arthropod Collection, Auckland (NZAC).
Morphology
Many of the characters of the North American species were discussed by Ferris (1955) and by
Miller & McKenzie (1967), resulting in definitions of the family for this area. The morphology
has also been discussed by Danzig (1980) and Hoy (1962) and a study of the labium was made by
Koteja (19740). Present studies of the British and other genera have revealed characters that
have not normally been used to separate species or genera and these characters are discussed
briefly here.
There is no generally accepted common name for the family but the names 'felted scale
insects' or 'felted coccids' have gained some acceptance and the former name could easily be
used. The name refers to the ovisac produced by most species from the dorsum and the ventral
margins. This ovisac, secreted from ducts, has a felted texture and completely covers the top of
the insect except for a hole at the posterior end, allowing first instars to escape. The felted sac
combines with mealy secretion on the venter, but on the dorsum rods of wax that seem to be
produced from the dorsal setae are also present. Many slide preparations show these rods
attached to the enlarged setae.
Body shape and segmentation. The shape of the body varies from globular to broadly-oval and
elongate, and although species may be found on leaves, stems and roots, the elongate species are
usually found on grasses. In many species the body is strongly nodulose and when this takes up
stain, it often masks minute characters. Numbering of the segments follows that given by Miller
(1984) for Eriococcidae and by Williams (19850) for Pseudococcidae. In this system the vulva is
situated between the 7th and 8th abdominal segments, so that the first abdominal segment on the
venter is represented by areas lateral to the 3rd coxae and the 2nd abdominal segment is
THE BRITISH AND SOME OTHER EUROPEAN ERIOCOCCIDAE 349
complete after the metathorax. On the dorsum the 1st abdominal segment is complete, but
because the segment containing the anal lobes is probably complex, it is here regarded as the 8th
or anal lobe segment.
Anal area. In many genera of Eriococcidae the posterior end of the body is produced into two
prominent anal lobes, each usually sclerotised and often conical with a long apical seta. Dorsally
there are frequently enlarged setae, varying in number, but in most species there are two such
setae on the inner edge and one on the outer edge, although this may be reduced to a much
smaller seta. Ventrally there are flagellate setae, often numbering 1-3, and near the inner corner
of each lobe there is a seta that Hoy (1962) named the suranal seta. This is usually slender but in
some genera it is flat or spatulate. In all the species discussed here the suranal setae are slender.
Sometimes the inner edges of the lobes are so strongly nodulose and sclerotised that they appear
to be toothed or serrate. When there are no definite anal lobes present the suranal setae are
represented by a pair just anterior to the anal ring on the venter. Species with anal lobes often
have a small dorsal lobe or plate projecting between the base of the lobes. It varies from being
only a slight prominence to triangular, well developed and nodulose. In some species it is heavily
sclerotised in the form of a definite plate, as in Eriococcus phyllanthi Ferris, and this led Ferris
(19576) to refer to it as a cauda. It is not certain whether this plate is homologous with the cauda
of the Aphidoidea and it is here called a dorsal plate no matter what the extent of its
development is. The anal ring, in species with well-developed anal lobes, is normally situated
vertically to the body between the dorsal plate and the apical margin of the venter between the
suranal setae. When prepared on microscope slides it often becomes distorted or is pushed
inwards, lying neither on the dorsum nor venter. Although the shape varies considerably, and
this is mentioned in the text when appropriate, a normal anal ring usually has 8 setae and a single
row of pores, except laterally, where there are a few pores forming a double row.
Antennae. These vary from being normal with 6 or 7 segments, to small stubs with 1 or 2
segments. A normal antenna appears to taper, with the terminal segment usually one of the
shortest and not wider than the previous segments.
Frontal lobes and frontal tubercles. Often just antero-medially to each basal antennal segment
there is a lobe-like structure that varies in size and length. They are much more pronounced in
distorted specimens when they can be observed to stick out from the surface. Normally they are
membranous but sometimes, as in Eriococcus buxi and its relative herein described, they tend to
be sclerotised. Their function is unknown. Signoret (1875) mentioned these lobes when
describing Acanthococcus aceris. Boratynski described them as eversible membranous tubercles
in A. munroi but there is no evidence that they are eversible. They were mentioned also by
Dziedicka & Koteja (1971) when describing Rhizococcus palustris. These structures seem to
have been ignored in most other works but Dr J. M. Cox and the present writer have observed
them in numerous species from all the zoogeographical regions. In at least one species from
Australia these lobes extend around and posterior to the basal antennal segment; for this reason
a more appropriate term would be frontal lobes and this term is used throughout this work.
Frontal tubercles are minute raised spots situated antero-medially to each antennal base.
They are difficult to see but once their normal position is located their presence or absence can
be noted. Again, their function is unknown but they seem to be homologous with the 'conical
disc pores' mentioned by Ben-Dov et al. (1975) in Protopulvinaria mangiferae (Green), and to
the 'preantennal spots' discussed by Ben-Dov (1979) in species of Kilifia. The species in these
genera belong to the family Coccidae and the presence of these tubercles may be much more
common throughout the group. They are certainly present in many species of Eriococcidae and
may have some taxonomic significance. In all species studied so far, frontal lobes and frontal
tubercles never occur together and one may have been derived from the other. Pseudochermes
fraxini, herein described, possesses frontal tubercles in all instars and this may be normal, but
some species have neither frontal lobes nor frontal tubercles.
Legs. Unlike species of Pseudococcidae, most of which have the tibia longer than the tarsus, the
tarsus in the Eriococcidae is often longer than the tibia. Also the trochanter pores in the
350 D. J. WILLIAMS
Eriococcidae form a line from the anterior lateral corner to about half-way along the posterior
edge, more or less dividing the trochanter into two equal parts. In the Pseudococcidae the
trochanter pores are situated towards the distal end.
Labium. The basal segment often has 2 pairs of setae but in some genera these are reduced to a
single pair. In Cryptococcusfagisuga, herein described, they are absent except in the first instar
when a single pair is present.
Setae. On the dorsum and ventral edges, there are often enlarged setae that are spine-like,
conical, pointed, blunt or truncate, but sometimes they are cylindrical, lanceolate or even
bulbous. Sometimes they are interspersed by quite slender setae. In some species, enlarged
setae are confined to the anal lobes or they are absent entirely, but usually the dorsal setae,
which may be minute, show some signs of being stiff and not flagellate. When the dorsal setae
tend to be small and slender in the adult, there are usually thicker setae in at least the first instar.
On the median part of the venter, what are here referred to as normal setae are flagellate setae
found in many groups of the Coccoidea. Laterally there are often other setae which, although
slender, are nevertheless stiff.
Macroducts. These have been discussed by Ferris (1955), Miller & McKenzie (1967), Miller
(1984) and by many other workers. They normally have the inner end reflexed in the form of a
cup and, although found in some other families, they represent one of the most important
characters of the Eriococcidae, although sometimes they are absent entirely.
Microducts. The significance of these minute ducts has been little understood but their structure
may indicate affinities and evolutionary paths. They have been discussed in some detail by Goux
(1948). Normally the inner end is bulbous, and the term ampulla is here adopted for it. It usually
bears a minute filament arising from the inner end. Between the ampulla and the external orifice
there is a tube varying from filamentous and slender to short and bulbous at the inner end, and
situated internally just next to the orifice there is a structure often wing-nut-shaped in profile but
which is here called a collar. Sometimes the orifice is bifid, as shown already by Miller &
Gonzalez (1975) in Eriococcus araucariae Maskell. The bifid orifice is external and may be seen
on microducts at the edge of the body where the bifid part protrudes from the surface of the
integument. They are not internal as shown by Hoy (1962) for E. araucariae. It is still not clear
whether the shape and type of the microduct has generic significance. Cryptococcus fagisuga
herein described possess microducts in the form of a double tube.
Enlarged ducts. These are illustrated here in E. buxi and E. sp. near buxi. They are usually larger
than the macroducts but differ in having a flat or slightly rounded inner end, instead of being
reflexed into a cup. Because Eriococcus possesses these ducts, Borchsenius (1949) separated this
genus from all others. The significance of these special ducts is still not clear because Miller &
Gonzalez (1975) discussed them in Exallococcus laureliae Miller & Gonzalez. Furthermore,
some Australian species currently assigned to Eriococcus also have them.
Pores. Although the disc pores are usually quinquelocular, there are often pores with up to 9
loculi. They are usually on the ventral surface but when present on the dorsum they have
sometimes 
