Here's why the brontosaurus got its dinosaur status back

Finally, the Portuguese diplodocid Dinheirosaurus lourinhanensis is found to belong to the same diplodocine clade as Australodocus from Tanzania and Supersaurus from the Morrison Formation. Tschopp et al. (2015) retain Australodocus as a separate taxon, but they opt to sink Dinheirosaurus into Supersaurus, meaning that this taxon is no longer exclusive to the USA but present in Portugal as well.

Several Jurassic dinosaurs first described from the Morrison Formation are now known from Portugal (including Stegosaurus, Ceratosaurus, Allosaurus and Torvosaurus), so this result isn't tremendously surprising.

So, these are the well known, much-discussed conclusions of this study. However, there are a few associated issues haven't been much discussed outside the technical dinosaur community, and I'd like to cover those things here. Some are pretty exciting.

There are (probably) yet more 'new' Morrison diplodocids to come

The first thing to say is that several diplodocid specimens occupy novel positions in Tschopp et al.'s (2015) cladograms and don't group within any named 'genus-level' clade (like Apatosaurus or Brontosaurus). Among these 'floating' specimens is AMNH 460… yes, the goddam 'Apatosaurus' on display in the American Museum of Natural History.

Another is FMNH 25112, Elmer Riggs's Field Museum 'Apatosaurus', previously intimated to be a distinct (as yet unnamed) species by Upchurch et al. (2004a). Not only do Tschopp et al. (2015) find 25112 to fall well away from Apatosaurus, they also find it to be a diplodocine in at least some of their results.

It should also be noted that a few diplodocine specimens (example: AMNH 7535) are found to be attached to named taxa (in this specific case, to Barosaurus), but are of ambiguous position as goes their species status, and more study (and more data) is needed before we definitely know what to do with them.

The clear conclusion from these results is that at least some of these 'floating' specimens are not just indeterminate specimens that can't be identified, but additional taxa that need new names. What, seriously… more Morrison diplodocids?

An oft-made complaint uttered whenever people start talking about yet more new Morrison sauropods can be paraphrased as "stop, stop, there are too many goddam Morrison sauropods already! The world can't take any more!!"

While, of course, there simply has to be some theoretical carrying limit for a 'chronofauna' just as there is for an extant ecosystem (and while there's good data indicating that many named species will turn out to be synonymous with others in time), we should always be sceptical of the implication that there are "too many sauropods", or even "too many dinosaurs".

(1) These complaints are virtually always wholly intuitive, and (2) they imply that we have a good handle on the carrying capacity and ecological dynamics of the 'chronofaunas' concerned. We really don't.

There aren't"too many dinosaurs" in view of the fact that we're talking about animal populations spread over millions of years of geological time. And, if your ideas about how many dinosaurs there 'should be' are in part inspired by megafaunal diversity in the modern world… well, you do know that the modern world is wholly atypical as goes megafaunal diversity, right?

Again, 'ontogenetic morphing' doesn't stack up

The second issue I want to highlight concerns recent suggestions that diplodocids and other diplodocoids and sauropods should be re-evaluated as the ontogenetic morphs of a lesser number of 'phenotypically flexible' taxa.

Anyone who knows anything about dinosaurs will be familiar with proposals that tyrannosaurids, pachycephalosaurs and ceratopsids underwent extreme 'ontogenetic morphing' – this is the seeming raison d'être of the Horner lab these days and it's generated an extraordinary amount of interest. Less well known is that there has been some push to extend this model to other Mesozoic dinosaurs as well.

A couple of years ago, Woodruff & Fowler (2012) proposed that the Morrison Formation sauropods Suuwassea, Haplocanthosaurus and Barosaurus are – like Torosaurus, Nanotyrannus and Dracorex– 'suspicious' taxa that might well deserve to be sunk into synonymy.

Additionally, they even implied that rebbachisaurids might be ontogenetic morphs of dicraeosaurids. I think these proposals were first aired at the Bonn meeting on sauropod biology, held in 2011, and they caused a reasonable amount of consternation at that event. A detailed evaluation of these claims by Wedel & Taylor (2013) found them wanting: Suuwassea, Haplocanthosaurus and Barosaurus differ from other taxa profoundly, in ways that just can't be interpreted as the result of ontogeny.

Is the Tschopp et al. (2015) study relevant to this debate?

Yes, because both their extraordinary cataloguing of new specimens and their identification of so many anatomical features provides strong additional support for the 'conventional' view of diplodocoid phylogeny and diversity (where dicraeosaurids and rebbachisaurids are successively more distantly related to diplodocids, where Haplocanthosaurus is nothing much to do with Flagellicaudata, and so on). (Flagellicaudata = the diplodocoid clade that includes the whip-tailed groups Dicraeosauridae and Diplodocidae.)

Furthermore, Tschopp et al. (2015) included over 10 diplodocid specimens that seem to be juveniles or subadults and (drumroll) found them to group with the adults of the taxa they seem to belong to. In other words, it seems that ontogenetic variation has little effect on phylogenetic position in these animals… or, in this analysis.

This is a very interesting result, because it actually conflicts with work on some non-bird dinosaurs where exactly the opposite has been asserted: that is, there is work on non-bird dinosaurs showing that ontogeny does have a significant influence on phylogenetic position, such that juveniles thrown into datasets tend to come out as 'basal' members of the respective lineage (Campione et al. 2013).

Why this conflict exists between Tschopp et al.'s (2015) conclusions and Campione et al.'s (2013) conclusions is not yet clear. Maybe patterns are simply different for different groups: maybe baby diplodocids always possess characters specific to their particular lineage, whereas baby other things – hadrosaurs, say – are more 'generic' and don't develop such features until later in ontogeny. Interesting.

We need to talk about taxonomic subjectivity

Finally, the one thing everyone knows about this study is its argument that Brontosaurus should be resurrected as a valid name. As the authors show, Brontosaurus and Apatosaurus are notably different in numerous characters, so much so that they belong to different branches of the tree.

As is so often the case, this conclusion is not a total surprise for specialists – people have been saying for a while that 'Brontosaurus-type' apatosaurines do look different from 'Apatosaurus-type' ones.

What hasn't been mentioned all that much is that B. excelsus (the type species of Brontosaurus) isn't a singleton in its lineage. On the contrary, the Brontosaurus clade includes quite a few specimens, including the holotypes of Eobrontosaurus yahnahpin (That species name is supposedly pronounced 'wha-na-pee'. It refers to a ceremonial necklace worn by the Sioux people.) and Elosaurus parvus, both of which Tschopp et al. (2015) lump into Brontosaurus.

I never did buy the idea that Eobrontosaurus is not a diplodocid, but a misidentified Camarasaurus specimen (Upchurch et al. (2004b), so it's good to see this. But why lump these animals together? Why not retain Eobrontosaurus and Elosaurus as separate 'genera'?

After all, Amphicoelias altus is also a member of the Brontosaurus clade, and yet gets to retain its genus-level name. Surely this is all subjective, and surely the view of multiple discrete genus-level names is more consistent with the rest of Mesozoic dinosaur taxonomy?

Actually, this issue is specifically addressed in the study: Tschopp et al. (2015) do their best to be quantitative about their taxonomic decisions, using counts of autapomorphies as a measure of where the lines might be drawn between 'genera' (autapomorphies are anatomical features unique to a given taxon).

This seems pretty sensible, and it might be interesting to look at phylogenetic studies of other dinosaur and animal groups to see if this concept might have any sort of consistency.

The caveat, of course, is that any decisions based on this method are provisional… views on taxonomic distance will (potentially) change as more data comes in.

But that's the whole point about taxonomy, isn't it. So long as binomial names are not just labels, but hypotheses about the affinities of an organism, we're stuck with a system whereby those names are liable to modification. Brontosaurus? Yeah, sure, it's good. For now.

There's a lot more to say about diplodocid diversity and phylogeny, of course. What of Amphicoelias 'brontodiplodocus'? What of the recent claim that Amphicoelias fragillimus… never… existed? Another time, another time.

Be sure to read the PeerJ interview with the authors. Also, check out Matt Wedel's excellent article on this study at SVPOW!