The evolution of sauropod dinosaurs from 1841 to 2008

6th May 2008

Michael P. Taylor
Palaeobiology Research Group, School of Earth and Environmental Sciences, University of Portsmouth, Burnaby Road, Portsmouth PO1 3QL, UK. <dino@miketaylor.org.uk>

The sauropod dinosaurs are among the most iconic of all prehistoric animals, with their distinctive small heads, long necks, huge torsos and long tails. The first genera now recognised as sauropods, Cardiodon and Cetiosaurus, were named by Owen in 1841, only 17 years after the first dinosaur was named and a year before the term "dinosaur" was coined. However, Cardiodon was based on a pair of uninformative teeth; and Cetiosaurus, from the scant remains then available, was initially interpreted as an immense aquatic crocodilian. The type humerus of Pelorosaurus was the first sauropod bone to be interpreted as that of a terrestrial animal, by Mantell in 1850, due to its possession of a medullary cavity, but it was not then recognised as related to Cetiosaurus. The two presacral vertebrae on which Seeley founded Ornithopsis in 1870 were thought to be pterosaurian due to their light construction and pneumatic foramina, features not then known in any terrestrial animal.

The much more complete remains of the new species Cetiosaurus oxoniensis, described by Phillips in 1871, began to cast light on the sauropod body plan, suggesting that sauropods were dinosaurs rather than crocodilians. But it was not until the description of the American sauropods Apatosaurus and Camarasaurus in 1877 that anything like a complete sauropod skeleton was found. In the same year, Lydekker named Titanosaurus on the basis of isolated tail vertebrae from India, becoming the first Gondwanan sauropod. Marsh named Diplodocus the next year, and in the same paper created the taxon Sauropoda to hold these genera.

Traditionally, sauropods were considered amphibious, requiring water to support their great mass, although Phillips suggested otherwise as early as 1871 and Riggs argued forcefully for terrestriality in 1904. Many independent lines of research now show that sauropods were predominantly terrestrial: they had compact feet unsuited for swamps, tall and relatively narrow torsos adapted for weight bearing, and pneumatised skeletons that greatly reduced their weight. Furthermore, sauropod fossils are found primarily in terrestrial sediments.

Early reconstructions of sauropod skeletons, such as Marsh's 1883 Brontosaurus, correctly positioned the legs upright. However, Tornier (1909) in Germany and Hay (1910) in America favoured a lizard-like sprawl in which the humerus and femur project at right-angles to the body -- a suggestion soundly refuted by Holland (1910). Neck posture, however, continues to be controversial. Although Marsh's Brontosaurus reconstruction showed a nearly horizontal neck, the neck of Christman's Camarasaurus in Osborn and Mook's 1921 monograph was inclined upward, and Janensch (1950) reconstructed the neck of Brachiosaurus brancai nearly vertical. Recent computer simulations of cervical joints favour horizontal neck postures; but these studies remain open to interpretation, and other lines of evidence, both mechanical and ecological, support high browsing.

New sauropod discoveries continue to expand our knowledge, with half of all known sauropod genera having been named in the last two decades. These include the spiked Amargasaurus, huge Argentinosaurus, wide-mouthed Nigersaurus and short-necked Brachytrachelopan, indicating that sauropods were far more diverse than previously recognised.

References

Ryder's (1877) reconstruction of Camarasaurus

Camarasaurus supremus Cope 1877, skeletal reconstruction by John A. Ryder. This reconstruction was executed in 1877, the same year that the first Camarasaurus remains were discovered, and is the earliest known attempt to reconstruct a sauropod skeleton. It was first exhibited at a meeting of The American Philosophical Society in Philadelphia, Pennsylvania, on December 21, 1877, and first published as a reduced figure by Mook (1914). Reproduced from Osborn and Mook (1921: plate LXXXII).

(The abstracts booklet for this conference was much better than for most: the abstracts are twice as long, and so contain actual information rather than just an advert for the talk, they are allowed to have references, and they get a figure, too. In fact, they're like tiny, tiny papers.)

[PDF Abstract] [Slides]

Feedback to <mike@miketaylor.org.uk> is welcome!