Re: DINOSAUR digest 2041

[jrhutch@stanford.edu]

Sorry, my browser here is having a hard time making out who is sending 
what e-mail; if you want me to reply to the list then it might be best 
to cc-me in the e-mail so I can reply to each message on its own, if 
necessary.  My apologies if I misattribute message senders below.  Some 
more quick replies:

=====

[I think Jaime said this:]
"<<  Actually, birds of all types have very, very large ilia. Many 
advanced quadrupedal dinosaurs, like titanosaurs and ankylosaurs, have 
very large ilia, but are still not considered swift runners.<<"

[Then Tracy said this?:]
"Yes, but large in one isn't the same as large in another. The ilia are
shaped differently so you really can't compare them to each other that 
well. Birds and theropods have different ilia in that they are large in 
different ways, thus the muscle attachment isn't the same, even though 
the same muscles are used. Pubis and ischia as well as the caudo-
femoralis muscle. If the femur was held the way Hutichinson & Gracia 
have it, how does the caudo-femoralis fit? Wouldn't that be stretched 
really far? I liken it to how different the wing muscles between 
pterosaurs and birds (well, maybe not that drastic but you should get 
the idea).

I respectfully disagree with Tracy here and agree more with Jaime's 
point.  Although no one has quantified this in a paper, to my 
knowledge, my impression is that neornithine ilia are proportionately 
larger than most non-avian dinosaur ilia.  I think the sacral count is 
a rough indicator of this as I noted in my 2001 ZJLS pelvis paper; 10-
22 sacrals in neornithines, ~5 in most non-maniraptorans, and stepwise 
evolution in between.  I don't see their shape as so drastically 
different to render comparisons inadequate; if they are then it needs 
to be demonstrated empirically rather than asserted as 
just "different."  Isn't that just a Martin/Feduccia et al. rhetorical 
strategy?

The femur orientation we use is just the same that other authors such 
as GS Paul have illustrated it, and I cannot rule out that it was 
possible.  I do not see a reason why the CFL would be overstretched.  
With muscle fibers arranged at angles to the central tendon, I would 
expect that the muscle would have a fairly wide range of action.
=========


[Thomas Miller said this?]
"<Was a 20-30% of mass even possible, say common 65 million years ago ?
Nobody really knowns.>"

[And Jaime replied:]
  "It is in some birds well over 30%, and such osteological features 
that indicate mass in these taxa indicate that similar values were 
present in cursorial theropods, like *Tyrannosaurus*.

[To which Tracy replied?:]
"Ok, to play devils advocate, 70% isn't."
===========

I'm not sure which birds/data Jaime is citing, and whether he is 
considering both legs and just extensor muscles (i.e. those with an 
extensor moment arm about a joint... and especially those that are 
actively contracting at mid-stance, which isn't all extensors 
necessarily).  So I'd have to say that I don't know of data that agree 
with his statement.  The ostrich data I've seen shows ~20% body mass as 
valid extensor muscles (both legs total).  More work does need to be 
done on how much extensor muscles are in living animals, and I am doing 
tons of that lately, but from what I've seen so far in large and small 
animals; mammals, crocs, lizards, birds, etc. 20% is toward an extreme 
end, especially for bipeds.

Keep in mind that (as Per Christiansen recently showed, and earlier 
work suggested) skeletal mass increases as a percentage of body mass, 
reaching ~20% mbody in larger tetrapods.  Pleurocoels, hollow long 
bones, etc. probably helped at least a little in theropods (compared to 
the absence of such things, which would have to add mass), but it's 
something to be considered in larger taxa.  The histogram in the NY 
Times article is a nice demonstration of this concept.
==============

>From: "Bruce E Shillinglaw" <SHILLINGLAWB@prodigy.net>
>Reply-To: SHILLINGLAWB@prodigy.net
>To: "Dinosaur Mailing List" <dinosaur@usc.edu>
>Subject: Re: New Tyrannosaurus paper
>Date: Fri, 1 Mar 2002 19:17:34 -0500

[snip]

>     Horses _also_ frequently suffer fatal damage if they fall while 
>running,
>as far too many thorobred owners and jockeys have discovered. And yet 
>horses
>still run! In fact, they're _specialized_ for it. Giraffes, as someone
>mentioned yesterday, also gallop at great speed, and if one ever fell, 
from
>_that_ height, getting up and walking away would be unlikely, to say 
the
>least.
>     Thus my question - why would the likelihood of serious damage be 
any
>more limiting to mega-theropods than it is to large, fast ungulates?
>             Bruce

Just to keep things straight, our Nature study does not discuss the 
risk of falling; that was originally proposed in JVP 1995 by Farlow et 
al., and more recently by Farlow et al. in American Zoologist 2000.  
Our current paper just deals with muscle moments and masses, which are 
needed to maintain equilibrium at mid-stance or else falling (or other 
hazards) would be a problem.  Don't give us credit for the "tripping 
tyrannosaur" idea.  :)

=======

>From: dexter dexter <dexter1647@caramail.com>

>Another thing to concider : They used 6,000 kg for MOR 555,
>which is heavy. In Currie's 2000 paper, he claims the
>Anderson method gives 4,160 kg for MOR 555. Would such a big
>(~1.45) difference in weight affect the outcome of their
>method ?

[Jordan Mallon replied:]
"According to the paper, actual body mass is irrelevent, and does not 
factor into the equation: "Although [body mass] is unknown for extinct 
taxa, our calculation expressed the required extensor muscle mass as a 
percentage of body mass (T), factoring out [body mass] (see Methods)."

First, it has been pointed out repeatedly by Farlow et al. 1995, 
Carrano, and others that the Anderson et al. equation likely 
overestimates body mass because it is based on femoral circumference, 
which has been shown to scale differently (more robust) in crown clade 
birds than in non-avian theropods, and so it is inappropriate to apply 
bird mass-circumference scaling data to much more basal theropods.  
6000kg is a lower end estimate for MOR 555; see a recent summary by 
Seebacker (sp?? in JVP?); also Henderson's 1999 Paleobiology paper.

On the model, it's more complicated than "actual body mass is 
irrelevant."  Clearly body mass is relevant, as Fig 3 in our paper 
shows, but it is a complex interaction between muscle force and moment 
arms, the ground-reaction force, and segment mass positions that helps 
produce this result.  I'll explain this more in a paper soon; it is a 
confusing part of our model for most people.

=====
[Dann Pigdon commented on this:]
MariusRomanus@aol.com wrote:
> 
> Face it... Bumble bees would not have been able to fly if we had not
> seen them do so.

"This was the example I was also thinking about. No amount of 
theoretical mathematics can fully account for the intricacies of 
biological structures or systems (try predicting the weather more than 
a few days in advance - even with a super computer)."
======

As I've mentioned before, this is not a criticism of much substance.  
It is a rhetorical argument that does not test any hypothesis or 
provide any scientific criticism.  The question is not whether 
theoretical mathematics can perfectly model data.  The question is 
whether it produces ADEQUATE results.  That IS testable!  For example, 
the same sorts of models I use are used by people at Stanford to model 
human surgery.  Modify a tendon in the model, run the model, and the 
biomechanics drives a 3D animation of a human walking.  It almost 
perfectly predicts the correct strategy for surgical treatment of human 
gait deformities!  People walk better because of it.  If theoretical 
models were inadequate, this approach would be ruining people's lives 
rather than making them much better.  Read up on research by Scott 
Delp's group if you're curious.  This is one of dozens and dozens of 
examples of theoretical models in biology that have been shown to work 
just fine.

A criticism that IS substantive would be a mathematical treatment that 
shows, using comparable or better methods and evidence that are 
replicable, that predicts very different results from our model.  This 
is science; casting aspersions without any contrary data is not.

[Dann Pigdon notes:]
"This is even more extreme in kangaroos than many other animals. I 
wonder whether a similar study into the musculature and skeletal 
structure of 'roos would predict their phenominal feats of leaping or 
stamina."

Haven't done it yet, but pretty sure it would based on other models 
I've done.  Stamina and leaping are not what we're modeling right now, 
but hopping is easily compared to running in our model.  An interesting 
side note is that many people might think hopping would be a great 
strategy for a biped like a tyrannosaur.  I get this question a lot 
from non-dinolisters; didn't tyrannosaurs hop?  One thing that people 
might not know is that hopping can be biomechanically disadvantageous; 
it can incur much larger forces than running.  Kangaroos are neat in 
that they are able to hop very efficiently (in a metabolic sense), but 
biomechanically (based on real experimental data I've seen) I expect 
that their maximum size is limited by the high forces that they 
experience during fast hopping.

[And again:]
"Not so much poor math: I'm sure the math's envolved in the tyrannosaur
paper were pretty good. It's the limitations of mathematics in their
ability to fully describe natural systems that is at fault. Nothing
beats first-hand experiments with actual specimens."
=====

We did our best.  I agree that first-hand experiments are great, and 
that's why I do a lot of them with living animals.  But if they are all 
that we can ever trust, paleontologists and evolutionary biologists, or 
anyone in historical sciences, should quit their jobs and go work for a 
dot-com, because they'd be wasting their lives.  All paleontology makes 
assumptions, whether it is strict anatomy, systematics, or functional 
biology. Assumptions about the biology of extinct animals are most 
rigorous when they are firmly grounded in data from extant animals, the 
more the better.  That's what we tried to do in the Tyrannosaurus 
running paper.  And the great thing is, we can be proven wrong or 
supported by later work.

Folks, if you disagree, then please explain to me why that sort of 
approach in paleontology is not valuable, and not the way science 
works.  Did Washington really cross the Potomac?  Oh, were you there?  
Basing science only on directly observable phenomena castrates it, as 
creationists try to do.

As long as theoretical models are at least indirectly testable (see 
post by Holtz for an example with our paper), I contend that they are 
important and cannot be dismissed merely by tiresome bumblebee (or 
tuna, flea, dolphin, etc...) rhetoric. You can test these theoretical 
models, and I am sure some people will try.  Shit or get off the pot.

Sorry if I sound testy, but I have little patience for weak arguments. 
If people have criticisms that are based on methods or evidence rather 
than just casting aspersions without doing any work or deep thought, I 
am glad to spend my research time responding to list e-mails.  
Otherwise, we could just keep sending messages saying "You're fulla 
baloney, bubba" back and forth?  Would that be fun?   :-)

With respects (and having a good time too despite my temperamental 
tone),
John