the first complete skeleton of megaloceros verticornis ...the first articulated skeleton of a young...
TRANSCRIPT
Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 5 (2002) 289-308 10.11.2002
The first complete skeleton of Megaloceros verticornis (Dawkins, 1868) Cervidae, Mammalia, from Bilshausen (Lower Saxony, Germany): description and phylogenetic implications
Thekla Pfeiffer
With 5 figures and 4 tables
Abstract
The first well preserved, articulated skeleton of a young male deer of Megaloceros verticornis (Dawkins, 1868) was excavated from early Middle Pleistocene sediments of the clay pit of Bilshausen (Unter-Eichsfeld, Lower Saxony). This find made it possible, for the first time, to establish, using cladistic techniques, the systematic position of Megaloceros verticornis among Pleistocene and Holocene plesiometacarpal and telemetacarpal cervids. By contrast to the antler and tooth characters, the postcranial characters, in particular, are suitable for phylogeny reconstruction. Megaloceros verticornb from Bilshausen shows great similarity with M. giganteus of the Upper Pleistocene of Europe in its skeletal morphology, and bootstrap values (BP = 100) show strong support for the monophyly of M. giganteus and M. verticornis.
The analysis yields no evidence, however, of a close relationship between Duma and Megaloceros, which has been widely discussed in the literature because of the presence of large, palmated antlers in both genera.
Key words: Pleistocene, vertebrate palaeontology, cervids, Megaloceros verticornis, skeletal morphology, phylogeny reconstruc- tion, cladistic analysis, taxonomy.
Zusammenfassung
Aus der Tongrube von Bilshausen (Unter-Eichsfeld, Niedersachsen) konnte das erste, vollstandige Skelett eines jungen Hirsches von Megaloceros verticornis (Dawkins, 1868) aus mittelpleistozanen Sedimentablagerungen geborgen werden. Dieser Fund ermoglichte es erstmalig, die systematische Stellung von Megaloceros verticornis im System plesiometacarpaler und tele- metacarpaler Hirsche des Pleistozans und Holozans auf breiter Basis zu untersuchen. Im Gegensatz zu den Geweih- und Zahnmerkmalen eignen sich die postcranialen Merkmale des Skelettes besonders gut fir eine phylogenetische Rekonstruk- tion der Hirsche.
Die Gemeinsamkeit grol3er Schaufelgeweihe bei Duma duma und dem Riesenhirsch Megaloceros giganteus hat dazu ge- fiihrt, beide in eine enge phylogenetische Beziehung zu setzen, was in der Literatur zu einer anhaltenden Kontroverse gefiihrt hat. Die Analyse der Morphologie der postcranialen Elemente zeigt jedoch, dass es keine enge Verwandtschaft zwischen Duma und Megaloceros gibt.
Schliisselworter: Pleistozan, Wirbeltierpalaontologie, Cerviden, Megaloceros verticornis, Skelettmorphologie, phylogenetische Rekonstruktion, kladistische Analyse, Taxonomie.
Introduction
In the early 20th century, Schmidt (1930, 1934) recovered a complete skeleton of AZces Zutifrons from the clay pit of Bilshausen (Lower Saxony, 20 km SE Gottingen), and several finds of articu- lated skeletal elements of large mammals were brought to light in the following years. In 1952 the complete extremities of another large cervid were found embedded in the clay, although the body was dislocated and destroyed. Scientific in-
vestigations ceased during the second world war, and some material was lost, but fortunately the best preserved find of the early years, the articu- lated skeleton of A. Zutifrons, was later discov- ered in the Museum of the Institute for Palaeon- tology at the University of Gottingen.
This museum also houses the well preserved skeleton of a young male deer of MeguZoceros verticornis from Bilshausen (Fig. 1) excavated by D. Meischner and J. Schneider (both at the University of Gottingen) in November 1964
I Gutswiese 21, D-38162 Cremlingen, Germany. Received April, accepted July 2002
290 F‘feiffer. T.. ComDlete skeleton of Mepaloceros
Fig. 1. The first articulated skeleton of a young male of Meguloceros verticornis (Dawkins, 1868) from Bilshausen.
(Meischner & Schneider 1967). Pollen analysis of Brockentuff of Karlich, 396,000 +/- 20,000 mea- the interglacial profile from Bilshausen was car- sured using the 40Ar/”Ar laser method by v. d. ried out by Muller (1965), and Gruger et al. Bogaard et al. (1989) was fully compatible with (1994) who made detailed studies of the Middle the dating suggested by Bittmann & Miiller Pleistocene warm phases of this area. Bittmann (1996) on the basis of the vegetational develop- & Miiller (1996) correlated the Karlich Inter- ment. The taphonomy of Bilshausen was inter- glacial site with Bilshausen. The age of the preted by Meischner (1995) as follows: “Bils-
Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 5 (2002) 291
hausen was during its existence a meromictic pond . . . fishes and mammals abound in the warved black shales. The skeleton of the giant deer showed no remnants of the skin, but the arrangement of the bones indicate that the corpse sank to the bottom, fell on the right side, and came to rest head-down on a slope. Bones maintained their original arrangement, but be- came disarticulated and were partly rotated, partly pulled-apart, by downslope movement of the still soft sediment.” The skeleton of the young male deer of Megaloceros verticornis is especially important because it is the first articu- lated skeleton found for this species.
Fossil and recent cervids have been described primarily on the basis of antler-, skull- and tooth characters, and additionally on body size. Cur- rent morphological taxonomy splits the family Cervidae on the basis of a single character: ab- sence (Hydropotinae) or presence (Odocoileinae + Cervinae) of antlers (Groves & Grubb 1987), even though the taxonomic value of antlers has been repeatedly questioned (see, for example, Scott & Janis 1993). Antlers show high varia- bility within a species, and this is dependent on on- togenetic growth patterns and habitat conditions.
Unfortunately, previous workers attempted to ascertain relationships of cervids based only on the comparison of antler and skull characters. An example of this problem is the unresolved discus- sion concerning the relationships of the Recent fallow deer (Duma duma) and the giant deer (Megaloceros giganteus) which stemms from the beginning of the last century. The presence of large, palmated antlers in Dama duma and Mega- loceros giganteus suggested a close relationship between these two taxa (Freudenberg 1914, Geist 1971, 1987, Lister 1984). By contrast Thenius (1958) and Kahlke (1994) held the view that the giant deer must have evolved from deer with an- tlers that lacked any palmation. Investigations by Pfeiffer (1999a, in press) that also include post- cranial characters, show that the genera Megulo- ceros and Duma belong to clearly separate lineages in the Pleistocene, and that their pal- mate antlers must have evolved independently.
The systematic position of Meguloceros verti- cornis has also been controversial (see below). On the basis of the skeleton from Bilshausen, and together with skulls, antlers, and selected postcranial skeletal material from adult indivi- duals of Meguloceros verticornis from the Thur- ingian localities SiiBenborn and Voigtstedt (Kahlke 1956, 1960, 1965, 1969) it is possible to determine the systematic position of this species
within the Pleistocene plesiometacarpal deer for the first time. Analysis of the morphology of the postcranial skeletal elements verifies a close rela- tionship between M. verticornis and M. gigun- teus.
Gould (1974) showed that the dramatic size reduction of spike antlers in the advanced South American genera Pudu and Muzania suggests that the reversal of morphological trends is pos- sible as a consequence of selection correlated with small body size. Some morphological char- acters seem therefore to be with highly homo- plastic in ungulates (Scott & Janis 1993), and are difficult to use for determining phylogene tic rela- tionships.
Material
The analysis conducted here was based on skeletons from males and females of a large number of Recent and fossil deer species, and the skeleton of Megaloceros verticornis from Bilshausen (Table 1). This taxon list includes the Re- cent telemetacarpal species Alces alces, Rangifer tarandus, Capreolus capreolus, C. pygargus, the Middle Pleistocene Roe deer C. suessenbomensis, the Lower Pleistocene plesio- metacarpal species Eucladoceros tegulensis, E. dicranios, Dama rhenana, Dama nestii, the Middle Pleistocene Megalo- ceros verticornis, M. giganteus, Cervus elaphus acoronatus, Dama darna clactoniana, D. duma geiselana, and the extant forms Cervus elaphus, C. nippon nippon, C. nippon hortulor- um, Axis axis, D. darna mesopotamica, and D. darna darna. Skeletons of the Recent Moschus rnoschiferus and skeletal material of the taxa Blastomeryx and Parablastomeryx from the Miocene of North America were added as outgroups.
In addition to skeletons, isolated bones of fossil taxa from different localities were also used. The scapula and pelvis, like the skull, were often broken. In the case of Megaloceros verticornis, the skeleton of the young stag from Bilshausen was analysed together with skulls, antlers and isolated post- cranial bones of adult individuals from SiiBenborn and Voigt- stedt. Because the skull of the deer from Bilshausen is not preserved and this specimen reprents a juvenile, it was neces- sary to study additional elements from adult individuals.
Method
A data matrix was coded for the 20 fossil and Recent deer listed above, and the antlerless musk deer, Moschus moschi- ferus, together with Blastomeryx and Parablastorneryx from the Miocene of North-America were used as outgroups. In current taxonomy, Moschus is commonly accepted as a close relative of the family Cervidae. Blastomeryx and Parablasto- meryx share many ancestral features with early cervids in- cluding, for example, large upper canines.
The one-hundred and twenty-two morphological characters based on postcranial bones, antlers, skulls, and teeth are ex- plained in Table 2, and listed in the data matrix (Table 3). The characters were coded as “unordered”, and equally weighted, while missing data was denoted by a question mark. Polymorphic character states, which occur within a spe- cies with approximately the same frequency, were coded with the equate format a = 0 + 2; b = 0 + 1. Phylogenetic recon- structions were obtained by the maximum parsimony (MP)
292 Pfeiffer, T., Complete skeleton of Megaloceros
method (PAUP 4.0, Swofford 1998), and the maximum-like- Wood (ML) method (quartet puzzling approach: Strimmer & von Haeseler 1996). The robustness of the phylogeny was assessed using the following approaches: the bootstrap per- centage (BT) (Felsenstein 1985) with lo00 resamplings, heur- istic search mode; the Jacknife with 50% deletion, heuristic search mode, lo00 replicates; the reliability percentages (RP), i.e. the number of times the group appears after 10000 ML puzzling steps (Strimmer & von Haeseler 1996).
Taxonomy
Megaloceros verticornis (Dawkins, 1868)
Family C e r v i d a e GoldfuB, 1820 Subfamily C e r v i n a e GoldfuB, 1820 Genus Megaloceros Brookes, 1828
1862 Megaceros Carnutorum - Laugel, A. 1868 Cervus verticornis - Dawkins, B. 1869 Cervus (euryceros) Belgrandi - Betgrand, E. 1882 Cervus dawkinsi - Newton, E. T. 1886 Cervus (Duma) priscus - Moullade, E. 1889 Cervus belgrandi - Harmer, F. 1892 Cervus pachygenis - Pomel, A. 1903 Cervus pliotarandoides - de Nessandri, G. 1920 Praemegaceros verticornis - Portis, A. 1927 Cervus megaceros mosbachensis - Soergel, W. 1953 Megaceros verticomis - Azzaroli, A. 1953 ‘Cervus’ obscurus - Azzaroli, A. 1956 Orthogonoceros verticomis - Kahlke, H.-D. 1956 Dolichodoryceros siifienbornensis - Kahlke, H.-D. 1992 Megaceroides verticomis - Azzaroli, A. & Mazza, F! 1993 Megaloceros verticomis - Lister, A.
The species M. verticornis was first described by Dawkins (1868), but initially referred to the genus Cervus.
Laugel (1862) described, with Megaceros car- nutorum from St. Prest, the antler characters of Megaloceros verticornis (Pfeiffer 1999b), but fig- ured teeth of Alces. Therefore, Heintz & Poplin (1981) used carnutorum for an alcine, Alces cur- nutorum. In any case the cervid material from St. Prest may belong to two or more species. Concerning Praernegaceros, Portis (1920) first used this genus name in connection with the spe- cies carnutorum. This cannot be regarded there- fore as a valid foundation for Praemegaceros.
Kahlke (1956) referred M. verticornis to a new genus, Orthogonoceros, presumed here to be a junior synonym of Megaloceros.
Azzaroli & Mazza (1992) included Megaloceros verticornis in the genus Megaceroides, which was described by Joleaud (1914) on the basis of a giant deer from Algeria. The antlers of M. algericus (Lydekker 1890), the type species of Megacer- oides, are poorly known, but the teeth of M. alger- I’cus, figured by Lydekker (1890) differ in their morphology from those of M. verticornis. Conse- quently, the systematic position of M. algericus within the giant deer has yet to be resolved.
Lister (1993, 1994) combined all giant deer species to a single genus Megaloceros following Azzaroli (1953). The name Megdoceros Brookes, 1828 (type species M. giganteus) has priority over Megaceros Owen 1842 (Lister 1987, International Commission on Zoological Nomen- clature 1989), and the author follows this opi- nion here.
Description of the skeleton of Megdoceros verticornis from Bilshausen
P r e s e r v a t i o n This young male individual of M. verticornis lay on its right side, legs parallel to one another (Fig. 1). The bones are dark brown, with well preserved surfaces, that usually permit detailed study of the morphological characters. The body is well preserved, but the cervical vertebral col- umn is disarticulated, and the skull, the anterior cervical vertebrae, the main part of the left an- tler, parts of the lumbar vertebral column, and the sacrum are missing (Fig. 1). Some isolated lower and upper molars are preserved, impor- tant among which are the left upper M3 and fragments of the lower m3. The last molars are completely developed, but show no (m3) or only very little abrasion (M3). At the age of 24 months the second dentition is complete in the extant genera Cervus (Habermehl 1961) and Dama (Rieck 1983). If M. verticornis developed the last lower and upper molars at the same on- togenetic stage, an age of two years is indicated for this stag at death. This observation is in ac- cordance with several long bones that show in- complete fusion of their epiphyses. The juvenile age also explains the relatively small degree of antler development (Fig. 2) in comparison to adult individuals from Suljenborn (Kahlke 1956, 1969). The right antler is nearly complete, while the left antler is only presented by fragments of the first and second anterior tines.
Both humeri, femora, tibiae, and the metapo- dials were broken in their proximal part, but are reconstructed. The fragmentary pelvis, sacrum and coccygeal vertebrae were reconstructed, but yield no informative morphological details. All anterior and posterior phalanges are complete and well preserved. As the carpal and tarsal bones were fixed in position during the recon- struction of the skeleton neither measurements, nor morphological details could be collected for these elements.
Mitt. Mus. Nat.kd. Berl.. Geowiss. Reihe 5 (2002) 293
Table 1 Skeletal material of the 20 cervid species studied in this investigation.
Recent: D. duma duma D. duma mesopotamica Axis axis C. nippon nippon C. nippon hortulorum Cervus elaphus Alces alces Rangifer tarandus Capreolus capreolus Capreolus pygargus
Archaeological sites: D. duma duma (Kastanas, Demircihuyiik) D. duma mesopotamica (Halawa, Kebara)
Late Pleistocene: Duma duma (Upper Rhine valley, Lebringen, Trafalgar Square) Megaloceros giganteus (Upper Rhine valley, Schlutup, Ireland) Alces alces (Upper Rhine, Valley, Magdeburg-Neustadt)
Middle Pleistocene D. duma geiselana (Neumark-Nord) Cervus elaphus (Neumark-Nord) D. duma clactoniana (Clacton, Swanscombe, Jaywick, Grays, Riano, Valdemino, Melpignano) Cervus elaphus (Clacton, Swanscornbe, Jaywick, Grays) Cervus elaphus acoronatus (Mosbach, SiiBenborn, Voigtstedt) Megaloceros verticornh (Mosbach, SiiBenborn, Voigtstedt, Bilshausen) Capreolus suessenbornensis (Mosbach, SiiBenborn, Miesenheim)
Lower Pleistocene Dama nestii (Tasso) Dama rhenana (Senkze, Tegelen) Eucladoceros tegulensis (SenBze, Tegelen) Eucladoceros dicranios (Tasso)
The teeth and skeletal elements of Megalo- ceros verticornis from Bilshausen are relatively large and within the size range of those of Mega- loceros verticornis from SiiBenborn, and the giant deer, M. giganteus, from other Upper Pleis- tocene German localities (Table 4).
A n t l e r The antlers of M. verticornis always show a char- acteristic medially-curved tine at a point 4 to 8 centimetres above the burr, and this can be ob- served at the antler of the young stag from Bils- hausen. In addition, some antlers of M. verticor- nis exhibits a growth near the base of the beam, ranging from a small bump to a strong tine, which is not developed on the antler from Bils- hausen. Due to this variation some specimens with a long tine in this position have been sepa- rated as species by previous workers, such as Cervus pliotarandoides and ‘Cervus obscurus’.
Measures of the preserved right antler are gi- ven in Fig. 2. The height of the right pedicle
6: 30 6: 12 6: 8 6: 11 6: 6 6: 12 6: 15 6: 5 6: 15 6: 2
619: 17 dl?: 6
$19: 6 8: 5-15 dl?: 7-11
6: 47 6: 17 619: 16-27
619: 2-15 $I$? 8-23 619: 2-10
619: 4-18
619: 12-29 6: 12-26 619: 12-29 619: 2-14
9: 28 ?: 12 9: 10 9: 11 9: 6 9: 10 9: 7 9: 4 $2: 8-15 9: 1
9: 2-7
9: 15 9: 1
9: 12-24
(HRst) is 55 mm, while its diameter (DRst), at 50.5 mm, is quite small. The burr is incomplete, with a preserved diameter (DR) of 56.5 mm, but must have been about 65 mm in life. These data support a juvenile age. The first anterior tine in- serts at a right angle about 60mm above the burr. The beam is rounded in its proximal part, with longitudinal ridges on its surface, but with- out pearls. It becomes flattened in the region where the second anterior tine branches off. A flattened posterior tine, now broken, also branched off 380mm above the burr. The distal part of the beam shows a weakly developed pal- mation and is directed anteriorly. The distal part of the antler is missing, and the preserved length is 747mm. The beam turns sidewards and up- wards above the burr.
U p p e r mola r s From the right upper jaw a fragment with M2 and M3 is preserved, while from the left side, a fragment of MI, and the complete, isolated M2
‘‘f
: : I . . i
Fig.2. Right antler of a two year old stag from Bilshausen indicating principal measurements (see expla- nation in the text). The develop- ment of the burr shows that the stag had shed an antler in the year before. The relatively small dimen-
v sions confirm a juvenile age.
and M3 are measured (Table 4). All molars show well developed entostyli, a small anterior cingu- lum, and a well expressed posterior cingulum. The anterior and posterior lingual lobes bifur- cate posteriorly (Fig. 3a).
adult megalocerines, and only slightly thickened. The isolated lower right and left ml, and the left m2 are preserved, the latter showing little abra- sion. An anterior cingulum is well developed, and the ectostylid is small (Fig. 3b). Fragments of the lower right and left m3 show no abrasion. This indicates that the individual is of about two years in age.
L o w e r m o l a r s A fragment of the left lower jaw is preserved, but lacks teeth. It is not pachygnathous, as in
Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 5 (2002) 295
Fig. 3. A. Left upper molars (M2, M3) in occlusal view, M3 shows little abrasion, indicating an individual age of two years at the time of death. All molars show well developed entostyli, a small anterior cingulum, and a well expressed posterior cingulum. The anterior and posterior lingual lobes are bifurcated posteriorly. B. Right lower molar (ml) in oc- clusal view. An anterior cingulum is well developed, the ec- tostylid is small.
V e r t e b r a l c o l u m n Axis: Only the cranial facet of the axis could be measured, its widths is 108.5mm (BFcr, v. d. Driesch 1976), and the distal epiphysis is not fused.
Only the corpi of the 3rd-7th cervical verteb- rae are preserved. Each epiphysis is separated from the corpus. Corpus lengths are in the size range typical for Megaloceros. The shape of the cranial epiphyses is similar to those of Megalo- ceros giganteus, flattened or concave dorsally, wide, and rounded, a diagnostic character for Megaloceros.
The corpus of each thoracic vertebrae is pre- served, but not one of the epiphyses is fused. By contrast, the proximal epiphysis of the first lum- bar vertebra is fused and shows a weakly de- pressed margin laterally.
S c a p u l a The cavitas glenoidalis is round and wider than long (cha. 12 in the description of character states, Table 2) and it has no V-shaped cavity in its articular surface (cha. 13). the tuberculum su- praglenoidale is short, coarse, and not medially curved (cha. 14) as in Megaloceros giganteus. M. verticornis has no foramen dorsally between the processus coracoideus and the cavitas glenoidalis (cha. 15) as in most species of deer, while M.
giganteus has a large foramen in this position. The scapula shaft is robust, but not especially widened as in adult individuals of M. giganteus (cha. 17). This character is strongly influenced by antler development. Consequently, the relatively juvenile age and the small antler of the indivi- dual from Bilshausen reflect the weaker expres- sion of this feature.
H u m e r u s The proximal epiphyses of both humeri are not fused, but the distal epiphyses were just begin- ning to fuse to the diaphysis at the time of death. Only characters of the distal part could be determined. The fossa radialis shows two rounded scores (cha. 21), a character that is diag- nostic for M. verticornis and M. giganteus. The morphology is generally similar to that of Mega- loceros giganteus, except that the epicondylus la- teralis has no pronounced bony knob (cha. 22).
R a d i u s a n d U l n a The caput olecrani is not fused to the ulna dia- physis, and the fusion of the distal epiphysis of the radius is incomplete. In addition, there is no bony connection between the diaphyses of the radius and ulna. This is consistent with an onto- genetic stage of two years. The attachment of the ligamentum collateralis laterale at the proximal end of the radius diaphysis projects out laterally, and slopes a little downwards (cha. 26). Megalo- ceros verticornis from Bilshausen shares this de- velopment with Megaloceros giganteus and the genus Cewus. Characters 34-37 could not be unequivocally assessed because the stag was rela- tively young.
M e t a c a r p u s I11 + I V The epiphyses of the right and left metacarpus I11 + IV are nearly completely fused. In most species of deer this epiphysis fuses in the end of the second year. The dimensions of these bones are similar to those of Megaloceros giganteus, moreover the specimen of M. verticornis from Bilshausen has also stout diaphyses of MC I11 + IV as is typical for the giant deer. The mor- phology of the distal end of MC I11 + IV is gen- erally similar to that of Megaloceros verticornis (cha. 43, 44, 45) and to most plesiometacarpal species of deer.
P e l v i s The bones of the pelvis are broken into numer- ous pieces, and no metric data could be ob- tained. The symphysis pelvina was not fused at the time of death.
Tabl
e 2
Des
crip
tion
of 12
2 sk
elet
al c
hara
cter
s and
thei
r ch
arac
ter s
tate
s fo
r va
riou
s cer
vids
use
d fo
r th
e co
mpu
ter-
aide
d ph
ylog
eny
reco
nstr
uctio
n.
0
1 2 3
0
Atla
s 0
Ala
estra
ight
19
skel
etal
4
elem
ent
8 D
escr
iptio
n of
cha
ract
er s
tate
s z
1 A
laec
onve
x
2 A
laec
onca
ve
Atla
s 0
Atla
s cr
ania
lly a
nd c
auda
lly n
arro
w
1 A
tlas
cran
ially
nar
row
, cau
dally
bro
ad
2 A
lae
cran
ially
and
cau
dally
bro
ad
14
-
Atla
s 0
Ala
e no
t elo
ngat
ed ca
udal
ly
1 A
lae
a lit
tle e
long
ated
caud
ally
2 A
lae
espe
cial
ly e
long
ated
cau
dally
3 A
lae
esD
ecia
llv e
lona
ated
and
wid
ened
cau
dallv
m u
0
0 1 2 3 4 0 1 2 3 4 0 1 2 0 1 2 0 1 2 0 1 2 3 0 1 2 0 1 2 3
4 5 0 1 2 3 4 5
I 1
S. d
. ins
erts
beh
ind
the
dens
in la
tera
l vie
w
7 ce
rvic
al-
0 pr
oxim
al e
piph
y
Des
crip
tion
of c
hara
cter
sta
tes
the
cav.
gle
noid
alis
show
s no
V-s
hape
d ca
vity
in it
s ar
ticul
ar s
urfa
ce
the
cav.
gle
noid
alis
show
s a
V-s
hape
d ca
vity
in a
few
cas
es
the
cav.
gle
noid
alis
show
s a
V-s
hape
d ca
vity
with
a fr
eque
ncy
up to
25%
the
cav.
gle
noid
alis
show
s a
V-s
hape
d ca
vity
with
a fr
eque
ncy
up to
50%
the
cav.
gle
noid
alis
sho
ws
a V
-sha
ped
cavi
ty w
ith a
freq
uenc
y up
to 8
5%
the
tube
rcul
um s
upra
glen
oida
le is
sle
nder
, med
ially
cur
ved
the
tub.
sup
. is
stro
ng, e
long
ated
, med
ially
cur
ved
the
tub.
sup
. is
espe
cial
ly lo
ng
the
tub.
sup
. is
espe
cial
ly s
trong
, lon
g, n
ot v
entra
lly c
urve
d
the
tub.
sup
. is
shor
t, co
arse
, not
ven
trally
cuw
ed
no fo
ram
en d
orsa
lly b
etw
een
proc
. cor
acoi
deus
and
cay
. gle
noid
alis
a la
rge
fora
men
dor
sally
bet
wee
n pr
oc. c
or. a
nd c
av. G
leno
idal
is
a ca
vity
dor
sally
bet
wee
n pr
oc. c
or. a
nd c
ay. g
leno
idal
is
the
spin
a sc
apul
ae e
nds
prox
imal
ly w
ith a
pea
k, fo
rmin
g an
ang
le o
f 80-90° w
ith th
e sh
aft
the
spin
a sc
apul
ae is
esp
ecia
lly e
long
ated
in th
e di
rect
ion
of th
e ca
v. g
leno
idal
is
the
spin
a sc
apul
ae e
nds
prox
imal
ly w
ith a
n an
gle
of a
bout
100
" with
the
shaf
t
the
scap
ula
shaf
t is
slen
der
the
scap
ula
shaf
t is
stro
ng
the
scap
ula
shaf
t is
espe
cial
ly w
iden
ed
the
tube
rcul
um m
ajus
is fl
atte
nd, w
ith n
o ca
vity
the
tub.
maj
us is
slig
htly
inde
nted
in a
few
cas
es
the
tub.
maj
us is
inde
nted
in m
ost c
ases
the
tub.
maj
us is
dee
ply
inde
nted
the
tub.
maj
us w
ithou
t vis
ible
mus
cle
atta
chm
ent l
ater
ally
the
tub.
maj
us w
ith a
wea
k m
uscl
e at
tach
men
t lat
eral
ly
the
tub.
maj
us w
ith a
mar
ked
mus
cle
atta
chm
ent l
ater
ally
the
epic
ondy
lus
hum
eri is
axi
ally
stro
nger
than
late
rally
the
epic
ondy
lus h
umer
i is a
xial
ly a
lso
slen
der i
n a
few
cas
es
the
epic
ondy
lus
hum
eri is
axi
ally
sle
nder
with
a fr
eque
ncy
up to
50%
the
epic
ondy
lus
hum
eri is
axi
ally
sle
nder
in m
ost c
ases
the
epic
ondy
lus
hum
eri is
axi
ally
esp
ecia
lly s
trong
in a
few
cas
es
the
epic
ondy
lus h
umer
i is a
xial
ly a
lway
s es
peci
ally
stro
ng
the
scor
es o
f the
foss
a ra
dial
is s
how
var
iabl
e sh
apes
the
foss
a ra
dial
is h
as tw
o w
edge
-sha
ped
scor
es
the
late
ral s
core
of t
he fo
ssa
radi
alis
is la
ckin
g in
a fe
w c
ases
the
late
ral s
core
of t
he fo
ssa
radi
alis
is la
ckin
g w
ith a
freq
uenc
y up
to 5
0%
the
foss
a ra
dial
is s
how
s la
tera
lly n
o sc
ore
the
foss
a ra
dial
is sh
ows
two
roun
ded
scor
es
3 pr
ox. e
piph
ysis
roun
ded,
8 ce
rvic
al-
0 no
tube
rcle
on
the
caud
al
4 5 6
I I v
erte
brae
I 1
la tu
berc
le o
n th
e ca
udal
pro
cess
I 1
9
Axi
s 0
Spi
na d
orsa
lis lo
w, s
traig
ht, c
rani
ally
poi
nted
1
5
1 S
.d. c
rani
ally
roun
ded,
caud
ally
risi
ng
2 S
.d. c
rani
ally
poi
nted
, cau
dally
risi
ng
3 S
.d. c
rani
ally
roun
ded,
caud
ally
ext
rem
ely
risin
g 16
-
Axi
s 0
Den
ssho
rt
1 D
ens
elon
gate
d
Axi
s 0
S.d.
ins
erts
abo
ve th
e de
ns in
late
ral v
iew
1
7
cerv
ical
- 0
trans
vers
e pr
oces
ses
shor
t, na
rrow
verte
brae
1
trans
vers
e pr
oces
ses s
hort,
bro
ad
19
1
I 2 I transv
erse
Dro
cess
es e
lona
ated
. nar
row
10
1 1
12
6th
lum
bar
0 di
stal
epi
phys
is w
ithou
t any
pro
cess
verte
bra
1 di
stal
epi
phys
is w
ith s
hort
proc
esse
s la
tera
lly
2 di
stal
epi
phys
is w
ith e
long
ated
pro
cess
es la
tera
lly
3 di
stal
epi
phys
is w
ith s
hort
proc
esse
s ve
ntra
l, la
tera
lly d
irect
ed
0 0
s sa
crum
and
6Ih
lum
b. v
erte
bra
with
a fl
exib
le c
onne
ctio
n 6t
h lu
mba
r
verte
bra
1 0
s sa
crum
and
Sth
lum
b. v
erte
bra
are
inte
rlock
ed
Sca
pula
0
the
cay.
gle
noid
alis
is o
val,
mor
e el
onga
ted
than
wid
e
II
the
cav.
gle
noid
alis
show
s a
knob
in d
irect
ion
to th
e sp
ine
in a
few
cas
es
the
knob
occ
urrs
with
a fr
eque
ncy
of 2
5% -
50%
the
knob
occ
urrs
with
a fr
eque
ncy
of m
ore
than
50%
I 4
the
cav.
gle
noid
alis
show
s a
knob
in d
irect
ion
of th
e tu
b. s
upra
glen
oida
le in
a fe
w c
ases
II
II
5
the
cav.
ole
noid
alis
is w
ider
than
lono
-
20
-
21
skel
etal
el
emen
t
Sca
pula
Sca
pula
Sca
Dul
a
Sca
pula
Sca
pula
Hum
erus
Hum
erus
Hum
erus
Hum
erus
EP -
ZP -
(ledie3eiawo!sald) Alleis!p paanpal sle!podeiaw
ap!s 0
ued io!iaisod si! u! uaweio4 paie6uola ue S
MO
~S
euln pue sn!p-ei uaamiaq ainins aqi
L
AI+III 3
W
8E -
9E adeqs paddoi-ielj 'ieln6U
e ue swio) '3alo 'qni aqi lo a6pa aqi z
uo!SS
aidap yeam e q
i!~
papunoi s! !uem
aio iaqni aqi 40 a6pa ayi
aJo3s padeqs-A e q
i!~
papunoi S
I !uemaio iaqni aqi 40 a6pa aqi o
sniawnH
skel
eta
2 I ele
men
a, a
0 0
6
skel
etal
el
emen
t 55
Ti
bia
Des
crip
tion
of c
hara
cter
stat
es
the
gap
betw
een
the
dist
al tr
ochl
eae
of M
C 11
1 and
IV fo
rms
a ke
y ho
le
51 I F
emur
di
stal
a, x 0 0 1 2
dist
al
prox
imal
prox
imal
Des
crip
tion o
f cha
ract
er s
tate
s
the
tube
rosi
ty fo
r the
liga
men
tum
cruc
iatu
m fo
rms
an e
long
ated
U-s
hape
the
tub.
for
the
lig. c
ruc.
form
s an
elo
ngat
ed o
r sh
ort U
-sha
pe
the
tub.
for t
he li
g. c
ruc.
is s
hort,
form
ing
a w
eak
U-s
hape
44
45
46
47
48
49
50
MC
III+
I
dist
al
MC
III+
I
dist
al
Fem
ur
prox
imal
Fem
ur
prox
imal
Fem
ur
prox
imal
Fem
ur
prox
imal
Fem
ur
1 th
e ga
p be
twee
n th
e di
stal
troc
hlea
e of
MC
111 a
nd IV
has
par
alle
l mar
gins
pr
oxim
al
1 th
e m
uscl
e at
t. la
tera
lly a
t the
t. m
ajor
form
s a
wea
k ke
el o
r a k
nob
2 I the m
. att.
late
rally
at t
he 1
. maj
or fo
rms
a kn
ob.
1 2 0 1 2 3 0 1 2 3 4 0 th
e pr
onou
nced
poi
nt is
enl
arge
d in
som
e ca
ses
the
poin
t at t
he c
ochl
ea ti
biae
is a
lway
s en
larg
ed a
nd v
ery
pron
ounc
ed
the
diap
hysi
s ha
s a
flatte
ned
surfa
ce p
oste
riorly
the
diap
hysi
s sh
ows
long
, stro
ng k
eels
pos
terio
rly
the
foss
a tro
chan
teric
a sh
ows
a w
eak,
U-s
hape
d de
pres
sion
in it
s m
argi
n
in m
ost c
ases
the
f. tro
chan
teric
a sh
ows
a de
ep U
-sha
ped
depr
essi
on in
its
mar
gin
the
f. tro
chan
teric
a sho
ws
alw
ays
a de
ep U
-sha
ped
depr
essi
on in
its
mar
gin
the
f. tro
chan
teric
a sh
ows
a fla
t, di
agon
ally
dire
cted
mar
gin
the
edge
of t
he tr
ocha
nter
maj
or fo
rms
a fla
t-top
ped
shap
e
the
edge
of t
he t.
maj
or is
risi
ng a
nter
iorly
in a
few
cas
es
the
edge
of t
he 1
. maj
or is
risi
ng a
nter
iorly
the
edge
of t
he 1.
maj
or s
how
s tw
o sm
all s
core
s
the
edge
of t
he t.
maj
or h
as o
ne d
eep
scor
e in
the
mid
dle
the
mus
cle
atta
chm
ent l
ater
ally
at t
he tr
ocha
nter
maj
or fo
rms
a w
eak
keel
.
3 th
e tu
b. fo
r the
lig.
cru
c. s
how
s a
depr
essi
on la
tera
lly
4 th
e tu
b. fo
r th
e lig
. cru
c. is
sho
rt, ro
unde
d, re
stric
ted
by a
sm
all r
idge
ant
erio
rly
0
the
att.
of th
e lig
. col
. tar
s. m
edia
lis is
hor
izon
tal,
form
ing
a pr
onou
nced
poi
nt a
nter
iorly
1 th
e at
t. of
the
lig. c
ol. t
ars.
med
ialis
is h
oriz
onta
lly d
irect
ed, e
ndin
g an
terio
rly
2 th
e at
t. of
the
lig. c
ol. t
ars.
med
ialis
turn
s up
war
ds, f
orm
ing
a cu
rve
3 th
e at
t. of
the
lig. c
ol. t
ars.
med
ialis
is ro
undi
sh, n
ot c
lear
cut
57
Tibi
a 0
the
dors
al fa
cet i
n co
ntac
t with
the
mal
leol
us la
tera
lis is
sm
all
56
Tibi
a
dist
al
dist
al
1 th
e do
rs. f
ac. i
n co
ntac
t with
the
mal
. lat
eral
is is
var
iabl
e in
siz
e
2 th
e do
rs. f
. in
cont
act w
ith th
e m
al. l
ater
alis
is v
ery
smal
l
3 th
e do
rs f
. in
cont
act w
ith th
e m
al. l
ater
alis
is e
nlar
ged
4 th
e do
rs. f
. in
cont
act w
ith th
e m
al. l
ater
alis
is e
nlar
ged
and
pron
ounc
ed
58
Tibi
a 0
the
late
ral p
art o
f the
coc
hlea
tibi
ae s
how
s a
smal
l, pr
onou
nced
poi
nt
3 4 0 1 2 th
e kn
ob is
usu
ally
acc
ompa
nied
by
a de
pres
sion
3 th
e at
t. of
the
m. v
ast.
Inte
rm. f
orm
s an
elo
ngat
ed k
eel
I the
m. a
tt. la
tera
lly a
t the
1. m
ajor
form
s a
roug
h, d
epre
ssed
scor
e
the
m. a
n. la
tera
lly a
t the
1. m
ajor
form
s a
mar
ked
keel
.
the
att.
of th
e m
. vas
tus
inte
rmed
ius
form
s a
wea
k kn
ob a
t the
pro
xim
al fe
mur
the
knob
is a
ccom
pani
ed b
y a
depr
essi
on in
som
e ca
ses
3 th
e la
tera
l par
t of t
he c
ochl
ea ti
biae
is n
early
flat
tene
d, fo
rmin
g a
wea
k cu
rve
0 th
e M
T 111
and
IV p
roje
ct a
ppro
xim
atel
y eq
ually
with
thei
r pro
xim
al fa
cets
1 th
e pr
oxim
al fa
cet o
f M
T 111
pro
ject
s cl
early
abo
ve th
e fa
cet o
f M
T IV
0 on
the
axia
l edg
e be
twee
n M
T 111
and
MT
II a
bul
ge e
xcee
ds th
e pr
oxim
al fa
cets
59
MT
III+I
V
prox
imal
MT
111 +
IV
60
1 th
e bu
lge
is e
spec
ially
pro
noun
ced
in m
ost c
ases
I I pro
xim
al I 2 I th
e bu
lge
is a
lway
s ve
ry p
rono
unce
d
4 0 1 0 1 0 1 2 0 1 2 0 1
the
att.
of th
e m
. vas
t. tn
term
. for
ms
a m
arke
d kn
ob d
ista
lly o
f the
troc
h. m
ajor
the
diap
hysi
s is
roun
ded
in c
ross
sec
tion
the
diap
hysi
s is
pos
terio
rly ke
eled
in c
ross
sec
tion
the
faci
es p
oplit
ea is
flat
tene
d, ro
ugh
the
f. po
plite
a fo
rms
a ro
ugh,
pro
min
ent t
uber
osity
the
foss
a in
terc
ondy
laris
is s
hallo
w, t
he d
ista
nce
of th
e co
ndyl
es is
sm
all
the
f. in
terc
ond.
varie
s in
its
dept
h, th
e di
stan
ce o
f the
con
dyle
s is
wid
e
the
f. in
terc
ond.
is d
eep,
the
dist
ance
of t
he c
ondy
les
is s
mal
l
the
inte
rnal
pea
k of
the
tub.
inte
rcon
dyla
re p
roje
cts
clea
rly b
eyon
d th
e ex
tern
al p
eak
the
tub.
inte
rcon
d. p
roje
cts
varia
bly
with
its
inte
rnal
and
ext
erna
l pea
k
the
tub.
inte
rcon
d. p
roje
cts
appr
oxim
atel
y eq
ually
with
its
inte
rnal
and
ext
erna
l pea
k
the
area
inte
rcon
d. ce
ntra
lis is
pos
t. sh
ort,
rest
ricte
d by
the
tub.
for t
he li
g. c
ruci
atum
the
a. i.
c. i
s la
tera
lly e
long
ated
, the
face
t of t
he la
tera
l con
dyle
is w
eakl
y el
onga
ted
3
the
bulg
e be
twee
n th
e fa
cets
of M
T 111
and
MT
II is
acc
ompa
nied
by
a sm
all g
ully
4 th
e bu
lge
is v
ery
smal
l, ac
com
pani
ed b
y a
gully
5 on
the
axia
l edg
e be
twee
n M
T 111
and
MT
I1 e
xist
s on
ly a
gul
ly
6 th
e pr
oxim
al fa
cets
of
MT
111 a
nd M
T II
bord
er o
n ea
ch o
ther
, the
re is
no
bulg
e or
gul
ly
0
the
shap
e of
the
post
erio
r fac
et in
con
tact
with
the
cubo
-nav
icul
ar b
one
is s
hort
1 th
e po
st. f
. is
mor
e el
onga
ted,
ove
rlapp
ing
with
the
face
t of t
he M
T 111
2 th
e po
st. f
. is
mor
e el
onga
ted,
ove
rlapp
ing
up to
the
mid
dle
of th
e fa
cet o
f the
MT
Ill
3 th
e po
st. f
. is
elon
gate
d, e
xcee
ding
the
mid
dle
of th
e fa
cet o
f the
MT
Ill
4 th
e po
st. f
. is
espe
cial
ly e
long
ated
, tra
nsve
rsal
ly d
irect
ed, f
latte
ned
0 th
e po
ster
ior f
acet
pro
ject
s sl
ight
ly a
bove
the
face
ts o
f MT
111 a
nd M
T IV
1 th
e po
st. f
. is
shor
t, an
d fo
rms
a pr
onou
nced
pea
k in
the
mid
dle
betw
en M
T 111
and
IV
2 th
e po
st. f
, is
elon
gate
d, w
ith a
sm
all p
eak
in th
e m
iddl
e be
twen
MT
111 a
nd IV
3
the
post
. f. i
s el
onga
ted
and
proj
ects
cle
arly
abo
ve th
e fa
cets
of M
T Ill
and
MT
IV
61
M
T 111
+ IV
prox
imal
-~
62
M
T 111
+ IV
prox
imal
2 th
e a.
i. c
. for
ms
a gu
lly, t
he fa
cet o
f the
lat.
cond
yle
is d
istin
ctly
elo
ngat
ed p
oste
riorly
I
4 th
e po
st. f
. is
flatte
nd, p
roje
ctin
g eq
ually
with
the
face
ts o
f MT
111 a
nd IV
skel
etal
e
lem
en
t
MT
Ill +
IV
prox
imal
I) 3 3 1
MT
111 +
IV
dist
al
orde
red
MT
III+I
V
Des
crip
tion
of c
hara
cter
sta
tes
lat.
a w
eak
groo
ve b
orde
rs th
e di
st. p
art o
f the
troc
h. in
con
tact
with
the
cubo
navi
c. pr
oc.
the
groo
ve is
mor
e st
eepl
v in
clin
ed, f
orm
ing
an e
xact
ly a
dius
ted
conn
ectio
n w
ith th
e c.
D.
MT
III+I
V
0 on
e sm
all,
post
erio
rly e
long
ated
fora
men
is lo
cate
d be
tw. t
he fa
cets
of M
T 111
and
IV
Cal
cane
us
73
Cal
cane
us
D 1 D 1 2 3 4 5 6 7 0 1 2 3
Cal
cane
us
Talu
s
corp
us
Talu
s
capu
t
smal
l bul
ge la
tera
lly b
etw
een
the
troch
lea
talis
pro
x. a
nd d
ist.
in a
nter
ior v
iew
pron
ounc
ed bu
lge
late
rally
bet
wee
n th
e tr
ochl
ea ta
lis p
rox.
and
dis
t. in
ant
erio
r vie
w
prox
imal
two
elon
gate
d bu
lges
axi
ally
and
late
rally
exc
eed
the
mid
dle
of th
e di
aphy
sis
the
late
ral b
ulge
is s
hort
er in
mos
t cas
es
the
axia
l bul
ge is
alw
ays
muc
h lo
nger
than
the
late
ral o
ne
the
expr
essi
on of
the
bulg
es va
ries
betw
een
0, 1
, and
4
the
axia
l, ve
ry p
rono
unce
d bu
lge
reac
hes
the
mid
dle
of th
e di
aph.
, a la
t. bu
lge
is la
ckin
g
the
expr
essi
on o
f the
bul
ges
varie
s be
twee
n 0,
7
expr
essi
on 0
is ra
re, e
xpre
ssio
n 7 o
ccur
s in
mos
t cas
es
the
prox
imal
bul
ges
are
stro
ng b
ut s
hort,
nev
er re
achi
ng th
e m
iddl
e of
the
diap
hysi
s
the
cros
s se
ctio
n of
the
diap
hysi
s is
som
ewha
t sm
alle
r tha
n th
e pr
ox. e
piph
ysis
the
diap
hysi
s is
sle
nder
, but
can
be
stro
nger
in a
dult
stag
s
the
diap
hysi
s is
alw
ays
slen
der
the
diap
hysi
s is
very
sle
nder
and
elo
ngat
ed
Talu
s
post
erio
r
view
0 1 2 3 gl 0 D
escr
iptio
n of
cha
ract
er s
tate
s In
the
diap
hysi
s has
a m
ediu
m d
iam
eter
the
diap
hysi
s is
slen
der
the
diap
hysi
s is
ver
y sl
ende
r
the
diap
hysi
s is
stro
ng
0 1 th
e fo
ram
en c
an h
ave
a sm
all o
r wid
er o
peni
ng
2 th
e fo
ram
en h
as a
wid
e, p
oste
riorly
elo
ngat
ed o
peni
ng
smal
l ste
p be
twee
n th
e ed
ge o
f the
sus
t. ta
li an
d th
e ar
t. fa
cet o
f the
cal
c. s
haft
77
3 th
e ca
nalis
met
atar
si ca
n en
d in
one
sm
all f
oram
en o
r in
two
smal
l for
amin
a
4 th
e ca
nalis
met
atar
si ca
n en
d in
one
wid
er fo
ram
en a
nd a
net
wor
k of
por
es
5 th
e ca
nalis
met
atar
si e
nds
alw
avs
in tw
o fo
ram
ina
1 2 3 4 5
the
U-s
hape
d sc
ore
is la
ckin
g in
a fe
w c
ases
the
U-s
hape
d sc
ore
occu
rs w
ith a
freq
uenc
y up
to 7
5%
the
U-s
hape
d sco
re o
ccur
s w
ith a
freq
uenc
y up
to 5
0%
the
U-s
hape
d sco
re is
rare
a U
-sha
oed s
core
is a
lwav
s la
ckin
a in
the
orox
imal
face
t
76
0 1 2 3 0 1 2 3 0 1 2 3 4 5
the
mar
gin
of th
e di
stal
troc
hlea
show
s a
wea
k V
-sha
pe
the
mar
gin
of th
e di
stal
troc
hlea
show
s a
wea
k w
avel
ine
the
mar
gin
of th
e di
stal
troc
hlea
is fl
at
the
mar
gin
of th
e di
st. t
r. is
wav
y, e
ndin
g in
an
elon
gate
d pe
ak a
xial
ly
The
1st p
hala
nx p
ost.
show
s tw
o sh
ort,
wea
kly
expr
esse
d bu
lges
ant
erio
rly
the
axia
l bul
ge is
elo
ngat
ed a
nd m
ore
pron
ounc
ed to
geth
er w
ith (
0)
toge
ther
with
cha
ract
er st
ate
(0) o
ccur
s (3
)
the
axia
l bul
ge is
wea
k bu
t rou
gh, r
each
ing
the
mid
dle
of th
e di
aphy
sis
the
tend
on a
ttach
men
t late
ral, o
n th
e m
iddl
e of
the
diap
hysi
s var
ies
in it
s ex
pres
sion
the
tend
. att.
Lat
., on
the
mid
. of t
he d
iaph
. is
expr
esse
d as
a tu
bero
sity
in m
ost c
ases
char
acte
r sta
te (1
) is
alw
ays
deve
lope
d
the
tend
. an.
is e
xpre
ssed
onl
y by
a ro
ugh
surf
ace
the
lat.
tend
. att.
on
the
diap
hysi
s is
loca
ted
very
dis
tally
two
mar
ked
tube
rosi
ties a
xial
lv a
nd la
tera
llv a
re lo
cate
d on
the
mid
dle
of th
e di
aohv
sis
0 th
e pr
oc. c
orac
oide
us is
sm
all
1 th
e pr
oc. c
orac
oide
us fo
rms
a pr
onou
nced
pea
k I
0 th
e la
tera
l ext
ensi
on o
f the
ridg
e at
the
corp
us ta
li va
ries
in it
s pr
oxim
al a
nd d
ista
l par
t 79
1
the
lat.
ext.
of th
e rid
ge a
t the
c. t
ali i
s eq
ual i
n its
pro
xim
al a
nd d
ista
l par
t
2 I the la
t. ex
t. of
the
ridge
is g
reat
er in
its
prox
imal
par
t tha
n in
its
dist
al p
art
3 0 1 2 3 4 0 1 2 3
the
lat.
ridge
at t
he c
. tal
i is
pron
ounc
ed in
its
prox
imal
par
t and
end
s in
the
dist
al p
art
troc
hlea
tali
dist
alis
with
out a
V-s
hape
d pi
t
a V
-sha
ped
pit i
n th
e tr
. tal
is d
ist.
occu
rrs
in a
few
cas
es
a V
-sha
ped
pit o
ccur
rs w
ith a
freq
uenc
y of
abo
ut 5
0%
a V
-sha
ped
pit o
ccur
rs in
mos
t cas
es
troc
hlea
tali
dist
alis
with
a fl
at U
-sha
ped p
it
troc
hlea
tali
dist
alis
with
a d
eep,
ver
tical
pit
a pi
t in
the
troc
hlea
tali
dist
alis
is la
ckin
g
troc
hlea
tali
dist
alis
with
a tr
ansv
ersa
lly d
irect
ed p
it
troc
hlea
tali
dist
alis
with
a s
mal
l dep
ress
ion
axia
lly
I 80
-
81
ske
leta
l e
lem
en
t
0 1 2
Tal
us
post
. vie
w
~
the
mar
gin
of t
he d
ista
l tro
chle
a is
wea
kly
V-s
hape
d
the
mar
gin
of th
e di
stal
troc
hlea
show
s (0
) and
(3) w
ith a
n eq
ual f
requ
ency
char
acte
r sta
te (3
) occ
urs
with
a h
ighe
r fre
quen
cy th
an (
0)
Tal
us
ant.
view
1st P
h. a
nt.
Dos
t.vie
w
1st P
h.an
t.
1st P
h.an
t.
top
view
1st P
h.an
t.
post
. vie
w
1st P
h. p
ost.
post
. vie
w
1st P
h. p
ost.
post
. vie
w
1st P
h. p
ost.
post
. vie
w
4 th
e di
aphy
sis i
s es
peci
ally
stro
ng
0 th
e 1s
t pha
. ant
. sho
ws
a de
ep U
-sha
ped s
core
pos
t., in
the
mid
dle
of th
e an
t. fa
cet
skel
etal
$
elem
ent
8 D
escr
iptio
n of
cha
ract
er s
tate
s 3
the
mar
gin
of th
e di
stal
troc
hlea
form
s a
wea
k w
avel
ine
the
mar
gin
of t
he d
ista
l tro
chle
a sh
ows
(0) a
nd (5
) with
equ
al fr
eque
ncy
the
mar
gin
of t
he d
ista
l tro
chle
a is
flat
tene
d
the
mar
ain
of t
he d
ista
l tro
chle
a is
flat
tene
d. b
ut w
avv
in a
few
cas
es
0
2
93
0)
'p 8 0 1 2 3 4
the
axia
l bul
ge is
esp
ecia
lly e
lava
ted
in m
ost c
ases
the
axia
l bul
ge is
alm
ost e
spec
ially
ela
vate
d
Des
crip
tion
of c
hara
cter
sta
tes
a br
owtin
e (A
t) is
lack
ing
the
brow
tine
inse
rts h
igh
abov
e th
e bu
rr, f
orm
ing
an a
cute
ang
le w
ith th
e be
am
the
brow
tine
inse
rts 4
cm
abo
ve th
e bu
rr, a
t rig
ht a
ngle
s to
the
beam
the
brow
tine
inse
rts d
irect
ly a
bove
the
burr
, for
min
g an
obt
use
angl
e w
tth th
e be
am
the
brow
tine
is fl
atte
ned,
with
furc
ated
end
s
197
1st P
h. p
ost.
0 th
e sa
gita
lly d
irect
ed s
core
in th
e di
stal
troc
hlea
is a
lway
s sy
mm
etric
al
post
. vie
w
1 th
e s.
dire
cted
sc.
in th
e di
st. t
. is
sym
m. o
r asy
mm
. with
an
equa
l fre
quen
cy
2 th
e sa
gita
lly d
irect
ed s
core
in th
e di
stal
troc
hlea
is a
lway
s as
ymm
etric
al
1st P
h. p
ost.
0 th
e la
tera
l edg
e of
the
1st p
ha. p
ost.
is s
light
ly b
ent i
nwar
ds
ant.
view
1
the
late
ral e
dge
of th
e 1s
t pha
. pos
t. is
slig
htly
ben
t inw
ards
in m
ost c
ases
2 th
e la
tera
l edg
e of
the
1st p
ha. p
ost
is s
traig
ht. n
ot b
ent i
nwar
ds
tst P
h. p
ost.
0 th
e di
aphy
sis
is s
trong
ant.
view
1
the
diap
hysi
s is
sle
nder
2 th
e di
aphy
sis
is v
ery
slen
der,
axia
lly c
urve
d in
war
d
3 th
e ax
ial b
ulge
is e
spec
ially
pro
noun
ced,
thic
kene
d
2nd
Ph.
ant
. 0
the
dist
al tr
ochl
ea fo
rms
an e
long
ated
pea
k ax
ially
-
94
-
95
-
96
the
dist
al tr
ochl
ea fo
rms
an e
long
ated
pea
k ax
ially
in m
ost c
ases
the
dist
al tr
ochl
ea s
how
s a
wea
kly
expr
esse
d pe
ak a
xial
ly
the
dist
al tr
ochl
ea s
how
s tw
o sh
ort D
eakS
axi
allv
and
late
rallv
5 th
e br
owtin
e is
alm
ost r
educ
ed
I 3r
d P
hal.
I 0 l
the
sole
of t
he 3
rd p
hala
nx is
axi
ally
edg
ed
0
botto
m si
de
3rd
Phal
,
1 th
e so
le o
f the
3rd
pha
lanx
is b
ent i
nwar
ds a
xial
ly, n
ot e
dged
in s
ever
al c
ases
la
t. pa
rt of
the
fac.
in c
onta
ct w
ith th
e 2n
d ph
al. a
s br
oad
as th
e ax
ial p
art o
f the
fac.
, el
onga
ted
prox
.
a se
cond
ant
erio
r tin
e (A
2) is
lack
ing
post
, vie
w
lat.
part
of t
he f.
in c
onta
ct w
ith th
e 2n
d ph
a. is
nar
row
, but
ver
y el
onga
ted
prox
., gi
rdin
g th
e ax
. par
t of t
he f.
pro
x.
3rd
Pha
l. 0
3rd
phal
anx
with
a tr
iang
ular
sha
pe in
late
ral v
iew
lat.
view
1
3rd
Dha
lanx
dor
sallv
arc
hed.
roun
ded
I 2 0 1 2
. _
-
the
A2 in
serts
dire
ctly
abo
ve th
e A
1
the
A2 in
serts
cle
arly
abo
ve th
e A
1
a se
cond
ant
erio
r tin
e (A
2) is
lack
ing
the
A2 is
wea
ker t
han
the
A1
the
A2 is
stro
naer
than
the
A1
antle
r
0 1 0 1
2 3 4 0 1 2 3 0 1 2 3 4 5 6 0 1 2 0 1 2 3 4 5 x
0 th
e be
am fo
rms
a si
mpl
e pe
ak
1 th
e be
am e
nds
with
a b
ifurc
atio
n
2 th
ree-
poin
t ant
ler
3 fo
ur-p
oint
ant
ler
4 fiv
e-po
int a
ntle
r
5 si
x po
ints
or m
ore
6 pa
lmat
ed a
ntle
r
7
tota
lly d
iffer
ent c
onst
ruct
ion
x no
beam
a th
ird a
nter
ior t
ine
(A3)
is la
ckin
g
A3
is d
evel
oped
post
erio
r tin
es a
re la
ckin
g
post
erio
r tin
es a
re ra
re, r
ound
ed in
cro
ss s
ectio
n
post
erio
r tin
es a
re lo
ng, f
latte
ned
post
erio
r tin
es a
re n
umer
ous
but v
ery
shor
t
post
erio
r tin
es a
re a
lmos
t red
uced
the
antle
rs s
how
no
palm
atio
n
the
palm
atio
n is
rest
ricte
d to
the
term
inal
fork
the
beam
and
the
tines
are
pal
mat
ed
a ro
unde
d be
am m
erge
s to
a s
trong
ly d
evel
oped
pal
m
term
inal
tine
s ar
e la
ckin
g
the
beam
end
s w
ith a
sim
ple
bifu
rcat
ion
term
inal
tine
s so
lely
orie
ntat
ed a
nter
iorly
term
inal
tine
s or
ient
ated
ant
erio
rly in
mos
t cas
es
term
inal
tine
s w
ell b
alan
ced
orie
ntat
ed a
nter
iorly
and
pos
terio
rly
term
inal
tine
s or
ient
ated
pos
terio
rly in
mos
t cas
es
term
inal
tine
s so
lely
orie
ntat
ed p
oste
riorly
bifu
rcat
ed tin
es a
re la
ckin
g
bifu
rcat
ed tin
es a
re r
are
bifu
rcat
ed ti
nes
are
frequ
ent
beam
sho
rt. r
ound
ed
beam
elo
ngat
ed, r
ound
ed
beam
elo
ngat
ed, f
latte
nded
beam
ext
rem
ely
long
, rou
nded
, mer
ging
into
a p
lam
beam
long
, rou
nded
, mer
ging
into
a p
lam
beam
ext
rem
ely
shor
t, m
ergi
ng in
to a
pla
m
nobe
am
-
101
2nd
Ph.
ant
. 0
the
dist
al tr
ochl
ea is
axi
ally
a li
ttle
broa
der t
han
late
rally
post
. vie
w
1 th
e di
stal
troc
hlea
is a
xial
ly c
lear
ly b
road
er th
an la
tera
lly
!nd
Ph.
ant
. 0
the
prox
imal
dia
phys
is s
how
s no
pit
late
rally
ant.
view
1
the
Dro
xim
al d
iaoh
vsis
sho
ws
a D
it la
tera
llv
skel
etal
el
emen
t br
owtin
e
A1
seco
nd a
nt.
tine
(A2)
A2
99
A3
post
erio
r
tines
palm
atio
n
dire
ctio
n
of t
erm
inal
tines
bifu
rcat
ion
of t
ines
beam
skel
etal
el
emen
t be
am
surfa
ce
beam
posi
tion
a,
D s 0
beam
leng
th
6
skel
etal
2
elem
ent
a pe
arly
sur
face
to th
e be
am is
lack
ing
11
0
ethm
oid
Des
crip
tion o
f cha
ract
er s
tate
s
Ded
icle
0,
-0
0 0 1 2
ront
al b
ont
Des
crip
tion
of c
hara
cter
sta
tes
ethm
oid
vacu
ity la
rge,
elo
ngat
ed
ethm
oid
vacu
ity s
mal
l
ethm
oid
vacu
ity n
early
con
nate
d
inea
nuc
hi
occi
pita
l
bone
lacr
imal
orifi
ces
1 2 3 0 1 2 3 4 5 x 0 1
a pe
arly
sur
face
is w
eakl
y ex
pres
sed
a pe
arly
sur
face
is s
trong
ly e
xpre
ssed
a pe
arly
sur
face
is m
ost e
xten
sive
ly e
xpre
ssed
11
1 sk
ull
the
beam
s ar
e di
rect
ed u
prig
ht, n
early
par
alle
l sid
ed
sutu
res
the
beam
s sl
ope
a lit
tle tr
ansv
ersa
lly
11
2
uppe
i th
e be
ams
slop
e a
little
tran
sver
sally
, are
elo
ngat
ed c
auda
lly, e
xcee
ding
the
neck
the
beam
s be
nd tr
ansv
ersa
lly, p
oste
riorly
and
are
cur
ved
uprig
ht
11
3
uppe
r th
e be
am is
cur
ved
horiz
onta
lly, t
he a
ntle
r is
broa
dene
d tra
nsve
rsal
ly
the
beam
inse
rts tr
ansv
ersa
lly a
t the
sku
ll, th
e an
tler i
s tra
nsve
rsal
ly d
irect
ed
114
uppe
r no
bea
m
mol
ars
the
beam
s ar
e ve
ry s
hort,
less
than
30
cm
leng
th o
f the
bea
m 3
0 - 8
0 cm
11
5 up
per
vacu
ity
-~
prem
olar
s
prem
olar
s
0 1 0 1
leng
th o
f the
bea
m u
p to
100
cm
leng
th o
f the
bea
m u
p to
150
cm
, lof
ty
sutu
res
of th
e sk
ull a
re a
lmos
t vis
ible
in a
dult
indi
vidu
als
sutu
res
of th
e sk
ull n
early
com
plet
e co
nnat
ed in
adu
lt in
divi
dual
s
no c
ingu
lum
occ
urs
on th
e up
per
P4
a w
eak
cina
ulum
occ
urs
in th
e u
me
r P4
chec
k-te
eth
0 1
0 1 2
ante
rior,
lingu
al w
ing
of th
e pr
emol
ars w
ithou
t fol
ds in
side
, or w
ith o
ne s
impl
e fo
ld
ante
rior,
lingu
al w
ing
of th
e pr
emol
ars
with
a re
ticul
ated
fold
insi
de
ento
styl
i are
stro
ngly
dev
elop
ed
ento
styl
i are
sm
all,
not a
lway
s ab
unda
nt
ento
styl
i are
lack
ing
0 1 0 1 2 3 0
uppe
r mol
ars
show
labi
ally
an
equa
l, un
iform
row
,
the
row
of t
he u
p. te
eth
has
a st
ep-li
ke o
utlin
e la
bial
ly, s
hifte
d fro
m a
nt. t
o po
st.
fold
s in
side
the
ante
rior w
ing
of th
e pr
otoc
one
are
lack
ing
sim
ple
fold
s in
side
the
ant.
win
g of
the
prot
ocon
e ar
e cl
early
dev
elop
ed
furc
ated
fold
s in
side
the
ant.
win
g of
the
prot
ocon
e fo
rm a
n is
land
insi
de (A
lces
)
fold
s in
side
the
ant.
win
g re
ticul
ated
(Ran
gife
r)
an u
pper
can
ine
is s
trong
ly d
evel
oped
in d
4 5 6 x 0
leng
th o
f the
bea
m u
p to
200
cm
, lof
ty
leng
th o
f the
bea
m u
p to
250
cm
, tra
nsve
rsal
ly d
irect
ed
leng
th o
f the
bea
m u
p to
400
cm
, tra
nsve
rsal
ly d
irect
ed
nobe
am
11
7
uppe
r
pedi
cle
very
long
, circ
ular
in c
ross
-sec
tion
cani
ne
uppe
r mol
ars
1 2 0
an u
pper
can
ine
is ru
dim
enta
rily
deve
lope
d in
d
an u
pper
can
ine
is a
lway
s la
ckin
g
crow
n of
the
upp
er M
3 po
ster
iorly
dee
ply
notc
hed
1 2 x 1
lcro
wn
of t
he u
me
r M3
Dos
terio
rlv c
onna
ted
pedi
cle
shor
t, ci
rcul
ar in
cro
ss-s
ectio
n
pedi
cle
very
sho
rt, a
nter
ior-
post
erio
rly e
long
ated
in c
ross
-sec
tion
no D
edic
le
0 1 0 1 2 3 4
front
al b
one
not t
hick
ened
bet
wee
n th
e pe
dicl
es
119
man
dibl
e
front
al b
one
thic
kene
d an
d ar
ched
bet
wee
n th
e pe
dicl
es
expr
esse
d as
sim
ple,
roun
ded
cres
t
nuch
al c
rest
late
rally
stre
ngth
ened
, and
bro
aden
ed
120
man
dibl
e
nuch
al c
rest
form
s a
S-s
hape
P3
P4
nuch
al c
rest
form
s a
doub
le-S
-sha
pe
nuch
al c
rest
is a
ngul
ar in
the
mid
dle
0 th
e he
ight
of t
he o
ccip
ital b
one
betw
een
the
nuch
al c
rest
and
the
cond
yles
is lo
w
1 th
e oc
cipi
tal p
art i
s hi
gh
121
2 th
e oc
cipi
tal p
art i
s hi
gh, t
he n
ucha
l cre
st p
roje
cts
clea
rly b
ehin
d th
e co
ndyl
es
0
two
lacr
imal
orif
ices
1 on
e la
crim
al o
rific
e 1
22
man
dibl
e
M3
man
dibl
e
0 1 2 0 1 2 3
4 0 1 2 0
P3
not m
olar
ised
P3
with
a p
ost-
met
acris
tid
post
ento
- and
hyp
ocris
dtid
are
con
nate
d
P4
not m
olar
ised
a pr
aem
etac
ristid
is d
evel
oped
ento
coni
d an
d m
etac
onid
win
gs fu
sed,
form
ing
a di
agon
al c
rest
(A/c
es, R
angi
fer)
P4
alm
ost m
olar
ised
P4
alw
ays
mol
aris
ed
the
last
lobe
of t
he M
3 is
ver
y sm
all
the
last
lobe
of M
3 is
stro
ngly
dev
elop
ed, p
oste
riorly
roun
ded
the
last
lobe
is s
trong
ly d
evel
oped
, with
a p
oint
ed e
nd p
oste
riorly
in a
few
cas
es.
the
man
dibl
e is
nar
row
in c
ross
-sec
tion,
not
thic
kene
d
1 2
the
man
dibl
e is
a li
ttle
thic
kend
in c
ross
-sec
tion,
(wea
kly
pach
ygna
thou
s)
the
man
dibl
e is
ext
rem
ely
thic
kene
d in
cro
ss-s
ectlo
n (p
achy
gnat
hous
)
302 Pfeiffer, T., Complete skeleton of Megaloceros
F e m u r location and development of the muscle attach- The caput and the great trochanter of each fe- ments (cha. 48) on the femur can be used to as- mur were not connected with the diaphyses, sign specimens to particular species in adult indi- while the distal epiphyses had just begun to fuse, viduals (Pfeiffer 1999a), but could not be indicating an age of less than three years. The determined exactly here. As in M. giganteus and
Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 5 (2002) 303
Cervus, the diaphysis is rounded in cross section (cha. SO) , the fossa intercondylaris is deep, and the distance of the condyles is small (cha. 52). The femur of M. verticornis from Bilshausen has a rough, but flattened facies poplitea (cha. 51) by contrast to M. giganteus, where the facies po- plitea forms a rough, prominent tuberosity. This character is influenced by muscle development and individual age. Consequently, the relatively juvenile age of the individual from Bilshausen reflect the weaker expression of this feature.
T ib i a The proximal epiphyses are not fused to the dia- physes both of which were broken and subse- quently reconstructed. The morphological char- acters of the proximal and distal ends of the tibia are similar to those of Megaloceros gigan- teus, thus the internal peak of the tuberculum intercondylare clearly projects beyond the exter- nal apex (cha. 53) (compare figure 8 in Pfeiffer 1999c) and, in particular, the development of the distal facets exhibits the derived character states found in Megaloceros verticornis and M. gigan- teus (cha. 56, 57, 58). The dorsal facet that con- tacts the malleolus lateralis is especially pro- nounced (cha. 57).
C a l c a n e u s a n d t a l u s The calcaneus and talus of the right and left side yielded morphological data. The tuber calcanei
was not fused to the corpus calcanei. A small step between the edge of the sustentaculum tali and the articular facet of the calcaneus shaft (Pfeiffer 1999a: fig. 75a, la) represents a plesio- morphic character state (cha. 67).
The talus provides more information for the phylogenetic analysis, because all the characters utilised here can be studied in juvenile indivi- duals. The morphology of the talus of Megalo- ceros verticornis is generally similar to that of M. giganteus. An important, derived character is an axially directed depression in the trochlea tali distalis (cha. 72).
M e t a t a r s u s I11 + I V Both left and right metatarsals are complete, although their proximal part is reconstructed. The distal epiphyses were beginning to fuse with the diaphyses at the time of death. This ontoge- netic stage is consistent with an individual age of two years. Due to the young age of the specimen the diaphyses of MT I11 + IV are not connected distally. Character states for the distal MT I11 + IV are similar to those for Megaloceros giganteus (Pfeiffer 1999c: fig. 9).
P h a l a n g e s All phalanges of M. verticornis from Bilshausen could be studied. The epiphyses of the first and second phalanges, which fuse at the beginning of the second year in deer, are all completely fused
Fig. 4. A. Fragments of the skull and antler of an adult male of Megalo- ceros verticornis from SiiBenborn. B. The fron- tal bones are particularly thick. The strategic align- ment and distribution of trabecular bone is primar- ily related to the direc- tion of normal loading. This orientation and the thickening of the frontals was also observed in the specimen lrom SiiRen- born (IQW SUB 7117).
304 Pfeiffer, T., Complete skeleton of Megaloceros
in the Bilshausen specimen. The first and second phalanges also have robust diaphyses and exhibit most morphological characters of Megaloceros (Pfeiffer 1999c: fig. 10).
Discussion and results
Many skeletal characters are influenced by onto- genetic growth patterns, sexual dimorphism, and functional adaptation, and may show high varia- bility within a cervid species (Pfeiffer 1999a). An important problem is that characters of antlers and skulls of juvenile individuals of one species with a complex antler morphology can appear similar to characters of antlers of adult stages of other species. As an example Cewus elaphus, and C. nippon e.g. can have a three-point antler stage in juvenile individuals that is comparable to the antler stage of an adult of Axis axis (Ta- ble 2, cha. 92). In most species of deer the su- tures of the skull are still almost visible in adult individuals, but are nearly completely erased in adult megalocerines (cha. 11 1). Juvenile megalo- cerines do not have pachygnathous mandibles as in most other species of deer, but they can be extremely pachygnathous in old individuals (Ta- ble 2, cha. 122). All antler- and skull characters in the data matrix (cha. 92-111, 122) are influ- enced by ontogenetic growth patterns. It is there- fore extremely important to specify the character development in adult individuals of all species in the phylogeny reconstruction.
Megaloceros verticornis from SiiBenborn and Voigtstedt (Kahlke 1956, 1960, 1965, 1969) shows, that palmation clearly increases in older individuals. The enormous weight of these an- tlers resulted in special adaptations in cranial morphology. Bone material had to be added at the linea nucha, and this became a nuchal crest in older stags. Adult male deer with wide spread antlers therefore show an extreme thickening of the frontal bones between the pedicles. The stra- tegic alignment and distribution of trabecular bone is primarily related to the principal direc- tion of normal loading. This orientation and thickening of the frontals was observed in an adult male of Megaloceros verticornis from SiiBenborn (Fig. 4). Similarly, an increase in an- tler development in Alces latifrons is related to the thickening of the frontals (Pfeiffer 1999b). Unfortunately, in the palaeontological literature the classification of the alcini is mainly based on adaptive characters (Azzaroli 1953, 1983, 1985, 1994). DIG& (2uul) & S C U S ~ the skull m f ; i c -
ters of the alcini in more detail noting the struc- tural adaptations made necessary by an increase of antler- and bodysize. The development of a nuchal crest, for example, could be observed in many cervid species with strong antlers, includ- ing E. tegulensis, Cewus elaphus, Dama dama, M. giganteus, and Alces latifrons. Such indepen- dently acquired characters seem to be highly homoplastic in cervids, and are difficult to use for determining phylogenetic relationships. This observation concords with the results of Scott & Janis (1993).
R e s u l t s of t h e c o m p u t e r - a i d e d c l ad i s - t i c a n a l y s i s The parsimony analysis (MP) using PAUP 4.0 re- sulted in a single most parsimonious tree (length 579, CI: 0.578). The robustness of the phylogeny was assessed using the bootstrap method (Fel-
Bootstrap
55
Parablastomeryx Blastomeryx Mosch. moschiferus C. suessenbornensis C. pygargus C. capreolus Rangif. tarandus Alces. alces Eucl. tegulensis Eucl. dicranios Megal. verticornis MegaLgiganteus C. el.acoronatus C. el. hippelaphus C. n. nippon C.n. hortulorum Axis. axis Dama. rhenana Dama. nestii 0. d. clactoniana 0. d. geiselana D.d. dama 0.d. mesopotarnica
Fig. 5. Majority rule consensus tree, after 1000 bootstrap re- samplings, describing the evolutionary relationships of 20 fos- sil and Recent cervids, and 3 outgroup artiodactyls as de- duced from 122 skeletal characters. At each node the bootstrap percentage (BP) is given. One single most parsimo- nious tree (MP) was found with the same topology (length = 579, CI = 0.5872). The monophyly of Megaloceros verticornis and M. giganteus ia w ~ l l q p ~ c d (BP ~ 1001.
305 Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 5 (2002)
senstein 1985), and the maximum likelihood method (ML). The majority-rule consensus tree after 1000 bootstrap resamplings, describing the evolutionary relationships between 20 fossil and Recent cervids, and three outgroup artiodactyls based on a comparison of 122 skeletal characters obtain has the same topology as the MP tree (Fig. 5). The tree computed by the (ML) method differs only in the position of Axis, which is lo- cated in a clade with Darna in the ML analysis. At each node, the bootstrap percentage (BP) is given above the branch.
The family Cervidae is defined by 32 charac- ters at the basal node of the MP tree, and there is strong support for cervid monophyly from the bootstrap values (BP = 100). This result agrees with the phylogenetic analysis computed from molecular data (Cronin et al. 1996, Douzery & Randi 1997, Randi et al. 1998). The subfamilies, Odocoileinae and Cervinae, occurred as mono- phyletic groups within the Cervidae. Fourteen characters (BP = 70) define the telemetacarpal
Table 4
odocoileine deer, and 19 characters (BP = 97) define the plesiometacarpal Cervinae.
Within the Odocoileinae the genus Capreolus is considerably differentiated from the other taxa, with 19 characters defining its branch on the MP tree (BP = 100). Capreolus suessenbor- nensis (= C. cusanoides Kahlke 2001) from the late Lower Pleistocene of UntermaBfeld and the Middle Pleistocene of Germany (Mosbach, Mie- senheim, SiiBenborn) is nearly identical to its Recent relatives C. capreolus and C. pygargus in its skeletal characters. Three characters define the position of C. capreolus and the clade of C. pygargus and C. suessenbornensis, while one character defines C. suessenbornensis. These taxa differ only in body size and antler development, the diagnostic antler characters of C. cusanoides described by Kahlke (2001) can be included in the range of antler variation exhibited by C. suessenbornensis from Mosbach. Therefore C. cu- sanoides is assumed here to be a junior synonym of C. suessenbornensis.
Metric data for preserved teeth and postcranial skeletal elements of Megaloceros verticomis from Bilshausen (measurements after v. d. Driesch 1976).
Upper/ lower molars Mi M2 M3 ml m2 sin dex sin dex sin dex sin sin
max. length at crown-base in mm 26.0 29.0 (30.0) 29.5 29.0 29.0 29.5 (30.2) max. width at crown-base in mm 29.5 30.0 (31 .O) 29.5 29.5 18.5 19.0 18.2
Scapula height along the spine (HS in mm): 392.0 greatest length at the processus articularis (GLP) 90.0 width of the cavitas glenoidalis (BG) 61.5 length of the cavitas glenoidalis (LG) 72.5
Humerus
max. proximal width (BP) 117.0
Radius and Ulna greatest length of the radius (GLR) max. proximal width of the radius (BP) max. distal width of the radius (BD )
386.0 93.0 80.5
Metacarpus III+IV greatest length (GL in mm) 343.0 BP 65.0 BD 65.5
39.5 min. width of the diaphysis (KD)
Femur BP BD KD
Tibia BP BD
Calcaneus and talus
GL 176.0 GB 64.3
calcaneus
Metatarsus III+IV GL BP BD KD
143.5 116.5 42.5
119.0 78.5
talus 81.8 53.5
377.0 58.5 69.0 40.2
Phalanges anterior posterior 1" Ph. 2nd Ph. 3rd Ph. 1" Ph. 2nd Ph. 3rd Ph.
GL 80.0 61 .O 76.0 85.5 64.0 72.0 BP 34.0 33.5 36.8 33.8 KD 27.5 26.5 26.5 26.9 max. height (GH) 48.0 46.0
306 Pfeiffer, T., Complete skeleton of Megabceros
The association of Rangifer in a clade with Alces (BP = 96) is in contrast to the results of a study by (Cronin et al. 1996) based on K-casein DNA-data. Both genera, for example, share a derived morphology of the lower p4, and the clade is well supported by 25 characters in the analysis reported here. However, Rangifer taran- dus is considerably differentiated from Alces alces: 24 characters define R. tarandus, while 20 characters support A. alces.
The plesiometacarpal deer split into a Dama- lineage, defined by 26 characters (BP = 99), and a clade consisting of the fossil genera Euclado- ceros, Megaloceros, Cervus (fossil and Recent species), and the Recent species Axis, defined by 12 characters. The bootstrap value is low for this internal node (BP = 65, DI = +2).
As demonstrated by Pfeiffer (1999a) Axis shares many postcranial characters with the genus Cewus, and on the other hand, also with Dama, but locating Axis in a single clade with Cervus, Eucladoceros, and Megaloceros, involves 12 additional steps (by contrast to the 14 addi- tional steps, necessary to unit Axis in a clade with Dama), and this is reflected in the relatively low consistency index of the data set (CI: 0.578).
Tree topology within the Dama-lineage re- flects the evolution of this genus from the Plio- Pleistocene to the Holocene. In the Plio-Pleisto- cene Dama rhenana had evolved a three-point antler stage, followed in the Lower Pleistocene by Dama nestzi, with a 4-point antler without pal- mation, and then, in the Middle Pleistocene to Recent times, of fallow deer with palmated an- tlers. Considering the rich skeletal material of the Dama-lineage it can be observed that more than one character state may occur in a single species for a particular character. Quite often, both apomorphic and plesiomorphic character states, may occur in the same species as a poly- morphic character. Such apparently variable characters can yield important phylogenetic in- formation. Within the Darna-lineage it was possi- ble to follow the shift from a plesiomorphic char- acter state that occurs with a high frequency in Dama rhenana, which lies at the basis of the lineage, to an apomorphic character state in the stratigraphically younger Darna-species (Pfeiffer 1999a, in press).
The clade including Eucladoceros, Megaloceros and Cewus is defined by 23 characters (BP = 90). Comparing this data in detail, the characters 2, 4, 5, 6 and 9 (concerning cervical vertebrae), 76, 77, 78, 82, 83, 84, 85, (concerning lSt and 2"d phalanges), 95, 96, (concerning an-
tlers), and 105, 107, (concerning the skull) all support this node. Except the Japanese sika deer C. nippon nippon, all taxa in this clade are ro- bust cervids, with large and heavy antlers. The morphology of the cervical vertebrae is extre- mely influenced by the morphology and weight of the antlers, as demonstrated by Lengsfeld (1975) and Pfeiffer (1999a), and the phalanges need special adaptations to support the body weight (Pfeiffer 1999a, b). Therefore it should be taken into account that the high bootstrap value for this clade may be the result of homoplasy.
The clade including Megaloceros and Cervus is defined by 17 characters (BP=66). In particu- lar, the shared morphology of the tuberculum su- praglenoidale of the scapula (cha. 14), the simi- lar morphology of radius and ulna (cha. 26, 28, 29, 39, and the morphology of the upper M3 (cha. 118) was not observed in other plesiometa- carpal deer genera, and seems to support their rela tionship.
As in other phylogeny reconstructions com- puted from molecular data (Cronin et al. 1996, Douzery & Randi 1997, Randi etal. 1998) the Cervus clade including red deer and sika deer is well supported (17 characters, BP = 88). Within Cervus, the East-Asian sika deer Cervus nippon (BP = 96) appears to have been distinct from the European Cervus elaphus since the Pliocene (BP = 85). A large group of consistent charac- ters are present in Cervus, and appear to have persisted for a long time span. Moreover a study of bone material of Pseudaxis grayi Zdansky 1925 from Lower Pleistocene localities of the Shansi Basin in China shows evidence that I? grayi is an early, true member of Cervus nippon (Pfeiffer, in prep.).
Douzery & Randi (1997) calibrated divergence times on the basis of molecular data. In their study divergences of 3.3-7.1 Myr within the genus Cervus are postulated, and 0.4-2.5 Myr within Cervus elaphus.
Cervus elaphus was first identified in Europe from early Middle Pleistocene localities in Great Britain (Cromer Forest-bed, Lister 1996) and Germany (Voigtstedt, SiiBenborn, Mosbach, Kahlke, 1956, 1960, 1965, Koenigswald & Hein- rich 1999, Pfeiffer 1999a) represented by the fos- sil subspecies C. elaphus acoronatus. The antlers of Cervus elaphus acoronatus (= Cervus acoro- natus Beninde 1937) from Mosbach always lack a crown in the distal part, although this species exhibits a nearly identical postcranial character set to that of extant individuals of Cervus ela- phus hippelaphus. Three characters define C. e.
Mitt. Mus. Nat.kd. Berl., Geowiss. Reihe 5 (2002) 307
acoronatus, while five define the extant C. ela- phus hippelaphus.
The monophyly of Megaloceros verticornis and M. giganteus is well supported (PB =loo, 27 steps), and these species share 15 postcranial character states do not occur in other plesiome- tacarpal deer species: the cranial facet of the 3rd and 4th cervical vertebrae (cha. 7), the shape of the cavitas glenoidalis of the scapula (cha. 12), the rounded scores in the fossa radialis of the humerus (cha. 21), the axially especially deep fossa olecrani (cha. 24), the development of the cranial facet of the radius (cha. 29, 30), character states of the distal tibia (cha. 56, 58), develop- ment of the proximal end of MC III+IV (cha. 61, 62), and the astragalus (cha. 72, 73). These characters seem to be diagnostic for giant deer, and support a close relationship between the two giant deer species which are combined in a sin- gle genus, Megaloceros, here.
Duma and Megaloceros clearly represent sepa- rate lineages within the main clade of plesiome- tacarpal deer, because they share relatively few postcranial skeletal character states. Moreover the topology of the MP tree supports the idea that the occurrence of palmate antlers in both genera must be the result of homoplasy.
Acknowledgements
I wish to thank Pro€ Dr. D. Meischner and Prof Dr. J. Schnei- der (Gottingen) for permission to study the giant deer from Bilshausen. Their kind support during my visits to Gottingen was of great help. Fig. 1 was provided by Prof Meischner. I wish to thank Prof. Dr. H.-D. Kahlke (Weimar), Prof. Dr. W. v. Koenigswald (Bonn), Dr. G. Bohme, and Prof. Dr. H.-P. Schultze (Berlin) for kindly discussing issues covered in this study. I am very grateful to various scientists and curators who allowed me to study material in their care: Dr. M. Ade, Dr. W.-D. Heinrich (Berlin), Prof. Dr. A. v. d. Driesch (Munich), Dr. W. Eckloff (Lubeck), Dr. A. Feiler (Dresden), Dr. D. Heinrich (Kiel), Dr. R. Hutterer (Bonn), Dr. H. Jung (Mainz), PD Dr. R.-D. Kahlke (Weimar), Prof. Dr. D. Mania, (Jena), Dr. U. Scheer (Essen), Dr. G. Gruber (Darmstadt), Dr. G. Storch, Dr. G. Plodowski, Prof. Dr. F. Schrenck (Frankfurt), Dr. R. Ziegler, Dr. E. Heizmann (Stuttgart), Mr. E. Zenker (Wiesbaden), Dr. L. Abbazzi, Prof Dr. D. Torre, Dr. L. Rook (Florence), Dr. L. Capasso Babato, Prof Dr. C. Petronio (Rome), Prof. Dr. B. Engesser (Basel), Prof. Dr. C. Guerin, Dr. M. FaurC (Lyon), Dr. S. Stuenes (Uppsala), Dr. A. Sutcliffe, Dr. J. Hooker, Dr. A. Currant (London), Prof. Dr. E. Tchernov, Dr. R. Rabinovich (Jerusalem), Dr. R. Tedford (New York), Dr. J. d. Vos (Leiden). I am grateful to Dr. J. Dunlop, and Dr. D. Unwin (Berlin) for help in cor- recting the English text, and to Mr. J.-P. Mendau for produ- cing Figs. 2 and 4, Mrs. E. Siebert (Berlin) for amending Fig. 1. I thank Prof. Dr. G. Arratia, Dr. D. Unwin, Prof Dr. W. v. Koenigswald, and Dr. R. Ziegler for detailed reviews. This investigation was made possible by financial support from the Deutsche Forschungsgemeinschaft (DFG).
References
Alessandri, G. de 1903. Sopra alcuni avanzi di cervidi plio- cenici del Piemonte. - Atti della R. Academia della Sci- ence Torino.
Azzaroli, A. 1953. The deer of the Weybourne Crag and Fo- rest bed of Norfolk. - Bulletin of the British Museum (Natural History), Geology 2: 3-96.
- 1983. On the Quaternary and Recent Cervid Genera Al- ces, Cewalces, Libralces. - Bolletino Societa Paleontolo- gia Italiana 2 0 147-154.
- 1985. Taxonomy of Quaternary Alcini (Cervidae, Mam- malia). - Acta Zoologica Fennica 170: 179-180.
- 1994. Forest bed elks and giant deer revisited. - Zoologi- cal Journal of the Linnean Society 112: 119-133.
Azzaroli, A. & Mazza, P. 1992. On the possible origin of the Giant Deer genus Meguceroides. - Atti Accademia Lin- cei, Rendiconti Scienze Fisiche e Naturali (gth series) 3:
Belgrand, E. 1869. Histoire generale de Paris. La Seine I: Basin parisien aux antehistoriques 3: 13. La Seine I. Ba- sin parisien, planch. palCont. S.13, Taf. 18 ff.
Beninde, J. 1937. Zur Naturgeschichte des Rothirsches. - Monographien der Wildsaugetiere 4: 1-223.
Bittmann, F. & Muller, H. 1996. The Karlich Interglacial site and its correlation with the Bilshausen sequence. In Tur- ner, C. (ed.). The early Middle Pleistocene in Europe: 187-193. Balkema, Rotterdam.
Bogaard, C. van den, Bogaard, P. van den & Schmincke, H.- U. 1989. Quartargeologisch-tephrostratigraphische Neu- aufnahme und Interpretation des Pleistozanprofils Kar- lich, F.R.G. - Eiszeitalter und Gegenwart 3 9 62-86.
Breda. M. 2001. The holotype of Cervalces gallicus (Azzaroli, 1952) from SenCze (Haute-Loire, France) with nomencla- tural implications and taxonomical-phylogenetic accounts. - Rivista Italiana di Paleontologia e Stratigrafia 107(3): 439-449.
Brookes, J. 1828. A Catalogue of the Anatomical and Zoolo- gical Museum Joseph Brookes, Esp., part 1 61.
Cronin, M. A., Stuart, R., Pierson, B. J. & Patton, J. C. 1996. K- casein gene phylogeny of higher ruminants (Pecora, Artio- dactyla). - Molecular Phylogenetic Evolution 6: 295-311.
Dawkins, W. B. 1868. On a new species of fossil Deer from Clacton. - Quarterly Journal of the Geological Society of London 24: 511-516.
Douzery, E. & Randi, E. 1997. The Mitochondria1 Control Region of Cervidae: Evolutionary Patterns and Phylo- genetic Content. - Society for Molecular Biology and Evolution 14 (11): 1154-1166.
Driesch, von den, A. 1976. Das Vermessen von Tierknochen aus vor- und friihgeschichtlichen Siedlungen. - Inst. Palaoanatomie, Domestikationsforschung und Geschichte der Tiermedizin Universitat Munchen: 1-114.
Felsenstein, J. 1985. Confidence limits on phylogenies: an ap- proach using the bootstrap. - Evolution 39: 783-791.
Freudenberg, W. 1914. Die Saugetiere des alteren Quartars von Mitteleuropa. - Geologische und Palaontologische Abhandlungen 12 (4/5): 455-670.
Geist, V. 1971. The relation of Social Evolution and Disper- sal in ungulates during the Pleistocene, with Emphasis on the Old World Deer and the genus Bison. - Quaternary Research l (3) : 283-315.
- 1987. On speciation in Ice Age mammals, with special reference to cervids and caprids. - Canadian Journal of
23-32.
Zoology 65: 1067-1084. GoldfuR, G. A. 1820. Handbuch der Zoologie 2: 20: 374 Gould, S. J. 1974. The origin and function of bizarre structu-
res: antler size and skull size in the ‘Irish elk’, Mega- loceros giganteus. - Evolution 28 191-220.
Groves, C. P. & Grubb, P. 1987. Relationships of living deer. In Wemmer, C. M. (ed.). Biology and management of the Cervidae, pp. 21-59, Smithsonian Institution Press, Wa- shington. D.C.
308 Pfeiffer. T., Comdete skeleton of Merraloceros ~
Griiger, E., Jordan, H., Meischner, D. & Schlie, P 1994. Mit- telpleistozane Warmzeiten in Gottingen, Bohrungen Otto- straBe und Akazienweg. - Geologisches Jahrbuch A 134:
Habermehl, K. H. 1961. Die Altersbestimmung bei Haus- tieren, Pelztieren und beim jagdbaren Wild. - Parey, Hamburg und Berlin.
Harmer, F. 1889. On a specimen of Cervus belgrandi (Lartet) (C. verlicornis Dawk.) from the Forest-Bed of East An- glia. -Transactions of the Zoological Society 15: London.
Heintz, E. & Poplin, F. 1981. Alces carnutorum (Laugel, 1862) du PICistockne de Saint-Prest (France). SystCmati- que et Cvolution des AIcinCs (Cervidae, Mammalia). - Quartarpalaontologie 4 105-122.
Joleaud, L. 1914. Cewus (Megaceroides) algericus Leydekker, 1890. - Recueil des Notices et MCmoires de la SociCtC ArchCologique du DCp. De Constantine 5 (5): 1-67.
Kahlke, H.-D. 1956. Die Cervidenreste aus den altpleisto- zanen Ilmkiesen von ShDenborn bei Weimar. - Bd. 1: 62 S., Akademie-Verlag, Berlin.
- 1960. Die Cervidenreste aus den altpleistozanen Sanden von Mosbach (Biebrich-Wiesbaden). - Abhandlungen der Deutschen Akademie der Wissenschaften, Klasse fur Chemie, Geologie und Biologie 7: 1-75.
- 1965. Die Cervidenreste aus den Tonen von Voigtstedt. - Palaontologische Abhandlungen A 2 (2/3): 379-425.
- 1969. Die Cerviden-Reste aus den Kiesen von SiiBenborn bei Weimar. - Palaontologische Abhandlungen A 3 (3/4):
- 2001. Neufunde von Cerviden-Resten aus dem Unterpleis- tozan von UntermaBfeld. - Monographien des Romisch- Germanischen Zentralmuseums Mainz 40 (2): 461 -482.
Kahlke, R.-D. 1994. Die Entstehungs-, Entwicklungs- und Verbreitungsgeschichte des oberpleistozanen Mummuthus - Coelodonta - Faunenkomplexes in Eurasien (GroB- sauger). - Abhandlungen der Senckenbergischen natur- forschenden Gesellschaft 546: 1-164.
Kishino, H. & Hasegawa, M. 1989. Evaluation of the maxi- mum likelihood estimate of the evolutionary tree topolo- gies from DNA sequence data, and the branching order in Hominoides. - Journal of Molecular Evolution 29:
Koenigswald, W. v. & Heinrich, W.-D. 1999. Mittelpleistozane Saugetierfaunen aus Mitteleuropa - der Versuch einer biostratigraphischen Zuordnung.- Kaupia 9 53-112.
Laugel, A. 1862. La faune de Saint-Prest, pr&s Chartres (Eure-et-Loir). - Bulletin SocietC GColique de France 2. Ser. 19 (1861-1862): 709-718.
Lengsfeld, K.-P. 1975. Uber den formenden EinfluB des Cer- videngeweihs auf Hinterhaupt und erste Halswirbel. - Dissertation Institut fiir Haustierkunde, Universitat Kiel.
Lister, A. M. 1984. Evolutionary and ecological origins of British deer. - Proceedings of the Royal Society of Edin- burgh 82B: 205-229. 1987. Megaloceros Brookes, 1828 (Mammalia, Artio- dactyla): proposed emendation of the original spelling. - Bulletin of Zoological Nomenclature 44 (4): 255-256. 1993. The stratigraphical significance of deer species in the Cromer Forest-bed Formation. - Journal of Quater- nary Science 8: 95-108. 1994. The evolution of giant deer, Megaloceros giganteus (Blurnenbach). - Zoological Journal of the Linnean Society 112: 65-100. 1996. The stratigraphical interpretation of large mammal remains from the Cromer Forest-bed Formation. In
167-210.
547-610.
170-179.
Meischner, D. & Schneider, J. 1967. 500000 Jahre alter Step- penhirsch gefunden. - Deutsche Jagerzeitung 2 2
Moullade, E. 1886. Note sur une nouvelle espkce de daim fossile. - MCm. SOC. Agr. Sci. H.te Loire 4: 304-306.
Muller, H. 1965. Die pollenanalytische Neubearbeitung des Interglazial-Profils von Bilshausen (Unter-Eichsfeld). - Geologisches Jahrbuch 83: 327-352.
Newton, E. T. 1882. The Vertebrata of the Forest bed Series of Norfolk and Suffolk. - Memoirs of the Geological Survey of English. Wales. HMSO, London.
Owen, R. 1842. Report on British Fossil Mamma1ia.- Re- ports of the British Association, p. 237.
Pfeiffer, T. 1999a. Die Stellung von Duma (Cervidae, Mam- malia) im System plesiometacarpaler Hirsche des Pleisto- zans - Phylogenetische Rekonstruktion - Metrische Analyse. - Courier Forschungsinstitut Senckenberg 211
- 1999b. Alces latifrons (Johnson, 1874) (Cervidae, Mamma- lia) aus den jungpleistozanen Kiesen der Oberrheinebene. - Neues Jahrbuch f i r Palaontologie, Abhandlungen 211
- 1999c. The Morphological Distinction of Limb Bones of Alces latifrons (Johnson, 1874) and Megaloceros giganteus (Blumenbach, 1799). - Kaupia, Darmstadter Beitrage zur Naturgeschichte 9: 113-126.
- in press. The position of Duma (Cervidae, Mammalia) in the system of fossil and living deer from Europe - Phylo- genetic analysis based on skeletal characters. - Quater- naire.
Pomel, A. 1892. Sur deux ruminants de l’epoque n”1ithique en Algerie: Cervus pachygenis et Antilope maupasi. - C.R. Acad. Sc. Paris, CXV 213.
Portis, A. 1920. Elenco delle specie di cervicorni fossili in Roma e attorno a Roma. - Bolletino della Societh Geo- logica Italiana 39: 132-139.
Randi, E., Mucci, N., Pierpaoli, M. & Douzery, E. 1998. New phylogenetic perspectives on the Cervidae (Artiodactyla) are provided by the mitochondria1 cytochrome b gene. - Proceedings of the Royal Society of London B 265:
Rieck, W. 1983. Damwildalter-Merkblatt. In Schalenwild- ausschul3 des Deutschen Jagdschutz-Verbandes e.V. (ed.), 3. Aufl. S. 1-7, Hoffmann, Mainz.
Schmidt, H. 1930. Ein Skelett von Alces latifrons aus der Ge- gend von Gottingen. - Palaontologische Zeitschrift 12: 135.
- 1934. Ein Skelett vom Riesenelch. - Forschungen und Fortschritte 10 (15): 198-199.
Scott, K. M. & Janis, C. M. 1993. Relationships of the Rumi- nantia (Artiodactyla) and an analysis of the characters used in ruminant taxonomy. In Szalay, F. S., Novacek, M. J. & Mc Kenna. M. C. (eds). Mammal Phylogeny: Pla- centals, pp. 282-302, Springer, New York.
Soergel, W. 1927. Cervus megaceros mosbachensis n. sp. und die Stammesgeschichte der Riesenhirsche. - Abhandlun- gen der Senckenbergischen Naturforschenden Gesell- schaft 39: 375-505.
Strimmer, K. & von Haeseler, A. 1996. Quartet puzzling: a quartet maximum-likelihood method for reconstructiing tree topologies.- Molecular Biology and Evolution 13:
Swofford, D. L. 1998. “PAUP*. Phylogenetic Analysis IJsing Parsimony (*and other methods).” Version 4. Sinauer As- sociates. Sunderland. MA.
870-872.
1-218.
(3): 291-327.
793 - 801.
964- 969.
lhmer, C. (ed.). The early Middle Pleistocene in Europe, pp. 25-44, Balkema, Rotterdam.
Lydekker, R. 1890. On a Cervine Jaw from Algeria. - Pro- ceedings of the Zoological Society of London for the year 1890. 602-604.
Meischner, D. 1995. Black Shales Models. - I11 EPA Work- shop, Dotternhausen.
Thenius, E.’ 1958. Geweihiugendstadien des eiszeitlichen Rie- senhirsches, Megaceros giganteus (Slum.) und ihre phylo- genetische Bedeutung. - Acta Zoologica Cracoviensia 2 (30): 707-721.
Zdansky, 0. 1925. Fossile Hirsche Chinas. - Palaeontologia Sinica C 2 (3): 1-94.