Content uploaded by Arnold H. Staniczek
Author content
All content in this area was uploaded by Arnold H. Staniczek on Dec 23, 2014
Content may be subject to copyright.
Revision of the giant pterygote insect Bojophlebia prokopi Kukalov
a-Peck, 1985
(Hydropalaeoptera: Bojophlebiidae) from the Carboniferous of the Czech
Republic, with the first cladistic analysis of fossil palaeopterous insects
Pavel Sroka
a
*, Arnold H. Staniczek
b
and G€
unter Bechly
c
a
Institute of Entomology, Biology Centre of the AS CR, Brani
sovsk
a 31, 37005
Cesk
e Bud
ejovice, Czech Republic;
b
Department of
Entomology, State Museum of Natural History, Rosenstein 1, 70191 Stuttgart, Germany;
c
Department of Palaeontology, State Museum
of Natural History, Rosenstein 1, 70191 Stuttgart, Germany
(Received 19 March 2014; accepted 14 October 2014)
The holotype is redescribed of the giant pterygote insect Bojophlebia prokopi Kukalov
a-Peck, 1985 from the
Pennsylvanian of the Czech Republic. Multiple errors in the original description are documented and corrected.
Bojophlebia prokopi has neither any visible traces of a costal brace nor an anal brace, but it does show triadic branchings
of MA, MP, CuA, and even, as rare a plesiomorphy, of CuP. It is therefore rejected as a fossil stem mayfly and attributed
as sister group of all other Hydropalaeoptera. The first cladistic analysis of fossil palaeopterous insects, including different
palaeodictyopterid groups, is presented. A revised phylogeny of Hydropalaeoptera and the stem line of Ephemeroptera are
suggested. Palaeodictyopterida is recognized as sister group of Neoptera; thus Palaeoptera s.l. is rejected as a paraphyletic
taxon. Four new higher taxa Paranotalia, Euhydropalaeoptera, Neopterygota and Litophlebioidea superfam. nov. are
introduced, as well as the new family Lithoneuridae.
http://zoobank.org/urn/urn:lsid:zoobank.org:pub:1B16D6AB-2B29-4891-A61D-A9212A85493
Keywords: Ephemerida; Palaeodictyopterida; Palaeoptera; Pterygota; Syntonopteroidea; phylogeny
Introduction
Bojophlebia prokopi Kukalov
a-Peck, 1985, from the
Pennsylvanian of the Czech Republic, was originally
described by Kukalov
a-Peck (1985), along with its
alleged larva. Both were attributed to Ephemerida (sensu
Kukalov
a-Peck 1985) as the only representatives of a
newly erected family Bojophlebiidae. The larval paratype
was later removed from Ephemerida, named Carbotri-
plura kukalovae Kluge, 1996, and attributed to Zygen-
toma (Kluge 1996). In a recent revision, C. kukalovae was
recognized as a putative sister group to Pterygota and
placed in a separate order, Carbotriplurida (Staniczek
et al. 2014).
Kukalov
a-Peck’s (1985) interpretation of the wing
venation of Bojophlebia prokopi was subsequently
adopted by several authors, who relied on the original
description and thus all placed Bojophlebia as the most
basal lineage within stem group mayflies (Rasnitsyn
2002; Grimaldi & Engel 2005; Willmann 2007a,b). More
recently, the description published by Kukalov
a-Peck
(1985) was reported to include several errors by authors
who had seen the holotype (Prokop et al. 2010; Staniczek
et al. 2011). This resulted in an uncertain phylogenetic
status for Bojophlebia, being considered as Pterygota
incertae sedis (Prokop et al. 2010) or Hydropalaeoptera
incertae sedis (Staniczek et al. 2011). However, despite
the fact that the interpretation of B. prokopi could be of
crucial importance to the phylogeny of basal pterygotes,
no detailed investigation of the holotype has been under-
taken since its original description.
The present study presents an in-depth re-examination
of the holotype to check the validity of characters stated
by Kukalov
a-Peck (1985) and to clarify the systematic
position of Bojophlebia prokopi within Pterygota. Based
on this reinvestigation, the first cladistic study of fossil
palaeopterous insects is here presented.
Material and methods
The holotype of Bojophlebia prokopi is housed in the
National Museum (N
arodn
ı muzeum), Prague, Czech
Republic, catalogue number 36338/1955. It was found in
tuffites of the middle Pennsylvanian/Moscovian (Silesian,
Westphalian C) near Vrapice, Czech Republic (Kukalov
a-
*Corresponding author. Email: pavel.sroka@centrum.cz
ÓThe Trustees of the Natural History Museum, London 2014. All Rights Reserved.
Journal of Systematic Palaeontology, 2014
http://dx.doi.org/10.1080/14772019.2014.987958
Peck 1985). The site of discovery was specified by
Kukalov
a-Peck (1985) as the President Anton
ın Z
apotock
y
Mine near Vrapice, in accordance with the original report
of its discovery (Z
azvorka 1956). Most probably it refers
to the coal mine shaft ‘Anton
ın IV’, also called ‘Prago IV’,
which is indeed located near Vrapice. This is part of a
larger mining facility that is known under different names:
originally called ‘Franti
sek Josef’, it became known as
‘Anton
ın Z
apotock
y’ in the 1950s, and later was also
named ‘Prago-Tragy’ (Roman H
ajek pers. comm.).
According to Prokop & Nel (2010), this deposit is the
so-called Whetstone Horizon from the continental basins
of the Bohemian Massif represents a peat mire ecosystem
with a shallow lake that was gradually filled by re-depos-
ited volcanic ashes from the Bolsovian of western Bohe-
mia (Czech Republic). The Whetstone Horizon has been
dated by Hess et al. (1985) with the Ar/Ar method as 309
§3.7 Ma.
The specimen was studied under an Olympus SZX 9
(Olympus Corporation, Tokyo) stereomicroscope and
photographed in different focal layers with a Nikon
60mm f/2.8 G ED AF-S Micro lens on a Nikon D800E
(Nikon Corporation, Tokyo) digital camera. It was
entirely covered with a layer of ethanol to achieve better
contrast of the fossilized structures. In Adobe Photoshop
version CS6 (Adobe Systems Incorporated, San Jose),
focus stacks of photographs were subsequently merged to
give compound images. These were sharpened and
adjusted in contrast and tonality.
Abbreviations for the wing veins used throughout the
text are as follows: C: costa; ScP: subcosta posterior; R:
radius; RA: radius anterior; RP: radius posterior; M:
media; MA: media anterior; MP: media posterior; Cu:
cubitus; CuA: cubitus anterior; CuP: cubitus posterior;
AA: analis anterior; I: intercalary vein. As the holotype of
Bojophlebia prokopi is preserved in ventral aspect, the
pleating is reversed and convex veins appear to be sunken
in folds, while concave veins appear to be on top of folds
on the photographs.
Systematic palaeontology
Class Insecta Linnaeus, 1758
Subclass Pterygota Lang, 1888
Infraclass Hydropalaeoptera Rohdendorf, 1969
Family Bojophlebiidae Kukalov
a-Peck, 1985
Genus Bojophlebia Kukalov
a-Peck, 1985
Bojophlebia prokopi Kukalov
a-Peck, 1985
(Figs 110)
Figure 1. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, overview (scale bar D4 cm). The elevations appear to be reversed due to
lighting from below, so that the legs only seem to be lower than the wing and ScP only seems to be concave (as in Figs 5, 6, 7 and 9A).
2 P. Sroka et al.
Material. The holotype and single known specimen is
rather poorly preserved, as already indicated by previous
authors who had briefly re-examined the fossil (Prokop
et al. 2010; Staniczek et al. 2011).
The fossil is visible from its ventral side, because legs
are superimposed on the wing veins, thoracic sternites are
clearly visible, and the corrugation of the wing veins is
reversed (concave veins are convex and vice versa).
Head, thorax, proximal part of abdomen, fragments of
legs, and basal parts of all four wings are preserved
(Fig. 1). For measurements see Table 1.
Description
Head. Prognathous. Antennae not preserved. Mouth parts
visible from ventral side (Fig. 2). The ventral closure of
the mouth parts in insects is formed by the labium, so the
most likely interpretation is given in Figure 2. The labium
is enlarged and obscures most of the remaining mouth
parts that are situated more dorsally. The two medioapi-
cal, paired, oval structures are interpreted here as large
glossae (gl); laterodorsally to these the smaller paraglos-
sae (pgl) are visible. The left paraglossa is only partly pre-
served. Posterior to glossae and paraglossae, a short
prementum (prm) and postmentum (pm) follow. On its
right side, the labium is somewhat distorted; in particular
the postmentum is medially cracked and, together with
the prementum, displaced and shifted anterolaterad. On
the right side, a three-segmented labial palp (plb) inserts
laterally at the prementum. The apical part of the corre-
sponding left labial palp is not preserved, only the first
segment is visible, but bent posteriorly. Basal part of right
maxilla preserved (ca, st). Other mouth parts not visible.
Thorax. Division into pro-, meso- and metathorax
clearly visible (Fig. 3). Prothorax smallest of all thoracic
segments, meso- and metathorax of approximately same
length. Prothoracic wings or paranota not present. The
thoracic sterna show multiple sclerites, for which we here
suggest a tentative homologization, based on a compari-
son with Recent mayflies (Tsui & Peters 1975).
Prothorax. With two sclerites that we interpret as basi-
and furcasternum (bs1, fs1).
Mesothorax. With several discernible sclerites. Antero-
laterally, the episternum extends to the basisternum (bs2).
The episternum is divided into an anterior anepisternum
(AES) and a posterior katepisternum (KES). The episterna
are medially separated by the basisternum (bs2).
Posteriorly to the basisternum, a large furcasternum (fs2)
is present. The lateral parts of the furcasternum are
bulged, forming a pair of furcasternal protuberances (fsp).
Laterally of the furcasternum, an epimeron (EM) is
visible.
Metathorax. With large, rectangular basisternum (bs3)
and furcasternal protuberances as in mesosternum. In
none of the three sterna could a separate spinasternum be
identified.
Legs. Long and slender; probably mainly used for cling-
ing to the vegetation (Fig. 1). For measurements see
Table 1.
In all visible legs, femora, tibiae, and tarsomeres are at
least partially preserved. Neither a separate patellar seg-
ment nor tarsal claws are visible in any of the preserved
legs (Fig. 4). A small number of scattered spines in a row
are discernable on the surface of femora and tibiae. Right
foreleg almost completely preserved except for tarsus and
Table 1. Measurements of Bojophlebia prokopi (holotype).
Measurement mm
Length of body
94.70
Length of head 14.78
Width of head 17.60
Length of labial palp 12.18
Length of prothorax 13.31
Width of prothorax 17.65
Length of mesothorax 19.64
Width of mesothorax 22.32
Length of metathorax 18.23
Width of metathorax 20.34
Length of right forewing
103.48
Width of right forewing 68.55
Length of left forewing
167.38
Width of left forewing 71.07
Length of right hind wing
69.38
Width of right hind wing
65.87
Length of left hind wing
168.12
Width of left hind wing 76.05
Length of right forefemur 21.39
Length of right foretibia 32.65
Length of right foretarsus
10.58
Length of left forefemur
11.74
Length of left foretibia
21.69
Length of left foretarsus
13.91
Length of right middle femur
10.53
Length of right middle tibia
37.06
Length of right middle tarsus
10.19
Length of left middle femur
9.97
Length of left middle tibia
25.70
Length of left middle tarsus
14.03
Length of right hind femur
22.82
Length of right hind tibia
21.74
Length of left hind tibia
9.30
Length of abdomen
30.75
Width of abdomen (segment I) 14.17
Preserved part.
Revision of the giant pterygote insect Bojophlebia prokopi 3
first two basal leg segments (coxa and trochanter). Region
of coxal insertion crushed, rounded sclerotized area may
represent either crushed coxa or coxal cavity; laterally to
it a deformed trochanter may be preserved (Fig. 3). Femur
clearly visible, forming an acute angle with long tibia.
First two tarsal segments and an isolated small tarsal frag-
ment preserved (Fig. 4E). Definite number of tarsal seg-
ments cannot be determined with certainty, but two tarsal
segments and the base of the third segment are clearly rec-
ognizable. However, the tarsus is not completely pre-
served. Tarsal claws not preserved.
Left foreleg only with distal part of femur, incomplete
tibia, and isolated tarsal fragment with two tarsomeres
preserved (Fig. 4F).
Right middle leg with fragmented coxa, distal part of
femur, probably almost complete tibia, and isolated frag-
ment with two partly preserved tarsomeres present
(Fig. 4G).
Left middle leg with incomplete coxa, distal part of
femur, proximal part of tibia, and isolated fragment with
two to three partly preserved tarsomeres (Fig. 4H).
Right hind leg only with distorted coxa, distal part of
femur, and proximal part of tibia visible.
Left hind leg almost completely missing except for pos-
sible coxa and isolated short tibial fragment.
Wings. Basal parts of right pair of wings and almost the
entire left pair of wings except for their apices are
Figure 2. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, head (scale bar D1 cm). A, photograph; B, line drawing.
4 P. Sroka et al.
preserved (Figs 1,5,6). Generally, most of the principal
venation pattern is preserved on the specimen, although
some veins are only fragmentary. Cross-veins are barely
visible except for the basal part of costal, subcostal, and
radial field in the left forewing (Fig. 7).
We frequently noticed red, green and yellow colour pig-
ments on the fossil (e.g. see Fig. 10), mostly along the main
wing veins, which appear to be remains of wax crayon
marks that had been applied directly onto the wing to mark
longitudinal veins, cross-veins and intercalary veins.
Left forewing (Figs 1,5,7,8).Oval in shape, basally
narrower than hind wing. Costal margin seems to be
slightly serrated. Costal brace (ScA sensu Kukalov
a-
Peck) absent. Costal field basally relatively broad. C pre-
served almost at entire length, except for some minor
interruptions, wing tip, and most basal part. ScP very thick
and long, approximately parallel to C and not shortened,
distally approaching C, probably ending near apex. In the
proximal half of preserved wing, RA and RP running very
close to each other with strongly pigmented wing mem-
brane preserved between these veins (Figs 5,7). Apical
part of wing in the area of RA and RP is very poorly pre-
served. Thus, it cannot be verified if RP is branched and if
IR1 and IR2 are present or not.
MA basally directly fused with RP. No basal vein con-
nection present between MA and MP (Fig. 8). At mid-
length of wing, the bifurcation of MA into MA1 and MA2
is not preserved, but apically two convex veins and one
concave vein can be distinguished that most probably rep-
resent the two branches of MA and in between one inter-
calary vein IMA.
MP basally approximated to RP, further distally MP
diverges from RP and is forked into MP1 and MP2,
Figure 3. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, body (scale bar D1 cm). A, photograph; B, line drawing.
Revision of the giant pterygote insect Bojophlebia prokopi 5
approximately at a length of 9.1 cm from the wing base.
One intercalary vein IMP present between MP1 and MP2.
An apparent vein closely parallel to MP2, which might be
misinterpreted as a branch of CuA, is actually only an
irregular crack that continues far beyond the area of the
wing even right into the hind wing.
CuA basally not preserved, only two isolated positive
veins visible between MP2 and CuP1 that most probably
represent CuA1 and CuA2. There is no indication whether
CuA is basally isolated, attached to MP, or attached to CuP.
The concave stem of CuP bifurcates basally into CuP1 and
CuP2. CuP2 is completely preserved, whereas CuP1 is only
visible as distal isolated vein fragment. Between CuP1 and
CuP2, an intercalary vein ICuP is preserved in the posterior
half of wing. Bases of R, M and Cu are very poorly pre-
served; possible fusions in that region are not excluded.
Anal veins mostly not preserved, only faint traces may
indicate the possible presence of three preserved anal
veins.
Left hind wing (Figs 1,6,9).Basally broader but not
much shorter than forewing. C preserved only as a small
fragment originating from wing base. C without row of
spines. Costal brace (ScA sensu Kukalov
a-Peck) not visi-
ble, most probably absent. ScP basally thickened, continu-
ing up to wing apex. RA and RP basally running parallel
and close to each other. RA not branched, entirely pre-
served except for a short distance in the basal third of
wing. RP furcated into RP1C2 and RP3C4 in the apical
third of wing, approximately 9.4 cm from wing base.
RP1C2 further divided into RP1 and RP2 subapically
from wing tip. Possible bifurcation of RP3C4 not visible,
as only a short fragment of this vein is preserved. No IR2
and IR1 visible between RP1C2 and RP3C4 and between
RP1 and RP2 respectively.
Basal stems of M, Cu and anal veins not preserved. MA
basally connected to RP in the basal third of preserved
wing. MA probably branched into MA1 and MA2 directly
below first RP furcation, but MA2 is only faintly pre-
served. MP branched into MP1C2 and MP3C4 at half
length of wing, about 7.4 cm from the wing base; MP3C4
posteriorly further divided into MP3 and MP4. Possible
bifurcation of MP1C2 not visible, as posterior part of
hind wing is missing.
Stem of Cu not preserved. CuA hardly discernible,
traces of two convex veins near the wing hind margin
between MP4 and CuP may represent CuA1 and CuA2. In
between these faintly preserved branches of CuA, a prom-
inent concave ICuA is present. Two further prominent
concave veins represent CuP1 and CuP2 that are basally
confluent, although an actual bifurcation is not preserved,
but can be inferred from the course of the two branches.
Large anal fan present with at least 15 alternating con-
vex and concave anal veins. Only posterior parts of these
veins well visible, their bases and branching pattern
mostly not preserved. Jugal veins are not to be expected,
as they only occur as synapomorphy in Phalloneoptera
(Paraneoptera CHolometabola) (Hamilton 1972), and
thus do not belong to the ground plan of Pterygota (contra
Kukalov
a-Peck 1983).
Figure 4. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, detailed photographs of legs (scale bars D1 cm). AD, femuro-tibial
joints of A, right foreleg, B, left foreleg, C, right middle leg and D, left middle leg. EH, preserved tarsomeres in E, right foreleg, F,
left foreleg, G, right middle leg and H, left middle leg.
6 P. Sroka et al.
Abdomen. For measurements see Table 1. Abdominal
segments IIV fully preserved, segment V distorted,
remaining segments almost entirely lost, only small frag-
ments visible (Figs 1,3). Genitalia and caudal filaments
not preserved.
Remarks on characters of Bojophlebia prokopi
mentioned in the original description
Prior to any consideration on the systematic placement of
Bojophlebia prokopi, it is necessary to reconsider and
discuss several of the characters described and depicted
by Kukalov
a-Peck (1985). Below we repeat in italics sev-
eral of Kukalov
a-Peck’s observations and interpretations
(as stated in her description and depicted in her fig. 2), fol-
lowed by our own observations and comments.
Head. “Head probably with relatively long and thick
antennae, large protruding eyes, and functional chewing
mouthparts” (Kukalov
a-Peck 1985, p. 936 and fig. 2, p.
935).
Figure 5. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, left forewing (scale bar D3 cm). A, photograph; B, line drawing; C,
detail of Kukalov
a-Peck (1985,fig. 2;Ó2008 Canadian Science Publishing or its licensors, reproduced with permission).
Revision of the giant pterygote insect Bojophlebia prokopi 7
Neither eyes nor antennae are visible or preserved in
the specimen. Kukalov
a-Peck (1985) herself stated in the
species description that these structures are actually not
preserved in Bojophlebia prokopi. In a footnote to the
generic description she revealed that she only assumed
the respective character states according to their (also
assumed?) occurrence in the putative related Syntonopter-
idae. However, in his revision of the family
Syntonopteridae Carpenter (1987, p. 385) mentioned that
“Antennae, mouthparts, legs, and cerci (are) unknown”.
Only a schematic drawing of the general body structure
of B. prokopi was provided by Kukalov
a-Peck (1985),
and we consider the head as drawn to be inaccurate, since
several important features are missing, such as the right
labial palp and the labium itself, which are clearly visible
in the fossil (see Fig. 2).
Figure 6. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, left hind wing (scale bar D3 cm). A, photograph; B, line drawing; C,
detail of Kukalov
a-Peck (1985,fig. 2;Ó2008 Canadian Science Publishing or its licensors, reproduced with permission).
8 P. Sroka et al.
Thorax. “prothoracic wings present as veined lateral
lobes” (Kukalov
a-Peck 1985, generic description, body of
adult, p. 936).
There are no lateral projections of the prothorax visible
(Figs 1,3). Kukalov
a-Peck herself did not even depict
‘prothoracic wings’ in her own figure (p. 935, fig. 2).
However, the thoracic sternal structures depicted there
were in general congruent with our own observations, but
she did not include the depicted characters in her text.
Legs. “legs probably long and comparatively stout;
patello-tibia longer and femur shorter than basitarsus,
tarsus and posttarsus combined; tarsus with four sub-
segments” (Kukalov
a-Peck 1985, generic description,
p. 936 and fig. 2, p. 935).
The presence of a separate patellar segment in Bojo-
phlebia prokopi is not substantiated (Fig. 4AD).
Kukalov
a-Peck only depicted a separate patella in the
right foreleg. A superficial view might indeed suggest a
short patellar segment (Fig. 4A). However, the corre-
sponding left foreleg (Fig. 4B), which is visible in ventral
aspect, does not support this assumption. If a true patella
were present, a respective segment border between patella
and tibia would also be visible from ventrally in the left
foreleg. Its absence leads us to assume that the alleged
tibio-patellar border in the right foreleg is instead the
result of a crack in the fossil rather than true segmentation.
In all other preserved legs there is also no indication for a
separate patella or even a superficial tibio-patellar suture
as present only in extant mayflies (see Kluge 2004).
Kukalov
a-Peck (1985) referred to the presence of a
‘posttarsus’ (generally referred to as pretarsus) but did not
mention the number of tarsal claws in the text. However,
two tarsal claws on the left foreleg are shown in her figure 2.
Rasnitsyn (2002) postulated the presence of single tarsal
claw in B. prokopi as a putative apomorphy of Bojophlebii-
dae. In fact, in none of the legs are tarsal claws preserved
(Fig. 4EH). In the right middle leg, the most distal pre-
served tarsal fragment resembles a claw in shape (Fig. 4G).
However, as this fragment clearly bears spines we can
definitely exclude its interpretation as claw.
Wings. “low arched and forked subcostal brace ScAC”
(Kukalov
a-Peck 1985, family diagnosis, p. 934);
“subcostal brace forming low broad arch, ScACforked,
ScA12 fusing with costal margin, ScA34 fusing with
ScP” (Kukalov
a-Peck 1985, generic description, p. 936).
Such a brace is actually visible neither in the bases of
forewings (Fig. 7) nor the hind wings (Fig. 9). Although
the posterior part of the costal field in the left forewing is
somewhat distorted, its anterior part is sufficiently pre-
served to suggest strongly that there is no costal brace
present at all. The costal field of the hind wing is very
small and contains no traces of an arched vein as
depicted by Kukalov
a-Peck. An alleged basal crossvein in
the costal field (Fig. 9A) is an artefact, which becomes
obvious under different illumination (Fig. 9B). Thus we
consider a costal brace in Bojophlebia prokopi to be
absent.
“MA vein ... is arched towards but not fused with the
RP vein” (Kukalov
a-Peck 1985, family diagnosis, p. 934;
generic description, p. 936).
The basal attachment of MA to RP is actually visible in
both left fore- and hind wings (Figs 8,10). MA
approaches RP at an angle that clearly implies a common
stem, and MA is not visible anywhere between basal RP
Figure 7. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, basal part of left forewing (scale bar D1 cm), photograph.
Revision of the giant pterygote insect Bojophlebia prokopi 9
and MP, even though the wing cuticle in that area is pre-
served (dark colour) and neighbouring veins (RP and MP)
are clearly visible. In our view, there is a direct connection
of both veins, which does not involve a strut as suggested
by Kukalov
a-Peck. Additionally, she depicted an oblique
veinal connection between the alleged strut and MP. How-
ever, this is also not visible in the specimen.
“CuA not fused to M but connected with M by a strut”
(Kukalov
a-Peck 1985, family diagnosis, p. 934).
The strut is actually not discernible on the specimen. In
the left forewing, the proximal part of CuA is not visible
at all; only distal parts can be tracked near the posterior
wing margin. Additionally, the entire area of the supposed
basal connection between CuP (CuP and AA1 in
Kukalov
a-Peck’s interpretation, but see below), CuA, and
M is obscured by the left foreleg (Fig. 7). Nevertheless,
CuP basally is bent just posterior of the foreleg fragment,
thus one might assume a connection to CuA (and possibly
to M) in that region. However, in all other wings, CuA is
also not preserved in its proximal part. Therefore, any
possible connection to M is only speculative.
At least for the hind wing, Kukalov
a-Peck (1985) indi-
cated basal parts of CuP, CuA and M by dashed lines
(including CuAM strut, see Fig. 6), which implies that
such a connection was only a presumption. But also the
entire hind wing CuA stem, bifurcation of CuA, and
Figure 8. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, photograph of left forewing (scale bar D1 cm). A, showing basal fusion
of MA and RP (indicated by an arrow); B, with overlay drawing of venation.
10 P. Sroka et al.
CuA2 as drawn by Kukalov
a-Peck (1985) are erroneous,
because the corresponding structures in the fossil clearly
are just patterns of cracks as is easily visible by their very
irregular course. Nearly identical cracks can also be
observed outside of the fossil.
“AA1 completely fused with CuP in both fore and hind
wings” (Kukalov
a-Peck 1985, family diagnosis, p. 934).
The wing sector around the veins CuP and AA1 is actu-
ally not preserved as depicted by Kukalov
a-Peck (1985).
Branching pattern and possible braces between the veins
posterior to CuP are not visible on any wing (see Figs 5,
6), and there is no evidence for a fusion of CuP and AA1.
Prokop et al. (2010) already doubted this fusion even on
the basis of the original figure of the wing venation.
“AA2 diverging from CuP in an arch” (Kukalov
a-Peck
1985, family diagnosis, p. 934).
The basal part of AA1 and AA2 is not preserved on any
wing of the holotype (Figs 7,9). Therefore, this character
Figure 9. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, basal part of left hind wing (scale bars D1 cm). A, B, photographs
under different illumination.
Revision of the giant pterygote insect Bojophlebia prokopi 11
cannot be verified. There is no evidence for an anal brace
in Bojophlebia prokopi, which was used by Kukalov
a-Peck
(1985), along with the costal brace, as important characters
for the attribution of this fossil taxon to Hydropalaeoptera
and Ephemerida. An anal brace is only developed in basal
Odonatoptera, but not in fossil or extant mayflies. Conse-
quently, an anal brace can no longer be considered as a
ground plan character of Hydropalaeoptera.
“richer branching of almost all veins” (Kukalov
a-Peck
1985, family diagnosis, p. 934).
There is no evidence for an unusually high degree of
vein branching in the preserved venational pattern of
Bojophlebia prokopi. The alleged repeated branching
pattern into four branches cannot be confirmed with the
possible exception of MP in the hind wing. There may be
more vein branches in the distal wing regions, but these
areas are almost not preserved at all.
“presence of irregular, weak, anastomosed crossveins
and archedictyon” (Kukalov
a-Peck 1985, family diagno-
sis, p. 934).
Cross-veins are actually barely visible. In basal fields
between C, Sc, RA and RP, an alternating pattern of some
stripes of lighter coloration is apparent, which might indi-
cate cross-veins. There is no trace of an archedictyon on
any wing. Irregular patterns that were apparently inter-
preted by Kukalov
a-Peck (1985) as archedictyon clearly
Figure 10. Bojophlebia prokopi Kukalov
a-Peck, 1985, holotype, photograph of left hind wing (scale bar D1 cm). A, showing basal
fusion of MA and RP (indicated by an arrow); B, with overlay drawing of venation.
12 P. Sroka et al.
represent structures of the rock matrix that are also visible
outside of the wing area.
Abdomen. “abdomen with two segments” (Kukalov
a-
Peck 1985, species description, p. 936, but see also fig. 2,
p. 935); “cerci and paracercus probably with short hair,
cerci shorter than paracercus” (Kukalov
a-Peck 1985,
generic description, p. 936).
There are not just two abdominal segments present as
described by Kukalov
a-Peck, but actually the first four
segments are complete and the fifth segment is partially
preserved. Shape and dimensions of the preserved abdom-
inal segments also differ strongly from the drawing of
Kukalov
a-Peck (1985), which appears to be very sche-
matic in this respect. The statement in the generic descrip-
tion regarding the terminal filaments again is a mere
assumption by Kukalov
a-Peck that is not supported by
any evidence (see Fig. 1).
Concluding remarks
It is obvious from the preceding paragraphs that Kukalov
a-
Peck’s original description suffers from erroneous observa-
tions and over-interpretations. This finding is reinforced by
the fact that other authors in their revisions of fossils
described by Kukalov
a-Peck found similar shortcomings
(Carpenter 1987; Rasnitsyn & Novokshonov 1997; Will-
mann 1999;B
ethoux & Briggs 2008;Staniczeket al.
2014). Several phylogenetically relevant character states
that were documented in the original description turned out
either not to be present in the actual fossil specimen at all,
or to have a different state than that described. Hence, the
conclusions suggested by Kukalov
a-Peck for the phyloge-
netic affinities of Bojophlebia prokopi and the evolution of
early mayflies are unsubstantiated.
Phylogenetic affinities of Bojophlebia prokopi
Hennigian phylogenetic systematics
Contrary to a common misconception, traditional Henni-
gian phylogenetic systematics does not involve a ‘manual’
but rather a ‘mental’ analysis of character patterns
(Bechly 2000). This includes a careful description and
comparison of the characters, as well as a weighting of
their phylogenetic relevance based on reasonable estima-
tions of their probability for homoplasy. Complex and
rare derived character states are weighted higher than sim-
ple and often independently evolved character states.
Like computer cladistics, Hennigian phylogenetic sys-
tematics relies on the principle of parsimony (‘Occam’s
razor’). However, it does not agree with computer cladis-
tics that the principle of parsimony is exhausted by the
procedure of minimizing the number of character transfor-
mations, but also considers the plausibility of the implied
homology hypotheses and evolutionary transitions. This
rejection of always preferring the shortest trees is right in
the spirit of Einstein’s famous quote ‘Everything should
be made as simple as possible, but not simpler’.
Character polarities are established by an a priori out-
group comparison (to be distinguished from a posteriori
outgroup rooting in computer cladistics), for which repre-
sentatives are preferably selected from several basal line-
ages of taxa that are certainly outside the group in study
but nevertheless closely related. This does not involve the
logical fallacy of circular reasoning, but an abductive
inference of the best explanation that was called ‘mutual
enlightenment’ by Willi Hennig and that resembles the
methodology of the hermeneutic spiral. Ground plans are
reconstructed by inferring all relevant plesiomorphic and
apomorphic character states for the most recent common
ancestor of a clade using the same procedure of outgroup
comparison.
Our study of Bojophlebia allows a revised diagnosis for
the family Bojophlebiidae: very large size with a wing
span of nearly 40 cm (autapomorphy); forewing with large
basal costal field (plesiomorphy); costal brace absent in
both wings (plesiomorphy); ScP very thick (autapomor-
phy); ScP and RA very long, nearly reaching apex (plesio-
morphy); RA and RP closely parallel, but remaining
separated basally (plesiomorphy); MA attached to RP with-
out anterior connection of MA to MP (autapomorphy, con-
vergent with Ephemerida and higher Odonatoptera); CuP
branched (plesiomorphy, except for a possible reversal in
Miracopteridae); MA, MP, CuA, CuP with intercalaries
developed as triadic branchings (synapomorphy with
Hydropalaeoptera, otherwise only present as a convergence
in the hind wing of the highly aberrant palaeodictyopteran
Lodetiella magnifica B
ethoux et al., 2007), thus as dichoto-
mous branching of a main longitudinal vein that is bracket-
ing a long and straight intercalary vein of opposite
corrugation; hind wing base broader than in forewing, with
large anal field (anal fan) composed of alternating concave
and convex veins (plesiomorphy); glossae developed as
enlarged plates (autapomorphy). The polarity of these char-
acters is based on an outgroup comparison with the ground
patterns as exhibited by the most basal lineages of Ephem-
erida, Odonatoptera, Palaeodictyopterida and Neoptera.
A remarkable feature of Bojophlebia is the presence of
a very thick ScP compared to RA. In most pterygote taxa,
RA is the thickest vein in the wing. This unique autapo-
morphy could suggest a different scheme of wing articula-
tion in Bojophlebia, which is unfortunately not preserved
in the single known fossil.
Even though the most distal parts of the wings are not pre-
served, the veins ScP and RA most probably reached close
to the wing apex. This character was mentioned by Kluge &
Sinitshenkova (2002) as one of the diagnostic characters of
Ephemerida, but it seems instead to be a symplesiomorphy
because it is also present in some basal Palaeodictyoptera
(e.g. Megaptilidae, Eugereonidae and Spilapteridae).
Revision of the giant pterygote insect Bojophlebia prokopi 13
Thus, most of the characters that can be reliably
observed on Bojophlebia are either unique autapomor-
phies, or plesiomorphies and homoplasies that are present
in various unrelated pterygote lineages. All of these char-
acters are not suitable when attempting to resolve the phy-
logenetic affinities of Bojophlebia.
However, several characters at least allow the exclusion
of some higher taxa. The presence of a branched MP in
both wings excludes a position within Odonatoptera, where
MP is always single (Bechly 1996). The absence of suck-
ing-piercing mouth parts excludes a position within Palaeo-
dictyopterida, where mouth parts always look very
different, even in species with a short haustellum. The
absence of the costal brace excludes a position in Panephe-
meroptera sensu Staniczek et al. (2011). Finally, the sepa-
rated origins of RA and RP exclude a position within
Neoptera. Of course, the absence of the diagnostic autapo-
morphies of the forementioned taxa would not exclude a
sister-group position, but only exclude an in-group position.
Due to the poor preservation and the scarcity of useful
characters, there is no general consensus in the available
literature concerning the systematic position of Bojophle-
bia: Kukalov
a (1985) considered it as an early stem group
mayfly, which was accepted by Grimaldi & Engel (2005),
Willmann (2007b), Klass (2007) and Engel et al. (2013).
Rasnitsyn (2002) classified Bojophlebia in the separate
order Syntonopterida as sister group of true mayflies. Hub-
bard (1987) and Kluge (2004) considered it as Syntonopter-
oidea incertae sedis, and Prokop et al. (2010)asa
Pterygota incertae sedis, while Staniczek et al. (2011)ten-
tatively considered it as a Hydropalaeoptera incertae sedis.
There is no doubt that Bojophlebia has some phenetic
similarity with other genera previously attributed to Syn-
tonopteroidea. However, to resolve finally the question of
its systematic position, two other issues must be addressed
first: (1) are the higher taxa Syntonopteridae, Sytonopter-
oidea and Syntonopterida monophyletic; and (2) what are
putative synapomorphies of the ‘syntonopterid’ genera
with other taxa?
Below we evaluate previous diagnoses of Syntonopteri-
dae given by different authors and provide a phylogenetic
assessment in square brackets for each character.
The diagnosis of Synotonopteridae according to Car-
penter (1987, p. 385) is: “Medium-sized to large insects
[uninformative]. Fore wing (incompletely known): costal
area relatively broad proximally, very narrow distally
[symplesiomorphy]; R without branches [symplesiomor-
phy]; stem of RS independent of stem R [symplesiomor-
phy]; stem of MA either close to RS or coalesced with it
for brief interval [symplesiomorphy, apomorphy of
Hydropalaeoptera]; RS, MA, MP, and CUA with interca-
lary, triad branches [symplesiomorphy, apomorphy for
Hydropalaeoptera]; cross veins numerous, but rarely
branched [symplesiomorphy]. Hind wing: broader than
fore wing basally [symplesiomorphy]; venation
essentially as in fore wing, but slightly modified by wing
shape [symplesiomorphy]. Body: little known [uninfor-
mative]. Pronotum apparently with small lateral lobes
[symplesiomorphy, only known in Lithoneura, but refuted
by Willmann 1999]. Antennae, mouthparts, legs, and cerci
unknown [uninformative].”
The main diagnostic characters of Syntonopteridae
according to Prokop et al. (2010) are: “wing corrugate
[symplesiomorphy]; no archedictyon but a simple and rel-
atively straight pattern of cross-veins between the longitu-
dinal veins [also present in Odonatoptera and some
Palaeodictyopterida (e.g. Breyeriidae)] [symplesiomor-
phy, apomorphy of Hydropalaeoptera]; CuP simple [sym-
plesiomorphy, apomorphy of Euhydropalaeoptera]; MA
with a strong anterior curve at its base, touching RP for a
short distance [only a synapomorphy of Lithoneura C
Anglolithoneura]; most anterior branch of AA with a
strong curve (‘zigzag’) and touching CuP in one point
[just a synapomorphy of Lithoneura CAnglolithoneura].”
Prokop & Nel (2011) featured an undescribed new
genus and species, and stated the diagnosis of Syntonop-
teridae as: “a strong corrugation of the main longitudinal
veins connected by mainly simple transverse crossveins
also present in Odonatoptera and some Palaeodictyopter-
ida (e.g., Breyeriidae) [symplesiomorphy], MA with a
strong anterior curve at its base, shortly connected with
RP distally [just a synapomorphy of Lithoneura,Angloli-
thoneura, and the undescribed genus and species]; CuA
with a short terminal twigging [symplesiomorphy] and
anal area with well-defined cell(s) [dubious character].”
Prokop et al. (2010) mentioned that the genera
“Syntonoptera,Lithoneura,andAnglolithoneura are similar
in wing morphology, and likely closely related as they also
share several apomorphies”. However, the four putative
synapomorphies mentioned in the three preceding para-
graphs (namely MA with a strong anterior curve at its
base, touching RP for a short distance; most anterior
branch of AA with a strong curve (‘zigzag’) and touching
CuP in one point; characteristic concave vein between the
twoconvexveinsAA1C2 and AA3C4;andconstrictionof
the area between the first and the second convex branches
of AA, with a strong brace between them) are only visible
in Lithoneura Carpenter, 1938 and Anglolithoneura Prokop
et al., 2010, while the relevant area of the wing is not pre-
served at all in Syntonoptera Handlirsch, 1911.
Consequently, there are no known synapomorphic char-
acters that could support the monophyly of Syntonopter-
ida and Syntonopteroidea. Bojophlebia and Syntonoptera
share only a general similarity in trivial or symplesiomor-
phic (e.g. triadic branchings) features with Lithoneura C
Anglolithoneura.
We therefore here establish a new family Lithoneuri-
dae fam. nov. (type genus: Lithoneura; main diagnostic
apomorphy: MA with conspicuous basal curve approxi-
mated or connected to RP) for Lithoneura and
14 P. Sroka et al.
Anglolithoneura, and the undescribed genus and species
featured by Prokop & Nel (2011). The family Miracopter-
idae has to be considered as closely related to Lithoneuri-
dae, based on the shared constriction and concave vein
between AA1C2 and AA3C4 (Prokop et al. 2010). We
restrict the family Syntonopteridae to its type genus
Syntonoptera.
The genus Gallolithoneura Garrouste et al., 2009 is
insufficiently preserved and is considered as a Hydropa-
laeoptera incertae sedis, which might belong to Syntonop-
teridae or Lithoneuridae, or even basal Ephemerida.
The genus Aedoeophasma Scudder, 1885, which was
attributed to Syntonopteridae by Demoulin (1954), was
considered by Prokop et al. (2010) as a Palaeoptera incer-
tae sedis. However, its wing venation was incorrectly
homologized by these authors in their figure 4. The side-
light in the photo shows the corrugation of the main wing
veins very well and proves that vein MA1C2 of Prokop
et al. (2010) is a branched RP3C4, IMA is an intercalary
between RP3 and RP4, MP1C2 is MA1, IMP is IMA,
MP3C4 is MA2, and the area CuA and CuP is unclear and
probably includes MP and IMP and or CuA. This new
homology also fits much better with the fact that the holo-
type fossil is an apical wing fragment. The available char-
acters allow an attribution of Aedoeophasma only to
Hydropalaeoptera incertae sedis.
The same holds for an unnamed new genus and species
of Syntonopteridae described by Prokop & Nel (2011)
from the Middle Permian of France, which in our cladistic
analyses is mostly resolved as sister group to Triblosobi-
dae CEphemerida.
Since a typical ephemerid costal brace is clearly present
in Lithoneura lameerei Carpenter, 1938 (the correspond-
ing wing base in L. mirifica Carpenter, 1944 and Angloli-
thoneura is not preserved), but other diagnostic
apomorphies of Ephemerida are absent, we suggest a
basal position of Lithoneuridae CMiracopteridae as sister
group to Ephemerida within Panephemeroptera.
The triadic branching (at least of MA, MP, CuA and
CuP) represents a unique and strong putative synapomor-
phy of all Hydropalaeoptera (including Bojophlebiidae,
Syntonopteridae, Miracopteridae, Lithoneuridae, Triplo-
sobidae, Ephemerida and Odonatoptera) because such a
branching pattern is known in all basal Ephemerida and
the most basal Odonatoptera (Eugeropteridae), while it is
always absent from Palaeodictyopterida (except the aber-
rant genus Lodetiella) and Neoptera. The alternative
assumption of a triadic branching pattern in the ground
plan of Pterygota would represent a less parsimonious
hypothesis.
The widely branched CuP of Bojophlebia is a striking
plesiomorphy among pterygote insects that is otherwise
only known from Miracopteron Novokshonov, 1993
(according to our cladistic study, the narrowly branched
CuP of Miracopteron most probably is a reversal), many
Palaeodictyopterida, and a few very ‘primitive’ extinct
Polyneoptera (‘Protorthoptera’). Therefore, this plesio-
morphy excludes a position in the clade that is formed by
all Hydropalaeoptera, which share an apomorphic
unbranched CuP (namely Triblosobidae, Syntonopteridae,
Lithoneuridae, Ephemerida and Odonatoptera). We pro-
pose the new taxon Euhydropalaeoptera for this clade,
and classify Bojophlebia as its sister group in a monotypic
family Bojophlebiidae.
Syntonopteridae sensu stricto would have to be consid-
ered as belonging to Euhydropalaeoptera incertae sedis
because it shares the unbranched CuP as a synapomorphy,
but the triadic branching is symplesiomorphic, and a cos-
tal brace is not preserved, so that a closer relationship
with Ephemerida cannot be documented. However, in our
cladistic analyses Syntonopteridae resolved between Mir-
acopteridae and Lithoneuridae as sister group of the latter,
so that Syntonopteroidea could be monophyletic after all.
Apart from Syntonopteridae and Odonatoptera, other
clades within Hydropalaeoptera that seem to lack a costal
brace are represented by the Carboniferous family Triplo-
sobidae, which was retransferred from Palaeodictyopter-
ida to Hydropalaeoptera by Staniczek et al. (2011), and
the Triassic mayfly families Litophlebiidae and
Tintorinidae.
Affinities of Litophlebia Hubbard & Riek, 1977 and
Tintorina Krzeminski & Lombardo, 2001 to Bojophlebia
are highly unlikely because they are very dissimilar with
no known synapomorphies, and Tintorinidae differs in the
autapomorphic shortening of vein ScP. Based on several
derived similarities in the wing venation, we concur with
Krzeminski & Lombardo (2001) that Tintorinidae and
Litophlebiidae are sister groups within the narrow-winged
Permoplectoptera, and thus consider their shared absence
of a costal brace as an apomorphic reversal within Ephem-
erida. This hypothesis is also strongly supported by our
cladistic analysis and we suggest uniting the two families
in a new superfamily Litophlebioidea superfam. nov.
Triplosobidae are characterized by a very strong costal
margin constituted by two veins fused or connected (prob-
ably C and ScA), with a series of small spines along the
anterior margin of both of them (Prokop & Nel 2009).
Such spines are also present in Permian stem mayflies
(Protereismatidae and Misthodotidae according to Will-
mann 2007b and Bechly unpublished). The same charac-
ter state was reported for Odonatoptera by Bechly (1996)
and for Diaphanoptera Brongniart, 1893 (Palaeodictyop-
terida) by B
ethoux & Nel (2003), where these two veins
diverged in the proximal part. The hypothetical presence
of such an arrangement of the costal margin in Bojophle-
bia could eventually explain the abrupt step in the costal
margin of the left forewing of the holotype. However, we
have no direct proof of the presence of two veins in
the costal margin of Bojophlebia. The wing venation of
Triplosoba Handlirsch, 1906 strongly differs from
Revision of the giant pterygote insect Bojophlebia prokopi 15
Bojophlebia in the arrangement of MA and RP in the hind
wing, these two veins being widely separated in Triplo-
soba whereas in Bojophlebia they are connected. Bojo-
phlebia also differs from Triplosoba in the shape of the
hind wing (with a more developed anal area in Bojophle-
bia), and the branched CuP in both wings. Many addi-
tional characters cannot be compared since crucial wing
regions (mainly stems of M, R and Cu) are too poorly pre-
served in Bojophlebia.
The Carboniferous genus Thesoneura Carpenter, 1944,
which was considered as Hydropalaeoptera incertae sedis
by Willmann (1999), was retransferred to Palaeodictyop-
tera by Staniczek et al. (2011). This position is also
strongly corroborated by our cladistic analysis.
The RA and RP veins remain separated until the wing
base in Bojophlebiidae and Eugeropteridae (Odonatop-
tera). These two veins are developed as a double-barrelled
structure up to the wing base in Lithoneuridae and Odona-
toptera (except Eugeropteridae), and are basally fused
only in Triplosobida and Ephemerida. This suggests that
Hydropalaeoptera is the sister group of all other pterygote
insects because Palaeodictyopterida and Neoptera share a
reduction of the terminal filament (epiproct) and a long
common stem of RA and RP, which only separate in the
distal half of the wing. Willmann (1999) earlier suggested
these putative synapomorphies. We propose redefining
Neopterygota Crampton, 1924 to accommodate the mono-
phylum Palaeodictyopterida CNeoptera. Consequently,
the taxon Palaeoptera s.l. is rejected by us as a paraphy-
letic group.
All of these phylogenetic inferences have been tested
through an extensive cladistic analysis that basically con-
firmed the results of our Hennigian study. Thus, we here
present the first profound phylogenetic tree of fossil palae-
opterous insects (Fig. 11), including all putative stem
group mayflies as well as several subgroups of Palaeodic-
tyoptera and the other three orders of Palaeodictyopterida,
which can serve as a basis for further phylogenetic studies
and the development of evolutionary scenarios for basal
pterygote insects.
Computer-assisted cladistic analysis
For the cladistic study two data matrices were assembled
with the Nexus Data Editor NDE 0.5.0 (http://taxonomy.
zoology.gla.ac.uk/rod/NDE/nde.html). The first data matrix
included 34 higher taxa (and a hypothetical ur-pterygote)
and 66 characters, which were coded as assumed ground
plan states for these higher taxa (inferred with the tradi-
tional Hennigian method described above). The use of
reconstructed ground plans of higher taxa allows a better
polarization of the characters, because obvious autapo-
morphies and reductions of individual species can be
eliminated, by comparison of representatives of different
species that are close to the base of the taxon. Further-
more, the data matrix can be more complete because
fragmentary data from different species can be united.
The two original Nexus data matrices (including detailed
descriptions of all characters, their states, and comments
about the selection of taxa etc.), two data matrices in TNT
(tree analysis using new technology) format, and three
trees as screen shots in the lossless portable network
graphics (PNG) format are available in the Online Supple-
mentary Material.
Theseconddatamatrixincluded35species(32fossil
species, two Recent mayfly species, and a hypothetical
pterygote ancestor) and 66 characters. The most completely
preserved species were selected for each fossil taxon, and
missing character states were partly supplemented by infor-
mation from congeneric or confamilial species, as docu-
mented in the Supplemental Material. The use of species
as terminal taxa has the advantage of greater objectivity as
it excludes potential bias and errors in reconstructed ground
plans. The hypothetical pterygote ancestor (hypanc) (com-
pare Bryant 1997) was coded based on a conceptual recon-
struction of the ground plan by including either those states
that are common in the main subclades (Palaeodictyopter-
ida, Neoptera, Ephemeroptera and Odonata), or states that
are obviously plesiomorphic based on general evolutionary
considerations (e.g. a smoothly curved course of the main
longitudinal veins with branching is certainly plesiomor-
phic compared to sharply bent or kinked courses or simple
veins without branching, or unfused veins are plesiomor-
phic compared to fused ones).
Even though the enigmatic Palaeozoic family Paoliidae
has previously been suggested to represent the stem group
of all other pterygote insects (Prokop et al. 2012), we did
not include it in our study because the most recent phylo-
genetic re-evaluation of this family by Prokop et al.
(2014) revealed that it is the sister group of Dictyoptera.
Furthermore, no species of paoliids are documented by
sufficiently well preserved fossils that allow for a reason-
ably complete coding of the characters as many crucial
character states are unknown (mouth parts, tarsi, genitalia,
etc.).
Both data matrices were converted from a NEXUS to a
TNT format with Mesquite v. 2.75 (Maddison & Maddi-
son 2011). The cladistic analysis was performed with
TNT v. 1.1 (Goloboff et al. 2008). The first data matrix
was analysed with the New Technology search option and
the standard parameters for Sect. Search/Ratchet/Drift/
Tree-fusing, which resulted in seven shortest trees of
length 192 (consistency index CI D0.474, retention index
RI D0.731) from which a majority rule consensus tree
(cut 50) was calculated. The second data matrix was ana-
lysed with the standard options and parameters of the
New Technology search and resulted in a single shortest
tree of length 214 (CI D0.430, RI D0.711), which
formed the basis for the phylogenetic tree shown in
Figure 11, which mostly agrees with the results of the
Hennigian phylogenetic analysis described above. An
alternative search with the Traditional search algorithm
16 P. Sroka et al.
resulted in 50 shortest trees of length 214 (CI D0.430, RI
D0.711), from which a majority rule consensus (cut 50)
was calculated that was less resolved but did not conflict
significantly with the first analysis. In all analyses the sil-
verfish Tricholepidion gertschi was used as outgroup, and
the rooting was improved by including Carbotriplura
kukalovae, the putative sister group of winged insects
(Staniczek et al. 2014), and a hypothetical pterygote
ancestor (ground plan states again inferred with the Hen-
nigian method described above). Removal of the hypo-
thetical pterygote ancestor resulted in a less resolved
consensus tree but again not in significantly conflicting
Figure 11. Phylogenetic tree of palaeopterous insects. Based on cladistic study for which the two data matrices and three trees are avail-
able in the Supplemental Material.
Revision of the giant pterygote insect Bojophlebia prokopi 17
topologies. No characters were weighted and no con-
straints were imposed in any of the cladistic analyses.
The main differences with our Hennigian analysis
above are the recovery of Syntonopteridae as the sister
group of Lithoneuridae, and Gallolithoneura as the sister
group of Odonatoptera (represented by Eugeropteridae),
which are both only very weakly supported by homoplas-
tic characters.
The relatively low consistency index of all our cladistic
analyses is due to homoplasy mainly in the wing venatio-
nal characters and numerous unknown character states in
the fossil taxa, which makes the tree topology sensitive to
changes of included taxa and characters. This explains the
few differences in tree topology of the two datasets.
Unfortunately, this is a general problem that will persist
because stronger character sets such as internal morphol-
ogy or genetic data will remain unavailable for fossil taxa,
which are mostly documented by, often fragmentary and
isolated, wings. However, the CIs of both cladograms is
far from being bad; Klassen et al. (1991, fig. 6) demon-
strated that the statistical support of the CI value strongly
correlates with the number of taxa, so that a CI of 0.43 is
highly statistically significant for a data matrix with 35
taxa, while it would be low for a matrix with much fewer
than 10 taxa.
List of main synapomorphies for the most impor-
tant nodes
Generally, only characters from the data matrix are
included, even though many more characters support
recent crown groups such as Pterygota, Neoptera, Odona-
toptera and Ephemeroptera. For the four newly proposed
taxa (Paranotalia, Neopterygota, Euhydropalaeoptera and
Litophlebioidea) we include other known evidence that
supports their monophyly. These nodes are named
because each reflects an important step in the early evolu-
tion of pterygote insects. Furthermore, they are supported
by several synapomorphies without convincing conflicting
evidence.
Paranotalia taxon nov. Laterally expanded paraterga on
thorax and abdomen. Further putative synapomorphies
were suggested by Staniczek et al. (2014): meso- and
metathorax enlarged; abdominal coxopodites reduced;
long legs with elongated femora, tibiae, and tarsi; large
compound eyes (maybe rather a symplesiomorphy).
Pterygota. Articulated wings on meso- and metathorax.
Neopteragota sens. nov. Larval and adult terminal filum
reduced; long common stem of RA and RP.
Palaeodictyopterida. Labium and labrum form beak
with stylet-like hypopharynx, mandibles, and maxillae;
labial palps reduced (one-segmented or absent); adult
cerci much longer than body.
Megasecoptera CPermothemistida CDiaphanop-
teroidea. Forewing costal field not very large and broad;
archaedictyon reduced.
Permothemistida CDiaphanopteroidea. Only three
tarsomeres; undulation of ScP and RA near wing base;
cross venation strongly reduced.
Neoptera. Subimaginal stages reduced; neopterous wing
folding with second muscle on third axillary sclerite; cos-
tal spines reduced; RA with oblique veinlets near wing
apex.
Hydropalaeoptera. MA approximated or fused to RP;
RP, MA, MP, CuA, and CuP with triadic branchings (con-
vergent to Lodetiella), incl. IR1 and IR2; pectinate
branching of RP1 reduced; archaedictyon reduced.
Bojophlebiidae. Gigantic size (wing span 40 cm); glos-
sae are enlarged plates; ScP thicker than RA.
Euhydropalaeoptera taxon nov. CuP unbranched
(reversed in Miracopteridae); hind wing with connection
between AA and CuP (anal brace); aquatic larvae (maybe
rather a convergence because the aquatic adaptations like
gills are not homologous). This taxon is also well sup-
ported by further wing venational characters (e.g. Ax0 of
odonates is probably homologous to the costal brace of
mayflies according to Bechly 1996), secondarily stiffened
wing articulation (Willkommen 2009), head characters
(Blanke et al. 2012,2013), and several modern phyloge-
nomic studies (Hovm€
oller et al. 2002; Kjer et al. 2006;
Regier et al. 2010; Ishiwata et al. 2011; Thomas et al.
2013), including first results from the 1KITE project
(Misof et al. 2014). It must be noted that some of the mor-
phological characters (e.g. bristle-like antennae) that were
suggested by Hennig (1981) and/or Kukalov
a-Peck (1983,
1985,1991) in favour of a monophyletic (Hydro)Palaeop-
tera are erroneous and refuted by the fossil evidence (e.g.
long antennae in Palaeozoic stem odonates and mayflies).
Odonatoptera. MP simple.
Panephemeroptera. At least a small costal brace; CuA
basally curved and touching stem of M.
Syntonopteroidea. Constriction and concave vein IN
between AA1C2 and AA3C4 (ambiguous because not
preserved in Syntonopteridae).
Triblosobida CEphemerida. Wing span much less than
10 cm; costal field not very large and broad; anal area of
hind wing only slightly wider than in forewing; ScP
reaches apex.
Triplosobida. MA and CuA simple.
Ephemerida. Costal brace elongate; MP-fork basal of
CuA-fork; both wings with curved veinal arch, from
which CuP and anal veins seem to originate.
18 P. Sroka et al.
Permoplectoptera. RP2 asymmetrically branched.
Litophlebioidea stat. nov. Costal brace reduced; RP
seems to originate on MA; CuA, CuP, and anal veins very
short (anal field largely reduced).
Heptabranchia. Larvae with abdominal gills only seg-
ments 17 (gills on 8 and 9 reduced); larvae only one tar-
sal segment with single claw; costal spines reduced.
Ephemeroptera. Forewing triangular; hind wing short-
ened (maximum half length of forewing); veinal arch,
from which CuP and anal veins seem to originate, is more
strongly curved.
Conclusions
A redescription of the giant Carboniferous pterygote Bojo-
phlebia prokopi has revealed several shortcomings and
errors in the original description by Kukalov
a-Peck
(1985), which are here corrected. B. prokopi is shown not
to be a stem mayfly. A revised phylogenetic position for
this taxon in the stem group of Hydropalaeoptera is sug-
gested and corroborated by a cladistic study of fossil and
extant palaeopterous insects.
Acknowledgements
We are indebted to Martin Valent (National Museum
Prague) for giving us access to study the specimen. We
acknowledge the Willi Hennig Society for generously
making the TNT software (Goloboff et al. 2008) freely
available. We thank Canadian Science Publishing for the
granted permission to reproduce figures originally pub-
lished by Kukalov
a-Peck (1985) in the Canadian Journal
of Zoology. We thank an anonymous reviewer for provid-
ing constructive comments to improve this manuscript.
This work was supported by European Community’s
Seventh Framework Programme [grant number 229518].
Supplemental material
Supplemental material for this article can be accessed
online http://dx.doi.org/10.1080/14772019.2014.987958.
References
Bechly, G. 1996. Morphologische Untersuchungen am
Fl€
ugelge€
ader der rezenten Libellen und deren Stammgrup-
penvertreter (Insecta; Pterygota; Odonata) unter besonderer
Ber€
ucksichtigung der Phylogenetischen Systematik und des
Grundplanes der
Odonata. Petalura, Special Volume, 2,
1402.
Bechly, G. 2000. Mainstream Cladistics versus Hennigian Phy-
logenetic Systematics. Stuttgarter Beitr€
age zur Naturkunde,
Serie A,613,111.
Blanke, A.,Wipfler, B.,Letsch, H.,Koch, M.,Beckmann, F.,
Beutel, R. &Misof, B. 2012. Revival of Palaeoptera head
characters support a monophyletic origin of Odonata and
Ephemeroptera (Insecta). Cladistics,28, 560581.
Blanke, A.,Greve, C.,Wipfler, B.,Beutel, R. G.,Holland, B.
R. &Misof, B. 2013. The identification of concerted conver-
gence in insect heads corroborates Palaeoptera. Systematic
Biology,62, 250263.
B
ethoux, O. &Briggs, D. E. G. 2008. How Gerarus lost its
head: stem-group Orthoptera and Paraneoptera revisited.
Systematic Entomology,33, 529547.
B
ethoux, O. &Nel, A. 2003. Revision of Diaphanoptera species
and new diagnosis of Diaphanopteridae (Palaeoptera: Diaph-
anopterodea). Journal of Paleontology,77, 10161020.
Bryant, H. N. 1997. Hypothetical ancestors and rooting in cla-
distic analysis. Cladistics,13, 337348.
Carpenter, F. M. 1987. Review of the extinct family Syntonop-
teridae (Order uncertain). Psyche,94, 373388.
Demoulin, G. 1954. Aedoeophasma anglica Scudder, syntonop-
teroide m
econnu (Insecta: Pal
eodictyopt
eres). Bulletin et
Annales de la Soci
et
e Royale de l’Entomologie de Belgique,
90, 278281.
Engel, M. S.,Davis, S. R. &Prokop, J. 2013. Insect Wings:
The Evolutionary Development of Nature’s First Flyers. Pp.
269298 in A. Minelli, G. Boxshall & G. Fusco (eds)
Arthropod Biology and Evolution. Springer, Berlin, ix C
532 pp.
Goloboff, P. A.,Farris, J. S. &Nixon, K. C. 2008. TNT, a free
program for phylogenetic analysis. Cladistics,24, 774786.
Grimaldi, D. &Engel, M. S. 2005. Evolution of the Insects.
Cambridge University Press, Cambridge, xv C755 pp.
Hamilton, K. G. A. 1972. The insect wing, part III. Venation of
the orders. Journal of the Kansas Entomological Society,45,
145162.
Hennig, W. 1981. Insect phylogeny. Wiley, New York, xxii
C514 pp.
Hess, J. C.,Lippolt, H. J.,Holub, V. M. &Pe
sek, J. 1985. Iso-
topic ages of two Westphalian C tuffs a contribution to
the Upper Carboniferous time scale. Terra Cognita,5,
236237.
Hovm€
oller, R.,Pape, T. &K€
allersj€
o, M. 2002. The Palaeop-
tera problem: basal pterygote phylogeny inferred from 18S
and 28S rDNA sequences. Cladistics,18, 313323.
Hubbard, M. D. 1987. Ephemeroptera. Pp. 199 in F. West-
phal (ed.) Fossilium Catalogus. I: Animalia. Pars 129.
Kugler Publications, Amsterdam.
Ishiwata, K.,Sasaki, G.,Ogawa, J.,Miyata, T. &Su, Z. H.
2011. Phylogenetic relationships among insect orders based
on three nuclear protein-coding gene sequences. Molecular
Phylogenetics and Evolution,58, 169180.
Kjer, K. M.,Carle, F. L.,Litman, J. &Ware, J. 2006. A
molecular phylogeny of Hexapoda. Arthropod Systematics
and Phylogeny,64,3544.
Klass, K.-D. 2007. Die Stammesgeschichte der Hexapoden: eine
kritische Diskussion neuerer Daten und Hypothesen. Deni-
sia,20, 413450.
Klassen, G. J.,Mooi, R. D. &Locke, A. 1991. Consistency
indices and random data. Systematic Zoology,40, 446457.
Kluge, N. J. 1996. A new suborder of Thysanura for the Carbon-
iferous insect originally described as larva of Bojophlebia,
with comments on characters of the orders Thysanura and
Ephemeroptera. Zoosystematica Rossica,4,7175.
Revision of the giant pterygote insect Bojophlebia prokopi 19
Kluge, N. 2004. The Phylogenetic System of Ephemeroptera.
Kluwer Academic Publishers, Dordrecht, xiii C442 pp.
Kluge N. J. &Sinitshenkova, N. D. 2002. Order Ephemerida Lat-
ereille, 1810. The True Mayflies (DEphemeroptera Hyatt et
Arms, 1891 (s.l.); DEuephemeroptera Kluge, 2000).
Pp. 8997 in A. P. Rasnitsyn & D. L. J. Quicke (eds) History
of insects. Kluwer Academic Publishers, Dordrecht, xii C
517 pp.
Krzeminsky, W. &Lombardo, C. 2001. New fossil Ephemer-
optera and Coleoptera from the Ladinian (Middle Triassic)
of Canton Ticino (Switzerland). Rivista Italiana di Paleon-
tologia e Stratigrafia,107,6978.
Kukalov
a-Peck, J. 1983. Origin of the insect wing and wing
articulation from the arthropodan leg. Canadian Journal of
Zoology,61, 16181669.
Kukalov
a-Peck, J. 1985. Ephemeroid wing venation based
upon new gigantic Carboniferous mayflies and basic mor-
phology, phylogeny, and metamorphosis of pterygote insects
(Insecta, Ephemerida). Canadian Journal of Zoology,63,
933955.
Kukalov
a-Peck, J. 1991. Fossil history and the evolution of
Hexapod structures. Pp. 141179 in CSIRO (ed.) The
Insects of Australia.2
nd
edition. Vol. 1. Melbourne Univer-
sity Press, Carlton.
Maddison,W.P.&Maddison,D.R.2011. Mesquite 2.75: a mod-
ular system for evolutionary analysis. http://mesquiteproject.org
Misof, B.,Liu, S.,Meusemman, K.,Peters, R. S.,Donath, A.,
Mayer, C.,Frendsen, P. B.,Ware, J.,Flouri, T.,Beutel,
R. G. et al. 2014. Phylogenomics resolves the timing and
pattern of insect evolution. Science,346, 763767.
Prokop, J.,Krzeminski, W.,Krzeminska, E. &Wojciechowski,
D. 2012. Paoliida, a putative stem-group of winged insects:
Morphology of new taxa from the Upper Carboniferous of
Poland. Acta Palaeontologica Polonica,57,161173.
Prokop, J.,Krzeminski, W.,Krzeminska, E.,H€
ornschemeyer,
T.,Ilger, J.-M.,Brauckmann, C.,Grandcolas, P. &Nel, A.
2014. Late Palaeozoic Paoliida is the sister group of Dictyop-
tera (Insecta: Neoptera). Journal of Systematic Palaeontology,
12,601622.
Prokop, J. &Nel, A. 2009. Systematic position of Triplosoba,
hitherto the oldest mayfly, from Upper Carboniferous of
Commentry in Central France (Insecta: Palaeodictyopter-
ida). Systematic Entomology,34, 610615.
Prokop, J. &Nel, A. 2010. New griffenfly, Bohemiatupus ele-
gans from the Late Carboniferous of western Bohemia in the
Czech Republic (Odonatoptera: Meganisoptera: Meganeuri-
dae). Annales de la Soci
et
e entomologique de France,46,
183188.
Prokop, J. &Nel, A. 2011. New Middle Permian palaeopteran
insects from Lod
eve Basin in southern France (Ephemeroptera,
Diaphanopterodea, Megasecoptera). Zookeys,130,4155.
Prokop, J.,Nel, A. &Tenny, A. 2010. On the phylogenetic
position of the palaeopteran Syntonopteroidea (Insecta:
Ephemeroptera), with a new species from the Upper
Carboniferous of England. Organisms Diversity & Evolu-
tion,10, 331340.
Rasnitsyn, A. P. 2002. Cohors Libelluliformes Laircharting,
1781 (= Subulicornes Latreille, 1807, DHydropalaeopalop-
tera Rohdendorf, 1968). Pp. 8589 in A. P. Rasnitsyn &
D. L. J. Quicke (eds) History of insects. Kluwer Academic
Publishers, Dordrecht, xii C517 pp.
Rasnitsyn, A. P. &Novokshonov, V. G. 1997. On the morphol-
ogy of Uralia maculata (Insecta: Diaphanopterida) from the
Early Permian (Kungurian) of Ural (Russia). Entomologica
Scandinavica,28,2738.
Regier, J. C.,Shultz, J. W.,Zwick, A.,Hussey, A.,Ball, B.,
Wetzer, R.,Martin, J. W. &Cunningham C. W. 2010.
Arthropod relationships revealed by phylogenomic analysis
of nuclear protein-coding sequences. Nature,463,
10791084.
Staniczek, A. H.,Bechly, G. &Godunko, R. J. 2011. Coxo-
plectoptera, a new fossil order of Palaeoptera (Arthropoda:
Insecta), with comments on the phylogeny of the stem group
of mayflies (Ephemeroptera). Insect Systematics & Evolu-
tion,42, 101138.
Staniczek, A. H.,Sroka, P. &Bechly, G. 2014. Neither silver-
fish nor fowl: The enigmatic Carboniferous Carbotriplura
kukalovae Kluge, 1996 (Insecta: Carbotriplurida) is the puta-
tive fossil sister group to winged insects (Insecta: Pterygota).
Systematic Entomology,39, 619632.
Thomas, J. A.,Trueman, J. W. H.,Rambaut, A. &Welch, J.
J. 2013. Relaxed Phylogenetics and the Palaeoptera Prob-
lem: Resolving Deep Ancestral Splits in the Insect Phylog-
eny. Systematic Biology,62, 285297.
Tsui, P. T. P. &Peters, W. L. 1975. The comparative morphol-
ogy and phylogeny of certain Gondwanian Leptophlebiidae
based on the thorax, tentorium, and abdominal terga
(Ephemeroptera). Transactions of the American Entomologi-
cal Society,101, 505595.
Willkommen, J. 2009. The tergal and pleural wing base sclerites
homologous within the basal branches of Pterygota?
Aquatic Insects,31(Suppl. 1), 443457.
Willmann, R. 1999. The Upper Carboniferous Lithoneura
lameerei (Insecta, Ephemeroptera?). Pal€
aontologische Zeits-
chrift,73, 289302.
Willmann, R. 2007a. Persisting-type stem group Ephemeroptera.
Pp. 154163 in Martill, D.M., Bechly, G. & Loveridge, R.F.
(eds.) The Crato Fossil Beds of Brazil: Window into an
Ancient World. Cambridge University Press, Cambridge, UK,
xvi C625 pp.
Willmann, R. 2007b. Die Stammgruppenvertreter der Ephemer-
optera und ihre systematische Stellung (Insecta). Species,
Phylogeny and Evolution,1, 109128.
Z
azvorka, V. 1956. N
alez svrchnokarbonsk
ev
a
zky z Kladensk
e
kamenouheln
ep
anve [Finding of an Upper Carboniferous
dragonfly in the Coal Basin of Kladno].
Casopis Pro Miner-
alogii a Geologii,1, 122126 [in Czech].
20 P. Sroka et al.
