HUNTERIA - Gregory S. Paul
HUNTERIA - Gregory S. Paul
HUNTERIA - Gregory S. Paul
Create successful ePaper yourself
Turn your PDF publications into a flip-book with our unique Google optimized e-Paper software.
\ \<br />
.,..,<br />
\<br />
vol. 2<br />
llo.3<br />
<strong>HUNTERIA</strong><br />
Jocit:ta&<br />
~<br />
(J~<br />
ISSN No. 0892-3701, University of Colorado<br />
Museum, Campus Box 315, Boulder, CO 80309-0315<br />
THE BRACHIOSAUR GIANTS OF<br />
THE MORRISON AND TENDAGURU WITH A DESCRIPTION<br />
OF A NEW SUBGENUS, GIRAFFATITAN, AND<br />
A COMPARISON OF THE WORLD'S LARGEST DINOSAURS<br />
Proceedings of the North American Paleontological Conference IV:<br />
The Golden Age of Dinosaurs - The Mid-Mesozoic Terrestrial Ecosystem of North America<br />
Field Trip and Colloquium.<br />
Host institutions: Museum of Western Colorado, Grand Junction, Colorado<br />
and University Museum, University of Colorado, Boulder.<br />
GREGORY S. PAUL<br />
3109 N. Calvert St. Side Apt.<br />
Baltimore, MD 21218<br />
ABSTRACT<br />
A new skeletal restoration of Brachiosaurus brancai shows that this gracile, giraffe-like taxon is a<br />
distinct subgenus from Brachiosaurus altithorax. Ultrasaurus macintoshi is a junior synonym of<br />
B. altithorax and is similar in size to the largest B. brancai specimen. A survey of exceptionally large<br />
sauropod remains indicates that the largest weighed about 50 tons in lean condition, but this size was<br />
probably not the ultimate limit of the group. HUltrasaurus" was not larger than the largest African<br />
brachiosaurs and published estimates of a body weight up to 190 tons are unwarranted exaggerations.<br />
INTRODUCTION<br />
Brachiosaurs are not only the largest of the Morrison<br />
dinosaurs, they are the largest terrestrial vertebrates of all time<br />
for which good remains are known. The first Morrison<br />
brachiosaur, Brachiosaurus altithorax, was discovered by<br />
Riggs (1901, 1903, 1904) in 1900. The east African Tendaguru<br />
quarries that contained Brachiosaurus brancai were<br />
discovered in 1907(Fraas, 1908;Janensch, 1914).Over the years·<br />
there has been a tendency to consider the North American<br />
--..----~~
GREGORY S. PAUL<br />
in lean condition. The temporal and biogeographic implications<br />
these taxa have for their respective formations is<br />
considered.<br />
Museum abbreviations - AMNH American Museum of<br />
Natural History, New York; BMNH British Museum of<br />
Natural History, London; BYU Brigham Young University,<br />
Provo; CM Carnegie Museum of Natural History, Chicago;<br />
HMN Humboldt Museum fur Naturkunde, East Berlin; MLP<br />
Museo La Plata, La Plata; MNDN Museum National<br />
D'histoire Naturelle, Paris; PEM Port Elizabeth Museum, Port<br />
Elizabeth; YMP Yale Peabody Museum, New Haven.<br />
BRACHIOSAURUS BRANCAI<br />
Brachiosaurus brancai ]anensch 1914is by far the best<br />
known of the brachiosaurs, so a consideration of the group<br />
starts with this species. Although no one complete individual<br />
exists, multiple specimens make knowledge of its osteology<br />
almost complete. A full understanding of this species has been<br />
hindered by the absence of an accurate, detailed skeletal<br />
restoration. The Berlin mount, HMN SIl, is a composite made<br />
of different sized individuals, plus some bones modeled in<br />
plaster. Janensch's restoration of SITis schematic and includes<br />
postural and proportional errors. The methodologies for restoring<br />
sauropods and other dinosaurs are discussed in <strong>Paul</strong> (1987)<br />
and <strong>Paul</strong> and Chase (in press).<br />
The basis for the new skeletal restoration in Fig. 1 is the<br />
holotype and best specimen, and the one on display in Berlin,<br />
HMN SII. This individual includes a partial skull (S116, in<br />
Janensch, 1935-36),all but the first three neck vertebrae, dorsal<br />
vertebrae 1-4, 8?, lO-12and parts of others, most of the dorsal<br />
ribs, a sternal, scapular material, a coracoid, a complete<br />
forelimb and hand, the pubes, a partial femur, a fibula, and<br />
hindfoot bones. Another important specimen. is BMNH M<br />
23, which includes a complete though poorly figured dorsal<br />
column (Migeod, 1931, most of this specimen has since been<br />
destroyed; McIntosh pers. comm.). The HMN specimens<br />
figured by Janensch (1950a, 1961)provide virtually all the rest<br />
of the missing elements, including a hip, sacrum, and tail<br />
(HMN Aa). Only a few hindfoot bones' are absent. In addition<br />
to the excellent figures of the various elements provided<br />
by Janensch, the author used photographs taken of the HMN<br />
mount and material - extreme care was taken in executing<br />
the profile of each element in the skeletal restoration. Duplication<br />
of parts in the HMN sample of elements allow different<br />
sized individuals to be scaled to the size of HMN SIT<br />
(Table 1); consequently the confidence in the morphology and<br />
proportions of the restoration of HMN Sll is high.<br />
The skull is the complete HMN tl, modified slightly and<br />
scaled up to S116/SIT. The ball and socket head-neck articulation<br />
was highly mobile, but since the occipital condyle points<br />
down and backwards, the head was usually carried at a sharp<br />
angle to the neck. The neck articulates in a gentle S curve,<br />
a basic dinosaurian adaptation. Ball and socket central facets<br />
and very large zygapophyses that remained articulated under<br />
~ a wide range of motion show that the neck was very flexible.<br />
The dorsals are wedge shaped and form a gentle arch, another<br />
<strong>HUNTERIA</strong> VOL. II, no. 3, pp. 2, February 19, [988<br />
general dinosaur character. Janensch did not have a complete<br />
dorsal column and did not explain why he gave B. brancai<br />
11 dorsals, but BMNH M 23 indicates that there were 12<br />
dorsals (Migeod, 1931,and McIntosh, pers. cornrn.; dorsal 1<br />
is considered to be the first relatively short-centrum vertebra<br />
that supports a long dorsal rib). Although the dorsal centra<br />
also have ball and socket central articulations, the following<br />
features show that the posterior dorsal region was stiff: partly<br />
ossified interspinal ligaments; expanded neural spine heads that<br />
supported enlarged interspinal ligaments; auxiliary zygapophyses<br />
that interlock transversely; and a tendency for the<br />
posterior dorsals to coossify.The ball and socket centra of the<br />
posterior dorsals added strength to the rigid column. In a<br />
similar manner, birds have stiff backs in association with<br />
saddle-shaped central articulations. The anterior dorsals may<br />
have been moderately mobile in the vertical plane (Bakker,<br />
pers. comm.). The tail-base vertebrae are also wedge shaped<br />
and form a gentle upwards arch (Gilmore, 1932, 1937;<br />
]anensch, 1950a); moderate sized. zygapophyses show a fair<br />
degree of flexibility. A multitude of sauropod trackways show<br />
that they rarely if ever dragged their tails on the ground (Bird,<br />
1985; Dutuit and Oazzou, 1980; Langston, 1974; Lockley et<br />
al., 1986).<br />
Articulated sauropod skeletons show that the anterior ribs<br />
are swept backwards relative to the main axis of the body.<br />
Proceeding backwards the ribs become more perpendicular<br />
relative to the main axis, so the ribs bunch together towards<br />
their ends. The anterior ribs are straight shafted and vertical<br />
in front view of the body, forming a narrow, slab-sided chest<br />
for articulation with the shoulder girdle. The posterior ribs<br />
are curved and arch far out to the side, creating a cavernous<br />
abdominal cavity.<br />
B. brancai has enormous sternals. With their postero-lateral<br />
corners attaching to the first long, robust dorsal rib (number<br />
two) via a short sternal rib, the sternals help determine the<br />
breadth of the chest. A cartilaginous anterior sternum is<br />
restored in front of the paired sternals as per Norman (1980).<br />
The anterior sternum has grooved edges along which the coracoid<br />
glided back and forth (Bakker, 1975;<strong>Paul</strong>, 1987).Observe<br />
that the scapulocoracoid does not perform anato~ical violations<br />
as it rotates, contrary to Bennett and Dalzell (1973).Also<br />
note that the backswept ribs set the sternum and coracoids<br />
a bit behind a perpendicular to the cervico-dorsal transition,<br />
not forward of it. The deep chest and short coracoid are compatible<br />
with an upright scapula seen in most tetrapods, instead<br />
of a bird-like horizontal one as often shown. This is the normal<br />
tetrapod condition - only protobirds and birds with their<br />
extremely long coracoids and scapulae have horizontal<br />
scapulae.<br />
The detailed limb joint articulations of B. brancai are<br />
discussed elsewhere (<strong>Paul</strong>, 1987). Forelimb posture is erect and<br />
the forefoot gauge is narrow as shown by morphology and<br />
trackways (Bakker, 1971a,b,c, 1974, 1975; Bird, 1985; Dutuit<br />
and Ouazzou, 1980; Langston, 1974; <strong>Paul</strong>; 1987). A downwardly<br />
facing shoulder glenoid and distally restricted humeral<br />
condyles show that the forelimb was also columnar. [anensch<br />
(1961)correctly restored the long, unguligrade, circular arcade<br />
.1<br />
'\<br />
I
'><br />
;;!<br />
OJ<br />
rm<br />
Lengths mm<br />
Skull t<br />
13 Cervicals c<br />
Longest Cervical<br />
12 Dorsals c <<br />
Dorsals 6-12 c <<br />
5 Sacrals<br />
55 Caudals c <<br />
Longest Rib<br />
Scapula<br />
Scapula-coracoid<br />
Sternal<br />
Humerus<br />
Radius<br />
Metacarpal II<br />
Ilium<br />
Pubis<br />
Ischium<br />
Femur<br />
Tibia-asrragulus<br />
Fibula<br />
Metatarsal n<br />
Toe Claw I<br />
Height mm<br />
Tallest Dorsal<br />
Dorsal 12<br />
Circumference mm<br />
'" .~<br />
~<br />
,g't;<br />
()"<br />
'-:'E 1'0..<br />
HMN<br />
XV2<br />
1300e<br />
Sll Y Ki24<br />
880<br />
8996p<br />
1155<br />
3781p<br />
1711p<br />
1150<br />
7404e<br />
2900e 2580<br />
2175e 192ge<br />
2660e 2360e<br />
1100<br />
2400e 2130<br />
1240<br />
635<br />
2350e<br />
1340<br />
1360e<br />
1150<br />
1195e<br />
2090e<br />
1150e<br />
1190<br />
276<br />
240e<br />
1320e 1170"<br />
790<br />
Femur 730<br />
Breadth mm<br />
Femoral Head<br />
Hir.dfooe Pad<br />
Meters<br />
Total Length'<br />
Total Height<br />
Rearing Height<br />
Shoulder Height<br />
Hip Height<br />
- 850e - 750p<br />
25e<br />
16e<br />
1ge<br />
6.8e<br />
5.4e<br />
22.Zp<br />
14.0<br />
17.0<br />
6.0<br />
4.8p<br />
1540 - 890<br />
-1090<br />
D<br />
1700 1600<br />
XVI<br />
2140 1550<br />
900<br />
940<br />
560<br />
Sa9 Aa<br />
1930<br />
J<br />
1050<br />
890<br />
925<br />
1610<br />
950<br />
960<br />
150<br />
130<br />
6530<br />
1190<br />
1 il<br />
6<br />
;.t:k<br />
c e<br />
" >-<br />
•..0<br />
E'3:S<br />
rri~<br />
FMNH BYU<br />
P25107<br />
4760p<br />
2475<br />
950<br />
From :<strong>Paul</strong> 1988a<br />
10'"<br />
;q E~<br />
1'0"<br />
~
of the hand. Contrary to previous reports, the sauropod hand<br />
is not at all elephant-like - in fact it is unique and has only<br />
a thumb claw and no hooves. As restored the femur/tibia<br />
or fibula ratio of SITis a little higher than in other specimens<br />
(Table 1), but as the femur is always just a little shorter than<br />
the humerus in brachiosaurs, this ratio must be correct. The<br />
hindlimb is also columnar, and the foot is unguligrade and<br />
very elephant-like (excepting the three or four big laterally<br />
splayed banana claws; the outer one or two toes are neither<br />
clawed nor hooved). The astragular distal articulation faces<br />
more downwards and less forwards in well preserved sauropod<br />
ankles, relative to those of digitigrade prosauropods; therefore<br />
the sauropod foot is posturally unguligrade instead of plantigrade<br />
as suggested by Cooper (1984). The extremely short<br />
metatarsus and toes backed by a big pad show that the ankle<br />
was nearly immobile. The animal is shown in an elephantlike<br />
amble (Muybridge, 1887), its fastest gait.<br />
The most important point about the mounted HMN Sll<br />
is that the presacrals on display are not the originals, they<br />
are plaster models. The centra of the dorsal models are<br />
significantly larger than those of the originals, and Janensch's<br />
(1950b)paper skeletal restoration includes the s.ameerror. Why<br />
the modeled centra are so long is not clear, for although the<br />
dorsals are moderately crushed and too fragile to mount, most<br />
of the centra appear to be little altered in length. With centra<br />
of proper length, the dorsal column of B. brancai is some<br />
20% shorter relative to the limbs than indicated by ]anensch,<br />
even though one more vertebra is included in the new restoration.<br />
Other errors in Janensch's restoration include vertical<br />
anterior dorsal ribs and a shoulder girdle that is consequently<br />
too far forward - as well as too high - on the ribcage; scapula<br />
and humerus too short, a sprawling forelimb, and a tail that<br />
is too long, too heavy, and droops. Burian's well known<br />
restoration of the species emphasizes these errors, and also<br />
shows the neck much too short (Spinar and Burian, 1972)~<br />
In addition the claws are incorrect. Because the new restoration<br />
has a scaled up HMN Aa tail that is longer than it is<br />
in the mount and a dorsal column that is shorter, the<br />
differences cancel each other and the new restoration and the<br />
mount share nearly the same length of over 22 m.<br />
A very unusual feature ofB. brancai is the extreme height<br />
of dorsal vertebra 4, especially the neural spine, relative to<br />
both the cervicals and posterior dorsals. Unfortunately the<br />
immediately surrounding neural spines are not preserved, but<br />
it appears that this sauropod had "withers", tall neural spines<br />
over the shoulders (Figs. 1, 2B). Rebbachisaurus garasbae,<br />
a species possibly assignable to the Brachiosauridae, may have<br />
even taller withers (Lavocat, 1952).Withers are fairly common<br />
among mammals, but are unknown among other dinosaurs<br />
except for the chasmosaurian ceratopsids. This feature suggests<br />
that nuchal ligaments helped to support the neck. The withers'<br />
modest height and the long neck suggest the ligaments were<br />
rather low, like a camel's (Knight, 1947; Dimery et al., 1985).<br />
The ossified cervical "ligaments" cited by Migeod (1931) and<br />
Alexander (1985)are more probably displaced ends of the long<br />
cervical ribs (McIntosh, pers. comm.). Also unusual is the small<br />
size of the posterior dorsals, especially the centra. HMN Sll<br />
<strong>HUNTERIA</strong> VOL. IT, no. 3, pp. 4, February 19, 1988<br />
GREGORY S. PAUL<br />
and BMNH M 26 have posterior centra that are only 9 inches<br />
long! Although brachiosaur posterior dorsals are very different<br />
from the great posterior dorsals and sacrals of diplodocids, it<br />
does not follow that brachiosaurs were weak in the back.<br />
Obviously these giants did perfectly well with the posterior<br />
dorsals they had. Brachiosaur dorsals were not as specialized<br />
for rearing up as were diplodocid dorsals (Bakker, 1971C,1978).<br />
On the other hand brachiosaurs were like all other dinosaurs<br />
in being hindlimb dominant - the center of gravity was<br />
towards the rear so the hindlimb was more robust and supported<br />
more weight than the forelimb. This weight distribution<br />
made it easier for brachiosaurs to rear in search of choice<br />
-I
,<br />
food items or to fight than it is for elephants, which, despite<br />
their forelimb dominance, also have the capacity to rear<br />
(Eltringham, 1982).Note that brachiosaurs are less hindlimb<br />
dominant than diplodocids (Anderson et al., 1985).<br />
BRACHlOSAUR GIANTS<br />
Corrected with shorter trunk, taller forelimbs, and withers<br />
B. brancai is even more giraffe-likethan previously realized.<br />
It is the only quadrupedal dinosaur which one would have<br />
to reach up to slap the belly as one walked under it! Most<br />
Figure l-Multiview skeletal restoration of Brachiosaurus (Giraffatitan) brancai holotype SII/S116 (skull), with some proportions and elements from<br />
other HMN specimens., Inset shows dorsal-sacral column and ilium ofB. (Brachiosaurus) altithorax holotype FMNH P25107 to same scale. Measurements<br />
in Table 1. '<br />
HUNTERlA VOL. II, no. 3, pp. 5, February 19, 1988
.",..---------~------------<br />
unusual for a tetrapod, much less a dinosaur, it is an<br />
exceptionally elegant and majestic design.<br />
A COMPARISON OF B. BRANCAI<br />
WITH B. ALTITHORAX<br />
AND THE DELTA GIANT<br />
The holorvpe of Brachiosaurus altithorax Riggs 1903<br />
FMNH, P25107 is the most complete North American<br />
brachiosaur specimen. It includes dorsals 6-12, ribs, sacrum,<br />
caudals 1 and 2, coracoid, humerus, ilium, and femur<br />
(Figs. 2A, 3F). This specimen is crushed to varying degrees,<br />
especiallyin the sacrum (dorsa-ventrally) and, to a lesserdegree<br />
in the dorsals (downwards to the right).<br />
,<br />
: If<br />
'- -- - " ,<br />
..•• -:...--- .•.<br />
A<br />
I,<br />
...J'~--"'J.-'"___ I<br />
•... ~/, .<br />
B<br />
GREGORY S. PAUL<br />
The holotype of the Delta or Uncompahgre giant<br />
Ultrasaurus macintoshi Jensen 1985is the crushed vertebra<br />
BYU 5000.(note that the specimen numbers applied to the<br />
Uncompahgre material by Jensen are sometimes contradictory,<br />
and hence those cited here are tentative). This vertebra has<br />
been identified as a posterior dorsal (Fig. 3E). However, its<br />
transversely narrow neural spine with a small head shows that<br />
it is an anterior dorsal. Clearly brachlosaurian in design, it<br />
is very similar to the anterior dorsals of B. altithorax. As<br />
far as can be told, BYU 5000 belongs to B. altithorax and<br />
is referred to it. Hence, U. macintoshi is not considered valid.<br />
The slenderness of the neural spine suggests that BYU 5000<br />
is forward of dorsal 6, so it bolsters our knowledge of the<br />
shoulder of B. altithorax. Also referable to B. altithorax<br />
is the extremely large (2690 mm long) Uncompahgre<br />
Figure 2~Dorsals 6-12 and sacraIs, longest rib and ilia of A) Brachiosaurus (Brachiosaurus) altithorax holotype FMNH P25107 and<br />
B) B. (Giraffatitan) brancai halo type HMN SIl and Aa, Drawn so that their respective humeri are to the same length, scale bars equal 1 m. Neither<br />
column is complete, the sacrals and ilia of HMN Aa are scaled up to SIl, the position of the longest dorsal rib is not certain in either, and some crushing<br />
is removed tram both examples.<br />
~HUNfERlA VOL. II, no. 3, pp. 6, February 19, 1988
scapulocoracoid (Jensen 1985a, BYU 5001, Fig 4A, popularly<br />
labeled "Ultrasaurus"). It has the short, rounded amerior<br />
scapular process, short rounded coracoid, narrow scapular<br />
neck, and broad, rounded blade fully compatible with a<br />
brachiosaur. Jensen suggests that the scapula blade was not<br />
as broad as in Brachiosaurus, but some B. brancai blades<br />
are similar in breadth.<br />
The caudals, scapula, coracoid, humerus, ilium, and femur<br />
ofB. altithorax and B. brancai are very similar, and though<br />
the limb bones of the former are a bit more robust these<br />
elements could belong to the same species. It is in the dorsal<br />
column and trunk that the significant differences occur. To<br />
start with, at any given point, the dorsal column of<br />
B. altithorax is about 25-30% longer relative to the humerus<br />
or femur than that ofB. brancai (Fig. 2, Table 1).Riggs (1903)<br />
commented on the unusually long ribs of the Morrison<br />
sauropod, and indeed the longest dorsal rib is some 10% longer<br />
relative to the humerus than in B. brancai.<br />
The dorsals of both species are crushed, hampering comparison,<br />
especially quantitative. The posterior dorsals appear<br />
to be fairly similar. However, all the dorsal centra of<br />
B. altithorax have pleurocoels that are about 50% larger than<br />
those ofB. brancai (Fig. 3E-G). The neural arches are taller<br />
A<br />
BRACHIOSAUR GIANTS<br />
and longer in B. altithorax, but are much narrower. The<br />
transverse processes form a shallow V in B. brancai; in<br />
B. altithorax they appear to be flatter.<br />
The anterior and mid dorsals differ the most. In B. brancai,<br />
dorsal 4 is very gracile, with very long, proximally deep, distally<br />
tapering transverse processes and a tall, slender neural spine.<br />
The centrum is rather small and short. Excepting the centrum,<br />
dorsal 4 differsgreatly from the posterior dorsals in being much<br />
taller and wider. In the upper portions, the anterior dorsals<br />
ofB. altithorax differ relatively little from the more posterior<br />
vertebrae - their neural spines and transverse processes are<br />
only a little longer. Compared to the posterior arches, the<br />
anterior arches are longer both vertically and fore and aft.<br />
Very notable is the length of the anterior dorsal centra<br />
relative to those of the posterior dorsals in B. altithorax.<br />
Incomplete columns, crushing, centra fusion, and incomplete<br />
measurements hamper quantitative comparisons. But in<br />
HMN SII and BMNH M 26, the anterior dorsals are about<br />
the same length as the posterior dorsals. In FMNH P25107<br />
the mid dorsal centra are about 50% longer than those of the<br />
posterior dorsals (Fig. 2). Such elongation of mid-anterior<br />
dorsals is most unusual for sauropods and dinosaurs - even<br />
in Diplodocus and Barosaurus only the first two dorsals<br />
F G<br />
Figure 3-Anterior dorsals in left lateral (top) and posterior (bottom) views, drawn to the same zygapophysis-to-bottom-of-centrum height. Dorsal 5 in<br />
A) Apatosaurus louisae CM 3018; B) Diplodocus carnegii CM 84; C) Camarasaurus grandis YPM 1901 or 1902; and D) Camarasaurus suprernus<br />
AMNH 5761. Anterior dorsal in E) B. (Brachiosaurus) altithorax referred BYU 5000; dorsal 6 in F) B. (B.) altithorax holotype FMNH P25107; and<br />
dorsal 4 in G) Brachiosaurus (Giraffatitan) brancai holotype HMN SI!. Data from Gilmore, 1937; Hatcher, 1901;Janensch, 1950a; Jensen, 1985a: Osborn<br />
and Mook, 1921; Ostrom and Mclntosh, 1966: Riggs, 1903.<br />
<strong>HUNTERIA</strong> VOL. II, no. 3, pp. 7, February 19, 1988
are elongated. The combination of lower anterior dorsal<br />
vertebrae and the apparent lack of a sharp change between<br />
the anterior and posterior dorsals shows that B. altithorax<br />
lacked the tall withers of B. brancai. If this is correct, then<br />
B. altithorax may have had a less developed, shorter neck<br />
than B. brancai.<br />
It appears that the two brachiosaurs are both derived relative<br />
to other sauropods in anterior dorsal design, but in different<br />
ways. B. brancai has tall withers but is normal in anterior<br />
dorsal centra length. B. altithorax is more normal in the<br />
shoulder spines, but has unusually long anterior dorsals. The<br />
latter is also longer, deeper, and much bulkier in the body<br />
relative to the limbs (see below).<br />
]anensch (1914)recognized the differencesbetween the North<br />
American and African forms and separated them at the species<br />
level. The next question is how do the differences between<br />
B. altithorax and B. brancai compare to those between<br />
other tetrapod genera and species. Simple proportional differences<br />
do not necessarily a genus make. A combination of<br />
proportional and morphological differences is more significant.<br />
In the context of other sauropods, the differences between<br />
these two brachiosaurs are more extreme than those found<br />
within other genera. In Camarasaurus supremus and<br />
C. grandis, dorsal 5 differs significantly, especially in the<br />
. neural arch (Fig.3C,D). The difference is great enough to question<br />
whether these species should be placed in the same genus,<br />
but- the differences appear to be less than between B.<br />
altithorax and B. brancai. Even Diplodocus and<br />
Apatosaurus show about the same degree of difference in<br />
their anterior dorsals as shown in the two brachiosaurs (Fig.<br />
3A,B). There is much less difference between the anterior dorsals<br />
of Diplodocus and Barosaurus (Lull, 1919).Indeed the<br />
latter may be considered subgenera, rather than full genera.<br />
Among other dinosaurs it is difficult to come up with a<br />
genus that shows as much variation in anterior dorsal design.<br />
For example Tyrannosaurus, Tarbosaurus and<br />
Dospietosaurus - which may form a single genus - do not.<br />
Indeed, such widely accepted large tetrapod genera as<br />
Varanus, Anas, and Canis show much more uniformity.<br />
An example of considerable variation in withers height is<br />
found in recent and living species of Bison. However, the<br />
withers are an important species-specificdisplay device in bison;<br />
the functional importance of differences in neck movement<br />
is reduced in these short necked animals. In brachiosaurs the<br />
long neck implies that differences in the withers were probably<br />
associated with differences in neck size and movement.<br />
A basic definition of anyone genus should include an important<br />
functional distinction from related species. The Morrison<br />
and Tendaguru brachiosaurs not only appear to show a significant<br />
functional difference, they appear to be phylogenetically<br />
derived in different ways. Since they also vary more from one<br />
another than most genera, it is considered probable that they<br />
represent different genera. Generic separation would also be<br />
useful in preventing the two taxa from being considered more<br />
similar than they really are. However, the incompleteness of<br />
the remains of B. altithorax makes it difficult to prove full<br />
generic separation, as does the small sample size of Morrison<br />
HUNTERlA VOL. IT, no. 3, pp. 8, February 19, 1988<br />
GREGORY S. PAUL<br />
and Tendaguru dorsal columns. Therefore, only a separation<br />
at the subgeneric level is proposed below, with the option of<br />
raising the separation to the generic level left open for future<br />
developments.<br />
SYSTEMATIC PALEONTOLOGY<br />
Family Brachiosauridae Riggs 1903<br />
Genus Brachiosaurus Riggs 1903<br />
Synonym - Ultrasaurus Jensen 1985<br />
Type species - B. altithorax Riggs 1903<br />
Diagnosis. as per Riggs (1903, 1904)and ]anensch (1914, 1929,<br />
1935-36, 1950a,b)<br />
Subgenus Brachiosaurus (Brachiosaurus) (Riggs 1903)<br />
Diagnosis. as for species.<br />
Brachiosaurus (Brachiosaurus) altithorax Riggs 1903<br />
Synonym - Ultrasau1'Usmacintoshi<br />
Holotype - FMNH F25107<br />
Referred specimens - USNM 21903,BYU 5000, BYU 5001,<br />
BYU Potter Creek vertebra.<br />
Diagnosis. Robust overall; mid dorsal centra much longer than.<br />
posterior dorsal centra, anterior dorsal spines and transverse<br />
processes not much taller or wider than those of posterior<br />
dorsals; neural arches long, tall and narrow; transverse<br />
processes flat; dorsal centra pleurocoels large; dorsal column<br />
over twice humerus length and very long relative to vertebrae<br />
height; body massive relative to limbs.<br />
Discussion. So far most Morrison brachiosaurs, including the<br />
USNM humerus 21903 and the BYU Potter Creek posterior<br />
dorsal (Jensen, 1985a), appear to be referable to this species.<br />
An Uncompahgre anterior caudal BYU 5002 was incorrectly<br />
referred to U. macintoshi (Jensen, 1985a); its cleft neural<br />
spine and handle bar transverse processes are clearly those<br />
of a diplodocid, not a brachiosaur.<br />
If the Uncompahgre fauna come from the uppermost<br />
Morrison (Jensen, 1985b), then it is possible that the fauna's<br />
brachiosaurs represent a distinct population of oversized<br />
animals (see below). These could either be a temporal<br />
subspecies, which does not require formal recognition, or a<br />
separate species. The Morrison brachiosaur material is much<br />
too limited to prove any of these options.<br />
The name Ultrasaurus has appeared informally a number<br />
of times since 1978 (Jensen, 1978, 1985a; Lambert, 1983;<br />
McWhirter and McWhirter, 1986; Norman, 1985). However,<br />
lacking a designated type species and formal technical description,<br />
it does not meet the International Code of Zoological<br />
Nomenclature criteria for availability. Ultrasaurus tabriensis<br />
Kim (1983) was applied to a medium sized and nondiagnostic<br />
Korean sauropod humeral head that was mistaken for a<br />
gigantic proximal ulna. Despite these problems, a holotype<br />
(DGBU-1973) was effectively designated and described in a<br />
public technical text, sufficient to give the generic and specific<br />
names a place in formal systematics. Since the type and<br />
referred material (a caudal? neural spine) are nondiagnostic,
they should be considered Sauropoda incertae sedis, and<br />
Ultrasaurus should be considered no longer valid. This result<br />
is unfortunate because it effectively bars Jensen's (1985a)<br />
subsequent formal use of Vltrasaurus to designate any<br />
Uncompahgre remains.<br />
Subgenus Brachiosaurus (Giraffatitan) n. subgen.<br />
Etymology - Loosely "gigantic giraffe",in recognition of the<br />
taxon's giraffe-likeform.<br />
Diagnosis. as for species.<br />
Discussion. Arkell (1956) suggested that Brachiosaurus<br />
(Giraffati tan) was first named Gigantosaurus. Actually<br />
Gigantosaurus megalonyx Seeley 1869was applied to some<br />
ambiguous European sauropod remains. Gigantosaurus<br />
robustus Fraas 1908 was used to accommodate Tendaguru<br />
remains that he could not place in other Tendaguru taxa.<br />
Much of this material was later placed in Tornieria robusta<br />
Sternfeld 1911.<br />
Brachiosaurus (Giraffatitan) brancai Janensch 1914<br />
Synonym - B. fraasi Janensch 1914<br />
Holotype - HMN Sll and Sl<br />
Referred specimens - As per [anensch (1914,1929,1935-36,<br />
1950a,b, 1961)<br />
Diagnosis. Gracile overall; 25 presacrals and 12 dorsals; mid<br />
dorsal centra about same length as posterior dorsal centra,<br />
anterior dorsal spines and transverse processesmuch taller and<br />
wider than those of posterior dorsals and form shoulder<br />
withers; neural arches short vertically and fore and aft;<br />
transversely broad transverse processes that form a shallow<br />
V; dorsal centra pleurocoels moderate in size;dorsal column<br />
less than twice humerus length and short relative to vertebra<br />
height; body mass relatively modest relative to limbs.<br />
Discussion. All Tendaguru brachiosaur specimens are referred<br />
to this species, including B. fraasi, the type<br />
(HMN Y) of which Janensch (1961)in later years listed under<br />
B. brancai. However, the possibility that some of the<br />
Tendaguru remains belong to another taxon remains open.<br />
THE MORRISON AND TENDAGURU<br />
FORMATIONS AND<br />
THEIR BRACHIOSAURS<br />
That B. (B.) altithorax and B. (G.) brancai are probably<br />
less similar than thought somewhat reduces the similarities<br />
between the Morrison and Tendaguru Formation faunas<br />
(Arkell, 1956; Galton, 1977, 1980, 1982).This does not, by<br />
any means, discredit the probability of a land connection<br />
between North America and Africa. One possibility is that<br />
habitat differences between the two formations may be<br />
responsiblefor the differencesin the brachiosaurs, even though<br />
both formations seem to have been seasonally dry (Dodson<br />
et al., 1980;Russell, 1980).That the two brachiosaurs are about<br />
equally advanced is compatible with their respective formations<br />
being roughly equal in age - within the Upper<br />
BRACHIOSAUR GIANTS<br />
Kimeridgian-Tithonian (Aitken, 1961;Arkell, 1956; Stokes,<br />
1985). Comparable age is suggested too by the very close<br />
similarity of some other elements of the two faunas (Galton,<br />
1977, 1980, 1982). Besides, it remains possible that B. (B.)<br />
altithorax or something like it lies unrecognized in the<br />
Tendaguru material, and that B. (G.) brancai is present in<br />
the Morrison.<br />
WHAT WAS THE BIGGEST?<br />
Riggs realized he had discovered an exceptionally great<br />
dinosaur, and the actual mass of these animals has aroused<br />
curiosity ever since. In order to determine their mass, which<br />
is a function of their volume, we must first consider their<br />
musculature. A detailed restoration of the contour muscles<br />
of B. (G.) brancai HMN SII is presented elsewhere (<strong>Paul</strong>,<br />
1987).The muscles are also profiled in solid black around the<br />
skeletal restoration given here.<br />
Generally B. (G.) brancai has been restored as a massively<br />
muscled animal wtih a heavy neck and tail and stout limbs<br />
(Colbert, 1962;Jackson and Matternes, 1972;Norman, 1985;<br />
Ostrom, 1978;Spinar and Burian, 1972;Watson and Zallinger,<br />
1960; BMNH commercial model). Burian's restoration is<br />
perhaps the epitome of this style. It is certainly incorrect. The<br />
intensely pneumatic and very bird-like neck vertebrae of<br />
sauropods were much lighter in life than they look as mineralized<br />
fossils, and the skulls they supported were small. This<br />
suggeststhat the cervical musculature was also light and rather<br />
bird-like, just sufficient to properly operate the head-neck<br />
system. The bulge of each neck vertebra was probably visible<br />
in life, as is the case in large ground birds, camels.and giraffes.<br />
The rigid, bird-like ribcage was lightly muscled also.<br />
However, like all herbivores, sauropods must have had big<br />
bellies that they kept filled with fermenting fodder to support<br />
the flourishing gut flora "necessary to break down plant<br />
materials. A ruminant-like system is, however, neither efficient<br />
or necessary for high digestiveefficiencyin animals of this size<br />
(Demmont and Soest, 1985).In ungulates and proboscideans<br />
the belly is as broad as it is deep (seerear views in Muybridge<br />
1887),and the arching posterior ribs ofB. (G.) brancai show<br />
this was the case here. Indeed it had an exceptionally large<br />
abdomen for a sauropod; only Camarasaurus matches it in<br />
this regard. The tail of B. (G.) brancai is very reduced, and<br />
muscles were probably extended only as far as the bones (as<br />
mummies show is true of hadrosaurs, <strong>Paul</strong>, 1987),except for<br />
the caudofemoralis, which bulges beyond the transverse<br />
processes. All of the author's sauropod restorations show the<br />
animals "lean", without fat reserves, as they would be toward<br />
the end of the dry season. Large herbivores sometimes carry<br />
an additional 15-33%of their lean mass as fat by the end of<br />
the heavy feeding season, although elephants are not noted<br />
for carrying large fat deposits (Carrington, 1959).Hence the<br />
maximum masses of each individual sauropod would likely<br />
have been about a sixth higher than listed in Table 1, perhaps<br />
as much as a third. Such fat was probably borne towards the<br />
base of the belly, and Bakker (pers. comm.) notes that<br />
crocodilians and adult monitor lizards have substantial fat<br />
<strong>HUNTERIA</strong> VOL. II, no. 3, pp. 9, February 19, 1988
...----------------<br />
deposits on their tail. Sauropods may have had the same in<br />
season. On the other hand, being terrestrial animals that<br />
carried their tails aloft as weapons, it is possible that the tail<br />
was never so burdened. Because the restored tail makes up<br />
only 5% of the total mass in brachiosaurs, and just 15% even<br />
in big tailed Apatosaurus, even doubling the tail mass will<br />
not alter the total mass value much.<br />
The small size and shortness of such muscle leverage attachments<br />
as the deltoid crest, olecranon process, ilium, cnemial<br />
crest, and hypotarsus relative to other dinosaurs indicates a<br />
modest limb musculature in B. (G.) brancai. That the foot<br />
was immobile confirms that only light shank muscles were<br />
needed to operate it. The lightness of sauropod limb muscles<br />
is logical since sauropods were e1ephantinely slow, and modern<br />
elephants also have slender limb muscles (Knight, 1947;<br />
Muybridge, 1887).Since brachiosaurs were as hindlimb dominant<br />
as other sauropods, the forelimb musculature was more<br />
slender than that of the hindlimb. All sauropod trackways<br />
show that the hand lacked a large central palm pad; instead<br />
it was hollow aft and half moon shaped (<strong>Paul</strong>, 1987).The hindfoot<br />
differs in having a great elephant-like heel pad.<br />
A model of HMN Sil was sculpted in plasticine and its<br />
volume measured by water immersion. Since most live animals<br />
float, the specific gravity was assumed to be 0.9. The exception<br />
was the highly pneumatic neck which was measured<br />
separately and assigned a tentative specific gravity of 0.6. In<br />
this species the neck was 13% of the total volume. HMN Sil<br />
proved to weigh about 32 tons. This is far less than Colbert's<br />
(1962) estimate of 78 tons for the same specimen. Why. there<br />
is so much difference is hard to judge because no plans of the<br />
model Colbert used were published. Colbert's museum model<br />
was presumably based on an overly long-bodied version of<br />
HMN Sil. The neck is too heavily muscled; even more so is<br />
the tail which is restored heavier at the base than it is in the<br />
accompanying Apatosaurus model. The reverse is actually<br />
the case. Both pairs of brachiosaur limbs are muscled much<br />
too heavily, especially the forelimb which appears more robust<br />
than the hindlimb. In summary so much flesh could not be .<br />
borne by even SIl's giant skeleton. Bakker's (1975) estimate<br />
of 50 tons for brachiosaurs is also excessive for average sized<br />
individuals.<br />
On the other hand, Russell et al.'s (1980) estimate, based<br />
on vaguely defined limb bone circumference/mass estimate,<br />
is far too low at 15 tons - so little flesh simply cannot be<br />
stretched over the animal's great frame. In a more rigorous<br />
analysis of limb element circumference relative to mass in living<br />
and fossil tetrapods, Anderson et al. (1985) arrived at a much<br />
more reasonable 29 ton estimate. However, the apparent agreement<br />
between this and my estimate means less than it appears.<br />
Bone robustness can give only a "ball park" estimate because<br />
in modern animals mass varies up to a factor of two at any<br />
given limb circumference (Anderson et al., 1985).Comparable<br />
variation was certainly present in sauropods also. For example<br />
the limb bone circumference of Apatosaurus louisae<br />
CM 3018 predicts a mass of 35 tons, bigger than Sil. However,<br />
the latter's skeleton is definitely more voluminous than<br />
CM 3018, and a model shows the apatosaur weighed only 18<br />
<strong>HUNTERIA</strong> VOL. II, no. 3, pp. 10, February 19, 1988<br />
GREGORY S. PAUL<br />
tons (Table 1). Clearly A. louisae was much more strong<br />
limbed relative to its mass than was B. (G.) brancai.<br />
The largest specimen of B. (G.) brancai is the fibula<br />
HMN XV2 (lanensch, 1950b, 1961);claims of bigger specimens<br />
are unsubstantiated. There is variability in fibula/humerus or<br />
femur ratios in various specimens, but assuming it is similar<br />
to Sil in proportions, XV2 is about 13% larger (Table 1). In<br />
this case this individual would weigh up to 45 tons.<br />
The mass ofB. (B.) altithorax FMNH P25107is lessreadily<br />
determined. To get a rough estimate it was assumed that the<br />
neck, tail, and limbs weighed the same relative to the femur<br />
as inB. (G.) brancai. This is a fairly safe assumption, except<br />
that the neck may actually be somewhat smaller. A new model<br />
of the trunk incorporating 12 dorsals was sculpted to determine<br />
this portion of the total mass. FMNH P25107 appears<br />
to have weighed about 35 tons; despite its shorter limbs it may<br />
have been heavier than HMN SIl.<br />
Jensen (1978, 1985a,b) suggests that the Uncompahgre<br />
brachicsaurs represent uniquely large individuals. However,<br />
the great scapulocoracoid BYU 5001 is not especially large<br />
because cross scaling indicates that B. (G.) brancai<br />
HMN XV2 also had a scapulocoracoid about this length<br />
(Table 1), and several other B. (G.) brancai scapulocoracoids<br />
are not much smaller (Fig. 4A,B). Presumably the gross dimensions<br />
of XV2 and 8YU 5001 were similar too. If so, then<br />
estimates that BYU 5001 weighed up to 190 tons (Lambert,<br />
1983; McWhirter and McWhirter, 1986; Norman, 1985) are<br />
overstated. Assuming that BYU 5001 belongs to<br />
B. (B.) altithorax, this individual was closer to 45-50 tons.<br />
Reported 3038 mm long Uncompahgre ribs (jensen, 1985b)<br />
may be brachiosaurid and are in the same range as estimated<br />
for 5001 and XV2 (Table 1).If the giant 1360mm long Uncompahgre<br />
cervical BYU 5002 actually is a single spined<br />
brachiosaur as Jensen (1985a) indicates (but see below), then<br />
it too has a size similar to that estimated for 5001 and XV2.<br />
The indefinite position of anterior dorsal BYU 5000 and the<br />
lack of comparable dorsals in FMNH P25107 makes it difficult<br />
to compare the two individuals' masses. Besides, vertebra<br />
height may be more variable relative to mass than long bone<br />
length. It is notable that B. (G.) brancai HMN XV2 probably<br />
had anterior dorsals of similar size to BYU 5000<br />
(Table 1). The Potter Creek posterior dorsal is also in the<br />
5000-XV2 size class. In summary, the fragmentary Uncompahgre<br />
remains indicate that B. (B.) altithorax was in much<br />
the same size class as - perhaps a little heavier than - the<br />
biggest B. (G.) brancai specimens. That "ultrasaurs" are<br />
unique species of truly exceptional size is unconfirmed.<br />
Having found that B. (G.) brancai and B. (B.) altithorax<br />
were similar in size, the question arises as to whether they<br />
were the largest known terrestrial tetrapods. The Uncompahgre<br />
holotype ofDystylosaurus edwini, dorsal 8YU 5750,<br />
may be brachiosaurian. It cannot belong to the same taxon<br />
as BYU 5000 because these two anterior dorsals are too<br />
different to belong within the dorsal series of one single species.<br />
In having a transversely broad neural arch, BYU 5750 is rather<br />
like B. (G.) brancai, but otherwise it is very different, especially<br />
in its short, broad spine. With a height of about 1100
mm, BYU 5750 is very large but not uniquely so and, without<br />
further data on the animal's form, its weight can not be<br />
determined more precisely.Among other brachiosaurs, Russell<br />
(1980) and Anderson et at. (1985) suggested that<br />
"Brachiosaurus" ataliensis was unusually large. If the<br />
specimen cited is the same one in Lapparent and Zbyszewski<br />
(1957), then the radius and tibia are shorter than those of<br />
HMN Sll (Table 1). The 1500 mm dorsals of the bizarre<br />
brachiosaur Rebbachisaurus garasbae are much taller than<br />
those of the biggest of the other brachiosaurs. This height is<br />
due to its unusually tall spines; the limb material shows the<br />
body is not overall very large (Lavocat, 1952).<br />
The largestof the PaluxyRiver prints, which may have been<br />
made by the brachiosaur Pleurocoelus (Bird, 1985),are in<br />
the very large sizeclass (Table 1).Russell (1980)notes that the<br />
Breviparopus trackways (Dutuit and Ouazzou, 1980) also<br />
indicate very large sauropods. Indeed they are about the size<br />
of the biggest brachiosaurs. That these trackways were made<br />
by animals unbuoyed by water, and that sauropods were<br />
terrestrial in habits, proves that animals much bigger than<br />
elephants and as big as most whales can be land creatures.<br />
Bakker (1971c) suggested that Morrison apatosaurs and<br />
camarasaurs were as large as brachiosaurs. The largest<br />
Apcroscurus specimen is an incomplete femur YFM 1840<br />
described by Marsh. He first estimated it to be over 2.4 m<br />
long (Marsh, 1878),but the final restoration is more correct<br />
at about 1950mm (Marsh, 1896).Scaling up from CM 3018,<br />
the animal should have weighed only some 23 tons (Table 1).<br />
Two proximal Apatosaurus femora, CM 83 and CM 33994<br />
(McIntosh, 1981),are not as large as YPM 1840;cM 83 is<br />
not as large as CM 3018.It is more difficult to calculate mass<br />
in Camarasaurus because a high fidelity skeletal model of<br />
a complete adult specimen has yet to be completed. Scaling<br />
up of complete juveniles suggeststhat the biggestknown adult<br />
specimens(in the Cope collection,including a 1800mm femur,<br />
Osborn and Mook, 1921)weighed 22 to 26 tons.<br />
At least one other Morrison sauropod may be as massive<br />
and much longer than the brachiosaurs. A 2440 mm long<br />
Uncompahgre scapulocoracoid (BYU 5500, popularly labeled<br />
"Supersaurus") was initially identified as a brachiosaur<br />
(Fig.4D). Jensen (1985a)no longer supports this identification.<br />
Instead its long, rectangular anterior scapular process with<br />
Figure 4-Scapulocoracoids drawn to the same scale of A) B. (Brachiosaurus)<br />
altithorax referred BYU 5001; B) B. (Giraffatitan) brancai<br />
HMN Sa 9 with coracoid scaled in from HMN Ki 74; C) Supersaurus<br />
vivanae referred BYU 5501; D) S. vivanae holotype BYU 5500;<br />
E) Diplodocus carnegii CM 84. Measurements in Table 1.<br />
BRACHIOSAUR GIANTS<br />
a sharp anterodorsal corner, long rectangular coracoid,<br />
fairly broad scapular neck and short, modestly expanded<br />
scapular blade indicate it is a gracile, Diplodocus- or<br />
Barosaurus-like diplodocid. A 2700 mm long Uncompahgre<br />
scapulocoracoid (BYU 5501)is very similar and was correctly<br />
referred to the same taxon as BYU 5500 by Jensen (Fig 4C).<br />
Jensen placed the two shoulder girdles in the new Supersaurus<br />
viviane Jensen 1985. However, except for a longer<br />
upper blade, they are so extremely similar to Diplodocus<br />
(Hatcher, 1901) that congeneric status is very possible<br />
(Fig. 4C-E). On the other hand, the giant neck vertebra<br />
BYU 5003 is reconstructed with diplodocid-type double neural<br />
spines.If this reconstruction iscorrect, then the vertebra's great<br />
length suggests that these diplodocids are long necked<br />
Barosaurus. Scapulocoracoid BYU 5501 may belong to an<br />
animal with a 2600 mm femur, if the proportions were<br />
diplodocid-like.But because such gracilediplodocids are lightly<br />
built they probably weighed "only" some 50 tons, as large as<br />
the biggest brachiosaurs. Length is another matter, for<br />
assuming that BYU 5501 had about the same total length to<br />
scapulocoracoid length ratio as Diplodocus (Table 1) then<br />
it was some 42 m long. At 1130mm the diplodocid anterior<br />
caudal BYU 5002 is not much taller than those of<br />
Diplodocus CM 84 and USNM 10865, so it is not an<br />
especially large individual. Likewise the 12 articulated<br />
diplodocid caudals (BYU 5502) have a length of 300 mm, no<br />
longer than those of CM 84. Since these Uncompahgre<br />
caudals are so small, it is very unlikely that they belong to<br />
the same individuals as do either of the scapulocoracoids.<br />
"Seismosaurus" has been based on some fragmentary<br />
remains from the Morrison Formation (Gillette, 1987).The<br />
tall neural spined caudals are clearly those of a gracile<br />
diplodocid. They are of unusually large size, but since truly<br />
large Supersaurus caudals are not available for comparison,<br />
it is difficult to tell just how much so. If the total length was<br />
37 m long as suggested by Gillette, then it may have been<br />
smaller than Supersaurus. Note that the tall neural spines<br />
of the Supersaurus and "Seismosaurus" caudals mean that<br />
they probably had the massive hips typical of diplodocids -<br />
a rearing-up adaptation (Bakker, 1971c, 1978) that indicates<br />
that they also had the short forelimbs found in all complete<br />
diplodocid skeletons.<br />
Amphicoelias altus is another large gracile diplodocid<br />
(Osborn and Mook, 1921; the straight instead of forwardsloping<br />
posterior dorsal neural spine indicates it is a different<br />
taxon from other Diplodocus and Barosaurus species).The<br />
longest 1770 mm femur indicates a 16 ton animal, substantially<br />
heavier than regular sizedDiplodocus and Barosaurus<br />
specimens (Table 1).It is possiblethat Amphicoelias, Super.<br />
saurus, and "Seismosaurus" all represent one giant species.<br />
All the specimens may be from the uppermost Mortison, and,<br />
because not enough parts are shared in common to determine<br />
their real taxonomic status, the new names may be premature.<br />
Outside the Morrison a number of sauropods from southerly<br />
latitudes, which may represent titanosaurs, exhibit a tendency<br />
towards gigantism. A broken and incomplete MLP femur<br />
measures nearly 2200 mm, but was much longer when com-<br />
. )<br />
,i3 ,<br />
<strong>HUNTERIA</strong> VOL. II, no. 3, pp. 11, February 19, 1988<br />
J<br />
I
plete (Table 1).Assuming a typical sauropod mass/femur length<br />
ratio it may have weighed over 50 tons. Lacking further data<br />
it is not possible to estimate the animal's total length. Another<br />
super-sized sauropod is the femur of the titanosaur Antarctosaurus<br />
giganteus (Huene, 1929; Van Valen, 1969;<br />
Table 1). It is in the same class as the biggest brachiosaur<br />
femora, but since the rest of the animal is poorly known, the<br />
best that can be said is that it was probably similar in mass<br />
too. The South African (Mclachlan and McMillan, 1976)and<br />
Laotian (Hoffet, 1942)femora Anderson et a1.(1985)cited are<br />
not from uniquely large individuals (Table 1).<br />
In the final analysis, B. (G.) brancai and B. (B.) altithorax<br />
were about as big as any other known sauropods. Just<br />
as importantly, B. (G.) brancai holds the record for the<br />
biggest species for which all of the skeleton is known. It was<br />
also the tallest, but not the longest. The largest known<br />
sauropods appear to cluster around 50 tons in lean condition,<br />
perhaps a third more in prime fat-bearing condition. This<br />
should not be taken as an ultimate limit. The sample of all<br />
known sauropods is a tiny fraction of the sampled populations<br />
available for many single species of living animals. Even<br />
larger sauropods certainly await discovery, and it is improbable<br />
that the largest were preserved in the fossil record. It is interesting<br />
that, in living tetrapods, extremely rare "world record"<br />
individuals are often twice as massive as average individuals<br />
(McWhirter and McWhirter, 1986).In this view sauropods of<br />
Aitken, W.G. 1961. Post Karroo. In: Quennell, -AM.,<br />
McKinlay, AC.M. and Aitken, W.G. (eds.). Summary of<br />
the geology of Tanganyika. Geological Survey of<br />
Tanganyika, Memoir 1. p. 153-256.<br />
Alexander, R.McN. 1985. Mechanics of posture and gait of<br />
some large dinosaurs. Zoological Journal Linnean Society<br />
83: 1-25.<br />
Anderson, J.F., Hall-Martin, A and Russell, D.A 1985.Longbone<br />
circumference and weight in mammals, birds and<br />
dinosaurs. Journal of Zoology 207:53-61.<br />
. Arkell, W.j. 1956.Jurassic Geology of the World. London:<br />
Oliver & Boyd Ltd.<br />
Bakker, R.T. 1971a. Dinosaur physiology and the origin of<br />
mammals. Evolution 25:636-658.<br />
Bakker, R.T. 1971b. Ecology of the brontosaurs. Nature<br />
229:172-174.<br />
Bakker, R.T. 1971c.Brontosaurs. In: McGraw Hill Yearbook<br />
of Science and Technology. New York: McGraw Hill. p.<br />
179-181.<br />
Bakker, R.T. 1974.Dinosaur bioenergetics: A reply to Bennett<br />
and Dalzell, and Feduccia. Evolution 28:497-502.<br />
Bakker, R.T. 1975.Dinosaur renaissance. Scientific American<br />
232 (4):58-78.<br />
Bakker, R.T. 1978. Dinosaur feeding behavior and the origin<br />
of flowering plants. Nature 248:168-172.<br />
Bennett, AF. and Dalzell, B. 1973. Dinosaur physiology: a<br />
critique. Evolution 27:170-174.<br />
Bird, R.T. 1985. Bones for Barnum Brown. Fort Worth:<br />
Texas Christian University Press.<br />
<strong>HUNTERIA</strong> VOL. U, no. 3, pp. 12, February 19, 1988<br />
GREGORY S. PAUL<br />
CITED REFERENCES<br />
100tons are not unrealistic, especially if bearing large amounts<br />
of seasonal fat.<br />
In comparison, Balaenoptera musculus typically weighs<br />
80-100 tons, and may reach 200 in feeding season (Ellis, 1980;<br />
McWhirter and McWhirter, 1986). Unfortunately there has<br />
never been a rigorous study of the mass of baluchitheres and<br />
the largest fossil proboscideans, which may rival each other<br />
as the biggest of terrestrial mammals. The baluchitheres and<br />
bigger mammoths appear to be rather gracile, and may not<br />
have been as massive as sometimes suggested. At perhaps 20<br />
tons or less, they certainly do not match the bigger sauropods<br />
in size. As for the greatest living land animal,Loxodonta<br />
africana bulls average 5 tons, often reach 7.5, and rarely reach<br />
about 10+ (Laws et al., 1975; McWhirter and McWhirter,<br />
1986).<br />
ACKNOWLEDGEMENTS<br />
Special thanks go to J. McIntosh and his bottomless store of Information<br />
on giant Mesozoic monsters. R. Bakker's many discussions on dinosaur<br />
anatomy were critical to this paper, as was his request that I prepare it.<br />
F. Boothman has provided important information and enjoyable discussion<br />
on the problem of big dinosaurs. K. Carpenter, M. Brett-Surman,<br />
D. Berman, D. Russell, R. Molnar, Dong Z., P: Currie and D. Gasparini<br />
have provided much helpful information and discussion on sauropods. Many<br />
thanks to H. Jaeger for his hospitality during my visit to the HMN. P. Sites<br />
helped reproduce the figures.<br />
Carrington, R. 1959. Elephants. New York: Basic Books.<br />
Colbert, E.H. 1962. The weights of dinosaurs. American<br />
Museum Novitates 2076:1-16.<br />
Cooper, M.R. 1984. A reassessment of Vulcanodon karibaensis<br />
Raath (Dinosauria: Saurischia) and the origin of<br />
the sauropods. Paleontologica Africana 25:203-231.<br />
Demmont, M.W. arid Soest, P.J. 1985. A nutritional explanation<br />
for body-size patterns of ruminant and nonruminant<br />
herbivores. American Naturalist 125:641-672.<br />
Dimery, N.j., Alexander, R.McN. and Deyst, K.A 1985.<br />
Mechanics of the ligamentum nuchae of some artiodactyls .<br />
Journal of Zoology 206:341-351.<br />
Dodson, P., Behrensmeyer, A.K., Bakker, R.T. and Mcintosh,<br />
J.H. 1980. Taphonomy and paleoecology of the dinosaur<br />
beds of the Jurassic Morrison Formation. Paleobiology<br />
6:208-232.<br />
Dutuit, J-M. and Ouazzou, A. 1980. Decouverte d'une piste<br />
de dinosaures sauropodes sur Ie site d'empreintes de<br />
Demnate (Haut-Arlas marocain). Memoirs Societe<br />
Geologie Francais 59 (139):95-102<br />
Ellis, R. 1980.The Book of Whales. New York: Alfred A<br />
Knopf.<br />
Eltringham, S.K. 1982. Elephants. Denver: Blanford Press.<br />
Fraas, E. 1908. Ostaftikanische Dinosaurier. Palaeontographica<br />
55:105-144.<br />
Galton, P.M. 1977. The ornithopod dinosaur Dryosaurus<br />
and a Laurasia-Gondwanaland connection in the Upper<br />
Jurassic. Nature 268:230-232.<br />
.:.
Galton, P.M. 1980.Dryosaurus and Camptosaurus, intercontinental<br />
genera of Upper Jurassic ornithopod dinosaurs.<br />
Memoirs Societe Geologie Francais 139:103-108<br />
Galton, P.M.· 1982. Elaphrosaurus, an ornithomimid<br />
dinosaur from the Upper Jurassic of orth America and<br />
Africa. Palaontologie Zeitschrift 56:265-275.<br />
Gillette, D.D. 1987. A giant sauropod from the [ackpile SS<br />
Member of the Morrison Formation. Journal of Vertebrate<br />
Paleontology 7 (suppl. to 3):16-17.<br />
Gilmore, C.W. 1932. A newly mounted skeleton of<br />
Diplodocus in the United States National Museum. Proceedings<br />
U.S. National Museum 81:1-21.<br />
Gilmore, CW. 1937. Osteology of Apatosaurus with special<br />
reference to specimens in the Carnegie Musuem. Memoirs<br />
Carnegie Museum 11:175-301.<br />
Hatcher, ].B. 1901. Diplodocus Marsh, its osteology, taxonomy,<br />
and probable habits, with a restoration of the<br />
skeleton. Memoirs Carnegie Museum 1:1-63.<br />
Hoffet, ].H. 1942. Description de quelques ossements de<br />
Titanosauriens du Senonien du Bas-Laos. C.R. Cons.<br />
Recherche Scientifique Indochine 1942:1-8.<br />
Huene, EY. von. 1929. Los saurisquios y ornitisquios del<br />
Cretaceo Argentino. Annales Museo La Plata 2 (3):1-196<br />
Jackson, K. and Matternes, J.H. 1972.Dinosaurs. Washington,<br />
D.C: National Geographic Society.<br />
Janensch, W. 1914.Die gliederung der Tendaguru-Schichten<br />
im Tendaguru-Gebeit und die entstehung der<br />
saurierlagerstatten. Arch. f. Biontologie 3:227-262<br />
[anensch, W. 1929.Material und formengehalt der sauropoden<br />
in der ausbeute der Tendaguru-expedition. Palaeontographica<br />
suppl. 7 (2):2-34.<br />
Janensch, W. 1935/1936. Die schadel der sauropoden<br />
Brachiosaurus, Barosaurus und Dicraeosaurus aus der<br />
Tendaguru-Schichten Deutsch-Ostafrikas. Palaeontographica<br />
suppl. 7 (2):145-298.<br />
Janensch, W. 1950a. Die wirbelsaule von Brachiosaurus<br />
brancai. Palaeontographica suppl. 7 (3):27-93.<br />
Janensch, W. 1950b. Die skelettrekonstruktion von<br />
Brachiosaurus brancai. Palaeontographica supp!. 7<br />
(3):97-102.<br />
[anensch, W. 1961.Die gliedmassen und gliedmassengurtel der<br />
sauropoden der Tendaguru-Schichren, Palaeontographica<br />
suppl. 7 (3):177-235.<br />
Jensen, J.A 1978. Ultrasaurus. Science News 116:84.<br />
Jensen, J.A. 1985a. Three new sauropod dinosaurs from the<br />
Upper Jurassic of Colorado. Great Basin Naturalist<br />
45:697-709.<br />
Jensen, J.A 1985b. Uncompahgre dinosaur fauna: a<br />
preliminary report. Great Basin Naturalist 45:710-720.<br />
Kim, H.M. 1983. Cretaceous dinosaurs from Korea. Journal<br />
Geology Society Korea 19 (3):115-126.<br />
Knight, CR. 1947.Animal Drawing: Anatomy and Action<br />
for Artists. New York: McGraw-Hill Book Company.<br />
Lambert, H. 1983.A Field Guide to Dinosaurs. New York:<br />
Avon.<br />
Lapparent, AF. de and Zbyszewski, G. 1957. Les dinosauriens<br />
du Portugal. Servo Geologiques du Portugal, Memoires<br />
2:1-63.<br />
Lavocat, R. 1952. Sur les dinosauriens du continental intercalaire<br />
des Kern-Kern de la Daoura. Congres Geologie<br />
Internationale d' Alger, fasc. XV:65-68.<br />
BRACHIOSAUR GIANTS<br />
Laws, R.M., Parker, l.S.C and Johnstone, R.CB. 1975.<br />
Elephants and Their Habitats. Oxford: Clarendon Press.<br />
Lockley, M., Houck, K.]. and Prince, N.K. 1986. North<br />
America's largest dinosaur trackway site: implications for<br />
Morrison Formation paleoecology. Geology Society of<br />
America Bulletin 97: 1163-1176.<br />
Lull, R.S. 1919.The sauropod dinosaur Barosaurus Marsh.<br />
Memoirs Connecticut Academy of Arts & Sciences<br />
VI: 1-42<br />
Mclntosh, ].S. 1981. Annotated catalogue of the dinosaurs<br />
(Reptilia, Archosauria) in the collections of the Carnegie<br />
Museum of Natural History. Bulletin Carnegie Museum<br />
of Natural History 18:1-67.<br />
McLachlan, LR. and McMillan, LK. 1976. Review and<br />
stratigraphic significance of southern Cape Mesozoic<br />
palaeontology. Transcripts Geology Society South Africa<br />
79:197-212.<br />
McWhirter, N. and McWhirter, R.M. 1986.Guinness 1986<br />
Book of World Records. New York: Bamam Books.<br />
Marsh, O.C 1878. Notice of new dinosaurian reptiles.<br />
American Journal of Science 15:241-244<br />
Marsh, O.C 1896. Dinosaurs of North America. Annual<br />
Report U.S. Geological Survey 16:133-244.<br />
Migeod, P.W.H. 1931. British Museum East Africa Expedition.<br />
Natural History Magazine 87-103.<br />
Muybridge, E. 1887.Animals in Motion. New York: Dover<br />
1957 reprint.<br />
Norman, D.B. 1980. On the ornithischian dinosaur Iguanodon<br />
bernissartensis from the Lower Cretaceous of Bernissart<br />
(Belgium). Institut Royale Scientifique Naturale<br />
Belgique Memoires 178:1-101.<br />
Norman, D.B. 1985. The Illustrated Encyclopedia of<br />
Dinosaurs. New York: Crescent Books.<br />
Osborn, H.P. and Mook, CO. 1921. Camarasaurus,<br />
Amphicoelias and other sauropods of Cope. Memoirs<br />
American Museum of Natural History 3 (30):249-387.<br />
Ostrom, J.H. 1978. New ideas about dinosaurs. National<br />
Geographic 154 (2):152-185.<br />
Ostrom, ].H. and Mclntosh, J.S. 1966. Marsh's Dinosaurs.<br />
Yale University Press, New Haven.<br />
<strong>Paul</strong>, G.S. 1987. The science and art of restoring the life<br />
appearance of dinosaurs and their relatives. In: Czerkas, S.<br />
and Olsen, E. (eds.). Dinosaurs Past and Present. Los<br />
Angeles: Natural History Museum of Los Angeles County.<br />
pp. 4-49.<br />
<strong>Paul</strong>, GS and Chase, T in press. Vertebrate paleontological<br />
reconstructions In: Hodges, E. (ed.).The Guild Handbook<br />
of Scientific Illustration. New York: Van Nostrand<br />
Reinhold.<br />
Riggs, E.R. 1901. The largest known dinosaur. Science<br />
13:549-550.<br />
Riggs, E.R. 1903. Brachiosaurus altithorax, the largest<br />
known dinosaur. American Journal of Science (4)<br />
15:299-306.<br />
Riggs, E.R. 1904.Structure and relationships of opisthocoelian<br />
dinosaurs, pt. II the Brachoisauridae. Field Columbian<br />
Museum, Publications in Geology 2 (6):229-247.<br />
Russell, D.A 1980. Reflections of the dinosaurian world In:<br />
Jacobs, LE. (ed.). Aspects of Vertebrate History. Flagstaff:<br />
Museum of Northern Arizona Press. p. 257-268.<br />
<strong>HUNTERIA</strong> VOL II, no. 3, pp. 13, February 19, 1988
Russell, D., Beland, P. and McIntosh, ].S. 1980. Paleoecology<br />
of the dinosaurs ofTendaguru (Tanzania). Memoirs Societe<br />
Geologie Francais 59 (139):169-175.<br />
Spinar, Z.v. and Burian, Z. 1972. Life Before Man. New<br />
York: American Heritage Press.<br />
Stokes, W.L. 1985. The Cleveland-Lloyd Dinosaur Quarry,<br />
window to the past. U.S. Government Printing Office<br />
1-27.<br />
The neck of B. (G.) brancai HMN JII is about 0.3 m<br />
shorter than the 9.3 m neck of the primitive, 9 ton diplodocid<br />
Mamenchisaurus hochuanensis PMNH 3 (Young and<br />
Chow, 1972). It is likely that Supersaurus also had a longer<br />
neck than the largest Brachiosaurus.<br />
Martin (1987) restores the neck of the sauropod<br />
Cetiosaurus as barely able to reach the ground, rise above<br />
the height of the shoulders, or swing to either side. That such<br />
a neck of 14 bird-like vertebrae would be far less flexible than<br />
the giraffe's neck of 7 vertebrae is untenable. A high number<br />
of vertebrae directly implies great flexibility, since a stiff long<br />
neck is better achieved by lengthening a few vertebrae, as in<br />
the giraffe. Errors in Martin's restoration include a maximum<br />
raised neck posture that is really its normal, neutral S curve,<br />
zygapophyses that are virtually immobile, and an average of<br />
only 1.4° of motion from the centerline between each cervical.<br />
Instead, the sauropod cervical combination of ball and socket<br />
centra articulations with large zygapophyses was designed to<br />
maintain articulation over a much greater range of motion<br />
than Martin shows, especially when the bony joint areas were<br />
expanded by cartilagenous surfaces. Exactly how much more<br />
I am not sure, what one can learn from dry bone manipulations<br />
or paper studies is useful, but not necessarily true to life.<br />
Sauropod and giraffe cervicals are remarkably similar; study<br />
of the latter might prove helpful to the problem. Note that<br />
a modest 12° or so of rotation between successive segments<br />
in 6 posterior cervicals would allow the sauropod head to reach<br />
GREGORY S. PAUL<br />
NOTES ADDED IN PROOF<br />
Van Valen, L. 1969. What was the largest dinosaur? Copeia<br />
1969:624-262<br />
Watson, ].W. and ZaUinger, R.F. 1960. Dinosaurs and Other<br />
Prehistoric Reptiles. New York: Golden Press.<br />
. high up. This mobility is plausible since many mammals can<br />
do the same with only 7 cervicals. The extreme neck<br />
inflexibility Martin restores in sauropods is also functionally<br />
illogical, since shorter necked ungulates can reach as far up<br />
and to the side relative to their size. Long necks are an extreme<br />
adaptation, and among land herbivores their only useful<br />
purpose is to increase the vertical reach of high browsers. Low<br />
browsers and grazers invariably have modest necks because<br />
they can always reach what they want by just taking a few<br />
steps towards it.<br />
Contrary to Martin (1987), the ilial pubic penducles of all<br />
sauropods were massive, buttressed from the front and inside<br />
by stout sacral ribs, and well able to bear the mass of a rearing<br />
individual. Note that sauropods only stood, or occasionally<br />
slowly walked, bipedally. They did not incur the stresses of<br />
fast bipedal motion.<br />
PMNH - Beijing Museum of Natural History, Beijing.<br />
Martin, ]. 1987. Mobility and feeding of Cetiosaurus<br />
(Saurischia, Sauropoda) - why the long neck? In: Currie,<br />
P.]. and Koster, E.H. (eds.). Fourth Symposium on<br />
Mesozoic Terrestrial Ecosystems. Drumheller: Tyrell<br />
Museum of Paleontology. pp. 150-155.<br />
Young c.c. and Chow X.]. 1972. On Mamenchisaurus<br />
hochuanensis. Paleontographica Sinica, A., Special Issue<br />
8:1-30<br />
Total cost of production of this paper was donated by Dinamation International, Chris Mays, President.<br />
<strong>HUNTERIA</strong> VOL. II, no. 3, pp. 14, February 19, 1988<br />
-- _