Dinosaurs, Birds, and Cladistics

Dinosaurs, Birds, and Cladistics

The evolution from theropod to bird, like all such evolutionary sequences (e.g. ungulate to whale) is a fascinating object of study. Yet perhaps one of the hardest things for the layperson to grasp in phylogenetic thinking is the statement "birds are dinosaurs". This is because such a statement requires not one but two cognitive shifts, from old style scaly dinosaurs to updated dinosaurs, and from evolutionary linnaean ancestor-descendant to cladistic sister group relationships. Yet the premise that "birds are dinosaurs" stands at the foundation of the modern revolution in evolutionary thinking as regards dinosaurs and prehistoric reptiles in general, including online popularisations. It is the result of us not only thinking in a different way about birds, but also about dinosaurs (and vice versa).

A big part of the problem is that if you say "dinosaur", people think of this:

"Brontosaurus" by Charles R Knight, via Wikipedia

This evocative painting by Charles R Knight, perhaps the greatest and most influential of all paleo artists, dates from 1897. For its time it was a totally accurate representation of what science knew of the animal called Brontosaurus. Generations of budding paleontologists, paleo artists, and lay enthusiasts (including the present writer) had their imagination captured by this awesome image. The massive bulk and tiny head (and by implication brain) of this massive swamp dweller, its enormous weight bouyed up by water, its slow movements suggested by the image, the hippo like skin with just a hint of scales reflectings its semi-aquatic lifestyle, the short stumpy legs, great arch of the torso and back, long neck, long dragging tail, the small rounded head on the lookout for its main enemy Allosaurus, from whom refuge can be sought by retreating into deeper water (for some reason, during the golden age of dinosaur art and discovery, theropods were considered to be frightened of water), the lush swampy vegetation and steamy primordial forest, all this screams "dinosaur".

Whereas if you say "bird", they think of this:

The usual response, not surprisinglty, is how could anything like the huge Brontosaurus, bigger than a house, evolve into a tiny little bird? (The author recalls many years arguing with a creationist who asked just such a thing.)

Even worse, how can anyone in their right mind call a duck or a sparrow a dinosaur? It looks nothing like the animal in the above painting! Even accepting evolution, it doesn't feel right.

Here we have the problem of what one roleplaying game blogger has very perceptively reffered to as the Dinosaur Lag Effect. Science has moved on, but the general populace still is stuck in images from decades or even a century or more past. From this point of view it can be the [anything] Lag Effect; it's just that dinosaurs make a pretty good case study.

Saltoposuchus from the Yale Mural by Rudolf Zallinger

Let's get back to the early to mid 20th century. At this time, dinosaurs were thought to have evolved from small bipedal reptiles called thecodonts (Heilmann 1926), which looked something like this (right). These small Triassic bipedal reptiles, called pseudosuchians ("false crocodiles", although some have since turned out to be actual stem- and proto-crocodiles after all, and many were neither small nor bipedal) were considered the stem group from which all other archosaurs evolved, as well as birds (which at this time were not referred to as archosaurs, since the latter were still included in the Class Reptilia). But if we look at reconstructions of peudosuchian thecodonts, such as the one on the right, or this one, it is clear that they look much more like birds than Brontosaurus, Triceratops or Tyrannosaurus rex

Then, during the 60s, 70s and 80s, three things happened that changed everything we knew and thought we knew about these prehistoric saurians.

First was the Dinosaur renaissance of the mid 70s onwards. John Ostrom of Yale University, who had, just a few years before, described the newly discovered coelurosaur Deinonychus Ostrom 1969), came to realise that it was actually very similar to early birds such as Archaeopteryx (Ostrom 1973, 1975, 1976). Ostrom's charismatic student Bob Bakker popularised and further developed the dinosaur bird link, emphasising dinosaurs as intelligent, active, warm-blooded animals (Bakker 1975, 1986). And Bakker's student Greg Paul not only further popularise these new discoveries and insights with beautiful artwork and field guides (Paul 1988, 2002 and 2010). This is the second cognitive shift referred to at the start of this essay, dinosaurs are not dim-witted behemoths (nor were they necessarily scaly, but we'll come back to that in a minute). Incidentally, Greg Paul is also the author of the radical suggestion (of enormous appeal to the lay paleo enthusiast, but mostly frowned upon by serious researchers) that advanced coelurosaurs, including Ostrom's Deinonychus were actually neo-flightless birds descended from Archaeopteryx, rather than the other way around. But that's a whole other story.

The second great revolution, simultaneous with the dinosaur renaissance, was the development of cladistics, as a means of empirically testing hypotheses regarding evolutionary relationships (called phylogeny), shown in branching tree-like diagrams called cladograms (the Palaeos and other interent (e.g. Wikipedia) dendrograms are popularised and informal versions therof). Cladistics began in the 1960s with the work of German entomologist Willi Hennig (1966). About two decades later it was applied to the evolution of reptiles and birds by vertebrate paleontologist Jacques Gauthier (Gauthier 1984, 1986). Using cladistics, Gauthier demonstrated that birds were a derived (i.e. an advanced or specialised) group of theropod dinosaurs, with the dinosaurs themselves representing a series of progressively more birdlike groups - called clades - (Theropoda, Tetanurae, Maniraptora...) with the dromaeosaurids (Deinonychus and co) as birds' nearest relatives. During this early period of cladistics, phylogeny was determined on the basis of a few obvious shared characteristics (called synapomorphies) organised in the most parsimonious manner. The following diagram by Carl Zimmer (from Zimmer 2009), maps out these various evolutionary stages, from ancestral theropod dinosaur to modern bird, including the main synapomorphies and timescale (note that actual cladograms, which are concerned with unique characteristics only, do not use timescales), and illustrating the changing structure of the bones of the hand, as the hand changes from locomotory organ (in quadrapedal walking), to grasping organ, to part of the wing (the blog page the diagram link points to includes useful discussion on definitions of dinosaurs, birds, etc, cladistics and so on).

Dendrogram showing the evolution of dinosaurs into birds by Carl Zimmer

Cladistics marked a radical change from the earlier evolutionarily systematics in one important way. It is no longer about which species or group evolves into which (ancestor-descendent relationship), but rather which species shares a common ancestor with which other species (sibling, or sister-group, relationship; this is the second cognitive shift referred to earlier). All species that share the same ancestor belong to the same clade, the same family group. Because birds share a common ancestor with different types of dinosaurs, they are included in the dinosaur clade. Therefore, cladistically speaking, birds are dinosaurs. This is illustrated in the above diagram where theropod dinosaurs are divided into various subgroups (clades), including birds. From this perspective of hierarchical categories we say that birds (bottom two clades) are maniraptors (includes the next two up - dromaeasaurs and oviraptorosaurs), maniraptors are coelurosaurs (includes the next two up - tyrannosaurs and compsognathids), coelurosaurs are theropods (includes the next three up - allosaurs, coelophysids (= podokesaurs) and Eoraptor), and theropods are dinosaurs.

By the 1990s, cheaper computing meant that cladistics was able to move from focusing only on a few personally chosen synapomorphies to statistical analysis of hundreds of anatomical traits, some quite minor and even arbitrary from dozens of species. As computer power increased further, more character traits could be keyed in and more species analysed in huge supermatrixes. Now we are finding that the previous parsimony-based algorithms are giving way to the more reliable but also more computationally intensive, maximum-likelihood statistical technique. But all of these analyses confirmed the dinosaurian origin of birds, and therefore retained birds as a subgroup of theropod dinosaurs (Senter 2007, Lee & Worthy 2011)

The third great discovery are fossilised feathers found in beautifully preserved dinosaurs found in China. The development of feathers, from simple filamentous structure sto proper flight feathers, matches both the cladistic series of increasingly more birdlike dinosaurs, and the evolutionary developmental biology of how feathers form in birds (contray to earlier theories, feathers are not modifoied scales) (Prum & Brush 2002, Prum & Brush 2003)

As a result, we no longer interpret dinosaurs as lumbering reptiles, but rather as agile proto-birds.

Even iconic Brontosaurus has changed in every way. It's now called Apatosaurus (due to a quirk of nomenclature), it's fully terrestrial like elephants, its legs are long and strong, its body lighter, because of avian-type air sacs distributed through the bones and soft tissue, and while still massive it is now more graceful in appearance, and its tail no longer dragged like the tail of a big lizard, but was held erect, waving in the air and snapping its tip at supersonic speeds. It had a complex social behaviour and lived in herds, and once it reached adult size and strength no theropod would be foolish enough to attack (so it didn't need to retreat to water, and anway, theropods could swim, and some were semi-aquatic). Even its head is different (for a long time, Brontosaurus/Apatosaurus had a Camarasaurus skull in place of its own, more pointed, Diplodocus-like head)

Brief mention should be made here of paleornithologist Alan Feduccia, one of the very few scientists who rejects a dinosaur origin for birds, as does fellow paleoornithologist Larry Martin, and argues instead that birds originate from small arboreal archosaurs in the Triassic (essentially, a modified version of the old thecodont origin theory). Cladistic evidence does not support the neo-flightless hypothesis (see e.g. Dinosaurs and Birds - an Update by Kevin Padian), nor have any plausible alternatives to the theropod origin been found. Feduccia cites Protoavis texensis, described by Sankar Chatterjee of Texas Tech University, allegedly an animal more advanced than Archaeopteryx, although it existed during the Norian (Chaterjee himself argues for the theropod origin of birds). Current consensus is that Protoavis is a chimera, a jumble of several different species. Exactly which types of animals is unclear, because of the poor preservation, but podokesaur ("ceratosaur") limbs and drepanosaur head and neck have been suggested (Wikipedia).

Nevertheless, while the theropod ancestry of birds is as solid as any scientific hypothesis could be, the semantic problem of saying "birds are dinosaurs" remains, even allowing for the new model Apatosaurus, Deinonychus, and co. Although cladistics has so swept the field that it is not possible to find a current reference to an ancestor descendant phylogeny that shows dinosaurs evolving into birds (i.e. because birds are dinosaurs, dinosaurs cannot evolve into birds, any more than mammals can evolve into bats, this doesn't mean that rival paradigms such as evolutionary systematics are incorrect. Simply that they have not been found as easily verifiable as cladistics. At the same time, a hierarchy of clades of sister-group relationships is not itself a phylogeny. It is a means of building hypothesis that can be applie dto phylogeny. Phylogeny itself involves the dynamic (sensu Heraclitus, flux and flow) transformation of species in deep time with one evolutionary stage being replaced by the next (Cladistics of course is also all about evolution, but it doesn't map evolution in this particular way). It is this difference between gradism and cladism that results in so much confusion, because gradism uses a stratigraphic (sequence according to the fossil record) deep time evolutionary ancestor-descendant (anagenetic) approach (birds evolved from dinosaurs), and cladism a hypothesis-testing sister group (cladogenetic) approach (birds are dinosaurs). For the present author (MAK), cladism and gradism, hypothesis building and phylogeny, are complementary. It is not that one is wrong and one is right, when actually they are both correct. And doubtless phylogenetic insights and hypothesis building will continue to develop and incorporat enew insighst and methodologies (for example cladism considers stratigraphy irrelevant, the exception being stratocladistics, which strangely very few standrad cladists seem to be interested in).

Summing up, we can confidently state that the hypothesis that birds evolved as (cladism) and from (gradism) theropod dinosaurs, through a number of transitional stages such as those described in the above diagram, is the only hypothesis that fits all the evidence: paleontological, stratigraphic, morphological, physiological, embryological, molecular, and more. Through discoveries such as these, we painstakingly yet confidently map out the evolutionary tree of life on Earth.