Content uploaded by Aryeh Grossman
All content in this area was uploaded by Aryeh Grossman
Content may be subject to copyright.
MIOSENGI BUTLERI, GEN. ET SP. NOV., (MACROSCELIDEA) FROM THE KALODIRR
MEMBER, LOTHIDOK FORMATION, EARLY MIOCENE OF KENYA
and PATRICIA A. HOLROYD
Department of Anatomy, Arizona College of Osteopathic Medicine,
Midwestern University, 19555 N. 59th Ave, Glendale, Arizona 85308, USA, firstname.lastname@example.org;
Museum of Paleontology,
1101 Valley Life Sciences Building, University of California, Berkeley, California 94720-4780, USA, email@example.com
The only described taxa of early Miocene macroscelideans
(afrotherians previously referred to as elephant-shrews, but bet-
ter dubbed sengis to clearly distinguish them from the unrelated
Soricidae or true shrews) are Myohyrax oswaldi from the Ken-
yan sites of Karungu, Chamtwara and Songhor, and Arrisdrift in
Namibia (Senut 2003), Protypotheroides beetzi also from Nami-
bia, Miorhynchocyon spp. from a large number of localities in
Kenya, and Hiwegicyon juvenalis from Rusinga Island in Kenya
(Butler, 1984, 1995). The first three taxa respectively represent
the macroscelidid subfamilies Myohyracinae (Myohyrax and
Protypotheroides) and Rhynchocyoninae (Miorhynchocyon),
while Hiwegicyon (a taxon only known from a juvenile speci-
men) is of unknown affinities. The Macroscelidinae, the subfam-
ily that comprises most living sengis, is a ghost lineage through
this time interval. Older representatives of the Neogene subfa-
milies are unknown, although macroscelideans noted from
the late Oligocene of the Rukwa Rift Basin in Tanzania may
relate to those of the Neogene (Stevens et al., 2006; Stevens
et al., 2008). The first appearance of macroscelidines is the
middle Miocene locality of Fort Ternan (Butler, 1984, 1995),
where Pronasalio ternanensis occurs at approximately 13.7 Ma
(Pickford et al., 2006). However, the split of macroscelidines
from other sengi lineages has been hypothesized to occur by
the beginning of the early Miocene, based on the presence by
this time of its putative sister lineages the Myohyracinae and
Rhynchocyoninae. Additionally, some molecular phylogenies
suggest rhynchocyonines had split off in the Eocene to early
Oligocene (42.74.8 Ma) and the presence of divisions within
the Macroscelidinae such that the group comprising the extant
Petrodromus,“Elephantulus” rozeti, and their sister taxon
Macroscelides diverged from the group containing other spe-
cies of Elephantulus in the late Oligocene to early Miocene
(21.33.3 Ma) (Douady et al., 2003). Thus, particularly given
the potential discrepancies of molecular clocks, early Miocene
representatives of Macroscelidinae are crucial to better under-
standing the evolution of the group.
Here we describe the first record of an early Miocene macro-
scelidine sengi that helps fill this gap in the subfamily’s fossil
record and provides insights into the morphological diversifica-
tion of the group. The new taxon is from the locality of Kalodirr,
located in northern Kenya about 30 kilometers west of the shore
of Lake Turkana and a few kilometers north of the road leading
from Lodwar to Kalakol (Fig. 1). A team from the National
Museums of Kenya led by Richard and Meave Leakey con-
ducted several field seasons at the site between 1983 and 1987.
They recovered a diverse mammalian fauna with a significant
small mammal component. The site is best known for its three
primate genera; Afropithecus turkanensis, Turkanapithecus kala-
kolensis, and Simiolus enjiessi (Leakey and Leakey, 1986a, b;
Leakey and Leakey, 1987). Although the primates have been
the subject of several studies (Leakey and Walker, 1997; 1988a;
Leakey et al., 1988b; Patel and Grossman, 2006; Rose et al.,
1992), the rest of the fauna is only now being formally described
(Grossman, unpubl. data). The new taxon represents the earliest
known member of the subfamily Macroscelidinae and is the only
macroscelidean known from the fauna.
Geological Setting and Age of the Fossils—the site of Kalo-
dirr (“Oryx” in the Turkana language) is located at the headwa-
ter of the Kalodirr River (3
45’E) between the
Lokipenata Ridge and the Basalt Hills to the east (Boschetto
et al. 1992; Leakey and Leakey 1986a, b). The fossils were recov-
ered from sediments within the latest early Miocene Kalodirr
Member of the Lothidok Formation (Boschetto, 1988; Boschetto
et al., 1992). The age of the Kalodirr fauna can be well con-
strained based on K/Ar dating of the underlying Kalodirr tuffs
(17.510.2 Ma) and the overlying Naserte tuffs (16.80.2 Ma),
which divide the Kalodirr Member from the younger Naserte
Member of the Lothidok Formation. The Miosengi fossils were
found during sieving and their exact location within the member
Sedimentological analyses indicate deposition via small mean-
dering stream channels (Boschetto, 1992). Environmental recon-
structions for the sites of Kalodirr and Moruorot based on the
mammalian fauna indicate that the site was probably a closed to
open woodland habitat (Grossman 2008a, b). This interpretation
is further supported by plant fossils from the site; these include
broad leaved forms, some possessing an acuminate tip ("drip
tip") indicating seasonally wet conditions (Boschetto, 1992). Ad-
ditionally, a number of fish, a pelomedusid turtle, and crocodilian
fossils also indicate the presence of lacustrine or fluvial water.
Institutional Abbreviations—KNM, Kenyan National Museums,
Nairobi, Kenya; KNM-WK, KNM specimens from the site of
Kalodirr, Kenya; KNM-FT, KNM specimens from the site of Fort
Ternan, Kenya; MVZ, Museum of Vertebrate Zoology, University
of California, Berkeley, California; UCMP, University of California
Museum of Paleontology, Berkeley, California.
Order MACROSCELIDEA Butler, 1956
Family MACROSCELIDIDAE Bonaparte, 1838
Subfamily MACROSCELIDINAE Bonaparte, 1838
MIOSENGI BUTLERI gen. et sp. nov.
(Figs. 2, 3)
Holotype—KNM-WK 17050, right mandible fragment with
m2 and the m3 alveolus (Figs. 2, 3).
Referred Material—KNM-WK 17326, right p2 or p3 (Fig. 3).
Type locality—Kalodirr, Kalodirr Member of the Lothidok
Formation, early Miocene, Kenya.
Etymology—The genus name is a combination of Mio- for the
Miocene and sengi, the word for macroscelidids in several Ban-
tu-based languages of central and southern Africa, and Swahili
of eastern Africa (Rathbun and Kingdon, 2006). Specific epithet
in honor of Dr. Percy Butler, whose work on sengis is unparal-
Diagnosis—Differs from other macroscelidines in relatively
lower crown height, and posterior wall of m2 talonid anteriorly
inclined and posteriorly-rounded (vs. vertical). Differs from
Journal of Vertebrate Paleontology 29(3):957–960, September 2009
2009 by the Society of Vertebrate Paleontology
Palaeothentoides and is similar to other macroscelidines in smal-
ler size; differs from Macroscelides, Petrodromus, and some
Elephantulus species in retaining m3; differs from Macroscelides
in having wider molars relative to length and a lower crown.
Differs from Pronasilio in having an m2 slightly mesiodistally
longer but with a lower crown; m2 lacking anterobuccal cingu-
lum; more gracile dentary. Miosengi butleri further differs from
Petrodromus, Paleothentoides, and Elephantulus, but is similar to
Macroscelides and Pronasilio, in retaining a salient hypoconulid.
Description—KNM WK 17050 is a right dentary fragment with
m2 and a small alveolus for a reduced, single-rooted m3. The m2
crown is of moderate height, with the talonid slightly lower than
the trigonid. It is mesiodistally longer than the m2 of Pronasilio
but has a lower crown (KNM WK 17050, length 2.24 mm, width
1.55 mm, height 1.60 mm vs. KNM FT 3409 [P. ternanensis],
length 1.87 mm, width 1.53 mm, height 2.11 mm). The trigonid is
dominated by the metaconid and a slightly lower protoconid.
These cusps are aligned mesiodistally and joined by a distinct
protolophid. The anterior margin of the tooth is formed by a
well-developed paracristid (or paralophid) arising from the pro-
toconid and sweeping mesiolingually to form a lingually open
trigonid. The paracristid is lower than the protolophid and termi-
nates in an abrupt dip slightly lingual of the tooth’s midline. The
paracristid of Miosengi butleri courses more lingually and
recurves. A small anterior cingulum is present low on the buccal
face of the paracristid, but is not as extensive as in Pronasilio.
The talonid is only very slightly lower than the trigonid. The
entoconid is slightly higher than the hypoconid. The hypoconulid
is a slight angulation of the hypolophid connecting the entoconid
and hypoconid, but is distinct, in contrast to younger macrosceli-
dines such as Paleothentoides, Petrodromus,orElephantulus. The
posterior profile of the crown is distinctly rounded and canted
anteriorly. This anterior slant is particularly discernable in lin-
gual view and is very different from the vertical posterior profile
of the m2 in Pronasilio or Paleothentoides or the anteriorly in-
clined surfaces formed by the posterior edge of the molars in
Petrodromus, Macroscelides, and Elephantulus. A prominent
cristid obliqua bends lingually from the hypoconid to join the
protolophid near the metaconid as in other macroscelidines. A
deep cleft opens the talonid basin lingually between the entoco-
nid and metaconid.
Direct comparisons with early Miocene Hiwegicyon are not
possible because it is only known from a deciduous dentition.
However, the m1 of H. juvenalis agrees in proportion with that
of P. ternanensis (Butler, 1984), and if these proportions hold
true for the m2 as they do in Pronasilio, this may serve to
distinguish between Hiwegicyon and Miosengi. Additionally,
Hiwegicyon lacks the rounded posterior margin on the lower
molar that is seen in Miosengi.
KNM-WK 17326 (Fig. 3) is a double-rooted premolar measur-
ing 1.64 mm in length, 0.76 mm in width, and 1.11 mm in height.
The protoconid is only slightly higher than the raised paracristid
FIGURE 1. Map showing occurrence of early and middle Miocene
faunas containing macroscelideans in Kenya. Localities in the Lake Vic-
toria region include early Miocene Songhor, Koru, Chamtwara, and
Rusinga, Maboku, and Mwafangano Islands, and middle Miocene Fort
Ternan (type locality of Pronasilio ternanensis).
FIGURE 2. Miosengi butleri, KNM WK-17050, holotype right mandi-
ble with m2 and alveolus of m3 in lateral, occlusal, and lingual views.
958 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 29, NO. 3, 2009
that extends almost directly anteriorly from its apex. The crown
has a short talonid formed by a raised crest, which is distinctly
lower than the trigonid. Comparing it with other macrosceli-
dines, the relative proportions of the trigonid and talonid and
the anteriorly directed paracristid are most similar to the p2 or
p3 of Palaeothentoides and to p3 in Elephantulus rupestris or
E. brachyrhynchus. It differs markedly from Petrodromus in
having a relatively high paracristid, suggesting that it may be
more closely related to the former taxa rather than to the Petro-
dromus-“E.” rozeti- Macroscelides complex, unless Douady
et al. (2003) are correct in their hypothesis that the morphology
of the clade comprising Elephantulus species excluding “E.”
rozeti represents the primitive condition for the entire subfamily.
The premolars anterior to p4 are not known in most Paleogene
macroscelideans; only from the late Eocene Herodotius (Simons
et al., 1991). However, in contrast to Herodotius, which also has
semi-molariform premolars, KNM-WK 17326 differs in having
the paraconid subsumed within a paracristid, rather than being a
distinct cusp. Macroscelidines (except Petrodromus) differ from
rhynchocyonines in having semi-molariform to molariform
p1through p3 in addition to the molariform p4 found in all sengis.
Pleistocene and extant macroscelidines share this trend for
molarization of the anterior premolar dentition with the Myohy-
racinae, where it is taken to hypsodont extremes. However, we
currently lack adequate sampling of the fossil record to establish
the timing and patterning of this transformational sequence.
We place Miosengi within crown Macroscelidea based on hav-
ing a well-developed lower molar paracristid with distinct para-
conid and a hypoconulid within a strong hypocristid (characters
12, 16 of Tabuce et al., 2001). It can also be included with the
unnamed clade comprising Macroscelidinae + Mylomygalinae
based on the shared possession of a lower molar cristid obliqua
that is directed towards the metaconid (character 15 of Tabuce
et al., 2001). However, it is important to note that at least some
specimens of Metoldobotes, which is placed outside of crown
Macroscelidea in Seiffert (2007), are reported to have well-de-
veloped lower molar paracristids and paraconids, distinct hypo-
conulids, and oblique cristids oriented toward the metaconids
(Characters 18426, 18430, 18441 in the supplemental information
of Seiffert, 2007).
We place Miosengi butleri within Macroscelidinae based on its
semi-molariform premolar (a condition shared with more
derived macroscelidines, excluding Petrodromus) and the lack
of any of the autapomorphies associated with the Myohyracinae
that can be evaluated in Miosengi (e.g., extreme hypsodonty).
Pronasilio has been regarded as the most primitive of the
macroscelidines due to its retention of a small m3 and having
molars in which the talonid is slightly lower than the trigonid
compared to later macroscelidines (Butler, 1984). Miosengi but-
leri differs from Pronasilio in the lower height of the m2 crown
and its rounded posterior margin suggesting that a lower crown
and rounded posterior margin may be more primitive. The pre-
molar assigned to M. butleri provides new data for the macro-
scelidines and suggests that the process of anterior premolar
molarization was well underway by the early Miocene. Although
we cannot assign the premolar to position with certainty, com-
parison of overall morphology with known Paleogene forms
demonstrates that M. butleri had slightly narrower yet more
semi-molariform premolars than observed in any Paleogene
macroscelidean. It also has a more pronounced paracristid and
talonid than seen in extant
Petrodromus, suggesting that it is
more derived in this feature. However, the relationships among
macroscelidines as well as myohyracines have not been formally
explored in a phylogenetic framework (and are beyond the
scope of the current paper), so hypotheses regarding dental
character polarities are tentative.
Pronasilio ternanensis is similar to the extant Elephantulus
brachyrhynchus in its overall dental row dimensions (Butler,
1984) suggesting that this animal weighed approximately 40-60g.
Miosengi butleri has a similarly sized lower second molar and
very likely was as small as P. ternanensis, both of which are
smaller than Paleogene macroscelideans such as Metoldobotes
and Herodotius. Macroscelidines are relatively smaller on aver-
age than other sengis, and M. butleri is one of the smallest
species within the group, demonstrating that the small size of
the Macroscelidinae was achieved early in the diversification of
The discovery of Miosengi butleri also bears on questions re-
garding the biogeographic origins of taxa in younger East Afri-
can sites. The middle Miocene localities of Maboko and Fort
Ternan are distinct from older localities in the Lake Victoria
region of western Kenya due to the introduction of putatively
allochthonous species (Butler, 1984), possibly correlated with
indications of a change towards drier climates in the region
(Andrews and Walker, 1976; Cerling et al., 1997; Kappelman,
1991; Shipman, 1982, 1986). The early Miocene site of Rusinga
Island in western Kenya on the shores of Lake Victoria is rough-
ly contemporaneous with Kalodirr (Drake et al., 1988). Macro-
scelidids from Rusinga Island comprise Miorhynchocyon and
Myohyrax oswaldi. These two taxa are also known from slightly
older early Miocene localities within the Tinderet sequence
(Meswa Bridge, Songhor, Koru, Legetet, Chamtwara; Butler,
1984, 1995). However, there are no remains of macroscelidines
at these localities, suggesting possible ecological differences
between Kalodirr and other early Miocene sites and similarities
to the younger Fort Ternan, as suggested by Harrison (1992)
in discussing small catarrhine monkey distributions. M. butleri
demonstrates that the subfamily Macroscelidinae was morphol-
FIGURE 3. Miosengi butleri, m2 of KNM WK-17050 in lateral, occlu-
sal, and lingual views; and KNM-WK 17326, anterior lower premolar, in
lateral, lingual and occlusal views.
SHORT COMMUNICATIONS 959
ogically distinct by the late early Miocene (16.6-17.7 Ma) and
likely lived several hundred kilometers farther north than the
localities of the Lake Victoria region. Thus, it appears that at
least some of the allochthonous faunal elements at Fort Ternan
in the middle Miocene (e.g., Pronasilio ternanensis) may have a
more geographically proximate origin within Africa, in contrast
to long distance dispersers that first appear in that fauna (e.g.,
We thank E. Mbua, F. K. Manthi, M. Muungu and the staff of
the Paleontology Department at the National Museums of
Kenya for their patient and excellent support of AG; M. G.
Leakey for her help in gaining access to this material and her
support, encouragement and help to AG; N. J. Stevens for addi-
tional photography; and J.G. Fleagle, E. Seiffert, and D. M.
Boyer for comments on this manuscript. J. H. Hutchison ably
executed Fig. 2. This research is partly funded by the Leakey
Foundation. This is UCMP contribution #1968.
Andrews, P., and Walker, A. 1976. The primate and other fauna from
Fort Ternan, Kenya, pp 279–306 in G. L. Isaac and E. R. McCown
(eds.), Human Origins: Louis Leakey and the East African evi-
dence. W. A. Benjamin, Menlo Park, California.
Bonaparte, C. L. 1838. Synopsis vertebratorum systematis. Nuovi Annali
delle Scienze Naturali 1:105–133.
Boschetto, H. B. 1988. Geology of the Lothidok Range. M.S. thesis,
University of Utah, Salt Lake City, Utah, 203 pp.
Boschetto, H. B., F. H. Brown, and I. McDougall. 1992. Stratigraphy of
the Lothidok Range, Northern Kenya, and K/Ar ages of Its Miocene
primates. Journal of Human Evolution 22:47–71.
Butler, P. M. 1956. The skull of Ictops and the classification of the
Insectivora. Proceedings of the Zoological Society of London
Butler, P. M. 1984. Macroscelidea, Insectivora, and Chiroptera from the
Miocene of East Africa. Palaeovertebrata 14:117–200.
Butler, P. M. 1995. Fossil Macroscelidea. Mammal Review 25:3–14.
Cerling, T. E., J. M. Harris, S. H. Ambrose, M. G. Leakey, and
N. Solounias. 1997. Dietary and environmental reconstruction with
stable isotope analyses of herbivore tooth enamel from the Miocene
locality of Fort Ternan, Kenya. Journal of Human Evolution
Douady, C. J., F. Catzeflis, J. Raman, M. S. Springer, and M. J. Stanhope.
2003. The Sahara as a vicariant agent, and the role of Miocene
climatic events, in the diversification of the mammalian order
Macroscelidea (elephant shrews). Proceedings of the National Acad-
emy of Science 100:8325–8330.
Drake, R. E., J. A. Vancouvering, M. H. Pickford, G. H. Curtis, and
J. A. Harris. 1988. New chronology for the early Miocene faunas of
Kisingiri, western Kenya. Journal of the Geological Society, London
Grossman, A. 2008a. Ecological and morphological diversity in catar-
rhine primates from the Miocene of Africa. Ph.D. dissertation,
Stony Brook University, Stony Brook, New York, 430 pp.
Grossman, A. 2008b. The Environment of Early Miocene catarrhines at
Kalodirr and Moruorot Hill. American Journal of Physical Anthro-
Harrison, T. 1992. A reassessment of the taxonomic and phylogenetic
affinitiesof the fossil catarrhines from Fort Ternan, Kenya. Primates
Kappelman, J. 1991. The paleoenvironment of Kenyapithecus at Fort
Ternan. Journal of Human Evolution 20:95–129.
Leakey, M., and A. Walker. 1997. Afropithecus: function and phylogeny;
pp. 225–240 in D. R. Begun, C. V. Ward, and M. D. Rose (eds.),
Function, Phylogeny, and Fossils: Miocene Hominoid Evolution and
Adaptations. Plenum Press, New York.
Leakey, R. E., and M. G. Leakey. 1986a. A new Miocene hominoid from
Kenya. Nature 324:143–146.
Leakey, R. E., and M. G. Leakey. 1986b. A second new Miocene homi-
noid from Kenya. Nature 324:146–148.
Leakey, R. E., and M. G. Leakey. 1987. A new Miocene small-bodied
ape from Kenya. Journal of Human Evolution 16:369–387.
Leakey, R. E., M. G. Leakey, and A. C. Walker. 1988a. Morphology of
Afropithecus turkanensis from Kenya. American Journal of Physical
Leakey, R. E., M. G. Leakey, and A. C. Walker. 1988b. Morphology of
Turkanapithecus kalakolensis from Kenya. American Journal of
Physical Anthropology 76:277–288.
Patel, B. A., and A. Grossman. 2006. Dental metric comparisons of
Morotopithecus and Afropithecus: Implications for the validity of
the genus Morotopithecus. Journal of Human Evolution 51:506–512.
Pickford, M., Y. Sawada, T. Tayama, Y. Matsuda, T. Itaya, H. Hyodo,
and B. Senut. 2006. Refinement of the age of the middle Miocene
Fort Ternan Beds, Western Kenya, and its implications for Old
World biochronology. Comptes Rendus Geosciences 338:545–555.
Rathbun, G., and J. Kingdon. 2006. The etymology of "SENGI".
Afrotherian Conservation 4:14–15.
Rose, M. D., M. G. Leakey, R. E. F. Leakey, and A. C. Walker. 1992.
Postcranial specimens of Simiolus enjiessi and other primitive catar-
rhines from the early Miocene of Lake Turkana, Kenya. Journal of
Human Evolution 22:171–237.
Seiffert, E. R. 2007. A new estimate of afrotherian phylogeny based on
simultaneous analysis of genomic, morphological, and fossil evi-
dence. BMC Evolutionary Biology 7:1–42.
Senut, B. 2003. The Macroscelididae from the Miocene of the Orange
River, Namibia. Memoirs of the Geological Survey of Namibia
Shipman, P. 1982. Taphonomy of Ramapithecus wickeri at Fort Ternan,
Kenya. Museum Brief 26. University of Missouri-Columbia, Colum-
Shipman, P. 1986. Paleoecology of Fort Ternan reconsidered. Journal of
Human Evolution 15:193–204.
Simons E. L., P. A. Holroyd, and T. M. Bown. 1991. Early Tertiary
elephant-shrews from Egypt and the origin of the Macroscelidea.
Proceedings of the National Academy of Sciences of the United
States of America 88:9734–9737.
Stevens, N. J., M. D. Gottfried, P. C. O’Connor, and E. M. Roberts. 2006.
Macroscelideans from the Oligocene of Southwestern Tanzania.
Journal of Vertebrate Paleontology 26:128A.
Stevens, N. J., M. D. Gottfried, E. M. Roberts, S. Ngasala, S. Kapilima,
and P. M. O’Connor. 2008. Paleontological Exploration in Africa: a
view from the Rukwa Rift Basin of Tanzania; pp. 159–180 in
J. G. Fleagle and C. C. Gilbert (eds.), Elwyn Simons: A Search for
Tabuce, R., B. Coiffait, P. E. Coiffait, M. Mahboubi, and J.-J. Jaeger.
2001. A new genus of Macroscelidea (Mammalia) from the Eocene
of Algeria: A possible origin for elephant-shrews. Journal of Verte-
brate Paleontology 21:535–546.
Submitted October 29, 2008; accepted December 11, 2008.
960 JOURNAL OF VERTEBRATE PALEONTOLOGY, VOL. 29, NO. 3, 2009