Monday, 27 August 2012

The Dragon Phylogeny - Introduction



The Dragon Phylogeny is an effort to map hypothetical evolutionary relationships among dragons. It’s true that these are mythical creatures, but the detail of their portrayal in historical art and literature makes them amenable to scientific analyses based on morphological variation.

This project began in the summer of 2009 as a diversion activity while writing my Ph. D. dissertation in Ecology and Evolutionary Biology at the University of Toronto, Canada. I collected images of 76 paintings, sketches, prints, and sculptures of dragons dating from the dark ages to the earliest 20th century, using images available on the internet. The most notable of these include prints of Asian dragons by Totoya Hokkei and Yashima Gakutei, Muslim depictions in Firdawsi’s Shahnama, a sketch by Leonardo Da Vinci, and a diversity of dragon species evident in various works featuring St. George’s dragon, the most famous of which are probably those by the Renaissance master Raphael.

What is a phylogeny? A phylogeny is a graphical representation of the hypothetical evolutionary relationships among different taxanomic groups, such as species, genera, families, etc. Most phylogenies today are based on DNA sequences, but dragon tissue is tough to acquire, and I think it’s also illegal under CITES or something. I could be wrong on that. Anyway, instead of DNA, I used the ‘old school’ approach that was used before the days of the polymerase chain reaction. It uses morphological instead of genetic variation.

How to build an ‘old school’ phylogeny: I identified 27 distinct traits that differentiated the 76 dragon art pieces (hereafter ‘species’) used in the study. These traits were then encoded into numbers so that they could be analyzed using a ‘neighbour-joining cluster analysis’. Just as the nucleotides of DNA (A,T,G and C) can be encoded as a single number (0-3), traits like skin texture, wing structure, and number of appendages can be numerically encoded. The neighbour-joining cluster analysis is a mathematical way of measuring the differences between the numbers representing the traits of interest, so that species with the fewest differences are joined together by the shortest branch distances. 

How to read a phylogeny: The output is sort of like a family tree, except that the tip of every branch represents a species and the lines joining the tips show how similar the species are to each other. The assumption is that species with more similar traits share a more recent common ancestor. The spatial arrangement of the tips doesn’t really matter – that’s why phylogenies can look like spirals, circles, or triangles. Instead, the length of the lines joining two species represents how distantly related they are. For example, in the phylogeny of mammals, the lines connecting humans and chimps are very short compared to the lines connecting humans and mice, because we share a common ancestor with chimps much more recently than we do with mice.

Key insights from the Dragon Phylogeny: The most obvious result is the evolutionary distance between the group that I’ve called Orientalia vs. the Eurasian dragons. Orientalia is an entire taxonomic class containing the Asian dragons or (a.k.a. lóng). If we follow the lines connecting all of the Orientalia species we see that they are more closely related to Mammalia than they are to the dragons found in Eurasia. This shows a deep evolutionary divergence with geography as the Asian lóng of the Eastern Hemisphere are biologically no more related to Eurasian dragons of the west than are unicorns and other mammals. It was therefore probably a mistake to translate ‘lóng’ as ‘dragon’, or at least there is no biological justification for this translation. If only those early western scholars had understood evolutionary phylogenetics as well as you do now.

To get to the common ancestor of Eurasian and Oriental dragons, we have to go way back to the time of the fish-like ancestors. If we follow the phylogeny down from Orientalia towards the center of the spiral we see three main branches. The first are the ancestral Actinopterygii, which include modern fish and are a distinct group from all of the land vertebrates. The second branch moving clockwise includes both mammals and lóng, indicating that they share the same tetrapod fish-like ancestor – something like the modern-day Coelacanth. In contrast, the Eurasian dragons must have shared an as yet unidentified hexapod fish-like ancestor, which gave rise to dragons with six appendages instead of four (i.e. four legs + two wings) – the Dracopteronidae and the Dracoverisidae. This assumes that complex appendages like wings or legs did not evolve out of nothing, which is probably a fair assumption. Continuing clockwise along the spiral we can see that the Wyvernidae gradually lost two legs and then two more legs in the Serpentidae ancestors, before losing wings in the more recently derived Serpentidae species at the top of the spiral. 

Conclusion: I know it’s a bit complicated but if you understand how to interpret these relationships you are already a step ahead of most first-year biology college students. Boom! you just learned all about phylogenetics.

This t-shirt design depicting the Dragon Phylogeny was submitted to threadless.com:
You can see it here: Dragon Phylogeny









NOV 15 UPDATE:
Dragon Phylogeny T-shirt now available at Threadless.com!
http://beta.threadless.com/product/4313/Dragon_Phylogeny/style,design  

3 comments:

  1. This comment has been removed by the author.

    ReplyDelete
  2. I love this, Though I disagree concerning oriental dragons. Asian dragons are nearly identical to European wyrms. The only major differences being a lack of wings. Some wyrms are portrayed with four legs and wings, suggesting they either evolved into or from flying hexapods. Thus the asian lung could easily have evolved from a a strain of wyrms that lost their wings and retained their aquatic serpentine shape.

    ReplyDelete
  3. What about dinosaurs and marine reptiles? Did you include them in your analyses :)?

    ReplyDelete