GeolCarp_Vol62_No5_463

www.geologicacarpathica.sk

GEOLOGICA CARPATHICA

, OCTOBER 2011, 62, 5, 463—475 doi: 10.2478/v10096-011-0033-8

Introduction

Chiapas  amber  is  peculiar  for  its  content  of  plant  remains,
occasional  fungi,  scarce  pulmonate  gastropods,  arachnids,
crustaceans,  relatively  abundant  insects,  amphibians,  rep-
tiles,  bird  feathers  and  mammal  hair  (Poinar  2003;  Engel
2004;  Solórzano-Kraemer  2007;  Vega  et  al.  2009a,b).  In-
sects  are  represented  by  235  species  all  of  which  belong  to
146  living  families  and  the  living  or  closely  related  (n= 3)
genera.  The  termite  Kalotermes  nigritus  Snyder,  1946
(Isoptera, Kalotermitidae), which still lives in South America
today (Zherikhin 1970) is famous.

It is worth mentioning that the earliest living insect species

are much older, originating from Eocene Baltic amber. These
species  are  Tetracha  carolina  (Linnaeus,  1766)  (Coleoptera,
Cicindelidae), Colasposoma metallicum Clark, 1865; Palaeo-
mymar  anomalum  (Blood  &  Kryger,  1922)  (Hymenoptera,
Mymaridae),  Cupes  tesselatus  (Motschulsky,  1856)  (Co-
leoptera, Cupedidae) and Setodes picescens Ulmer, 1912 (Tri-
choptera,  Leptoceridae).  Drosophila  rubrostriata  Becker,
1908 is known from the Oligocene of Fontainebleau.

The specimen reported here was collected from the Los

Pocitos locality (Fig. 1) of the amber-bearing Mazantic Shale,

Afro-Asian cockroach from Chiapas amber and the lost

Tertiary American entomofauna

PETER VRŠANSKÝ

1,2

, PAULINA CIFUENTES-RUIZ

, UBOMÍR VIDLIČKA

4,5

, FEDOR ČIAMPOR JR

and FRANCISCO J. VEGA

Geological Institute, Slovak Academy of Sciences, Dúbravská cesta 9, P.O. BOX 106, 840 05 Bratislava, Slovak Republic;

geolvrsa@savba.sk

Paleontological Institute, Russian Academy of Sciences, Profsoyuznaya 123, 117868 Moscow, Russia; lab@palaeoentomolog.ru

Instituto de Geología, UNAM, Ciudad Universitaria, Coyoacán, 04510 México D.F., Mexico; vegver@unam.mx

Institute of Zoology, Slovak Academy of Sciences, Dúbravská cesta 9, 845 06 Bratislava, Slovak Republic;

uzaevidl@savba.sk; f.ciampor@savba.sk

Department of Biology, Faculty of Education, Comenius University, Moskovská 3, 813 34 Bratislava, Slovak Republic;

vidlicka@fedu.uniba.sk

(Manuscript received January 28, 2011; accepted in revised form March 17, 2011)

Abstract: Cockroach genera with synanthropic species (Blattella, Ectobius, Supella, Periplaneta, Diploptera and ?Blatta),
as well as other insects such as honeybees, although natively limited to certain continents nowadays, had circumtropic
distribution in the past. The ease of their reintroduction into their former range suggests a post-Early Miocene environ-
mental stress which led to the extinction of cosmopolitan Tertiary entomofauna in the Americas, whilst in Eurasia,
Africa and Australia this fauna survived. This phenomenon is demonstrated here on a low diversity (10 spp.) living
cockroach genus Supella, which is peculiar for the circumtropical synanthropic brownbanded cockroach S. longipalpa
and also for its exclusively free-living cavicolous species restricted to Africa. S. (Nemosupella) miocenica sp. nov. from
the Miocene amber of Chiapas in Mexico is a sister species to the living S. mirabilis from the Lower Guinea forests and
adjacent savannas. The difference is restricted to the shape of the central macula on the pronotum, and size, which may
indicate the around-Miocene origin of the living, extremely polymorphic Supella species and possibly also the iso-
chronic invasion into the Americas. The species also has a number of characteristics of the Asian (and possibly also
Australian) uniform genus Allacta (falling within the generic variability of Supella) suggesting Supella is a direct
ancestor of the former. The present species is the first significant evidence for incomplete hiati between well defined
cockroach genera – a result of the extensive fossil record of the group. The reported specimen is covered by a mycelium
of a parasitic fungus Cordyceps or Entomophthora.

Key words: Mexico, synanthropic, fossil insects, parasitic fungi, Allacta, Supella miocenica sp. nov.

overlain by the Balumtum Sandstone (Fig. 1). Both lithostrati-
graphic bodies are regarded as informal units. The age of the
Chiapas amber has been a matter of debate. A Late Oligocene
to Early Miocene age has been proposed by Langenheim
(1966), Tomasini-Ortíz & Martínez-Hernández (1984), Santi-
ago-Blay & Poinar (1993), Bousfield & Poinar (1994), Poinar
& Brown (2002), Poinar (2003), Engel (2004), Castan

~eda-

Posadas  &  Cevallos-Ferriz  (2007).  Ferrusquía-Villafranca
(2006) described an artiodactyl from the Los Pocitos locality,
and  considered  a  Late  Oligocene  age  for  these  sediments,
based  on  previous  biostratigraphic  interpretations  of  Frost  &
Langenheim  (1974)  and  unreferenced  paleomagnetic  studies.
He suggested that the age of the Los Pocitos strata falls within
the  28—26 Ma  (Ferrusquía-Villafranca  2006,  p. 993).  Other
authors suggest that the amber-bearing stratigraphic units are
of  Middle  Miocene  age,  and  thus  correlatives  with  the  units
that  produce  amber  in  the  Dominican  Republic  (Meneses-
Rocha  2001;  Solórzano-Kraemer  2007;  Solórzano-Kraemer
& Mohrig 2007; Perrilliat et al. 2010). At the Los Pocitos lo-
cality  (Fig. 1),  dark  grey  shales  of  the  Mazantic  Shale  con-
tain  amber,  benthic  foraminifera,  gastropods,  bivalves  and
crustaceans.  Based  on

Sr/

Sr measurements taken from a

well-preserved shell of Turbinella maya from Los Pocitos,

464

VRŠANSKÝ, CIFUENTES-RUIZ, VIDLIČKA, ČIAMPOR Jr. and VEGA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

Vega et al. (2009a, p. 53) obtained an absolute age of 23 Ma
for  the  Mazantic  Shale,  placing  it  right  on  the  boundary  be-
tween  the  Oligocene  and  Miocene.  The  Mexican  amber  has
been interpreted as the resinous exudates of Hymenaea sp., a
leguminose  tree  whose  communities  developed  near  the  an-
cient  coast,  in  estuarine  environments,  very  similar  to  man-
groves (Poinar 1992). It is worth mentioning, that the closest
relative of amber producing H. mexicana is the relictuous H.
verrucosa from East Africa (Poinar & Brown 2002). Another
amber  species,  H.  allendis  is  a  relative  to  H.  courbaril  of
Americas (Calvillo-Canadell et al. 2009). The amber was con-
sequently transported to a shallow marine environment (Lan-
genheim 1995; García-Villafuerte 2008).

The native representatives of the genus Supella are recent-

ly limited to Africa and are considered to have diverged ear-
ly in the phylogeny of the family, with the genus being the
second basalmost offshoot (after Nahublattella) of the Blat-
tellidae (Klass 1997). Nevertheless, their position is obscure
in some other analyses (see Ware et al. 2008), and there is a
single  plesiomorphy  recorded  in  respect  to  most  Symploce
Hebard,  1916  (M  and  CuA  longitudinal),  which  is  consid-
ered to be the most primitive blattellid in our previous stud-
ies  (Vršanský  1997).  Thus  Supella  is  most  likely  derived
from this genus (Vršanský et al. 2011) or its predecessors.

Thanks to the synanthropic Supella longipalpa Fabricius,

1798  (circumtropic  –  introduced  by  commerce),  the  mor-
phology  of  the  present  genus  is  well  studied  at  the  micro-
structural  level  (Roonwal  &  Rathore  1983).  For  example,
olfactory,  gustatory,  and  mechanosensory  receptors  with  a
density of 73.700 sensilla/mm

were found on the most dis-

tal segments of maxillary as well as labial palps (Prakash et
al. 1995).

The ecology of the genus is much less known. S. longipalpa

reproduces in temperatures between 25—33 °C (Tsai & Chi

2007). It can be a carrier of human intestinal parasites (Kinfu
&  Erko  2008),  Salmonella  (Fathpour  et  al.  2003),  perhaps  a
cryptic  species  related  to  Gregarina  blattarum  von  Siebold,
1839  (Clopton  &  Gold  1996)  and  an  allergen  source
(Tungtrongchitr et al. 2004). Its ootheca may often be parasit-
ized  (over  19 %)  by  host-specific  hymenopterans  Comperia
merceti (Compere, 1938) (Encyrtidae) and Anastatus tenuipes
Bolivar  &  Pieltain,  1925  (Eupelmidae)  (for  details  see
Narasimham  1992).  Individuals  infected  by  archiacanthoce-
phalan  Moniliformis  moniliformis  (Bremser  in  Rudolphi
1819) spend more time in the shade (Moore & Gottelli 1992).

S. longipalpa nymphs may be more capable of producing

and utilizing extra metabolic water from food than Blattella
germanica (Linnaeus, 1767) nymphs (Melton 1995) and its
average motility and dispersion ability is also significantly
higher than those of B. germanica (Khrustalyova 1993).

The average life cycle takes 161 days.
The extremely close relation of the species described here

with the exclusively African S. mirabilis (Shelford, 1908) may
also be an indication of the possibility of living (at least mor-
pho-) species of cockroaches occurring in the Miocene. Living
cockroach species are still unknown in the fossil record, possi-
bly because of rapid phylogeny at their species level (Vršan-
ský 2008). Living genera of cockroaches (both primitive and
advanced) are known since the Eocene (Vršanský et al. 2011).

What is notable is the identity of the pronotum colouration

with  numerous  representatives  of  another  unrelated  genus
that includes synanthropic species, the genus Periplaneta of
the  Blattidae,  for  example,  circumtropical  P.  australasiae
(Fabricius, 1775). Additionally, Periplaneta such as P. indica
Karny,  1908  from  China  can  be  smaller  (forewing  length
16 mm) (Karny 1908) and superficially could be easily con-
fused  with  Nemosupella  Rehn,  1947.  Periplaneta  is  known
from Eurasia, starting from the Middle Eocene – P. eocaenica
Meunier, 1921, P. relicta Meunier, 1921 and possibly others
occur in Messel sediments (Schmied 2009). Miocene repre-
sentatives  are  P.  hylecoeta  Zhang,  1989  and  P.  lacera
Zhang, 1989 from the Shanwang and P. sphodra Zhang, Sun
& Zhou, 1994 from Shandong in China. P. houlberti Piton,
1940  is  reported  from  the  Late  Oligocene  of  Menat  in
France. The living genus and even some species have a cos-
mopolitan  distribution.  The  origin  of  the  synanthropic  P.
americana is currently anticipated to have shifted to South,
Central  and  south  of  North  America  with  the  slave  ships
from tropical Africa (Rehn 1945). Notably, both P. americana
and  P.  australasiae  live  both  synanthropically  and  free
(Vidlička 2001). Supella can be easily discriminated from all
Periplaneta by the characteristic pale “band” and simplified
venation.  The  identical,  but  homoplastic  habitus  of  cock-
roaches  significantly  differing  in  size  (ca.  10 mm  in  amber
specimen;  up  to  30 mm  in  living  Nemosupella  and  up  to
60 mm in living Periplaneta) is obscure.

Another similar and related genus Allacta Saussure &

Zehntner, 1895 is really problematic, because the only signif-
icant  difference  of  species  placed  in  this  genus  is  the  geo-
graphical  distribution  which  is  limited  to  Asia,  shape  and
colouration  of  the  pronotum  and  shape  of  subgenital  plate.
The vast majority of characters are overlapping between Su-
pella and Allacta and most likely these two genera represent

Fig. 1. Location and stratigraphic column of the Los Pocitos locali-
ty of the Chiapas amber, Mexico.

465

AFRO-ASIAN COCKROACH AND LOST TERTIARY AMERICAN ENTOMOFAUNA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

a single holophyletic taxon. Roth (1993) noticed the extreme
similarity  of  S.  longipalpa  and  A.  diluta.  Nevertheless,  al-
though we have studied all Allacta and Supella species, it is
not the aim of the present publication to revise both taxa. Be-
cause  both,  and  especially  Supella  is  extremely  polymor-
phic,  we  retain  the  determination  of  the  present  species
within  Supella  and  its  subgenus  Nemosupella,  but  it  is  evi-
dent Supella is a direct ancestor of Allacta.

The taxon is evidence of the rich cosmopolitan Tertiary

entomofauna,  which  went  extinct  in  the  Americas,  but  still
survives on other continents, in this case in Africa. The Afri-
can  taxa  apparently  also  gave  birth  to  the  rich  but  uniform
genus  Allacta,  radiated  in  the  whole  of  Asia,  and  compara-
tively  recently  also  in  Australia.  Allacta  itself  was  also
present in America during the Eocene.

It follows that the horse was not the only group which

went extinct in the Americas, but could easily be reintro-
duced there by humans.

Material and methods

A single completely preserved specimen was collected by

Luis Zú±iga Miganjos in the Los Pocitos locality and is depos-
ited in Museo Comuntario del Ambar, Simojovel, Chiapas
(MUCAS, supported by Instituto Nacional de Antropología e
Historia – INAH). It was photographed using a Canon EOS
Mark II and drawn with a camera lucida from an Olympus
SZH10. Wing terminology follows Vršanský (1997). The liv-
ing undescribed Allacta or Supella sp. provided in Fig. 3 was

collected in Central Laos, Bolikhamsai province, Ban Nape-
Kaew Nua Pass [N 18°22.3

’/ E 105°09.1’], by Ondrej Šauša

and Eduard Jendek (deposited in IZ SAS Bratislava).

Parsimony analysis was performed using PAUP* software

version 4.0b10 (Swofford 2002), with a TBR heuristic search
of 1,000,000 replicates and the option ‘save multiple trees’ ac-
tivated. All characters were treated as unordered (0 – plesio-
morphic, 1, 2, 3 – apomorphic states). MaxTrees option was
set to 1000. Characters were weighted regarding their evolu-
tionary  relevance  (characters  9,  12  –  weight = 10,  character
13  –  weigh = 1,  remaining  characters  –  weight = 5).  De-
tails of the procedure of each character are explicitely stated
in the character analysis below. A heuristic search produced
1000 equally parsimonious trees with length 368 (consisten-
cy index CI= 0.29, retention index RI= 0.77). Majority con-
sensus  tree  was  constructed  and  post-edited  (coloured)  in
MESQUITE software version 2.6. Only clades with frequency
> 50 % were retained.

Results

Systematic entomology

Order: Blattida Latreille, 1810 (= Blattaria Latreille,

1810 = Blattodea Brunner von Wattenwyl, 1882)

Family: Blattellidae Karny, 1908

Subfamily: Pseudophyllodromiinae Hebard, 1929

Supella Shelford, 1911

Fig. 2. Supella miocenica sp. nov. from the Miocene Chiapas amber of Mexico. Holotype MUCAS-001. a – Dorsal view; b – Ventral
view; c – Head; d – Fore leg; e – Fore tarsus. Original by PC and FV. Total specimen length (from the head to the end of wings) 10 mm.

466

VRŠANSKÝ, CIFUENTES-RUIZ, VIDLIČKA, ČIAMPOR Jr. and VEGA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

T y p e: Blatta longipalpa Fabricius, 1798 (circumtropical,

native in Africa).

= Blatta supellectilium (Serville, 1839)
= Blatta incisa (Walker, 1868)
= Ischnoptera quadriplaga (Walker, 1868)
= Blatta extenuata (Walker, 1868)
and numerous other synonyms of diverse specimens from the West
Indies (see Rehn 1947).

C o m p o s i t i o n: Besides the type; S. vicina Chopard,

1958  [Comores  islands];  S.  abbotti  Rehn,  1947  [Kenya,
Tanzania, Malawi]; S. dimidiata Gerstaecker, 1869 [Kenya,
Congo, Angola, Rhodesia, Malawi, Mosambique, Botswana,
Natal,  Transvaal];  S.  orientalis  Grandcolas,  1994  [Saudi
Arabia];  S.  (Mombutia)  chapini  Rehn,  1947  [Congo];  S.
(Nemosupella) gemma Rehn, 1947 [Ghana]; S. (Nemosupella)
mirabilis  [Cameroon,  Gabon,  Congo,  Uganda,  Kenya,
Tanzania];  S.  (Nemosupella)  occidentalis  Princis,  1963
[Guinea];  S.  (Nemosupella)  tchadiana  Roth,  1987  [Chad].
All extant in Africa.

D i a g n o s i s (Rehn 1947, in part): “Pronotum ovate sub-

trapezoideal  in  outline.  Tegmina  of  female  varying  in  length
from covering but half the abdomen to surpassing the abdomi-
nal apex by a distance equal to the pronotal length, in outline
ranging from 1.5 times as long as broad. Apex well-rounded.
Costal  veins  numerous,  straight  oblique,  several  of  the  more
distal ones usually ramose, reduced in number in the abbrevi-
ated  tegmined  forms;  discoidal  sectors  oblique,  tending  to-
ward  sublongitudinal  in  males  of S. abbotti  and S. mirabilis.
Anal field pyriform, anal veins five or more, regular; discoidal
sectors similarly developed, anal field always fully indicated.

Cephalic femora with ventro-cephalic margin bearing a

regular  series  of  spines,  evenly  reducing  in  length  and
strength  distad  (sometimes  replaced  by  setae  (e.g.  in  one
limb)  –  e.g.  in S.  mirabilis),  apical  spines  of  same  margin
two-three in number, the terminal one much the longer; me-
dian and caudal femora with ventral margins spined; caudal
tarsi  with  metatarsus  in  length  surpassing  the  other  articles
combined.  Arolia  well  developed;  tarsal  claw  of  equal
length, their margins unarmed, simple.”

Subgenus: Nemosupella Rehn, 1947

T y p e : Phyllodromia mirabilis Shelford, 1908.
C o m p o s i t i o n: Supella mirabilis, S. gemma, S. tchadi-

ana, S. occidentalis.

D i a g n o s i s (ex Rehn 1947 in part – only relevant

characters): “Females more robust with tegmina and wings
broader  and  in  length  less  markedly  surpassing  the  apex  of
the abdomen, apex well rounded. Head pyriform in outline,
transverse  facial  ridge  nearly  straight  transverse;  palpi  with
penultimate  and  antepenultimate  articles  elongate.  Caudal

tarsi moderately slender, metatarsus in length somewhat ex-
ceeding that of the remaining tarsal articles combined”.

Supella (Nemosupella) miocenica sp. nov.

(Figs. 2a—c, 3a—c)

H o l o t y p e: MUCAS-001. A complete female.
T y p e l o c a l i t y: Los Pocitos, Simojovel de Allende,

Chiapas amber.

T y p e h o r i z o n: Lower Miocene, Mazantic Shale, Tertiary.
D i f f e r e n t i a l d i a g n o s i s: The present species differs

from  its  consubgeners,  S.  mirabilis  in  being  smaller  (total
body  length  with  wings  ca. 10 mm  vs.  16—25.5 mm  in  S.
mirabilis),  in  having  discoidal  sectors  oblique  (oblique  to
sublongitudinal in S. mirabilis), and in having the central dark
pronotal macula divided into two parts; from S. gemma in size
(similar  as  S.  mirabilis),  in  having  wings  more  coloured  and
pronotal central macula smaller; from S. tchadiana of a com-
parable size (12 mm), in having pronotum without markings;
and from S. occidentalis in colouration and size.

D e s c r i p t i o n: Very small species (overall body length

without  wings  about  9 mm).  Head  small  (length  to  width:
1.6/1.3 mm) with very fine antenna covered by a row (basal
segments) or up to four rows with four short (roughly corre-
sponding  to  segments’  width)  sensilla  chaetica  in  each.
Pronotum  ovate  subtrapezoidal  in  outline,  significantly
vaulted  (1.7/2.9 mm),  pale,  with  dark  macula  covering  the
whole posterior margin and central macula, divided into two
separate parts.

Body slender, sterna (especially the posteriormost ones)

widely  carved,  cerci  with  up  to  16  segments,  very  long
(1.8 mm) with dense fine sensilla chaetica of diverse length
(0.2—2 times as long as the width of the median cercal seg-
ment). Legs slender, long (including fore legs), cursorial. Fore
legs  terminated  with  claw  and  arolium;  femora  (1.7/0.1 mm)
with  dense  chaetica,  tibia  (0.9/0.07 mm)  with  at  least  5  fine
spurs (arrangement of spines along the tibiae in 3 rows), tarsi
5-segmented (0.7, 0.2, 0.1, 0.05, 0.2 mm). Front femur Type
B2,  with  four  proximal  stout  spines  succeeded  by  a  row  of
uniform  piliform  spinules,  terminating  in  two  large  spines;
pulvilli present only on the fourth tarsomere, tarsal claws sym-
metrical and unspecialized (simple), simple arolia present.

Mid femora wide and with numerous sensilla (about 26

spines  in  two  rows)  (2.1/0.6 mm),  tibia  also  robust  (1.6/
0.2 mm), with long fine spurs (10 or more); tarsi curious, ex-
tremely short (0.7/0.1 mm), with an indistinct claw and aroli-
um.  Hind  legs  long,  femora  robust  (2.6/0.8 mm),  with  two
rows with numerous (about 19 2) fine spurs and terminal two
fine spurs; tibia long (2.9/0.2 mm), with numerous (up to 30)
fine  spines;  tarsi  long  (1.3,  0.3,  0.1,  0.1,  0.3 mm),  densely
haired, with distinct claw and arolium. First third of the first

Fig. 3. a—d – Supella miocenica sp. nov. from the Miocene Chiapas amber of Mexico. Holotype MUCAS-001. Ventral view. a – Com-
plete specimen with curculionid beetle; b – detail on head with mycelia of parasitic fungus Cordyceps or Entomophthora (white “bub-
bles”); c – detail on cercus; d – dorsal view, total specimen length (from the head to the end of wings) 10 mm. e – Undescribed male of
Allacta or Supella sp. from Central Laos. Total length ca. 12 mm. Nemosupella is even more similar in having the nearly identical pronotal
shape and colouration (the central macula is divided into 2 parts only in S. miocenica). Photograph of the sister species, S. (Nemosupella)
mirabilis is available free on the web (FOW).

468

VRŠANSKÝ, CIFUENTES-RUIZ, VIDLIČKA, ČIAMPOR Jr. and VEGA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

tarsal segment sparsely (5) haired, the rest with at least 33 sen-
silla; second segment with 9, third with two, fourth with six
and terminal one with at least two sensilla.

Forewing dark, with characteristic pale anterior margins

and central stripe making the illusion of the body being sepa-
rated into two halves. Apex rounded. Venation simple, with
minimum  deformities.  Total  number  of  veins  meeting  the
margin  (without  A)  ca. 30.  Sc  simple,  short,  R  nearly
straight,  R  branches  simple  except  for  the  few  terminal
branches (18 R veins meet margin). RS not clearly differenti-
ated.  Discoidal  sectors  oblique.  M  secondarily  branched,
with about 10 branches at the margin. CuA is fused with M
and  perhaps  consists  of  a  single  terminally  dichotomized
branch. Anal field pyriform, fully indicated, anal veins sim-
ple,  regular,  apparently  five  or  more.  Intercalaries  locally
distinct (in basal R and CuA) and probably common.

R e m a r k s: For comparison and details see Discussion.
Derivation of name: miocenica is after the Miocene epoch.
C h a r a c t e r a n a l y s i s: (0 – plesiomorphy; 1—3 – apo-

morphies relative to other species within genera Supella and
Allacta,  and/or  Cariblattoides  Rehn  &  Hebard,  1927  and
Symploce  (and  fossil  Piniblattella  Vršanský,  1997),  which
were  chosen  as  outgroups  based  on  high  similarity,  and
based on retension of all original states of characters due to
standard habitus (not derived like in Supella and Allacta):

1. Head with interocular (IO) space roughly identical with

the distance between antennal sockets: apomorphy; plesio-
morphic is partially reduced IO space (as in Symploce and
most regular cockroaches); IO space nearly absent is a strong
apomorphy (2), eyes very small and remote (3).

2. Pronotum large ovate subtrapezoidal in outline, signifi-

cantly vaulted: plesiomorphic (as in most cockroaches in-
cluding Symploce and Piniblattella); apomorphic states are
1) small and round; 2) subelliptical with margins parallel.

3. Pronotal colouration with basal dark stripe and two cen-

tral maculas: plesiomorphic (central maculas are present in
most primitive blattellids such as Piniblattella), derived states
are alternative colourations (dark with cental pale macula).

4. Subgenital plate long: apomorphy (plate is plesiomor-

phically of normal length in Symploce, Piniblattella and
most other blattellids).

5. Forewing with apex of radial area reduced to mostly

simple branches (RS indistinct): apomorphy (RS distinct in
primitive Symploce and Piniblattella).

6. Forewing M and CuA branched and curved: plesiomor-

phy (as in Mesozoic cockroaches); these branches are apo-
morphically longitudinal even in some Symploce; serrate (2).

7. Forewing colouration with characteristic transversal

stripes:  apomorphy  at  level  of  common  ancestor  of  Supella
and  Allacta.  Plesiomorphic  state  is  colouration  uniform  but
not  strong  as  in  Cariblattoides,  and  other  primitive  blattel-
lids (Nahublattella, Neoblattella etc.).

8. Colouration of wings soft: plesiomorphy; pronotum

strongly dark, with sophistic pale stripes on forewing is apo-
morphic (funebris spp. group).

9. Colouration of wings and pronotum with continuous co-

louration: plesiomorphy; derived apomorphic state is colou-
ration in dark dots, lines and blotches. This character was
found as a global irreversive reorganization of morphology

and colouration and thus has been given higher weight in
the cladistic analysis.

10. Hindwing R1 distinct: plesiomorhy (as in most primi-

tive blattellids including Symploce and Piniblattella); R1 is
apomorphically reduced to a single vein.

11. Fore tarsi of B-type: apomorphy (tarsi are plesiomor-

phically A-type in most primitive blattelids including most
Symploce).

12. Pulvilli exclusively on 4

tarsomere: apomorphy (pulvilli

are plesiomorphically on 4 tarsomeres in primitive blattellids
including Symploce). In addition to diagnosis of Allacta (Saus-
sure & Zehntner, 1895), this character was found unique, never
occurring homoplastically in any other group (additionally un-
related with respect to size changes) and thus has been given
higher weight in the cladistic analysis.

13. 1 or 2 terminal fore femoral spurs: apomorphy (plesio-

morphic  state  is  with  3  spines  in  both  A-  and  B-types  of
Symploce).  This  character  was  found  polymorphic  within
species and even on one specimen (L/R sides – Roth 1991,
1993,  1996  and  our  observation)  and  thus  has  been  given
lower weight in the cladistic analysis.

14. Habitus robust: plesiomorphy (as in Piniblattella), de-

rived apomorphic states are slender (1), extremely fragile (2)
and extremely elongated (3).

Allacta Saussure & Zehntner, 1895

= Abrodiaeta Brunner von Wattenwyl, 1893
= Pseudochorisoblatta Bruijning, 1948
= Arublatta Bruijning, 1947

T y p e: Abrodiaeta modesta Brunner de Wattenwyl, 1893

from Carin Ghecu in Burma, by selection.

C o m p o s i t i o n: Funebris species group (sensu Roth

1993):  basivittata  (Bruijning,  1947)  [New  Guinea,  Aroe  and
Aru  Islands],  bipunctata  (Walker,  1869)  [Celebes,  Aru  Is-
lands, New Guinea], funebris (Walker, 1868) [Borneo] (Roth
1993);  grandcolasi  Roth,  1995  [Irian-Jaya],  megamaculata
Roth,  1995  [Papua  New  Guinea],  straatmani  Roth,  1995
[Papua New Guinea] (Roth 1995), diagrammatica (Hanitsch,
1923) [Malacca, Singapore, Mentawai islands, Sumatra, Java].

Hamifera

species group (sensu Roth 1993): bimaculata

Bey-Bienko, 1969 [China], diluta (Saussure, 1863) [Ceylon,
India], figurata (Walker, 1871) [Ceylon, India], hamifera
(Walker, 1868) [Malacca, Java, Borneo, Philippinen], inter-
rupta (Hanitsch, 1925) [Borneo], luteomarginata (Hanitsch,
1923) [Singapore], maculicollis (Hanitsch, 1927) [Vietnam],
parva Shelford, 1906 [Borneo], pantherina (Hanitsch, 1933)
[Borneo] (Roth 1993); svensonorum Roth, 1995 [Malaysia,
Borneo] (Roth 1995).

Polygrapha

species group (sensu Roth 1993): fascia Roth,

1993 [Indonesia], immunda (Brunner von Wattenwyl, 1893)
[Burma,  Malacca],  polygrapha  (Walker,  1868)  [Thailand,
Malacca, Singapore, Sumatra, Borneo], picturata (Shelford,
1907)  [Singapore,  Sumatra,  Thailand,  Malaysia,  Borneo],
marmorata  Walker,  1869  [Burma,  Sumatra,  Malaysia],
mcgavini Roth, 1991b [Indonesia], robusta Bey-Bienko, 1969
[China],  transversa  Bey-Bienko,  1969  [Vietnam];  arborifera
(Walker,  1868)  [Malaysia,  Java,  Borneo,  Mentawai  Islands],

469

AFRO-ASIAN COCKROACH AND LOST TERTIARY AMERICAN ENTOMOFAUNA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

Cariblattoides with A-type and pulvilli on 4 tarsomeres, sup-
porting this state as ancestral.

Phylogenetically obscure is also the distribution of knee

spines at the distal end of the femora – 2 in present species
and some Allacta (see character matrix) and Supella, but more
often 3 (also in figured most closely related Allacta sp.).

The present amber species clearly differs from representa-

tives  of  the  A.  funebris  species  group,  which  are  very  dark
including the pronotum (eventually with a narrow pale mar-
gin only) and with basal narrow pale stripes. A. polygrapha
spp.  group  differs  in  having  a  large  symmetrical  pronotal
pattern  of  dark  dots,  lines  and  blotches  and  forewings  che-
quered  with  dark  cells  between  veinlets  and  with  larger
blotches  (A.  confluens  placed  here  is  somewhat  different
from other representatives of this species group, it has more
coherent colouration and pronotum colouration most similar
to the amber specimen. Nevertheless, the shape of the prono-
tum of S. miocenica is different: dissimilar to any described
Allacta  spp.).  A.  puncticollis  is  also  completely  different,
with a subparabolic reddish brown pronotum with small yel-
lowish spots, but has pulvilli limited to the 4

tarsomere. A.

crassivenosa is categorized as incertae sedis and most likely
belongs to another genus. The most related to the present
amber species within Allacta is the A. hamifera spp. group,
which differs in having smaller pronota with a different colou-
ration pattern and in the shape of the subgenital plate which is
never as long in Allacta. Even more similar, and hardly recog-
nizable from the Nemosupella spp. is an undescribed represen-
tative putatively attributed to Allacta on the basis of identical
legs (pulvilli limited to the 4

tarsomere – Figure 2b). Alter-

natively it can mean that this species belongs to Supella and
is its only Asian representative, but more likely it is a transi-
tional taxon leading to Allacta. The figured undescribed spe-
cies  (Fig. 3e)  cannot  be  placed  into  any  spp.  groups  of
Allacta,  but  is  most  closely  related  to  the  A.  hamifera  spp.
group in the shape of the head and underived colouration. It
is  apparent  that  A.  funebris  and  A.  polygrapha  spp.  groups
were derived much later, the latter apparently derived via A.
confluens,  which  has  similar  pronotum  colouration  and  un-
derived forewing colouration.

It is clear, that there is no strict hiatus between these two

genera, but this cannot be used as a reason for their synony-
mization or for the erection of additional genera. The problem
is that there are known “missing link” taxa in all cockroach
families  and  we  also  know  a half-cockroach—half  termite
(Vršanský 2010) and also half-cockroach—half mantodean. All
studied living cockroach genera with fossil records have these
transitional stages too, and the present taxon is no exception.
This  is  a half-Supella—half-Allacta,  but  better  Supella  than
Allacta. So splitting or erecting does not have a proper place
here just because the group is well studied. Paraphyletic taxa
and incomplete hiati are present in the vast majority of studied
cockroaches,  which  is  a result  of  the  extensive  fossil  record
with ca. 100,000 specimens. So our specimen has synapomor-
phies of Allacta+Supella (colouration, venation), autapomor-
phies  of  Allacta  (extremities),  but  major  autapomorphies  of
Supella (pronotum, subgenital plate).

The cladogram (Fig. 4) supports all the above-mentioned in-

ferences, with nearly ideal separation of all spp. groups, but it

australiensis  Roth,  1991  [Queensland],  confluens  (Hanitsch,
1925)  [Borneo],  labyrinthica  (Hanitsch,  1927)  [Vietnam],
loconti  Roth,  1993  [Indonesia],  megaspila  (Walker,  1868)
[Malacca, Mentawai Islands, Java, Borneo], ornata Bey-Bien-
ko,  1969  [China],  modesta  (Brunner  von  Wattenwyl,  1893)
[Burma;  type],  karnyi  (Hanitsch,  1928)  [Mentawai  Islands,
Sumatra]  (Roth  1993),  brossuti  Roth,  1995  [Irian-Jaya],
deleportei  Roth,  1995  [Papua  New  Guinea],  gautieri  Roth,
1995 [Papua New Guinea], nalepae Roth, 1995 [Papua New
Guinea],  persoonsi  Roth,  1995  [Papua  New  Guinea],  srengi
Roth, 1995 [Papua New Guinea] (Roth 1995).

A. puncticollis (Brunner von Wattenwyl, 1898) [Borneo]

(not placed sensu Roth 1993) and A. crassivenosa Bolívar,
1897 [India] (incertae sedis sensu Roth 1993). Except for
problematic A. australiensis, all are extant in Asia.

D i a g n o s i s: Front femur Type B [B2 or B3 according to

Roth 1993, or C (right and left femur can differ in type (B2-B3
or even B-C) according to Roth (1991, 1996))]; pulvilli
present on fourth tarsomere only, tarsal claws simple, sym-
metrical, arolia present (Roth 1995).

Discussion

Supella-Allacta complex

Because there is an immense similarity and relation be-

tween the genera Supella and Allacta, it is necessary to pro-
vide arguments for the categorization of the present fossil
within Supella.

These genera are clearly distinguished among other blat-

tellids by autapomorphies including the characteristic colou-
ration  with  a  pale  stripe  appearing  to  divide  the  body  into
two (or, in combination with the pronotal colouration into 3)
separate  parts;  M  and  CuA  descending  in  an  obtuse  angle;
mostly  simple  R  branches  with  indistinct  RS  (homoplastic
with  Pseudomops  Serville,  1831  and  Ectobius  Stephens,
1835); and in other characters unseen in the present fossil (see
Rehn 1947). Supella restricted to Africa is much more diverse
and  polymorphic  (including  forms  identical  with  Allacta)
which  suggests  its  direct  ancestral  position  in  respect  to
Asian  and  Australian  Allacta  (the  latter  is  restricted  to
a single species A. australiensis from Queensland, which has
an indicated hindwing R1 as in Supella, but is very different
from  both  Supella  and  Allacta  in  having  eyes  nearly  con-
nected,  and  in  subelliptical  –  with  anterior  and  posterior
margins  parallel  –  form  of  pronotum;  thus  it  can  simply
represent Supella or a different genus).

The present amber species has fore legs of B type identical

(!including  the  number  of  proximal  spines  and  terminal
spurs) with Allacta, dissimilar to most Supella (A-type), but
these  types  can  be  polymorphic  (L/R)  in  a single  specimen
of Allacta (B2/B3 or even B/C – see above), and the Supella
subgenus Nemosupella can have this B-type pattern too. The
most  primitive  living  blattellid,  Symploce  has  mostly  A3
type,  but  ocassionally  B3,  thus  it  is  likely  A  is  the  original
type, but the above-mentioned polymorphisms are evidence
for a convergent nature of this character changes. Closely re-
lated  and  perhaps  derived  from  Supella  is  also  the  genus

470

VRŠANSKÝ, CIFUENTES-RUIZ, VIDLIČKA, ČIAMPOR Jr. and VEGA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

was  weak  in  the  position  of  A.  interrupta  (well  nested  within
hamifera group by intuition (Roth 1993), very closely related
to A. hamifera) within the polygrapha group. The second prob-
lem is the terminal position of A. confluens which is expected to
be  the  basalmost  polygrapha  group  stem  (it  has  a  very  basal
position when all characters were equally weighted). Third, A.
bimaculata+A. pantherina (hamifera spp. group) appear nest-
ed within Supella. Notably, all three problems are absent in the
cladogram with the normal weight of character 13 (weight 1 in
contrast to weight 5 of all other characters except weight 10 of
characters 9 and 12), which is polymorphic (and thus has been
given a lower weight), A. bimaculata + A. pantherina (hamifera
spp. group) are nested (as basalmost offshots) within hamifera
spp.  group  using  the  same  weights  but  using  different  (500)
number of maximal search trees. A. puncticollis (out of any spp.
group) appears to be a sister taxon to A. svensonorum and A. lu-
teomarginata  (trichotomy  at  cladogram).  When  all  characters
were weighted normally and in many other options tried, S. di-
midiata were often nested within Allacta. S. dimidiata is a good
Supella, and this placement is evidence for numerous homopla-
sies within the group. Generally, the homoplasies wihin cock-
roaches are enormous.

Paleogeographically, the basalmost blattellids were pre-

served  in  Asia,  but  their  ancestors  within  the  Mesoblat-
tinidae  were  also  common  in  Europe  (Vršanský  &  Ansorge
2007). The basalmost Supella is up to recently clearly nested
within Africa, but the present American species is clear evi-
dence for the past circumtropic distribution. Allacta was pos-
sibly  derived  in  Africa  but  radiation  was  apparently  not
limited  to  Asia  (except  for  the  mentioned  synanthrope,  Al-
lacta  is  present  in  sediments  of  the  Green  River  in  Colo-
rado).  According  to  this  cladogram  A.  australiensis  was
derived  quite  recently  from  one  of  the  species  in  the
polygrapha spp. group.

From the most primitive blattellids, Nahublattella and

Symploce  (for  position  of  these  genera  within  Blattellidae
see also Klass & Meyer (2006)), there is a significant reduc-
tion and simplification of venation, which is evidence of the
very  early  divergence  of  the  whole  complex  from  the  main
blattellid  stem.  Another  eventual  conclusion  considers  the
small size of the present species, which may be a plesiomor-
phic character, which would explain the significant simplifi-
cation  of  venation  even  in  large  living  species.  Notably,  in
the  derived  genus  Allacta,  terminal  radial  veins  are  simpli-
fied only in the smallest species (A. parva).

Notwithstanding, some living Supella and Nemosupella in

particular have more primitive traits and likely diverged be-
fore  the  speciation  of  S.  miocenica.  Its  single  insignificant
deformity  (not  clearly  visible,  insignificantly  changing  the
wing  geometry)  supports  the  fact  that  Eocene  and  Miocene
species have few accumulated wing deformities and their oc-
currence in living species represents support for them being
inheritable mutations (see Vršanský 2005).

Extinct American

Supella

In spite of the close relation of its 3 known subgenera, the

general habitus of the respective taxa in Supella is very di-
verse. S. longipalpa is a slender, fine cockroach, while S.

orientalis  has  extremely  elongated  wings  (—1 : 4.5),  and  the
subgenus  Nemosupella  is  clearly  differentiated  by  robust
habitus with a robust unplain pronotum and more or less nor-
mal longitudinal veins. The present species share all the au-
tapomorhies  of  the  genus  and  subgenus  Nemosupella  and
can be safely categorized within this taxon.

The colouration and general appearance is hardly recog-

nizable  from  females  of  S.  mirabilis,  its  sister  species,  al-
though  the  shape  of  the  pronotum  is  somewhat  transitional
between  its  males  and  females.  Except  for  the  significantly
smaller  size  of  the  present  new  species,  the  sole  difference
between the two taxa is the divided central pronotal macula.
There  are  no  additional  plesiomorphies,  which  indicate  the
present as well as the living species of Nemosupella diverged
near the Mid-Miocene. The other three representatives of the
subgenus are closely related to S. mirabilis, but are dissimi-
lar to the present fossil due to different colouration.

On the other hand, other representatives of the genus such

as S. longipalpa reveal significant divergence from the main
morphological standard and suggest rapid phylogeny at the
subgeneric level.

The similarity with the undescribed representatives of the

genus Allacta (Fig. 3e) is so striking (and involves size – Ne-
mosupella are much larger), that it is apparent Allacta is de-
rived  from  Nemosupella  via  the  predecessors  of  this
undescribed taxon and also via predecessors of Supella mio-
cenica. Its direct ancestry can be excluded based on the de-
rived  pronotum  of  S.  miocenica.  While  there  is  a standard
Allacta  placed  within  a living  spp.  group  (polygrapha)
present  in  much  older  Eocene  Green  River  sediments,  the
Chiapas is apparently another case of the presence of primitive
species in amber, when compared to isochronous sedimentary
record. Relic character of amber cockroach (and all insect to
some  extent)  assemblages  is  characteristic  also  for  the  only
two studied Mesozoic ambers (Lebanon and Archingeay, but
also in Baltic amber), which might either be caused by dif-
ferent  methods  of  dating,  or  by  the  more  humid,  dark  and
colder source microclimates of amber forests. This is in con-
trast  to  Cretaceous  ecosystems,  where  primitive  cockroach
forms of the Jurassic type are restricted to younger, but dry
to semiarid ecosystems (Vršanský et al. 2002).

Very little can be said about the ecology of the present spe-

cies.  Generally  the  ecosystems  of  Chiapas  were  perhaps  di-
verse, ranging from lowland tropical dry forest tending toward
open  forest  and  mangroves  (Solórzano-Kraemer  2007).  The
wide range of ecosystems from the rainforest down to savan-
nas,  of  the  closely  related  S.  mirabilis,  indicate  this  genus  is
highly adaptable to diverse conditions. This ecological plastic-
ity could have resulted in invasions of early Supella (Nemosu-
pella)  into  the  Americas  before  the  Mid-Miocene  (and  later
into Asia as the genus Allacta). Living species of Supella are
cavicolous  (Grandcolas  1994)  and  the  genus  most  likely  also
originated in Africa because the most primitive blattellid, Sym-
ploce Hebard, 1916 is circumtropic, but rare in America. On the
other hand, the genus Nahublattella Bruijning, 1959 considered
to be even more primitive by Klass (1997) is native to Central
and South America, which could indicate the opposite.

The parasitic (or predatory) fungus Entomophthora or

Cordyceps is indeterminable, but the mycelium is richest in

471

AFRO-ASIAN COCKROACH AND LOST TERTIARY AMERICAN ENTOMOFAUNA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

Fig. 4. Parsimony analysis of all the known Supella and Allacta species (excluding obscure A. crassivenosa, S. occidentalis, S. vicina, and
including 3 undescribed Allacta and/or Supella species) with their geographic distribution. S. miocenica and Piniblattella vitimica are ex-
tinct. Position of A. interrupta (top) is illusory due to numerous homoplasies with A. robusta and other polygrapha spp. group species, as
this species apparently belongs to the hamifera spp. group. 50% majority consensus tree from 1000 equally parsimonious trees was gained
with maximum parsimony search (PAUP). Numbers above branches represent clade frequencies in %.

the junction of head and pronotum and could have its epicen-
tre  in  the  head.  This  fungus  provides  a  contribution  to  the
poorly known microorganisms of the Chiapas amber. Only a
?Bacillus-like  cell  and  two  types  of  budding-bacteria-like
microrganisms  were  reported  previously  (Veiga-Crespo  et
al. 2007).

Comments on synanthropism in cockroaches

The fifty species of synanthropic cockroaches comprise

only an insignificant fraction of the total of about 5000 (Bell
et al. 2007) described species of living cockroaches. Neverthe-
less, they are important for their number and ecological signif-
icance. Most of the species had their genera recently limited to

certain continents and only nowadays have become cosmopol-
itan (or circumtropical). On the contrary, their history on a
geological scale is much richer than we would expect and
their original distribution was also circumtropical.

The genus Blattella was until very recently limited to Africa

(26  species),  Asia  and  the  Pacific  islands  (23  species)  (Roth
1985),  and  the  synanthropic  species  B.  germanica  spread  to
the whole world from east Asia (Roth 1985). The occurrence
of  this  genus  in  the  Mesozoic  of  Europe  as  a  single  nymph
(adult  could  eventually  differ)  (Vršanský  2008)  and  in  the
Eocene of the USA (Green River, Colorado) is thus surprising.

Free Ectobius is limited to Europe, but in the Tertiary it was

cosmopolitan. The same situation is found with the present
Supella introduced from North Africa to Central America on

472

VRŠANSKÝ, CIFUENTES-RUIZ, VIDLIČKA, ČIAMPOR Jr. and VEGA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

slave ships (Rehn 1947). The Miocene of Mexico is quite dis-
tant from the recent distribution limited to Africa.

According to Princis (1954), Blatta originates from the

Near and Middle East – the native place of the closely relat-
ed Shelfordella Adelung, 1910, but the other predicted origin
of this genus is North Africa (Rehn 1945; Cornwell 1968),
and Blatta furcata Bohn, 1985 is known from the Near East
and North Africa (Bohn 1985). Their close relative also oc-
curred in the Eocene of the USA (Green River).

The last significant synanthrop is Periplaneta, recently

limited to Asia.

The only viviparous cockroach group met the same fate:

the Diplopteridae, now restricted to two genera (one African,
one Asiatic), were common in America during the Eocene
(in Green River). This family is also reported from the
Eocene of Quilchena (Archibald & Methewes 2000), but the
figured specimen (Q-0040) is very different from all known
Diplopteridae (details to be provided elsewhere).

Thus it is apparent that all the synanthropic species belong

to cosmopolitan genera (cosmopolitan genera are otherwise
rare), very likely with a broad environmental tolerance – and
thus pre-adapted for synanthropism.

Lost Tertiary American entomofauna

The occurrence of some cosmopolitan synanthropic spe-

cies in the Americas and their absence prior to re-introduc-
tion, triggered the present discussions about the causes of the
extinction of these entomofaunas during the Tertiary.

It was Eocene Ectobius from the Green River – a member

of an extinct genus, but which was extremely easily reintro-
duced in North America several times (with 3 species),
which concentrated our efforts on the search for this fauna in
2006. It was very rapidly supported by the discovery of a
honey bee in the Miocene sediments of Nevada (Engel et al.
2009), a genus extinct in the Americas.

Poinar et al. (1999) also noticed the Early Tertiary North

American extinctions of species of living tropical ant genera
Technomyrmex  Mayr,  1872,  Leptothorax  Mayr,  1855  and
Dolichoderus Lund, 1831, recorded in the Eocene of British
Columbia.  These  records  comprise  only  the  species  level,
which is insignificant on the present time scale, but Techno-
myrmex is now, with the exception of a single Central Amer-
ican  species  (and  its  abundance  in  the  Dominican  amber),
limited to the tropics of the old world; Leptothorax is holarc-
tic today; Dolichoderus is cosmopolitan.

The Eocene of the Okanagan Highlands reveals a represen-

tative of the Myrmeciinae, currently limited to the Australian
region (Archibald et al. 2006). The only genus – determined
hemipteran from Quilchena, Megymenum is today found in
only in the Oriental biotic region and Australia (G. Gross,
personal communication in Archibald & Mathewes 2000).

The post-Miocene cooling was unlikely to be a reason for

this extinction, as both Ectobius and Apis Linnaeus, 1758
occur in Northern Europe today.

The loss of another taxon from North America is now ap-

parent – Supella, in which case cooling could be the reason
as nowadays this genus is restricted to Africa. On the other
hand it is hardly possible that this taxon went extinct in

warm Central and South America. Their historical absence in
South America is also difficult to anticipate as other cock-
roach taxa from the Dominican amber are present in South
America (see below).

Thus, of fourteen studied cockroach genera (all still living)

from the Eocene—Miocene of North America only two (Cari-
blattoides, Sigmella) survive nowadays in (South and Central)
America and only Cariblattoides is characteristic for Central
America (although it occurs in Brazil). It is perhaps not inci-
dental that a representative of Sigmella was dominant during
the Eocene and also in the present Mexican amber. An addi-
tional taxon reported from the Mexican amber is Ischnoptera
sp.,  currently  distributed  in  Central  and  South  America,  but
this  determination  is  obscure  (determination  may  be  correct,
but  no  diagnostic  characteristics  for  the  genus  are  provided,
and  the  species  (Ischnoptera  sp. 1  in  Solórzano-Kraemer
2007)  may  well  belong  to  Supella  or  some  other  blattellid
taxon). All things being equal, the diversity of cockroaches in
Chiapas  amber  was  certainly  high:  7  specimens  belong  to
7 different genera and species of the family Blattellidae.

Another cockroach genus, exclusively African today,

known from the Tertiary of North America is Namablatta
Rehn, 1937.

The closely related termites are ubiquitous in the present

context. While Kalotermes nigritus still lives in South Amer-
ica, the whole cosmopolitan family Mastotermitidae (present
as Mastotermes electromexicus Krishna & Emerson, 1983
and Mastotermes electrodominicus Krishna & Grimaldi,
1991 occurring in the Dominican amber according to Solór-
zano-Kraemer (2007)) went extinct in the Americas and sur-
vives only in Australia.

Some other insect groups from the Mexican amber (Solór-

zano-Kraemer 2007) and Green River (our data) reveal a
similar pattern, to be analysed in detail elsewhere.

It is of special consideration that the Dominican amber

shows a very different pattern in respect to the distribution of
cockroach genera. If the determinations of Arillo & Ortun

(2005) are correct, then there are no shared taxa (even on the
generic  level)  between  the  Mexican  and  the  Dominican  am-
ber, and all Dominican amber cockroach genera are not only
highly advanced, but with the exception of the circumtropical
Anaplecta  Burmeister,  1838  all  –  Euthlastoblatta  Hebard,
1917,  Pseudosymploce  Rehn  &  Hebard,  1927,  Plectoptera
Saussure, 1864, Cariblatta Hebard, 1916, Holocompsa Bur-
meister,  1838  (a single  species  (H.  debilis  (Walker,  1868))
also  occurs  in  Ceylon,  Java,  Sumatra,  Borneo  and  Phil-
lipines) – are characteristic of Central and/or South America.
Taxa  described  by  Gorochov  (2007),  including  obscure
Agrabtoblatta Gorochov, 2007 and  Erucoblatta  Gorochov &
Anisyutkin, 2007, also appear limited to South America.

Taking all this preliminary information together, it is ap-

parent that sometime after the Mid-Miocene some extensive
environmental  change  influenced  North  and  probably  also
Central  and  South  America,  resulting  in  the  loss  of  cosmo-
politan Early Tertiary entomofaunas. Judging from the mod-
ern  composition  of  the  Dominican  amber,  this  may  (if  the
abovementioned determinations are correct) mean a recovery
occurred during the time between the Early Miocene Mexi-
can  amber  (23—7.1  Ma)  and  the  Dominican  amber  times

473

AFRO-ASIAN COCKROACH AND LOST TERTIARY AMERICAN ENTOMOFAUNA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

(20.5—16.4 Ma). The dating of both of ambers is still uncer-
tain (the abovementioned datings are after EDNA database),
more counterbalanced by the Late Barstovian (14.5—14 Ma)
dating of the Nevada (with honeybee) sediments. Recently,
the age of 23 Ma was designated for the basalmost amber
bearing strata of Chiapas (Vega et al. 2009a).

As it is very difficult to imagine some geological or eco-

logical process which would be able to trigger such an exten-
sive change (aridization and/or cooling would not influence
some of the cockroaches), the change was probably biologi-
cal – either caused by diversification of cockroach parasites
which  were  consequently  reduced  (as  reintroduction  has
been  easy),  or  diversification  of  more  progressive  insect
groups. The parasite hypothesis may be valid in the present
case, as  Comperia merceti eradicated populations of synan-
thropic  cockroaches  in  Europe,  even  when  its  function  as  a
control  of  Supella  is  still  not  validated  (Goudey-Perriere
1991)  and  Encyrtidae  and  Eupelmidae  parasitizing  ootheca
have  Tertiary  origin  (A.P.  Rasnitsyn,  personal  communica-
tion  2010),  known  only  from  Europe  starting  with  the
Eocene  Baltic  amber  (Trjapitzin  1963).  (These  parasites
could  also  cause  extinctions  of  external  ovipositor  bearing
cockroaches which did not lay eggs in ootheca.)

The occurrence of advanced taxa in the Dominican amber

(isochronous with the Mexican amber according to Solórza-
no-Kraemer  (2007),  however  see  above)  would  favour  the
diversification and radiation of the modern South-American
cockroach  taxa  hypothesis.  Nevertheless,  Diploptera  is  the
most advanced cockroach that has ever lived, and thus its ex-
tinction  in  the  Americas  falsifies  the  latter  hypothesis.  On
the other hand, it is possible that its viviparity evolved only
in the common ancestor of Asian and African species.

Plants were perhaps not as influenced as fauna, as Eocene

flora  of  British  Columbia  in  Canada  is  characteristic  of  the
modern eastern North American deciduous forest zone, prin-
cipally  the  mixed  mesophytic  forest,  but  also  including  ex-
tinct taxa: taxa known only from eastern Asian mesothermal
forests, and a small number of taxa restricted to the present-
day  North  American  west  coast  coniferous  biome  (Green-
wood  et  al.  2005).  Also,  according  to  Solórzano-Kraemer
(2007),  all  plants  from  the  Chiapas  amber  are  currently
present in Pacific coastal forest.

Conclusions

The genus Supella with S. miocenica sp. nov. was native to

America  during  the  Miocene  time  of  the  Chiapas  amber.  It
represents  another  case  of  rich  cosmopolitan  Early  Tertiary
entomofauna, which suddenly went extinct in America some-
where  around  the  Miocene  (but  which  still  survives  in  other
continents).  Supella/Allacta  complex  (Allacta  was  derived
from  Supella)  is  another  case  of  the  genera  which  now  in-
cludes synanthropic species, which were natively circumtrop-
ic,  and  can  be  easily  reintroduced  in  America  nowadays.  S.
(Nemosupella) miocenica sp. nov. is the earliest known cock-
roach  which  can  be  categorized  within  the  living  subgenus
and also the first published direct evidence of transitional spe-
cies (and thus incomplete hiatus) at the level of living genera.

Acknowledgments: We thank the Museo Comunitario del
Ámbar de Simojovel, Chiapas for making this material avail-
able for study; the generous support of Luis Zún

~iga and Jorge

Balcázar is greatly appreciated; Sonia Fraga Lopes (National
Musem, Rio de Janeiro), Dong Ren (Capital Normal Universi-
ty, Beijing), Maria del Carmen Perrilliat (Institute of Geology,
Mexico), Jozef Michalík (Geological Institute, Slovak Repub-
lic)  and  five  anonymous  reviewers  for  revision  of  the  manu-
script, Ladislav Roller (Institute of Zoology SAS, Bratislava)
for  literature  supply,  Ivona  Kautmanová  (Slovak  National
Museum,  Bratislava)  for  fungus  determination  and  Graeme
Butler (Bratislava) for linguistic correction. Supported by the
UNESCO, AMBA/NUUR, VEGA 6002, 2/0125/09, 2/0167/9,
MVTS and the Literary Fund.

References

Archibald S.B. & Mathewes R.W. 2000: Early Eocene insects from

Quilchena, British Columbia, and their paleoclimatic implica-
tions. Canad. J. Zool. 78, 8, 1441—1462.

Archibald S.B., Cover S.P. & Moreau C.S. 2006: Bulldog ants of

the Eocene Okanagan Highlands and history of the subfamily
(Hymenoptera: Formicidae: Myrmeciinae). Ann. Entomol. Soc.
Amer. 99, 3, 487—523.

Arillo A. & Ortun

~o V.M. 2005: Catalogue of fossil insect species

described from Dominican amber (Miocene). Stuttgarter Beitr.
Naturkunde B 352, 1—68.

Bell W.J., Roth L.M. & Nalepa C.A. 2007: Cockroaches: ecology,

behaviour, and natural history. Johns Hopkins University Press,
Baltimore, 1—230.

Bohn H. 1985: Blatta furcata (Karny), the nearest relative of the

Oriental cockroach (Blatta orientalis L.) (Insecta: Blattodea:
Blattidae). Israel J. Zool. 33, 1—2, 39—50.

Bousfield E.L. & Poinar G.O. Jr. 1994: New terrestrial amphipod

from Tertiary amber deposits of Chiapas province, southern
Mexico. Historical Biology 7, 105—114.

Calvillo-Canadell L., Cevallos-Ferriz S. & Rico-Arce L. 2009: Mio-

cene Hymenaea flowers preserved in amber from Simojovel de
Allende, Chiapas, Mexico. Rev. Palaebot. Palynol. 160, 3—4,
126—134.

Castan

~eda-Posadas C. & Cevallos-Ferriz C.R.S. 2007: Swietenia

(Meliaceae) flower in Late Oligocene—Early Miocene amber
from Simojovel de Allende, Chiapas, Mexico. Amer. J. Botany
94, 1821—1827.

Clopton R.E. & Gold R.E. 1996: Host specificity of Gregarina blat-

tarum von Siebold, 1839 (Apicomplexa: Eugregarinida)
among five species of domiciliary cockroaches. J. Invertebr.
Pathol. 67, 219—223.

Cornwell P.B. 1968: The Cockroach. Vol. 1. A laboratory insect

and an industrial pest. Hutchinson and Co., London, 1—391.

Engel M.S. 2004: Arthropods in Mexican amber. In: Llorente-

Bousquets J.E., Morrone J.J., Yán

~ez-Ordón~ez O. & Vargas-

Fernández I. (Eds.): Biodiversidad, Taxonomía, y Biogeografía
de Artrópodos, v. IV. UNAM/CONABIO First Edition, México
D.F., 175—186.

Engel M.S., Hinojosa-Diaz I.A. & Rasnitsyn A.P. 2009: A honey

bee from the Miocene of Nevada and the biogeography of Apis
(Hymenoptera: Apidae: Apini). Proc. Cal. Acad. Sci., Ser. 4,
60, 1—20.

Fabricius J.C. 1775: Systema entomologiae, sistems insectorum

classes, ordines, genera, species, adiectis synonymis, locis, de-
scriptionibus, observationibus. Officina Libraria Kortii, Flens-
burgi et Lipsiae, 1—832.

474

VRŠANSKÝ, CIFUENTES-RUIZ, VIDLIČKA, ČIAMPOR Jr. and VEGA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

Fathpour H., Emtiazi G. & Ghasemi E. 2003: Cockroaches as reser-

voirs and vectors of drug resistant Salmonella spp. Fresenius
Envi. Bull. 12, 7, 724—727.

Ferrusquía-Villafranca I. 2006: The first Paleogene mammal record

of middle America: Simojovelhyus pocitosense (Helohyidae,
Artiodactyla). J. Vertebrate Paleontology 26, 989—1001.

Frost S.H. & Langenheim R.L. 1974: Cenozoic reef biofacies; Ter-

tiary larger foraminifera and scleractinian corals from Chiapas,
Mexico. Northern Illinois University Press, De Kalb, 1—388.

García-Villafuerte M.A. 2008: First fossil record of the genus

Hemirraghus (Araneae, Theraphosidae) from the Tertiary am-
ber of, Chiapas, México. Rev. Ibérica. Aracnología 16, 43—47
(in Spanish).

Gorochov A.V. 2007: New and little known Orthopteroid Insects

(Polyneoptera) from fossil resins: Communication 2. Paleon-
tological J. 2, 39—50.

Goudey-Perriere F. 1991: Comperia merceti (Hymenoptera, Chalci-

doidea, Encyrtidae) parasitizes ootheca of the cockroach Supel-
la longipalpa (Dictyoptera, Blattellidae). Bull. Soc. Zool.,
France 116, 3—4, 219—228.

Grandcolas P. 1994: Blattaria (Insecta: Dictyoptera) of Saudi Arabia:

a preliminary report. Fauna of Saudi Arabia, 14, 40—58. In:
Büttiker W. & Krupp F. (Eds.): Fauna of Saudi Arabia. Riyadh,
Basle. National Commission for Wildlife Conservation and
Development, Pro Entomol., 454.

Greenwood D.R., Archibald S.B., Mathewes R.W. & Moss P.T.

2005: Fossil biotas from the Okanagan Highlands, southern
British Columbia and northeastern Washington State: climates
and ecosystems across an Eocene landscape. Canad. J. Earth
Sci. 42, 2, 167—185.

Karny H. 1908: Orthoptera. A. Dictyoptera. In: Wissenschaftliche

Ergebnisse  der  Expedition  Filchner  nach  China  und  Tibet.
Vol. 10,  Part 1,  Section 1.  Zool.  Sammlungen  Hexapoda,
1903—05, Berlin,  1—56, pls. 1—2, figs. 1—25.

Khrustalyova N.A. 1993: Study of the motility and distributional

capacity of synanthropic cockroaches obtained from the natu-
ral biotops and kept together in laboratory conditions. Zool.
Zh. 72, 10, 36—40.

Kinfu A. & Erko B. 2008: Cockroaches as carriers of human intesti-

nal parasites in two localities in Ethiopia. Trans. Roy. Soc.
Trop. Med. H 102, 11, 1143—1147.

Klass K.-D. 1997: The external male genitalia and the phylogeny of

Blattaria and Mantodea. Bonn Zool. Monogr. 42, 1—341.

Klass K.-D. & Meyer R. 2006: A phylogenetic analysis of Dictyo-

ptera based on morphological characters. Entomol. Abh. 63, 1—
2, 3—50.

Langenheim J. 1966: Botanical source of amber from Chiapas,

Mexico. Ciencia 24, 201—211.

Langenheim J.H. 1995: Biology of amber-producing trees: focus on

case studies of Hymenea and Agathis. In: Anderson K.B. &
Krelling J.C. (Eds.): ACS symposium series 617: Amber, resin-
ite, and fossil resins. Amer. Chem. Soc., Washington, 1—31.

Melton R.H. 1995: Differential adaptation to water-deprivation in

first instar nymphs of the German cockroach (Blattella ger-
manica) and the brown-banded cockroach (Supella longipalpa).
Entomol. Exp. Appl. 77, 1, 61—68.

Meneses-Rocha J.J. 2001: Tectonic evolution of the Ixtapa graben,

an example of a strike-slip basin in southeastern Mexico: im-
plications  for  regional  petroleum  systems.  In:  Bartolini  C.,
Buffler  R.T.  &  Cantú-Chapa  A.  (Eds.):  The  Western  Gulf  of
Mexico  Basin:  Tectonics,  sedimentary  basins,  and  petroleum
systems. Amer. Assoc. Petrol. Geol., Mem. 75, 183—216.

Moore J. & Gottelli N.J. 1992: Moniliformis-moniliformis increas-

es cryptic behaviours in the cockroach Supella longipalpa. J.
Parasitol. 78, 1, 49—53.

Narasimham A.U. 1992: Comparative biological parameters of

Comperia merceti (Compere) (Hym., Encyrtidae) and Anasta-
tus tenuipes Bolivar (Hym., Eupelmidae), oothecal parasitoids
of the cockroach Supella longipalpa (Fab.) Biol. Control. 2,
73—77.

Perrilliat M.C., Vega F.J. & Coutin

~o M.A. 2010: Miocene mollusks

from the Simojovel area in Chiapas, southwestern Mexico. J.
South Amer. Earth Sci. 30, 111—119.

Poinar G. Jr. 1992: Life in Amber. Stanford University Press, Stan-

ford, 1—350.

Poinar G. Jr. 2003: Coelomycetes in Dominican and Mexican am-

ber. Mycol Res. 107, 1, 117—122.

Poinar G. Jr. & Brown A.E. 2002: Hymenaea mexicana sp. nov.

(Leguminosae: Caesalpiniodeae) from Mexican amber indi-
cates Old World connections. Botanical J. Linnean Soc. 139,
125—132.

Poinar G., Archibald B. & Brown A. 1999: New amber deposit pro-

vides evidence of Early Paleogene extinctions, paleoclimates,
and past distributions. Canad. Entomol. 131, 171—177.

Prakash S., Mendki M.J., Rao K.M., Singh K. & Singh R.N. 1995:

Sensilla on the maxillary and labial palps of the cockroach Su-
pella longipalpa Fabricius (Dictyoptera: Blattellidae). Int. J.
Insect. Morphol. Embryol. 24, 1, 13—34.

Princis K. 1954: Wo ist die Urheimat von Blatta orientalis L. zu

suchen? Opuscula Entomol. 19, 202—204.

Rehn J.A.G. 1945: Man’s uninvited fellow traveller – the cock-

roach. Sci. Mon. 61, 265—276.

Rehn J.A.G. 1947: African and Malagasy Blattidae (Orthoptera).

Part IV. Proc. Acad. Nat. Sci. Philadelphia 99, 59—92.

Roonwal M.L. & Rathore N.S. 1983: Wing micro-sculpturing in the

small house cockroach, Supella longipalpa (Dictyoptera, Blat-
tidae). Proc. Indian. Acad. Sci. (Anim. Sci.) 92, 333—342.

Roth L.M. 1985: A taxonomic revision of the genus Blattella Cau-

del (Dictyoptera, Blattaria: Blattellidae). Entomol. Scand.,
Suppl. 22, 1—221.

Roth L.M. 1991: New combinations, synonymies, redescriptions,

and new species of cockroaches, mostly Indo-Australian Blat-
tellidae. Invertebr. Taxon 5, 953—1021.

Roth L.M. 1993: The cockroach genus Allacta Saussure & Zehntner

(Blattaria, Blattellidae: Pseudophyllodromiinae). Entomol. Scand.
23, 4, 361—389.

Roth L.M. 1995: New species of Allacta Saussure and Zehntner

from Papua New Guinea, Irian Jaya and Sarawak (Blattaria,
Blattellidae: Pseudophyllodromiinae). PNG J. Agric, For and
Fish 38, 1, 51—71.

Roth L.M. 1996: The cockroach genera Sundablatta Hebard, Pseudo-

phyllodromia Brunner, and Allacta Saussure & Zehntner (Blat-
taria: Blattellidae, Pseudophyllodromiinae). Tijdschrift Entomol.
139, 215—242.

Rudolphi C.A. 1819: Entozoorum Synopsis cui accedunt mantissa

duplex et indice locupletissimi. Berolini, Sumtibus Augusti
Rücker, 1—811.

Santiago-Blay J.A. & Poinar G. Jr. 1993: First scorpion (Buthidae:

Centruroides) from Mexican amber (lower Miocene to upper
Oligocene). J. Arachnology 21, 147—151.

Saussure H. & Zehntner L. 1895: Histoire naturelle des Blattides et

Mantides. In: Grandidier A. (Ed.): Histoire physique, naturelle
et politique de Madagascar. Vol. XXIII: Histoire Naturele des
Orthopt res 1

partie. Blattides et Mantides. Paris, 1—244.

Schmied H. 2009: Cockroaches (Blattodea) of the middle Eocene of

Messel (Germany). Diploma Thesis, University of Bonn.

Serville J.G.A. 1839: Histoire naturelle des insectes Orthopteres.

Acc. de planches. Roret, Paris, 1—776.

Solórzano-Kraemer M.M.S. 2007: Systematic, palaeoecology, and

palaeobiogeography of the insect fauna from Mexican amber.
Paleontographica Abt. A 282, 1—6, 1—133.

Solórzano-Kraemer M.M. & Mohrig W. 2007: Schwenckfeldina

475

AFRO-ASIAN COCKROACH AND LOST TERTIARY AMERICAN ENTOMOFAUNA

GEOLOGICA CARPATHICA

GEOLOGICA CARPATHICA, 2011, 62, 5, 463—475

archoica n. sp. (Diptera, Sciaridae) from the middle Miocene
Mexican amber. Alavesia 1, 105—108.

Swofford D.L. 2002: PAUP*. Phylogenetic analysis using parsimo-

ny (* and other methods), version 4.0b10. Sinauer, Sunder-
land, MA.

Tomasini-Ortíz A.C. & Martínez-Hernández E. 1984: Palinology of

the Eocene-Oligocene of Simojovel, Chiapas. Univ. Nacional
Autónoma de México, Inst. Geol. Paleont. Mexicana, 50, 60 (in
Spanish).

Trjapitzin V.A. 1963: A new hymenopterous genus from Baltic am-

ber. Paleontol. J. 3, 89—95 (in Russian).

Tsai T.J. & Chi H. 2007: Temperature-dependent demography of

Supella longipalpa (Blattodea: Blattellidae). J. Med. Entomol.
44, 5, 772—778.

Tungtrongchitr A., Sookrung N., Munkong N., Mahakittikun V.,

Chinabut P., Chaicumpa W., Bunnag C. & Vichyanond P. 2004:
The levels of cockroach allergen in relation to cockroach species
and allergic diseases in Thai patients. Asian Pac. J. Allerg.
Immunol. 22, 2—3, 115—121.

Vega F.J.T., Nyborg T., Coutin

~o M.A., Solé J. & Hernández-Monzón

O. 2009a: Neogene Crustacea from southeastern Mexico. Bull.
Mizunami Fossil. Mus. 35, 51—69.

Vega F.J., Zún

~iga L. & Pimentel F. 2009b: First formal report of a

crab in amber from the Mocene of Chiapas and other uncom-
mon Crustacea. Geol. Soc. Amer. Abstr. Programs 41, 7, 631.

Veiga-Crespo P., Blasco L., Poza M. & Villa T.G. 2007: Putative

ancient microorganisms from amber nuggets. Int. Microbiology
10, 117—122.

Vidlička . 2001: Blattaria – cockroaches, Mantodea – mantises

(Insecta: Orthopteroidea). Fauna of Slovakia. [Fauna Sloven-
ska.] Veda, Vydavate stvo SAV, Bratislava, 1—171 (in Slovak).

Vršanský P. 1997: Piniblattella gen. nov. – the most ancient genus

of the family Blattellidae (Blattodae) from the Lower Creta-
ceous of Siberia. Entomol. Probl. 28, 1, 67—79.

Vršanský P. 2005: Mass mutations of insects at the Jurassic/Creta-

ceous boundary? Geol. Carpathica 56, 6, 473—781.

Vršanský P. 2008: New blattarians and a review of dictyopteran as-

semblages from the Lower Cretaceous of Mongolia. Acta
Palaeont. Pol. 53, 1, 129—136.

Vršanský P. 2010: Cockroach as the earliest eusocial animal. Acta

Geol. Sin. (EE) 84, 4, 793—808.

Vršanský P. & Ansorge J. 2007: Lower Jurassic cockroaches (In-

secta: Blattaria) from Germany and England. Afr. Invertebr.
48, 1, 103—126.

Vršanský P., Mostovski M.B., Bazylev B.A. & Bugdaeva E. 2002:

Early Cretaceous climate changes suggested on the Basis of
Cockroach Wing Variations. Proceedings of the XVII. Con-
gress of Carpathian-Balkan Geological Association, Bratislava,
September 1st—4th 2002 (CD). Geol. Carpathica, (Spec. Issue)
33, 1—5.

Vršanský P., Vidlička ., Čiampor F. & Marsh F. 2011: Derived, still

living cockroach genus Cariblattoides (Blattida: Blattellidae)
from the Eocene sediments of Green River in Colorado, U.S.A.
Ins. Sci. 18, 2. DOI: 10.1111/j.1744-7917.2010.01390.x.

Walker F. 1868: Catalogue of the specimens of Blattariae in the col-

lection of the British Museum. British Mus. (Nat. Hist.), Lon-
don, 1—239.

Ware J., Litman J., Klass K.-D. & Spearman L.A. 2008: Relation-

ships among the major lineages of Dictyoptera: the effect of
outgroup selection on dictyopteran tree topology. Syst. Ento-
mol. 33, 429—450.

Zherikhin V.V. 1970: Zoogeographical relation of Paleogene insects.

Doklady na 22. Ježegodnom čtenii pamâti N.A. Kholodkowsko-
go, 14th April 1969. Nauka, Leningrad, 29—88 (in Russian).