Di­nosaur fam­ily tree might have to be to­tally re­drawn


NOR­MALLY the di­nosaurian world is rocked by a new fos­sil – the biggest, fastest, or tooth­iest. But the lat­est di­nosaur re­search threat­ens to change our un­der­stand­ing of how di­nosaurs evolved at a much deeper level, and blow aside 130 years of agree­ment on the topic.

A new paper pub­lished in the journal Na­ture sug­gests that sci­en­tists need to re­or­gan­ise the ma­jor groups used to clas­sify di­nosaurs. This means we may have to re­visit what we think we know about the first di­nosaurs, what fea­tures they evolved first, and where in the world they came from.

The way we clas­sify di­nosaurs goes back to the 19th cen­tury. In 1887, Harry Govier See­ley, a clas­sic, hard-work­ing Vic­to­rian palaeon­tol­o­gist, di­vided di­nosaurs into two ma­jor sub­or­ders based pri­mar­ily on their hip struc­ture. Sau­rischia com­prises the flesh-eat­ing theropods such as Tyran­nosaurus and the pon­der­ous, long-necked sauropodomorphs such as Di­plodocus. Or­nithis­chia com­prises all the rest, in­clud­ing the two-legged Iguan­odon, and the ar­moured, four-legged Stegosaurus, Tricer­atops, and Anky­losaurus.

This or­der­ing of di­nosaurs has stood the test of time for 130 years, weather­ing the on­slaught of cladis­tics in the 1980s, when palaeon­tol­o­gists be­gan us­ing com­put­ers to an­a­lyse and cat­e­gorise groups of an­i­mals based on fea­tures that pointed to a com­mon an­ces­tor. There are now thou­sands of di­a­grams (clado­grams) of di­nosaur sub­groups, and ever-grow­ing data ma­tri­ces, that closely doc­u­ment the anatom­i­cal fea­tures of each species.

The new paper com­pletely dis­rupts the con­sen­sus over See­ley’s cat­e­gories. The re­searchers ran a cladis­tic anal­y­sis of 457 char­ac­ter­is­tics across 74 species (that is a data ma­trix of 33 818 bits of in­for­ma­tion recorded from skele­tons). They con­cluded that, based on 21 unique char­ac­ter­is­tics of the fos­sils, the theropods were more closely re­lated to the Or­nithis­chia group and should be moved into that cat­e­gory. This would cre­ate a new group named Or­nithoscel­ida and leave be­hind the Sau­ropodomor­pha.

The trick in cladis­tics is to find a unique anatom­i­cal fea­ture that evolved at a spe­cific time and can in­di­cate a par­tic­u­lar sub­group. For ex­am­ple, See­ley noted that the hip bones of or­nithis­chi­ans were ar­ranged with pu­bis and is­chium run­ning back­wards (su­per­fi­cially, like mod­ern-day birds).

This sug­gests the two groups split from a com­mon an­ces­tor and evolved dif­fer­ent hip shapes.

This was a mas­sive anatom­i­cal change or nov­elty, and palaeon­tol­o­gists un­til now have as­sumed that it hap­pened only once in evo­lu­tion­ary his­tory.

Group­ing the theropods with the or­nithis­chi­ans sug­gests that the hip change oc­curred later and raises the ques­tion of whether some early theropods had this fea­ture.

The re­searchers also sug­gest that the new anal­y­sis can re­set our un­der­stand­ing of where di­nosaurs orig­i­nated and what their diet was. The clas­sic view was that the first di­nosaur was a car­ni­vore liv­ing in what is now South Amer­ica. The new anal­y­sis makes this more of an open ques­tion and sug­gests they might have evolved as om­ni­vores in the north­ern hemi­sphere.

Tree of life

None of this changes what we know for sure about what di­nosaurs evolved which traits and when. But the key point is that ac­cu­rately de­pict­ing the tree of life mat­ters. If you care about mod­ern bio­di­ver­sity, it’s im­por­tant that all species are not equal.

Some are more dis­tinc­tive than oth­ers, pos­sess­ing more unique fea­tures, and hav­ing a longer in­de­pen­dent his­tory. Work­ing this out re­quires an ac­cu­rate tree.

On a broader scale, get­ting the tree right af­fects our cal­cu­la­tions of rates of trait evo­lu­tion, ex­tinc­tion and post-ex­tinc­tion re­cov­er­ies.

We will never find the very first di­nosaur but we can es­tab­lish some things about it by es­ti­mat­ing the an­ces­tral states of dif­fer­ent species from a cor­rect tree.

We in­vest enor­mous ef­forts into con­struct­ing testable sys­tems for cat­e­goris­ing dif­fer­ent species, and their size is in­creas­ing as com­put­ing power grows.

When I ran my first clado­gram in 1982, I had to use punch cards on a main­frame com­puter, and I could in­clude only 10 or 12 species and 50 or so char­ac­ter­is­tics.

To­day, I was able to run all the data for this new paper through my desk­top com­puter and get an an­swer in 33.21 sec­onds, while writ­ing this ar­ti­cle at the same time.

Re­cent pub­li­ca­tions have sported trees of all 10 000 species of birds, and even sum­mary trees of all life.

The dream is to run such trees with all 1.5m named species, us­ing data about both genes and phys­i­cal shape.

Is this new paper the true an­swer for the evo­lu­tion­ary ori­gins of di­nosaurs?

The data we have is rid­dled with ques­tion marks, and so the al­go­rithms still strug­gle to cal­cu­late the one true tree.

This is no crit­i­cism of the re­searchers, just a state­ment of the prac­ti­cal­i­ties.

We don’t know yet whether we can see the wood for the trees.

New re­search threat­ens to change our un­der­stand­ing of how di­nosaurs evolved.

Newspapers in English

Newspapers from South Africa

© PressReader. All rights reserved.