Content uploaded by Louis L. Jacobs
Author content
All content in this area was uploaded by Louis L. Jacobs on Nov 08, 2015
Content may be subject to copyright.
ARTICLE
LENDE CHIWETA, A NEW THERAPSID FROM MALAWI, AND ITS INFLUENCE ON
BURNETIAMORPH PHYLOGENY AND BIOGEOGRAPHY
ASHLEY KRUGER,*
,1
BRUCE S. RUBIDGE,
1
FERNANDO ABDALA,
1
ELIZABETH GOMANI CHINDEBVU,
2
and LOUIS L. JACOBS
3
1
Evolutionary Studies Institute and School of Geosciences, University of the Witwatersrand, Johannesburg, WITS 2050,
South Africa; ashleykruger@gmail.com; bruce.rubidge@wits.ac.za; nestor.abdala@wits.ac.za;
2
Ministry of Sports and Culture, Private Bag 384, Lilongwe 3, Malawi, egomanichindebvu@yahoo.com;
3
Roy M. Huffington Department of Earth Sciences, Southern Methodist University, Dallas, Texas 75275, U.S.A.,
jacobs@mail.smu.edu
ABSTRACT—The Chiweta Beds of Malawi have yielded a diverse late Permian fossil tetrapod fauna that correlates with
that of the Cistecephalus Assemblage Zone of the South African Karoo Supergroup. Amongst the fossil therapsids from the
Chiweta Beds is the well-preserved skull and lower jaw of a burnetiamorph, a group of biarmosuchians with numerous bosses
and swellings on the skull. This specimen was reported in a preliminary paper in 2005 as the first burnetiamorph described
outside of South Africa and Russia. Reanalysis of the morphology and phylogeny of this specimen places Lende chiweta, gen.
et sp. nov., as the sister taxon to the clade formed by Proburnetia (Paraburnetia (Pachydectes, Bullacephalus, Burnetia,
Niuksenitia)). The greatest diversity of this basal therapsid group is from South Africa, with six of nine described genera and
a stratigraphic range that extends from the middle Permian Tapinocephalus Assemblage Zone to the upper Permian
Dicynodon Assemblage Zone. Bearing in mind the constraints that govern fossil preservation, current data suggest that what
is now southern Africa may have been the area of origin for burnetiamorphs. Under this premise, what is now central Africa
represented a corridor that allowed migration of representatives of the group between the southern and northern portions of
Pangea during the late Permian.
http://zoobank.org/urn:lsid:zoobank.org:pub:BB094736-24AE-4B98-A33CBD7B7668DC2F
SUPPLEMENTAL DATA—Supplemental materials are available for this article for free at www.tandfonline.com/UJVP
Citation for this article: Kruger, A., B. S. Rubidge, F. Abdala, E. Gomani Chindebvu, and L. L. Jacobs. 2015. Lende chiweta,a
new therapsid from Malawi, and its influence on burnetiamorph phylogeny and biogeography. Journal of Vertebrate
Paleontology. DOI: 10.1080/02724634.2015.1008698.
INTRODUCTION
The basal therapsid clade Biarmosuchia was first proposed by
Hopson and Barghusen (1986), but until that time, most genera
of this taxon were considered as part of the Gorgonopsia. Sigog-
neau (1970), in her systematic revision of the Gorgonopsia, rec-
ognized five genera (Hipposaurus, Ictidorhinus, Lycaenodon,
Lemurosaurus, and Rubidgina) within the family Ictidorhinidae.
After subsequent taxonomic revision of the ‘ictidorhinids,’
Sigogneau-Russell (1989) recognized four families within the
Biarmosuchia: Biarmosuchidae (including Biarmosuchus), Hip-
posauridae (including Hipposaurus and Lycaenodon), Ictidorhi-
nidae (including Ictidorhinus, Lemurosaurus, and Rubidgina),
and Burnetiidae (including Burnetia, Proburnetia, and Styracoce-
phalus). Rubidge and van den Heever (1997) subsequently iden-
tified Styracocephalus as a dinocephalian.
Hopson and Barghusen (1986) considered the Biarmosuchia to
be characterized by the possession of primitive therapsid charac-
ters but were unsure whether the group was mono- or paraphy-
letic. Hopson (1991) later proposed several characters
supporting the monophyly of biarmosuchians. Over the past two
decades, several new biarmosuchian genera have been described
(Sidor, 2000; Rubidge and Kitching, 2003; Sidor and Welman,
2003; Sidor et al., 2004; Jacobs et al., 2005; Sidor and Rubidge,
2006; Smith et al., 2006; Sidor and Smith, 2007). Many of these
belong to the Burnetiamorpha, which has become the most spe-
cies-rich subclade of biarmosuchians.
Originally, the family Burnetiidae was erected for Burnetia
from South Africa (Broom, 1923). Recent discoveries of
more specimens, however, prompted new studies on burnetia-
morphs, which have recognized these therapsids as a distinct
monophyletic group within Biarmosuchia (Sidor and
Rubidge, 2006; Kemp, 2012), which includes all taxa more
closely related to Burnetia than to Ictidorhinus or Hipposau-
rus (Sidor, 2000; Sidor and Welman, 2003). Currently, the
Burnetiamorpha comprise nine genera: Bullacephalus, Burne-
tia, Lemurosaurus, Lobalopex, Lophorhinus, Paraburnetia,
and Pachydectes from South Africa, and Niuksenitia and Pro-
burnetia from Russia. In addition, Sidor et al. (2010) recently
described a partial skull roof including the dorsal margin of
orbits and parietal foramen of an unnamed burnetiid from
the upper Permian of Tanzania, and Sidor et al. (2014) noted
the presence of a burnetiid in the middle Permian of Zambia.
Despite the fact that several genera have been described,
burnetiamorphs are relatively rare, with each genus being
represented by one or two specimens, and most are from the
South African Permian beds of the Beaufort Group.
*Corresponding author.
Journal of Vertebrate Paleontology e1008698 (10 pages)
!by the Society of Vertebrate Paleontology
DOI: 10.1080/02724634.2015.1008698
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
Jacobs et al. (2005), in a report on the late Permian fauna of
the Chiweta Beds of northern Malawi, which they correlated
with the Cistecephalus Assemblage Zone of the South African
Karoo, briefly described an unnamed burnetiamorph skull
(MAL 290). Their preliminary phylogenetic analysis recovered it
as the sister taxon to Proburnetia, Burnetia, and Bullacephalus.
Subsequent preparation has exposed additional portions of the
skull and allowed for a full description and a more comprehen-
sive phylogenetic analysis.
Institutional Abbreviations—BP, Evolutionary Studies Insti-
tute (formerly the Bernard Price Institute for Palaeontological
Research), University of the Witwatersrand, Johannesburg,
South Africa; CGP, and also; WB123, Council for Geosciences
(formerly the Geological Survey of South Africa), Pretoria,
South Africa; MAL, Malawi Department of Antiquities Collec-
tion, Lilongwe and Nguludi, Malawi; NHMUK, Natural History
Museum, London, U.K.; NMQR, National Museum, Bloemfon-
tein, South Africa; NMT, National Museum of Tanzania, Dar es
Salaam, Tanzania; RC, Rubidge Collection, Graaff-Reinet,
South Africa; SAM, Iziko South African Museum, Cape Town,
South Africa.
Anatomical Abbreviations—a, articular; bo, basioccipital; bsp,
basisphenoid; d, dentary; eam, external auditory meatus; ec,
ectopterygoid; f, frontal; fb, frontal boss; fm, foramen magnum;
j, jugal; lac, lacrimal; m, maxilla; n, nasal; nb, nasal boss; op, opis-
thotic; p, parietal; pal, palatine; pf, postfrontal; pin, pineal fora-
men; pm, premaxilla; po, postorbital; pp, postparietal; prf,
prefrontal; ps, parasphenoid; pt, pterygoid; q, quadrate; qpt,
quadrate ramus of pterygoid; sm, septomaxilla; so, supraoccipi-
tal; sp, splenial; sq, squamosal; sur, surangular; t, tabular; v,
vomer.
MATERIALS AND METHODS
The specimen MAL 290 was discovered in 1992 as a result of
collaborative field work by members of the Malawi Department
of Antiquities and Southern Methodist University during a day
trip to the Permian Karoo Supergroup rocks near Chiweta
(Jacobs et al., 2005). Following a preliminary description, the
specimen was further prepared at the Evolutionary Studies Insti-
tute, where a detailed study was undertaken.
For comparative purposes, the following specimens were
examined: BP/1/816, holotype of Lemurosaurus pricei; BP/1/
3924, referred specimen of Herpetoskylax hopsoni; BP/1/5387,
holotype of Bullacephalus jacksoni; BP/1/5735, holotype of
Pachydectes elsi; CGP/1/61, holotype of Lobalopex mordax;
CGP/1/67, holotype of Herpetoskylax hopsoni; NHMUK R5700,
holotype of Lycaenodon longiceps; cast of NHMUK R5397,
holotype of Burnetia mirabilis; NMQR 1702, referred specimen
of Lemurosaurus pricei; NMT RB4, unnamed burnetiamorph;
RC55, holotype of Rubidgina angusticeps; SAM-PK-K6655,
holotype of Lophorhinus willodenensis; and WB123, undescribed
specimen of Hipposaurus boonstrai. Information on Ictidorhinus
and Niuksenitia was obtained after Boonstra (1935), Tatarinov
(1977), Sigogneau-Russell (1989), and Ivakhnenko (2002).
A cladistic analysis was undertaken based on a data matrix of
36 characters and 15 taxa (see Supplementary Data). Twenty-
three characters in the matrix were previously used by Rubidge
et al. (2006); six characters were also included in previous phy-
logenies (e.g., Rubidge and Kitching, 2003; Sidor and Welman,
2003), and seven characters are proposed here for the first time.
Taxa included in the analysis are Biarmosuchus, Bullacephalus,
Burnetia, Herpetoskylax, Hipposaurus, Ictidorhinus, Lemurosau-
rus, Lobalopex, Lophorhinus, Lycaenodon, Niuksenitia, Pachy-
dectes, Paraburnetia, and Proburnetia, with Biarmosuchus
representing the outgroup used for character polarization. The
main intention of the analysis is to explore the monophyly and
internal groups of Burnetiamorpha.
As knowledge of the postcranium of the vast majority of biar-
mosuchian taxa is poor, the data matrix incorporates only cranio-
dental characters. The data matrix used in the analysis is freely
available on Morphobank as project P1190.
The program TNT (Goloboff et al., 2003) was used to search
for the most parsimonious trees. All characters were given equal
weight, and multistate characters with states representing mor-
phologically adjacent conditions were considered ordered (Lips-
comb, 1992). Considering the small size of the data matrix, a
search strategy of implicit enumeration, guaranteeing the recov-
ery of the shortest most parsimonious tree(s), was used. Bremer
support and resampling were calculated to evaluate the reliabil-
ity of the monophyletic groups recovered. Sensitivity analyses
implying alteration in additivity of characters and using implied
weights (Goloboff, 1993) were also considered, and the resulting
most parsimonious trees compared with those of our original
analysis.
SYSTEMATIC PALEONTOLOGY
THERAPSIDA Broom, 1905
BIARMOSUCHIA Sigogneau-Russell, 1989
BURNETIAMORPHA Broom, 1923
Diagnosis—Supraorbital boss present, antorbital fossa or pit
on lateral surface of lacrimal; median frontal ridge present but
variably expressed; boss on ventral surface of squamosal lateral
to level of quadrate; boss present on lower margin of zygomatic
arch at level of postorbital bar (Sidor and Welman, 2003).
LENDE CHIWETA, gen. et sp. nov.
(Figs. 1–3)
Holotype—MAL 290, almost complete skull with lower jaw.
Etymology—MAL 290 was informally referred to as the ‘head
of the devil’ because of reminiscent excrescences on the skull. In
recognition of that whimsical reference and more seriously in
honor of the culture of Malawi, the generic name is given for a
member of Gule wa Mkulu, masked dancers that bridge the spirit
world in Chewa culture (Malawi transliterated spelling;
UNESCO, 2014). ‘Lende’ is a notorious figure in the Gule wa
Mkulu who appears at night and sometimes disappears from
view. The specific epithet refers to the type locality and should
be treated as a noun in apposition.
Diagnosis—Small burnetiamorph, with short snout; high and
long ridge-like midline nasal boss extending posteriorly up to the
anterior margin of the orbit; very robust and high frontal boss
(as high as the supraorbital boss); flat to slightly concave dorsal
surface of parietal surrounding parietal foramen, lacking chim-
ney and high angulation (105!) between the preorbital and post-
orbital portions of the cranial roof; midline ridge of postparietal
does not extend dorsally onto parietal.
Locality and Age—Lower Bone Bed (B1) of the Chiweta
Beds, Malawi. Correlated with the Cistecephalus Assemblage
Zone of South Africa, and therefore also the Usili Formation
of Tanzania, and the Upper Madumabisa Mudstone Forma-
tion of Zambia (Jacobs et al., 2005; Sidor et al., 2013). Late
Permian.
Comments—The holotype was prepared in a way that new
information, especially from the palate, is now available. Consid-
ering the exceptionally large orbit, it is possible that the speci-
men represents a juvenile individual and that diagnostic
characters such as snout length may prove to be linked with age.
This is also the case with the dorsal skull angulation, because
there is evidence that change in the dorsal profile of the skull
(calvaria of living mammals) is linked with deflection of the ros-
trum in relation to the basicranium during ontogeny (Lieberman
et al., 2008; Flores et al., 2010).
Kruger et al.—A new burnetiamorph from Malawi (e1008698-2)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
DESCRIPTION
The skull and lower jaw of Lende chiweta is well preserved,
only slightly distorted on the left side and almost complete
except for damage to the front of the snout and the anterior end
of the right ramus of the lower jaw. Because of the pachyostotic
nature of the bone, it is difficult to identify sutures. The high and
narrow snout of Lende makes up approximately one-third of the
skull length.
In lateral view, Lende exhibits a prominent parietal boss post-
erodorsal to the temporal opening, a triangular supraorbital
boss, and large, elongated frontal and nasal bosses. No parietal
chimney is present. The temporal fenestra is positioned behind
and somewhat below the orbit, and the postorbital region of the
skull is short compared with the length of the snout, features that
are reminiscent of pelycosaur-grade synapsids and basal
therapsids.
Skull Roof
The premaxilla, which forms the tip of the snout, is damaged at
its anterior end and has a relatively short dorsal process, which
does not extend far on the snout. This differs from the basal biar-
mosuchians Hipposaurus and Biarmosuchus, which exhibit a rel-
atively long dorsal process of the premaxilla that reaches
halfway between the tip of the snout and the orbit (Sidor and
Rubidge, 2006). Although five incisors are typical for
burnetiamorpha, three small incisors are preserved on the left
premaxilla and two are visible on the right side (Fig. 1).
Only a small triangular sliver of the septomaxilla facial process
is preserved and is directed posteriorly between the maxilla and
the nasal on the left side of the specimen.
The maxilla is a large bone that makes up most of the lateral
surface of the snout (Fig. 1B, D). Because of the weathered
nature of the anterior end of the skull, it is not possible to recog-
nize the suture between maxilla and premaxilla. The maxilla has
a long, dorsal sutural contact with the nasal, a pointed postero-
dorsal contact with the prefrontal, and a vertical posterior con-
tact with the lacrimal. On the left side of the skull, a large ovoid
fossa is present on the contact between the maxilla, the lacrimal,
and the jugal, but it is absent on the right side. The maxilla makes
up the ventral margin of the snout and forms a long, tapering
posterodorsal sutural contact with the jugal until half way along
the length of the orbit, where it appears to touch a thin antero-
ventral extension of the squamosal on the ventral margin of the
skull. The jugal is a prominent bone of the zygomatic arch and
almost forms the entire ventral margin of the orbit. Its posterior
contact with the postorbital is unclear, and it appears to form the
anteroventral margin of the temporal fenestra.
The nasals cover a large area on the dorsal side of the snout.
Anteriorly they form the roof of the external nares; they have a
long lateral contact with the maxilla and respectively meet the
prefrontal and frontal posterolaterally and posteromedially. The
FIGURE 1. Lende chiweta, gen. et sp. nov., MAL 290, holotype, skull and mandible. A, photograph, and B, illustration of specimen in left lateral
view; C, photograph, and D, illustration of specimen in right lateral view. Scale bar equals 1 cm.
Kruger et al.—A new burnetiamorph from Malawi (e1008698-3)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
nasal has only a short, pointed contact with the lacrimal, which is
visible on the left side. A narrow elongated but prominent nasal
crest is present and is similar to that of Lobalopex (Sidor et al.,
2004) and Lophorhinus (Sidor and Smith, 2007). Anteriorly, it
terminates immediately behind the external naris, and posteri-
orly above the anterior margin of the orbit.
In lateral view, the lacrimal is rectangular and has a clear dor-
sal sutural contact with the prefrontal. The anterior margin con-
tacts the maxilla, and posteroventrally it meets the jugal. A fossa
is present on the dorsal margin of the lacrimal, but it is larger on
the right side of the skull than it is on the left.
The prefrontal is a large, broadly triangular, flattened bone,
which forms the anterodorsal margin of the orbit but does not
contribute to thickening of the orbital margin. It contacts the
maxilla anteroventrally, the nasal dorsally and anteriorly, and
the frontals at the basal margin of the supraorbital boss. As is
evident on the left side of the skull (Fig. 1B), the prefrontals con-
tact the lacrimal ventrally by means of a long, horizontally ori-
ented suture positioned half way down the anterior orbital
margin.
The paired frontals make up most of the skull roof (Fig. 2A,
B). Anteriorly they meet the nasals but the suture is unclear, and
they meet the prefrontals anterolaterally. The nasals make up a
large part of the dorsal margin of the orbit and have a long
curved posterolateral contact with the postfrontals. The nature
of the posterior contact with the parietals is uncertain. Laterally
the frontals are thickened and contribute to two slightly pachyos-
tosed supraorbital bosses on either side of the skull, in contrast to
the condition in other burnetiamorphs such as Lophorhinus and
Lemurosaurus, where the prefrontals contribute to the formation
of the supraorbital boss (Sidor and Welman, 2003; Sidor and
Smith, 2007). Between the supraorbital bosses of the frontals, a
large elongated midline frontal boss is present. This boss is of
comparable height to the supraorbital bosses and is slightly
pachyostotic. In dorsal view, the midline boss is broadly triangu-
lar, displaying its greatest width posteriorly and tapering anteri-
orly to a point that is in line with the anterior margin of the
orbits. The highest point of the frontal boss is situated in line
with the midpoint of the orbits and is relatively much larger than
that of Lemurosaurus. At the posterior end, the frontal boss flat-
tens without secondary thickening and apparently terminates in
front of the parietal, but the contact between the frontal and the
parietal is not clear. The postfrontals form a small portion of the
posterodorsal margin of the orbit, contacting the frontals and
contributing to the pachyostosed supraorbital boss in a similar
way to that of Lobalopex (Sidor et al., 2004). This protuberance
is of comparable height to the median frontal boss. The postfron-
tals extend posteriorly as they form the posterodorsal margin of
the orbit and have a sutural contact with the postorbitals poste-
rior to the orbit.
The postorbital is a large bone forming a substantial part of
the posterior margin of the orbit and the anterior and dorsal
margins of the temporal fenestra. In lateral view, the postorbital
contacts the jugal on the posteroventral border of the orbit and
the postfrontal on its posterodorsal margin. Dorsomedially the
postorbital contacts the parietal on the skull roof. The tabular
and squamosal bones contact the postorbital posteriorly.
The squamosal is a large bone forming the posterior and ven-
tral margins of the temporal opening. Anteromedially, it is in
sutural contact with the postfrontal and likely the postorbital. In
lateral and occipital views, the squamosal forms a ‘squamosal
horn’ (Figs. 1, 3), located on the posterodorsal margin of the
temporal fenestra. Below the temporal fenestra, the squamosal
forms a prominent boss on the ventral margin of the zygomatic
arch. Anterodorsally it meets the jugal by means of a horizontal
suture below the postorbital bar, and it appears to extend anteri-
orly as a thin wedge along the ventral margin of the zygomatic
arch to contact the maxilla below the orbit.
As in most other biarmosuchians, the presence of a prepar-
ietal is difficult to determine because of the pachyostotic
nature of the bones of the skull roof. In many biarmosu-
chians, this bone is believed to form the anterior margin of
the parietal foramen, being part of the raised chimney (Sidor
and Welman, 2003, Sidor et al., 2004). In Lende, the pineal
foramen is located posterior to the base of the frontal boss,
but no pineal boss is present. In Paraburnetia, Proburnetia,
and Bullacephalus, the parietals create a broad swelling
around the pineal foramen but do not create a chimney,
whereas Lemurosaurus, Hipposaurus,andHerpetoskylax
have a distinctive parietal chimney (Rubidge and Sidor, 2002;
Rubidge and Kitching, 2003; Sidor and Welman, 2003; Sidor
and Rubidge, 2006).
The parietal makes up a large portion of the posterodorsal sur-
face of the skull and quite certainly extends onto the occiput,
although the sutures of the bony elements on the occipital area
are very difficult to trace (Fig. 2A, B). Anteriorly the parietal
contacts the frontals and postfrontals, whereas laterally it con-
tacts the postorbitals. In dorsal view, it is evident that the parietal
is situated between the postorbital and the dorsal margin of the
tabular.
Palate
The palate of MAL 290 (Fig. 2C) is reasonably preserved;
however, it is difficult to place palatal sutures with confidence.
The vomer is a long, thin bone with an elongated, anteroposter-
iorly extending trough, produced by the downturned lateral mar-
gins. Because no midline suture is visible, it appears that the
vomer is unpaired. Posteriorly it narrows to meet the palatine
behind the internal nares, but the nature of this contact is
unclear.
The paired palatine forms a large proportion of the palatal sur-
face. Anterolaterally, it overlies the maxilla and extends posteri-
orly from the level of the canine along the posterolateral and
posterior margins of the internal nares to meet the vomer medi-
ally. The left and right palatines are separated by a long trough
between the prominent dentigerous palatine bosses. Three and
four small teeth are visible on the left and right bosses, respec-
tively. These teeth are difficult to see, and there is no apparent
pattern of tooth distribution. On its lateral side, the palatine boss
slopes in a dorsal direction and forms a horizontal platform close
to its lateral contact with the maxilla.
Posterolaterally, the palatine meets a small triangular
ectopterygoid, which is positioned on the anterior flank of the
lateral process of the pterygoid. The tripartite pterygoid com-
prises an anterior medial process, the lateral processes, and
the posterior process, which extends posterolaterally to form
the quadrate ramus. The anterior process makes up the den-
tigerous pterygoid boss, which is confluent with the palatine
boss but is much smaller. The suture between the palatine
and pterygoid is not visible. Four teeth are present on the
pterygoid boss, two on each side, but no pattern can be ascer-
tained. A broad anterolaterally oriented trough is present
between the pterygoid boss and the lateral process. In ventral
view, the edentulous lateral process of the pterygoid is broad
laterally and tapers posteromedially to form a narrow ridge
that curves posteromedially to the interpterygoidal vacuity,
from where it runs posteriorly, parallel to the midline, to the
posterior end of the pterygoid. Another ridge extends down
the lateral side of the pterygoid. Anteriorly, it begins as a low
ridge immediately behind the pterygoid process, and becomes
more accentuated as it extends onto the quadrate ramus. Only
the anterior portions of the quadrate rami on either side are
preserved, and at this point the quadrate ramus is a flat hori-
zontally oriented bone. A narrow interpterygoid vacuity is
present, with its anterior margin in line with the anterior
Kruger et al.—A new burnetiamorph from Malawi (e1008698-4)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
FIGURE 2. Lende chiweta, gen. et sp. nov., MAL 290, holotype, skull and mandible. A, photograph, and B, illustration of specimen in dorsal view;
C, photograph, and D, illustration of specimen in ventral view. Scale bar equals 1 cm.
Kruger et al.—A new burnetiamorph from Malawi (e1008698-5)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
margin of the lateral process, and a further vacuity is present
at the posterior end of the pterygoid.
The basisphenoid is a relatively small quadrangular bone,
which has a bulbous midline, pointed contact with the pterygoid,
and posteriorly has a broad transverse sutural contact with the
basioccipital half way along the fenestra ovalis. The stapes is not
preserved. A broad trough extends down the midline of the
basisphenoid.
Occiput
The occiput of Lende (Fig. 3) is transversely rectangular, ori-
ented vertically, and curves posterolaterally. Sutural borders of
the postparietal, parietal, tabular, and supraoccipital bones are
clearly delineated. Both quadrates are preserved. The single
postparietal has the form of an inverted triangle with the apex
pointing down the midline of the occipital surface. It overlies the
FIGURE 3. Lende chiweta, gen. et sp. nov.,
MAL 290, holotype, skull and mandible.
A, photograph, and B, illustration of speci-
men in occipital view. Scale bar equals 1 cm.
Kruger et al.—A new burnetiamorph from Malawi (e1008698-6)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
supraoccipital and extends ventrally to a point immediately dor-
sal to the foramen magnum. Dorsally the postparietal has a long
contact with the parietal on the occipital surface, and laterally it
meets the tabular. A prominent midline ridge extends the entire
length of the postparietal but not onto the parietal. This differs
from the situation in other biarmosuchians, where the ridge
extends along the parietal from the posterior surface of the pari-
etal chimney.
The tabulars are square, flat bones, but the sutures are not
clear. Dorsomedially the tabular is in contact with the posterior
side of the parietal, whereas dorsolaterally, laterally, and ventro-
laterally it has a long contact with the squamosal.
The supraoccipital is a large, transverse bone at the base of the
occipital surface and forms the dorsal and a portion of the lateral
margin of the foramen magnum. Dorsolaterally, it has a short
contact with the tabular.
The opisthotic is preserved on the left side by its paroccipital
process, which is a flat and transversely triangular bone, posi-
tioned ventral to the supraoccipital. The paroccipital process of
the opisthotic contacts the basioccipital medially.
The exoccipitals could not be discerned, and the basioccipital,
which forms the entire ventral margin of the foramen magnum,
meets the basisphenoid anteriorly on the ventral surface of the
skull. What appears to have been a single occipital condyle has
been largely eroded away.
The quadrate and quadratojugal are preserved on the left
side but are not well exposed. In occipital view (Fig. 3), the
quadrate is dorsally broad and partially overlain dorsally
by the squamosal, and thins ventrally to form the quadrate
condyles.
Lower Jaw
The left lower jaw is complete, but the right ramus is damaged
anteriorly and weathered posteriorly such that portions of the
dentary are not preserved. Accordingly, this description is based
largely on the left ramus (Fig. 1A, B). The mandible is deep pos-
teriorly and shallows towards its thinnest point just behind the
canine tooth.
The dentary forms most of the lateral surface of the mandible.
In lateral view, the canine and postcanine teeth are not visible,
but three incisors are present toward the anterior end. The lat-
eral surface of the dentary is flat and smooth, but its surface
changes ventral to the orbit, where the dorsal margin swells and
forms an overhanging dentary ledge that extends posteriorly
onto the surangular. This feature is also present in Proburnetia
(Rubidge and Sidor, 2002), Herpetoskylax (Sidor and Rubidge,
2006), and Lemurosaurus (Sidor and Welman, 2003). The poste-
rior margin of the dentary slopes anteroventrally. Posterodor-
sally the dentary contacts the surangular, which makes up the
posterodorsal end of the lower jaw. Medially the dentary is
largely overlain by the splenial anteroventrally and by the prear-
ticular slightly more dorsally.
The angular is a relatively small bone forming the posteroven-
tral portion of the lateral side of the mandible and contacting the
dentary anteriorly. A small, reflected lamina is present and its
lateral surface is devoid of ornamentation (Fig. 1A, B).
The splenial is a relatively large flat bone on the anterome-
dial side of the lower jaw. It is wide anteriorly and thins pos-
teriorly to wedge out on the ventral margin of the jaw ramus
posterior to the lateral process of the pterygoid. The articular
is not preserved on the left ramus and on the right side it is
not visible, but the prearticular is well exposed on the left
ramus, where it extends as a thin flat rod from the postero-
ventral margin of the lower jaw to wedge out along the dor-
sal margin of the splenial at the level of the anterior margin
of the palatine boss.
Dentition
The left upper canine is complete, whereas the right one
exhibits a damaged tip (Figs. 1, 2). The long recurved canines,
which are oval in cross-section and lack serrations, extend almost
as far as the ventral margin of the dentary when the jaws are
occluded. Six small, pointed upper postcanine teeth, which lack
serrations, are present on the left side, but none are preserved on
the right (Fig. 1). Three incisors, which also lack serrations, are
preserved on the left premaxilla, and two slightly eroded incisors
are present on the right. Only the basal portion of the lower inci-
sors is preserved, and as a result it is not possible to determine
whether intermeshing of the incisors occurred during occlusion.
DISCUSSION
Because of the presence of the following characters, it is evi-
dent that Lende is a burnetiamorph biarmosuchian (Jacobs et al.,
2005): lateral surface of lacrimal bears one or more deep fossae;
dorsal surface of frontals are thickened; presence of bosses on
ventral surface of zygomatic arch and suborbital bar; the pres-
ence of a squamosal boss; large midline frontal ridge; and small
boss on anterior margin of squamosal.
Phylogenetic Analysis
Our analysis yielded 32 most parsimonious trees (MPTs), with
the majority-rule consensus shown in Figure 4A. In the majority-
rule tree, Lende is sister to the (Proburnetia (Paraburnetia (Bur-
netia, Bullacephalus, Pachydectes, Niuksenitia))) clade. Herpe-
toskylax and Lycaenodon form a polytomy with burnetiamorphs.
Most of the lack of resolution in the majority-rule tree is concen-
trated within the Burnetiamorpha. Indeed, there are two polyto-
mies, one including Lophorhinus, Lobalopex, and the remaining
burnetiamorphs (excluding the basal Lemurosaurus), and a sec-
ond that includes Burnetia, Bullacephalus, Pachydectes, and
Niuksenitia (Fig. 4A). Bremer support is 1 for all the groups, and
there are no bootstrapping values higher than 40. Therefore,
monophyletic groups are poorly supported in the most parsimo-
nious trees.
Placement of the Malawian taxon in this analysis is consis-
tent with the results of the cladistic analysis of Jacobs et al.
(2005). The differences are the result of enhanced sampling,
e.g., Paraburnetia and Pachydectes were not yet described at
the time of Jacobs et al. (2005), and Bullacephalus is recov-
ered as sister taxon of Burnetia in the analysis of those
authors.
Four MPTs resulted from an analysis in which all multistate
characters were considered as unordered (Fig. 4B). The resolu-
tion of the consensus is enhanced with the recovery of a mono-
phyletic clade including Lophorhinus and Lobalopex, and
resolution of the major polytomy in the previous analysis:
Niuksenitia (Burnetia (Pachydectes, Bullacephalus)). There is no
further resolution of polytomies in non-burnetiamorph biarmo-
suchians. Similar to Rubidge et al. (2006), our analysis indicates
that the clade of Pachydectes and Bullacephalus is nested deeply
in the burnetiamorphs.
Analysis using implied weights (K function D3) with ordered
multistate characters resulted in three MPTs. With regard to bur-
netiamorphs, the consensus tree in this analysis is similar to that
of the main analysis, but Herpetoskylax is recovered as the sister
taxon to burnetiamorphs.
Finally, the consensus of the three MPTs obtained in an analy-
sis with implied weights (K function D3) and unordered multi-
state characters shows some differences to the other results, viz.,
the sister-group relationship between Lende and Lobalopex and
the polytomy represented by node ‘a’ (Fig. 4C).
Kruger et al.—A new burnetiamorph from Malawi (e1008698-7)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
Biogeography
Burnetiamorphs have long been known from South Africa and
Russia, and more recently a fragment of skull roof (NMT RB4)
was described from Tanzania (Sidor et al., 2010). Lende is the
first burnetiamorph reported from Malawi (Jacobs et al., 2005),
but as pointed out by Sidor et al. (2010:702), the faunal similarity
and geographic proximity of the Tanzanian and Malawian locali-
ties suggests that the “two sets of exposures might represent
parts of the same original basin.” The unnamed Tanzanian speci-
men (NMT RB4; Sidor et al., 2010) is a larger individual than
the Lende holotype and has two larger and more pachyostosed
supraorbital bosses, but surprisingly, the frontal boss of Lende is
much larger than that of the Tanzanian specimen. The Tanza-
nian NMT RB4 thus appears to belong to a different taxon to
Lende, but because of the fragmentary nature of the latter speci-
men, further comparison between the two skulls is not possible
at this stage and is dependent on finding better-preserved
specimens.
The results of our phylogenetic analysis and the paleodistribu-
tion of burnetiamorph taxa suggest that what is now southern
Africa was a potential area of origin for burnetiamorphs. This
proposal is supported by the following observations: (1) the most
basal taxon in our phylogeny (Fig. 4B, C), and those identified in
previous phylogenetic analyses of burnetiamorphs (Sidor and
Wellman, 2003; Rubidge et al., 2006; Sidor and Smith, 2007), is
known from South Africa; (2) the oldest burnetiamorphs are
known from South Africa; and (3) the majority of the known spe-
cies (six of nine) are from the Karoo Basin. Alternatively,
because burnetiamorphs are now known from South Africa,
Malawi, Tanzania, Zambia, and Russia, they could have had a
Pangea-wide distribution and the biogeographic pattern evident
from the cladogram results from biased sampling because of
problems relating to non-preservation in the fossil record, as was
suggested by Sidor and Welman (2003).
The South African Karoo Basin preserves the longest tempo-
ral distribution of the Burnetiamorpha, which ranges from the
middle Permian Tapinocephalus Assemblage Zone (AZ) until
the late Permian Dicynodon AZ. Each of the described genera
(apart from Lemurosaurus) is represented by only a single speci-
men. Currently, the only named middle Permian genera (Bulla-
cephalus and Pachydectes) are from South Africa, and Sidor
et al. (2014) mentioned the presence of this group in middle
Permian deposits of Zambia, but this material has not been
described. Of the late Permian taxa, five (Lobalopex, Lophorhi-
nus, Niuksenitia, Proburnetia, and Paraburnetia) are from the
stratigraphic equivalent of the Tropidostoma AZ; two (Lemuro-
saurus and Lende) are from the Cistecephalus AZ; and one (Bur-
netia) is from the Dicynodon AZ. Therefore, at this stage of
knowledge, the heyday of burnetiamorphs was in Tropidostoma
AZ times, at which stage they were widely distributed in Pangea.
Of importance is the fact that the phylogeny presented in this
work is not in agreement with the geological record of burnetia-
morphs, because the basal representative Lemurosaurus is from
the Cistecephalus AZ (Sidor and Welman, 2003), whereas Pachy-
dectes and Bullacephalus from the older Tapinocephalus AZ are
sister taxa in a polytomy following Lemurosaurus (see Fig. 4).
The record of the large burnetiamorph Pachydectes (skull length
of 318 mm) in the Tapinocephalus AZ indicates that in the mid-
dle Permian, burnetiamorphs and basal biarmosuchians (of the
large size of Hipposaurus) were members of the large predator
guild of the time, in conjunction with anteosaurid dinocephalians
and basal therocephalians. The middle Permian burnetiamorphs
have strongly pachyostotic cranial bones, a condition that is
greatly reduced in the medium to small burnetiamorphs from the
Tropidostoma and Cistecephalus AZs. Cranial pachyostosis
regains importance in the latest Permian Burnetia from the
Dicynodon AZ.
Discovery of late Permian burnetiamorphs in at least four
basins in Africa (Jacobs et al., 2005; Sidor et al., 2010, 2014; this
work) suggests that this geographic region was part of a corridor
permitting the migration of the clade between ‘Gondwana’ and
FIGURE 4. A, majority-rule consensus tree of the main analysis (equally
weighted and ordered multistate characters). Numbers indicate the fre-
quency a clade is represented in the most parsimonious trees; B, strict
consensus tree from analysis with unordered multistate characters;
C, strict consensus tree from analysis with unordered multistate characters
and implied weights. The ‘a’ indicates the polytomy discussed in the text.
Only burnetiamorph taxa are represented in Band C. Asterisks denote
Laurasian taxa.
Kruger et al.—A new burnetiamorph from Malawi (e1008698-8)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
‘Laurasia,’ where the group is represented by two Russian spe-
cies during the late Permian. This supports the idea of a broadly
distributed and highly connected therapsid community during
the Permian (Sidor et al., 2013).
CONCLUSIONS
Lende chiweta, a new genus and species, is the first burnetia-
morph from Malawi, and because the specimen is relatively well
preserved, it adds to morphological, stratigraphic, and biogeo-
graphic understanding of the taxon. Burnetiamorph relationships
are poorly understood, and this can be attributed to three main
reasons. The paucity of specimens means that, of the known gen-
era, only one (Lemurosaurus) is represented by more than one
specimen. Secondly, because of the pachyostosed nature of the
skull roof, sutural contacts are difficult to place with certainty
(Rubidge and Sidor, 2002; Sidor and Welman, 2003). Finally,
knowledge of some species (e.g., Pachydectes elsi) is based on
extremely poorly preserved material that provides low-quality
anatomical information.
Our phylogeny (Fig. 4) confirms that Lemurosaurus is the
most basal burnetiamorph and that Lende is member of that
group. Considering the poor support of monophyletic groups in
our hypothesis, it is expected that the phylogenetic position of
Lende, as well as other members of the group, will change as
additional biarmosuchian specimens are discovered and improve
the quality of information. The occurrence of the oldest burne-
tiamorphs, an enhanced diversity, as well as results of our phy-
logeny, suggests that southern Africa could be the area of origin
for the lineage. The presence of this clade together with other
therapsid genera in different Permian localities in what is now
central Africa as well as in eastern Europe (Sues and Boy, 1988;
Angielczyk and Kurkin, 2003; Kammerer et al., 2011; Sidor
et al., 2013) suggests that central Africa was part of a corridor
that allowed migration of burnetiids and other therapsids
between southern Gondwana and Laurasia.
ACKNOWLEDGMENTS
Funding for field work in Malawi was provided mainly by five
grants from the National Geographic Society Committee for
Research and Exploration (2882-84; 3578-87; 4114-89; 4329-90;
4676-91). In South Africa, funding was provided by the Palaeon-
tological Scientific Trust (PAST), the National Research Foun-
dation (NRF), the NRF/DST Centre of Excellence for
Palaeosciences, and the University of the Witwatersrand. We are
grateful to all of the field crews in Malawi over the years, fully
listed in Jacobs (1993), and especially those friends and col-
leagues who are with us in spirit: R. Jerry Britt, W. R. Downs III,
Z. M. Kaufulu, J. Khomu, G. Mgomazulu, and F. Morocco. We
thank our stalwart friend Y. Juwayeyi for his continuing advice.
We acknowledge the assistance of P. Webb in making this
description possible and to Charlton Dube for preparation. Our
gratitude is extended to A. Huttenlocker, S. Modesto, C. Kam-
merer, and C. Sidor for very thorough and helpful review com-
ments that greatly enhanced our manuscript.
LITERATURE CITED
Angielczyk, K. D., and A. A. Kurkin. 2003. Phylogenetic analysis of Rus-
sian Permian dicynodonts (Therapsida: Anomodontia): implications
for Permian biostratigraphy and Pangaean biogeography. Zoologi-
cal Journal of the Linnean Society 139:157–212.
Boonstra, L. D. 1935. On the South African gorgonopsian reptiles pre-
served in the American Museum of Natural History. American
Museum Novitates 772:1–14.
Broom, R. 1905. On the use of the term Anomodontia. Records of the
Albany Museum 1:266–269.
Broom, R. 1923. On the structure of the skull in the carnivorous dinoce-
phalian reptiles. Proceedings of the Zoological Society of London
2:661–684.
Flores, D. A., F. Abdala, and N. Giannini. 2010. Cranial ontogeny of
Caluromys philander (Didelphidae: Caluromyinae): a qualitative
and quantitative approach. Journal of Mammalogy 91:539–550.
Goloboff, P. A. 1993. Character optimization and calculation of tree
lengths. Cladistics 9:433–436.
Goloboff, P. A., J. S. Farris, and K. C. Nixon. 2003. TNT (Tree analysis
using New Technology). Available at www.zmuck.dk./public/
phylogeny. Accessed May 15, 2013.
Hopson, J. A. 1991. Systematics of the nonmammalian Synapsida and
implications for patterns of evolution in synapsids; pp. 635–693 in
H. P. Schultze, and L. Trueb (eds.), Origins of the Higher Groups of
Tetrapods: Controversy and Consensus. Cornell University Press,
Ithaca (New York) and London.
Hopson, J. A., and H. Barghusen. 1986. An analysis of therapsid relation-
ships; pp. 83–106 in N. Hotton, P. D. Maclean, J. J. Roth, and E. C.
Roth (eds.), The Ecology and Biology of Mammal-Like Reptiles.
Smithsonian Institution Press, Washington, D.C.
Ivakhnenko, M. F. 2002. Taxonomy of east European gorgonopians
(Therapsida). Paleontological Journal 36:283–292.
Jacobs, L. L. 1993. Quest for the African Dinosaurs. Villard Books, New
York, 316pp. (Reissued with new Introduction by Johns Hopkins
University Press, Baltimore, Maryland, 2000.)
Jacobs, L. L., D. A. Winkler, K. D. Newman, E. M. Gomani, and A.
Deino. 2005. Therapsids from the Permian Chiweta Beds and the
age of the Karoo Supergroup in Malawi. Palaeontologia Electronica
8:1–23.
Kammerer, C. F., K. D. Angielczyk, and J. Fr€
obisch. 2011. A comprehen-
sive taxonomic revision of Dicynodon (Therapsida, Anomodontia)
and its implications for dicynodont phylogeny, biogeography, and
biostratigraphy. Society of Vertebrate Paleontology Memoir 11:1–
158.
Kemp, T. S. 2012. The origin and radiation of therapsids; pp. 3¡30 in A.
Chinsamy-Turan (ed.), Forerunners of Mammals: Radiation, His-
tology, Biology. Indiana University Press, Bloomington, Indiana.
Lieberman, D. E., B. Hallgrimsson, W. Liu, T. E. Parsons, and H. A.
Jamniczky. 2008. Spatial packing, cranial base angulation and cra-
niofacial shape variation in the mammalian skull: testing a new
model using mice. Journal of Anatomy 212:720¡735.
Lipscomb, D. L. 1992. Parsimony, homology and the analysis of multi-
state characters. Cladistics 8:45–65.
Rubidge, B. S., and J. W. Kitching. 2003. A new burnetiamorph (Therap-
sida: Biarmosuchia) from the lower Beaufort Group of South
Africa. Palaeontology 46:199¡210.
Rubidge, B. S., and C. A. Sidor. 2002. On the cranial morphology of the
basal therapsids Burnetia and Proburnetia (Therapsida: Buretiidae).
Journal of Vertebrate Paleontology 22:257¡267.
Rubidge, B. S., and J. A. van den Heever. 1997. Morphology and system-
atic position of the dinocephalian Styracocephalus platyrhychus.
Lethaia 30:157¡168.
Rubidge, B. S., C. A. Sidor, and S. P. Modesto. 2006. A new burnetia-
morph (Therapsida: Biarmosuchia) from the middle Permian of
South Africa. Journal of Paleontology 80:740¡749.
Sidor, C. A. 2000. Evolutionary trends and relationships within the Syn-
apsida. Unpublished Ph.D. dissertation, University of Chicago, Chi-
cago, Illinois, 370 pp.
Sidor, C. A., and B. S. Rubidge. 2006. Herpetoskylax hopsoni, a new biar-
mosuchian (Therapsida: Biarmosuchia) from the Beaufort Group
of South Africa; pp. 76¡113 in M. T. Carrano, T. Gaudin, R. Blob,
and J. Wible (eds.), Amniote Paleobiology: Perspectives on the
Evolution of Mammals, Birds and Reptiles. University of Chicago
Press, Chicago, Illinois.
Sidor, C. A., and R. M. H. Smith. 2007. A second burnetiamorph therap-
sid from the Permian Teekloof Formation of South Africa and its
associated fauna. Journal of Vertebrate Paleontology 27:420¡430.
Sidor, C. A., and J. Welman. 2003. A second specimen of Lemurosaurus
pricei (Therapsida: Burnetiamorpha). Journal of Vertebrate Pale-
ontology 23:631¡642.
Sidor, C. A., J. A. Hopson, and A. W. Keyser. 2004. A new burnetia-
morph therapsid from the Teekloof Formation, Permian, of South
Africa. Journal of Vertebrate Paleontology 24:938¡950.
Sidor, C. A., K. D. Angielczyk, D. M. Weide, R. M. H. Smith, S. J. Nes-
bitt, and L. A. Tsuji. 2010. Tetrapod fauna of the lowermost Usili
Kruger et al.—A new burnetiamorph from Malawi (e1008698-9)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
Formation (Songea Group, Ruhuhu Basin) of southern Tanzania,
with a new burnetiid record. Journal of Vertebrate Paleontology
30:696–703.
Sidor, C. A., K. D. Angielczyk, R. M. H. Smith, A. K. Goulding, S. J.
Nesbitt, B. R. Peecook, J. S. Steyer, and S. Tolan. 2014. Tapinoce-
phalids (Therapsida:Dinocephalia) from the Permian Madumabisa
Mudstone Formation (Lower Karoo, Mid-Zambezi Basin) of south-
ern Zambia. Journal of Vertebrate Paleontology 34:980–986.
Sidor, C. A., D. A. Vilhena, K. D. Angielczyk, A. K. Huttenlocker, S. J.
Nesbitt, B. R. Peecook, J. S. Steyer, R. M. H. Smith, and L. A. Tsuji.
2013. Provincialization of terrestrial faunas following the end-Perm-
ian mass extinction. Proceedings of the National Academy of Scien-
ces of the United States of America 110:8129–8133.
Sigogneau, D. 1970. R"
evision syst"
ematique des gorgonopsiens Sud-Afri-
cains. Cahiers de Pal"
eontologie, Paris, 414 pp.
Sigogneau-Russell, D. 1989. Theriodontia I; pp. 1¡127 in P. Wellnhofer
(ed.), Encyclopedia of Paleoherpetology, Part 17B. Gustav Fischer,
Stuttgart.
Smith, R. M. H., B. S. Rubidge, and C. A. Sidor. 2006. A new burnetiid
(Therapsid: Biarmosuchia) from the upper Permian of South Africa
and its biogeographic implications. Journal of Vertebrate Paleontol-
ogy 26:331¡343.
Sues, H.-D., and J. A. Boy. 1988. A procynosuchid cynodont from central
Europe. Nature 331:523–524.
Tatarinov, L. P. 1977. A new gorgonops from the Upper Cretaceous
deposits of Vologda Province. Paleontological Journal 2:227–232.
UNESCO (United Nations Educational, Scientific and Cultural Organi-
zation). 2014. Third proclamation of masterpieces of the oral and
intangible heritage of humanity. Available at www.unesco.org/cul-
ture/intangible-heritage/22afr_uk.htm. Accessed September 24,
2014.
Submitted October 8, 2013; revisions received December 9, 2014;
accepted January 11, 2015.
Handling editor: Sean Modesto.
Kruger et al.—A new burnetiamorph from Malawi (e1008698-10)
Downloaded by [The Library, University of Witwatersrand] at 23:29 29 October 2015
