[Date Prev][Date Next][Thread Prev][Thread Next][Date Index][Thread Index][Subject Index][Author Index]

Re: Cope's Rule in Dinos

----- Original Message -----
From: "T. Michael Keesey" <keesey@gmail.com>
Sent: Friday, January 27, 2006 10:51 PM

Forwarded comments from David Hone on Hone et al. 2005 (Macroevolutionary
trends in the Dinosauria: Cope's rule).

I apologize for the delay (which is even greater than you can see -- I had been sent this e-mail offlist several days before). I've been working too hard.

---------- Forwarded message ----------
From: David Hone <d.hone@lrz.uni-muenchen.de>

Just a few short (ish) rejoinders to some of David's comments. These
have been truncated (partly for space) so I would advise would be
readers to read all his comments in full first:

[Keesey - here is the original post: http://dml.cmnh.org/2006Jan/msg00317.html ]

David Marjanovic wrote:
IMHO (i) would not be the operation of any "rule" but the inevitable
result of body size diversification restricted by a minimum
size near which the ancestor is. So Cope's rule seems to be
defined as "everything that, however superficially, looks
like Cope's rule".

Yep! This one was overstated a little. In fact it should be termed Cope's Rule sensu lato (a soft interpretation). This will be corrected in a paper I and colleagues have in review on a similar study on Mesozoic birds. Therefore, dinosaurs as a whole, can be seen to be diverging rather than just increasing.

I see.

As stated, some lineages DO show
Cope's Rule sensu stricto, notably the tyrannosaurs,

Looks like it.

(although the
discovery of new, small tyrannosaurs may have 'broken' that one too).

You mean if *Nanotyrannus* is real?

However, given the (mostly) huge difference in sizes I'm not sure this
make any real difference to the outcome. After all, for the purposes of
our comparisons, its really irrelevant if a sauropod is estimated at 15
or 30m in length if we are comparing it to one that is only 10m. You
will still get a very positive result and a correct interpretation of
how those organisms are evolving.


While laudably taking phylogeny into account, it does not
include any attempt to infer the body size of any ancestor.
Instead it is forced to assume that the old taxa are direct
ancestors of their closest known young relatives.

Again, I'm not sure this is a problem. One cannot identify ancestors, so

one needs to infer their sizes based on their descendants and other relatives.

a proxy is required. However, the above is slightly
misrepresentative: the older taxa are assumed to be sister-taxa to the
ancestors of the recent taxa. This is still a large assumption, but it
is fair to allow that sister-taxa are usually very similar.

If the branch lengths involved are short enough.

Also, by definition, that is how we are inferring
the body size of the unknown 'ancestor'.

The program I'm using gets around this constraint
and even takes branch length into account (after all a lot of
evolution can happen in tens of millions of years).

True, but also very little can happen at times.

Squared-change parsimony is the method the program uses. It _tells_ you how big an ancestor was. This is better than any other proxy! If you do a little extra work, it gives you confidence intervals on the size of selected ancestors.

Squared-change parsimony takes branch length into account. Yates (2004) used it, but McClade, which he used, regards all branch lengths as identical...

> using a recent supertree of 272 genera (Pisani et al., 2002),
> the most inclusive dinosaur phylogeny available.

A supertree is a time-averaged consensus of opinions about
phylogeny, no matter (in this case) how well supported those
opinions were when they were produced, let alone how well
supported they are now after the discovery of new data.

I'm not sure if this is a comment on my paper or not, however, I will say that we can only use the best data at the time, however bad we think it is now.

My point is that it was not the best available data at that or any time. For example its sources include precladistic speculations, it is biased toward those characters that the most analyses use, and it lacks any bias towards bigger or newer phylogenetic hypotheses, treating all the same.

Matt Carrano's much more detailed Cope's Rule in dinos analysis (2005)
produced very similar results to mine suggesting reliability of our

They may be precise... they may not be accurate... :-)

> A comparison in which the later taxon is larger than the
> earlier taxon supports Cope's rule. Under the null hypothesis
> of no body size trend, around half of such comparisons should
> show a size increase and the remainder a size decrease, with
> the mean change across all comparisons not differing
> significantly from zero.

This null hypothesis works only if there's no minimal body
size. IMHO there are several different such minima. For
examples theropods may have
been limited by competition with mammals, lizards and so on.

If we assume a theoretical minimal size of say, 50cm then this is not a problem.

For theropods maybe. But for sauropods the minimum could be around the size of a small elephant (*Bonitasaura*...).

So if they could shrink, grow or stay the same and they
grew then this is positive evidence for a trend towards large size.

Or for diversification with subsequent extinction of the small ones.

p. 589:
> "In comparisons where the bauplan is varied (e.g. within the > Stegosauria)"
> Stegosauria???

Yes, but we are talking about length as a mass proxy. The bodies of the stegosaurids became increasingly robust for a given length (i.e. scaling factors would have to be drastically altered) and this had to be allowed for in our comparisons.


>Take-home message: "Nothing makes sense in evolution, except
in the light of a good phylogeny!"

True! But how do you know the phylogeny IS good? I'll leave that one for the philosophers. And, yes, I am a cladist, but there are lots of problems with just about every published tree somewhere.

Of course, but bigger analyses are generally better than smaller ones, newer analyses (which are usually bigger, and which have gone through a few rounds of learning about correlated characters and the like) are generally better than older ones, and analyses are almost automatically better than non-analyses.

p. 590: "Phylogenetic analyses of the fossil data require that
both the phylogeny and the fossil record are adequate for the
purpose. We used the method of Benton (1995) to test the
goodness of fit between the phylogeny and the stratigraphy."

This is a really good and simple method. Plus you can (as we did) test it at multiple levels. It really doesn't get enough attention and I think would greatly add to an awful lot of "evolution based on phylogeny" studies. Off topic but worth mentioning.

I agree.

"[...] we tested both orders and each family and superfamily
independently." A bit too much reliance for my taste, but it can't matter much
-- all those taxa are monophyletic after all.

Sorry, wanted to write "reliance on ranks".

However, a) a few or even a lot of spelling errors are not the end of
the world,

I never said they were.

(It does, however, tend to give readers the impression that whoever wrote it saw most of those names for the first time...)

b) as stated above, we were perhaps overly reliant on
non-primary data sources.

Compendia like Glut's encyclopedia (1997, 2000, 2002, 2003) and The Dinosauria (I) (1990 -- phylogenetics-poor, but at least it contains a few skeletal reconstructions with a scale bar) have been very useful to me. Glut's copies of illustrations are often bigger and thus easier to measure than the originals.

- As mentioned above many of these taxa are so incomplete that
I wouldn't dare estimate a total length. *Unenlagia*,
*Therizinosaurus* and *Antarctosaurus* for instance.

Someone did, thats where we got our data (I'm not going through my records for just this post). As I have said above (and will reiterate here, lucky you) time forced our hand on some issues. Not a great excuse, but honest. I simply could not cross reference every single point from multiple data sources.

I think you could have taken The Dinosauria I and looked up how much material is known for each genus... of course it's too old to include *Unenlagia*... Besides, how much time did you have for your M. Sc. thesis? Half a year?

- The size of *Dilophosaurus* is not known. The only known
specimen is subadult. Thus *D.* was most probably not slightly
smaller but quite a bit bigger than *Liliensternus*.

See above.

Well, this is a case where both may have been somewhere around 6 or 7 meters, instead of one being 50 to 150 % longer than the other.

 - Comparing *Alvarezsaurus* with *Avimimus* relies on a
 probably wrong phylogeny (see above).

See above above: you can only use what is available. This and a few other comments of a similar nature should really be directed at the supertree itself, rather than our use of it.

_Because_ the supertree has such flaws you should IMHO not have used it.

My stratigraphic analysis (Benton 1995)
suggests that it is very conservative at all levels,

This is probably hard to avoid. :-) I recently compiled a supertree that contained all pre-Pliocene lissamphibians, and despite the thoroughly mind-boggling ghost lineages especially in the Mesozoic, the tree is more consistent with the stratigraphy than at least 9999 random trees with the same taxa.

and it was composed of a total of 126 published phylogenies.

This is actually a disadvantage -- see above.

>The silhouette of *Argentinosaurus* is a fantasy portrait,
except for a few things like the length of the lower legs. I think it was
made for *Apatosaurus* and scaled up. See above for *Alamosaurus*

Yep! But it looks pretty, which is why the editors asked me to add them.


However, as a final defence of my work, I would point to the fact that
essentially this paper includes a number of very different analyses
(phylogeny, stratigraphy, oldest vs youngest, biggest vs smallest,
Jablonski polygons, basic scatter graph) and all of them point to the
same answer.

I agree. And indeed, a VERY preliminary one-second analysis of my data points to just the same consistent size increase across Dinosauria. But, like your analyses, it does not distinguish the effects of time and phylogeny. This is where we could (!) get a surprise.