Although this seems to be simple question, the answer turns out to be very complicated, and beset by ``if''s and ``but''s. The right way to slice up the dinosaur genera that we know of is perhaps the single most contentious issue in contemporary palaeontology, and certainly the one in which the State Of The Art changes most quickly. As will become apparent below, even the most basic classification of all dinosaurs into two groups is now questioned, so any answer given here can only be an approximation towards the ever-moving target of current interpretation.
That said, here is a brief survey of the major dinosaurian groups. The following table of contents can also be read as a basic family tree, with derived groups below and to the right of their parents.
1. Saurischia
1.1. Sauropodomorpha
1.1.1. ``Prosauropods''
1.1.2. Sauropoda
1.1.2.1. Diplodocidae
1.1.2.2. Macronaria
1.1.2.2.1. Brachiosauridae
1.1.2.2.2. Titanosauria
1.2. Theropoda
1.2.1. Ceratosauria
1.2.1.1. Coelophysoidea
1.2.1.2. Neoceratosauria
1.2.2. Tetanurae
1.2.2.1. Spinosauria
1.2.2.2. Carnosauria
1.2.2.3. Coelurosauria
1.2.2.3.1. Tyrannosauria
1.2.2.3.2. Ornithomimosauria
1.2.2.3.3. Oviraptorosauria
1.2.2.3.4. Therizinosauria
1.2.2.3.5. Alvarezsauria
1.2.2.3.6. Troodontidae
1.2.2.3.7. Deinonychosauria
1.2.2.3.8. Avialae
2. Ornithischia
2.1. Thyreophora
2.1.1. Stegosauria
2.1.2. Ankylosauria
2.2. Marginocephalia
2.2.1. Ceratopsia
2.2.1.1. Centrosaurinae
2.2.1.2. Chasmosaurinae
2.2.2. Pachycephalosauria
2.3. Ornithopoda
2.3.1. Iguanodontia
2.3.1.1. Hadrosauridae
2.3.1.1.1. Hadrosaurinae
2.3.1.1.2. Lambeosaurinae
Since the earliest days of dinosaur palaeontology, dinosaurs have been split into two very broad groups - the Saurischia and Ornithischia (meaning lizard-hipped and bird-hipped respectively.) Indeed for some time it was believed that these two groups had no common ancestor, so that the term ``dinosaur'' was not a taxonomically sound one. However, most palaeontologists now believe that there was a single common ancestor of all dinosaurs, and the grouping Dinosauria is restored.
In recent times, though, even the basic Saurischia/Ornithischia division - like almost everything - has been questioned. Some people now hold that the two main divisions within Saurischia - the Sauropoda and Theropoda - are actually separate suborders, yielding a three-fold division; while others would now classify the sauropods together with the Ornithischia in a grouping called Phytodinosauria, meaning ``plant(-eating) dinosaurs''. The jury seems still to be out on this one.
Rather unnecessarily, life is made yet more confusing by those who wish to rename the Ornithischia as the Predentata (named after the predentary, a bone at the front of the lower jaw which is unique to this group of dinosaurs.) In this document, we use the older, better established term. If we were in the business of assigning new names to existing taxa, there are plenty with more inappropriate names than the Ornithischia (for example, the theropods and ornithopods seem to have been given each other's names.) My feeling is that since there's so much scope for getting legitimately confused when studying dinosaurs, it's not necessary to create artificial new confusion. Or as Henry Spencer has put it, ``Your creativity is better used in solving problems than in creating beautiful new impediments to understanding.''
New theories notwithstanding, current orthodoxy has two main groupings in the Saurischia: the sauropods are the largest of all dinosaurs, with heavy bodies and long necks and tails; and the theropods are the bipedal meat-eaters. The Therizinosauria (previously known as Segnosauria) have sometimes been considered as a candidate third group within the Saurischia (and have sometimes been considered as ornithischians!) but current consensus buries them deep within the theropods. More on these beasties later.
In one of the naming foul-ups that so persistently characterise dinosaurian classification, it turns out that birds are saurischian dinosaurs - that is, they are ``lizard-hipped'' rather than ``bird-hipped''. This just goes to show that it's a dangerous thing to name groups of animals before you know much about them. But then what alternative did the early palaeontologists have? They could hardly wait around for a century or so to see what would be discovered subsequently before naming their groups. The long and short of it is that we're stuck with the name Saurischia, and the best thing to do is just shrug and accept it.
The sauropods together with the ``prosauropods'' and few more basal animals are collectively known as the Sauropodomorpha.
The name ``prosauropod'' means ``before sauropods'', reflecting the old belief that they were the ancestors of the sauropods. However, current orthodoxy holds that prosauropods were a branch off the evolutionary line that led to the sauropods; or even that they are the aggregate of several such branches, and thus not a valid phylogenetic grouping at all. Typified by Plateosaurus, they were broadly similar in body shape to the sauropods, but rather smaller. Some genera may have been primarily or exclusively bipedal. The consensus is that they were herbivorous, but some argue for an omnivorous or carnivorous diet.
The two main groups within the sauropods are the Diplodocidae and Macronaria; in addition to these groups are numerous genera which don't fall into either camp, among them Cetiosaurus, the first sauropod to be discovered, and Mamenchisaurus, the dinosaur with the proportionally longest neck.
This grouping contains Diplodocus itself along with larger close relatives such as Supersaurus and Seismosaurus, as well as slightly more distant relatives including Apatosaurus (``Brotosaurus''.)
Apart from some of the larger Diplodocidae, all the largest (in the senses of longest and heaviest) dinosaurs fall into the Macronaria. In addition to basal genera such as Camarasaurus, the two major groupings within the Macronaria are the Brachiosauridae and the Titanosauria.
These are tall, heavily built sauropods characterised by their front legs being longer than their back legs, and therefore sloping upwards from the hips to the shoulders (hence the name Brachiosaurus, meaning ``arm lizard''.)
There is some controversy over how the various ``Brachiosaur'' genera are related. Some people place Brachiosaurus itself, Giraffatitan (if it's valid), Sauroposeidon, Cedarosaurus and Sonorasaurus in the group; others exclude the latter two (some of them considing Sonorasaurus to be a late-surviving species of Brachiosaurus), and yet others think that only Brachiosaurus belongs there.
For those who don't recognise the family Brachiosauridae at all, all these genera, along with a grab-bag of others, are left lurking guiltily around in the Titanosauriformes, a sub-macronarian group which also contains the Titanosauria proper. Perhaps we can look forward to clarification as newer genera like Sauroposeidon are more fully explored.
Unusually among Sauropods, at least some of the titanosaurs were armoured with bony plates embedded in the skin, in a somewhat similar manner to ankylosaurs. Recent work in the southern hemisphere, particularly Argentina, seems to be turning up new, massive, titanosaurs every five minutes or so. As I write, the titanosaur Argentinosaurus is widely considered the largest properly-described dinosaur, at perhaps 45m in length and 100 metric tonnes in weight; but already, new and as yet unnamed Argentinian titanosaurs have been discovered for which greater size is claimed in every dimension.
Traditionally, the theropods have been divided into just two groups: the big ones were called carnosaurs, and the small ones coelurosaurs. This grouping is appealing in its simplicity, but a moment's thought shows that it makes no sense - it's like classifying wolves together with leopards because they're both big, and chihuahuas together with Siamese cats because they're both small.
In recent times, this scheme has been thrown up in the air, although the names of both the old groups live on with new definitions. Amusingly, Tyrannosaurus rex, the quintessential carnosaur, is now classified as a coelurosaur. (You can't help thinking that the film Carnosaur! wouldn't have sounded quite so scary had it been named Coelurosaur!)
The basic division now is into two groups: the Ceratosauria and Tetanurae (the latter group containing both the new carnosaurs and coelurosaurs). Outside of these two major groups fall a few primitive theropods, notably Eoraptor (generally considered the oldest known dinosaur) and probably the Herrerasauria - although some consider them not to be dinosaurs at all, but to be the sister group to the Dinosauria.
This group is broken in two: the older Coelophysoidea and the more recent Neoceratosauria (from the early Jurassic and mid Cretaceous respectively.)
These are early predators such as Coelophysis (the first dinosaur in space[1]) and Dilophosaurus - the latter restored rather idiosyncratically in the film version of Jurassic Park, with a neck frill and poison-spitting ability neither of which are so much as hinted at in the fossil record.
This group consists of a few genera such as Ceratosaurus together with an enigmatic group called the Abelisauroidea, which contains, among others, the weird Carnotaurus. Many neoceratosaurs, including these two genera, had crests, horns and other such displays on their skulls.
Some palaeontologists have considered the massive carcharodontosaurines to be abelisauroid, but current consensus has them in with the allosaurs.
The name means ``stiffened tail'', which is among the characteristics of this very varied group. It is generally more advanced than its sister group the Ceratosauria.
This group contains Megalosaurus, the first dinosaur ever described, together with three main sub-groups: the spinosaurs, the carnosaurs and the coelurosaurs. The two latter groups (in the new senses of these terms) together with a few stray genera make up the Neotetanurae.
These are large animals, conjecturally piscivorous and characterised by long, crocodile-like snouts and sails on their backs. They include Spinosaurus (the rumoured star of Jurassic Park III), the charmingly named Irritator, and Baryonyx (which I have a soft spot for as it was the last exciting dinosaur found in England. STOP PRESS: that's not true now - we have Eotyrannus!)
Some people consider the spinosaurs to be advanced coelophysids, but this is a minority view.
This group now excludes many genera which it used to include, such as Carnotaurus (now a ceratosaur) and Tyrannosaurus (now a coelurosaur). In its new sense, it has been stripped down to basically the allosaurs plus a few bin-ends.
Of course, that's enough to be getting on with. As well as Allosarus itself, the archetypal Jurassic predator, this group is generally considered to contain the Carcharodontosaurines, a group represented by genera such as Giganotosaurus, perhaps the largest predator of all time. (Some people think they belong with the Abelisaurids, but this is a minority view.)
This is an extremely varied grouping, containing a vast selection of very different carnivores. Among the more basal coelurosauria are Ornitholestes, Compsognathus and Dryptosaurus (the most likely identity of the specimens named as ``Laelaps'' in the early decades of dinosaur palaeontology.)
The more advanced coelurosaurs are known as Maniraptoriformes. Below this level, classification is open to a great deal of interpretation, and orthodoxy seems to be in constant flux. However, for now, we can postulate the following relationships:
Coelurosauria
|--Tyrannosauria
`--Maniraptoriformes
|--Ornithomimosauria
`--Maniraptora
|--Alvarezsauria
|--+--Oviraptorosauria
| `--Therizinosauria
`--Paraves
|--Avialae
`--+--Troodontidae
`--Dromaeosauridae
If this phylogeny is correct, then the advanced coelurosaurs are the Tyrannosauria and Maniraptoriformes, with the latter consisting of Ornithomimosauria and Maniraptora; the Maniraptora consist of the Alvarezsauria an Oviraptorosauria-and-Therizinosauria clade known as the enigmosauria, plus Paraves; Paraves consists of the Avialae (including modern birds), and a clade consisting of Troodontidae and Dromaeosauridae
But remember that all of this is subject to change.
Tyrannosaurus rex and his buddies were the most advanced group of large carnivores, with specialisations including much larger and more robust skulls than other theropods, and tiny vestigial forelimbs with only two fingers on each hand.
T. rex is perhaps the most studied of all dinosaur species, yet controversy still abounds in almost every possible area of its study. Was it a hunter or a scavenger? Could it run fast, or was it restricted to slower speeds by its size? Were the males larger or smaller than the females? We still don't know, though study continues unabated.
This group contains the ``bird-mimics'' such as Ornithomimus, Struthiomimus and the mysterious Deinocheirus, of which only a gigantic pair of arms is known. The ornithomimosaurs had long necks, small heads with toothless, beak-like mouths, and long legs, with the distal parts much longer than the femur - an adaptation for fast running. The result of these characteristics is that they somewhat resemble large ostriches with tails, hence the name of the group.
Unlike other theropods, these animals may have been omnivorous or perhaps even herbivorous.
These specialised, medium-sized animals are characterised by a peculiar skull with a toothless beak and a nasal crest. Some adaptations of the skull suggest an egg diet, hence the name of the type genus, meaning ``egg stealer''. (Although ironically, the eggs that the type specimen of Oviraptor was thought to be stealing turn out to be its own, which it was actually brooding.)
The most notable feature of these weirdos is the very long claws on their hands - of the order of a meter in the case of Therizinosaurus. They also depart from theropod normality in many other ways: for example, they have four toes on each foot rather than usual three; and the pubis points backwards rather than forwards. These and other oddities have meant that in the past they have have been classified with the ``prosauropods'' and even within the Ornithischia!
These are very odd smallish dinosaurs, with long, slender hind legs and short, powerful arms. The latter suggest that the alvarezsaurs may have been strong diggers, but this seems to be at odds with the cursorial lifestyle suggested by the hind legs.
Troodon and its kin were small, agile dinosaurs with large forward-facing eyes (hence, probably, steroscopic vision) and grasping claws. They also had the largest brains (relative to body size) of all the dinosaurs, and can thus be considered to have been the most intelligent dinosaurs. Troodontids, like deinonychosaurs, are equipped with a single, large, retractable claw on each foot; however, this is considered to be an example of convergence, rather than a trait derived from a shared ancestor.
Troodontids have in the past been considered to belong inside Deinonychosauria as a sister group to the dromaeosaurids, and in several other positions in the family tree. Like so much, the exact postioning is still open to interpretation; however, pretty much everyone agrees that they are at least Maniraptoriformes.
Mostly smallish predators with grasping hands, stiff tails, and relatively large brains. The ``Velociraptors'' (actually more like Deinonychus) of Jurassic Park are typical of this group. Their most obvious feature is an enlarged recurved claw on the innermost toe of each foot, which was carried clear of the ground to keep it sharp.
Birds (including modern birds) and their close relatives. The line between advanced non-birds and primitive birds is so blurred that there are plenty of genera which may or may not be birds.
Some workers invert pretty much all of the coelurosaurian family tree, considering that birds came first (the so-called BCF hypothesis), and that many or all of the coelurosaurs were actually secondarily flightless birds - like penguins.
The Ornithischia are so named because their hip bones superficially resemble those of modern birds, in that the public bone points downwards and backwards, rather than down and forwards as in most (though not all) saurischians.
With no exceptions so far as we know, the ornithischians were exclusively herbivorous, and the more derived members of the group are characterised in part by successively more efficient chewing apparatus for dealing with tough plant material.
The ornithischians can be divided into three major sub-groups: the Thyreophora or ``armoured dinosaurs'', the Marginocephalia (``margin heads'') and the Ornithopoda (``boring dinosaurs''[2], [3]).
All of the Thyreophora are believed to derive from an animal like Scelidosaurus, a relatively unexciting quadropod with small bony plates embedded in its back. From these unpromising beginnings came two of the more spectacular dinosaur lineages.
These are the plated dinosaurs typified by Stegosaurus itself, the largest of the group, which had a series of eighteen plates arranged along its spine in two alternating rows, together with four spikes on the tail - two on each side. Other members of the group had different configurations of the same basic plan - for example Kentrosaurus had spikes along its spine instead of plates, and various members of the group also had spikes on their shoulders.
These are very heavily armoured dinosaurs with low, wide bodies, and consequently an unusually wide gait. In these animals, the small bony plates of the ancestral forms had developed into a solid, thick covering for the whole of the upper body, and in some cases the lower body too.
Within the Anylosauria are two major sub-groups: the Ankylosauridae have a bony ``club'' on the end of the tail (which they somehow managed nevertheless to carry clear of the ground!), which the Nodosauridae lack.
This grouping is characterised by a shelf (or ``margin'') on the top of the back of the skull, which overhangs the point on connection to the spinal column. This is much more obvious in the Ceratopsia, but is also clearly present in the Pachycephalosauria, along with other morphological indications that they are more closely related to the Ceratopsia than to the Ornithopods, as had previously been thought.
The name means ``horned face'', a reference to the most prominent feature of the more advanced members of the group - the Centrosaurinae and Chasmosaurinae, collectively known as the Ceratopsidae. All the members of this subgroup were quadropedal, and tended to be large.
All of the Ceratopsia share with this subgroup a substantial frill projecting backwards from the top of the back of the skull - an extreme form of the ``margin'' in Marginocephalia. Some of the non-ceratopsid Ceratopsia - such as Protoceratops - were quadropedal; other, more basal, members - such as Psittacosaurus were bipedal.
Of the two major ceratopsian subgroups, the Centrosaurinae are characterised by a large horn on the nose with the horns above the eyes being smaller or completely absent. Centrosaurines have short, deep snouts, and smaller frills than chasmosaurines.
By contrast, the chasmosaurines have larger brow horns than nasal horns, and long, shallow, narrow, snouts. In general, their frills are larger than those of the Centrosaurinae, and tend to have larger openings in the frill bones (although these openings would have been covered in tough skin in life). The exception to this generalisation is Triceratops which, uniquely among the Chasmosaurinae, has a completely solid frill.
The name means thick-headed lizards, which is an appropriate description. These quadropedal animals had thickly reinforced skulls, which must have been effective weapons. Because of their bipedal stance, they were once considered more closely allied to the ornithopods than the ceratopsians, but closer analysis reveals their true affinities.
The other unusual characteristic of the pachycephalosaurs is their very long and wide gut, extending as far back as the base of the tail. Pelvic modifications allowed the gut to pass right though the hip area.
Basal ornithopods include small to medium bipedal forms such as Thescelosaurus and Hypsilophodon. More advanced ornithopods, collectively known as the Iguanodontia, tend to be larger, and facultatively (that is, optionally) quadropedal.
Iguanodon and its relatives are characterised by a hand in which the first finger (thumb) has become a spike - presumably for defense - and the fifth finger is opposable. Iguanodon itself has the dubious distinction of having been restored in the bewildering variety of different stances, from Owen's original fat quadroped, through the first bipedal stance of the Bernissart skeletons and the all-but-vertical posture of Knight's paintings to the current, presumably correct, bipedal stance with the spine held horizontally.
The so-called ``duck-bills'' were very abundant by the late Cretaceous, so we have excellent fossil records including complete growth series for some genera. They are notable for their well-developed chewing mechanisms, including batteries of up to two thousand teeth in some genera.
In the past, the hadrosaurs were considered to be aquatic because their deep tails were considered good for swimming. It's now recognised that the tails were far far too stiff to be used in this way, and that the hadrosaurs were primarily terrestrial.
These hadrosaurs did not in general have crests on their heads. The group includes the original Hadrosaurus, the first dinosaur described from America, and the massive Shantungosaurus - the largest known non-sauropod dinosaur.
Finally we come to the crested duck-bills, which sported a variety of hollow bony crests, now considered most likely to be primarily display structures. (In the past, they were interpreted as snorkels!)
For more detailed and scientifically rigorous classification overview, Thomas Holtz's excellent series of cladograms at www.geol.umd.edu/~tholtz/G104/10417what.htm (and following pages) strike a fine balance between detail and exposition/comprehensibility; and Mike Keesey's dinosauricon website does a great job of tracking current thought in more detail: dinosauricon.com/taxa/index.html
Note 2. OK, I admit that Ornithopoda does not really mean ``boring dinosaurs''. It means ``bird-foot'', so named because ornithopods' feet look like those of ``beasts'', or mammals. No, that can't be right. That should mean that they're called ``theropods''. Oh no, wait - theropods are carnivores, so named because their feet resemble those of birds. Oh dear. I think this is where we came in ... [Back]
Note 3. Someone (well, several people) on the Dinosaur Mailing list objected to my non-literal translation of the term ``Ornithopoda'', and claimed that ducks are much more boring. To which Darren Naish <darren.naish@port.ac.uk> replied:
Errm... the evolution of carpal spurs and knobs, extreme pugnacity and territoriality, nest parasitism, creching behaviour, parental carrying of young both in the water and (!) in the air, monogamous pair-bonding, underwater copulation and the (?)reinvention of the penis, major sexual variation in tracheal structure, grass-eating and 20-minute gut carrying time, niche partitioning according to intestine size, carrion feeding on Subantarctic islands, the evolution of fern-eating, island giantism, island dwarfism, crepuscularity, serrated bill margins, filter feeding with buccal lamellae, deep-diving, species where males are flightless but females flighted, coevolution of browsing forms with spiky lobelioideaens, repeated increases and decreases in body size during phylogeny, the annual transportation of TONNES of sand... and, pant pant pant, quacking[4].Which just goes to show that everything is interesting if you take the time to learn about it. [Back]How *ON EARTH* can ducks be boring?????
Note 4. All that and he forgot to mention hoi-sin sauce and very thin pancakes! [Back]