There are more than 10,000 species of living, breathing dinosaurs on Earth today. It’s just that we call them birds. And while a chicken might seem like a measly ancestor for the enormous T-rex, modern birds can teach us a lot about dinosaur evolution. A huge genome sequencing project, which recently culminated in the publication of nearly 50 genome sequences and the most accurate tree of bird evolution to date, has further blurred the line between bird and dinosaur. Spurring a plethora of studies into the origins of our modern feathered, singing friends, the December 2014 edition of Science taught us that the transition from dinosaurs to birds was gradual and began long before the Dinosaurs were gone. It taught us that it involved multiple independent origins of bird song, a characteristic that now dominates around 10% of the genome. And it taught us that the evolution of flight was facilitated by new genes and new gene regulation, but also by the loss of genes.

Making a Chicken from a T-Rex

Last October, the Therapod Working Group constructed a new phylogeny (family tree) for Dinosaurs, based on morphological characteristics measured in fossil remains of over 150 different species. The new tree revealed fascinating insights into the nature and pace of avian evolution.

The phylogeny showed much more rapid evolution in early birds than other dinosaurs, consistent with previous studies that revealed over 150 million years of ecological innovation in the early ancestors of birds. However, the study also revealed a much slower accumulation of ‘bird-like’ traits in dinosaur lineages prior to this rapid evolution, which the authors claim was essential for birds to flourish in the wake the dinosaurs.

In order to investigate the speed and direction of evolution during the emergence of modern birds, researchers the University of Edinburgh and the American Museum of Natural History mapped anatomical features of living and fossil birds and dinosaurs, such as feathers, wings, wishbones and bills, onto the newly resolved phylogeny. This showed that the appearance of birds 150 million years ago was preceded by nearly 100 million years of gradual evolution accumulating avian traits in dinosaur lineages, followed by one final, explosive evolutionary radiation around 65 million years ago that produced most of the modern bird species we see today.

This pattern is consistent with the theory of punctuated equilibrium, which proposes that evolution proceeds very slowly, but is punctuated by short bursts of rapid evolution as a new innovation or major environmental change facilitates adaptation to new niches.

Meanwhile, a consortium of more than 30 researchers, across 20 different universities, had set about the most comprehensive study of avian genomics ever attempted. Their goal – to sequence at least one genome from every avian order. This was no small task; scientists have recently proposed 35 orders of birds. But the consortium succeeded in sequencing 48 different species spanning 32 of those orders (that’s 92%).

Avian Diversity

Using the wealth of sequence data generated by the consortium, Jarvis and colleagues at the Howard Hughes Medical Institute and China National GeneBank were able to construct the most accurate family tree for the birds (and their dinosaur ancestors) to date. This phylogeny revealed convergent evolution of Flamingos, Grebes and waterfowl with core water birds such as pelicans, cranes and penguins, and suggests the ancestor of modern land birds was probably a bird of prey. Further, it confirmed a rapid period of evolutionary change in birds shortly after the extinction of the dinosaurs, with 36 new lineages evolving in just 10 – 15 million years (that’s pocket change, in evolutionary terms). Thus, the new avian phylogeny appears to agree with Brusatte’s dinosaur phylogeny, with generally slow evolution of dinosaurs into birds followed by an explosion of new species around 65 million years ago.

The authors note that, despite the wealth of genetic sequence data used, the earliest branches of the tree leading to modern birds were difficult to resolve (put together), and blame convergent sequence evolution during the rapid period of evolution immediately following the extinction of the dinosaurs 66 million years ago.

Genomic Features

The new genome sequences confirmed that birds have unusually small and fragmented genomes compared to other vertebrates, a characteristic that the authors suggest might be due to the erosion of junk DNA, as well as loss of functional genes. However, they also found that bird genes tended to be very short. In fact, protein-coding genes in birds are on average 30 – 50% shorter than genes in reptiles and mammals. Birds achieve this by having denser genes. Genes contain ‘introns’, which don’t directly code for a protein, but are important in gene regulation, and a process called alternative splicing which allows one gene to code for multiple related proteins. Birds tend to have smaller introns and shorter regions between their genes.

Interestingly, one mammal group shows a very similar trend – bats. Therefore it seems likely that short, dense genes and smaller regulatory regions may be an adaptation to flying – Zhang and colleagues suggest it may facilitate more efficient gene regulation, necessary to meet the power demands of flight.

The Power to Fly

Powered flight (that’s the flapping kind, rather than the gliding kind seen in some insects and the occasional lizard or squirrel) needs a lot of adaptations – a strong, lightweight skeleton, a fast metabolism, and feathers, to name just a few. Zhang looked at evolution in specific genes related to these functions and found evidence for gene sequence changes, as well the evolution of new genes and the loss of old ones. For example, over half of genes involved in bone ossification showed evidence of positive selection (twice as many as in mammals), adapting the avian skeleton to flight.

In order to provide enough oxygen to power that fast metabolism, birds had to evolve different lungs. Lungs in birds are volume-constant, enabling them to absorb oxygen and release carbon dioxide more efficiency. The evolution of a new ventilation system left some redundant genes behind, and Zhang found that five genes known to be important in mammalian lung development have been lost in birds.

Finally, feathers are important for flight aerodynamics (as well as a host of other things from insulation to mating), and in order to achieve these roles bird feathers have utilised different proteins to construct them. Feathers, like fur, hair and nails, are made of keratin. But feathers are composed of a mixture of α- and β-keratins, whereas mammalian fur is just α-keratin (reptiles and birds use β-keratin, but as a structural protein under the skin). Concordant with all this, Zhang showed that the β-keratin gene family has grown considerably in birds, while the α-keratin gene family has shrunk. However, these changes happened early in bird evolution – the β-keratin genes are shared across all modern birds. Colour, as well as structure, is important when it comes to feathers – colourful plumage attracts mates and provides camouflage. Genes thought to be involved in plumage colour in birds show rapid evolution in 8 of the bird lineages analysed.

Toothless

Another major change in the emergence of birds was the replacement of teeth with beaks. This transition has proved difficult to resolve in the past, with some scientists arguing the loss occurred once in the ancestor of all birds and others claiming there were multiple independent losses during avian evolution. Meredith and colleagues from Montclaire State University and the University of California showed that toothlessness, also known as edentulism, most likely evolved just once, simmulataneously with the development of the beak. This loss likely occurred around 115 million years ago, in one of the earliest ancestors of modern birds.

Learning to Sing

Bird song, and vocal learning in particular, has received intense attention from scientists because of it’s parallels to human speech. Vocal learning is found in three groups of birds – the songbirds, parrots and hummingbirds. Jarvis’s molecular phylogeny of the birds suggested two possible evolutionary trajectories for vocal learning of bird song – it may have evolved independently three separate times, or it may have evolved twice and been subsequently lost in two lineages.

The emergence of vocal learning, as with any new trait, came with a host of genetic sequence changes. Zhang and colleagues found that 227 genes (2%) showed accelerated sequence evolution for vocal-learning bird species but not for non-learners. Most (73%) of these genes were expressed in the songbird brain, and around 20% of these have been shown to be involved in singing. These genes were mostly involved in neural development, connectivity and metabolism.

Other members of the consortium went on to investigate bird song further. Whitney and colleagues investigated brain gene expression patterns in singing zebra finches, and found that around 10% of the genome is differentially expressed in relation to bird song. Almost all of these genes showed different patterns of expression in the four major song nuclei, giving each nuclei a distinct profile of expressed genes. The authors found a cascade of regulatory genes triggered by the onset of singing, which in turn stimulated differential expression of protein-coding genes in different brain regions which production of extremely complex songs.

And a study by Pfenning revealed remarkable similarity between human speech and vocal learning in birds. They compared gene expression patterns in the brains of humans with songbirds, parrots and hummingbirds, and non-song-learning birds such as the dove and quail. And they found similar patterns of gene expression in genes controlling motor functions and brain connectivity in humans and birds that sing, which were not shared with non-singing birds.

This suggests that through multiple convergent evolutionary appearances of vocalisation, be it speech in humans or song in birds, the same genes have been co-opted to perform similar roles. Evolution often reuses the same genes for similar functions; Hox genes, which determine the overall body plan during development, and are shared across almost all living things, are a classic example of this, however a set of toolkit genes to control certain characteristics has also been proposed for social behaviour in insects, and this phenomenon may be even more widespread.

Across more than twenty new papers published in the last few months of 2014, we gained an enormous amount of information about how birds evolved from dinosaurs and radiated to produce the huge variety of birds we see today, from chickens to penguins. Further, these studies reveal deeper insights into the process of evolution itself, from finding support for the concept of punctuated equilibrium, where slow change is interspersed with sporadic periods of rapid evolution, to highlighting new examples of convergent gene evolution and revealing the importance of new genes, lost genes and gene regulation in the emergence of new traits such as flight. And the new genome sequences we now have still, no doubt, have many more stories to tell.

Want to Know More?

• Zhang, Jarvis & Gilbert (2014) A flock of genomes Science
• Zhang et al (2014) Comparative genomics reveals insights into avian genome evolution and adaptation Science
• Jarvis et al (2014) Whole-genome analyses resolve early branches in the tree of life of modern birds Science
• Pfenning et al (2014) Convergent transcriptional specializations in the brains of humans and song-learning birds Science
• Brusatte, Lloyd, Wang & Norell (2014) Gradual Assembly of Avian Body Plan Culminated in Rapid Rates of Evolution across the Dinosaur-Bird Transition Current Biology
• Ksepka (2014) Evolution: A Rapid Flight towards Birds Current Biology
• Benson et al (2014) Rates of Dinosaur Body Mass Evolution Indicate 170 Million Years of Sustained Ecological Innovation on the Avian Stem Lineage PLOS Biology
• Meredith et al (2014) Evidence for a single loss of mineralized teeth in the common avian ancestor Science