Re: Pelagornis chilensis

[Mike Habib <habib@jhmi.edu>]

On Sep 18, 2010, at 8:32 AM, evelyn sobielski wrote:

> Not having handled any material personally, I can only say that 
> Pelagornithidae bones are indeed *extremely* thin-walled judging from the 
> usual state of published material (long bones are usually crushed to the 
> point of being barely recognizable - indeed, some have been assigned to 
> Pelagornithidae more due to their size and bad preservation than due to any 
> recognizable feature, because the latter were all but obliterated). 

Indeed.  I have handled much of the North American material personally, and 
though the bones are extremely thin-walled, there is no indication that the 
overall skeleton was any lighter than would be expected for a bird of that 
wingspan.  For comparison, modern pelicans have similar bone wall thickness, 
and the ratio of skeletal weight to body weight in pelicans is not appreciably 
different from that of a mammal or non-volant bird.  Furthermore, other systems 
(such as feathers) account for a greater percentage of the mass than the 
skeleton, especially in large birds, and so even if the skeleton is slightly 
lighter than typical, it will not drop the body mass by a huge margin - note 
that the mass reported in the Mayr and Rubilar (2010) is just over half of what 
would be expected if Pelagornis had the same span loading as an albatross.  
Even if this turns out to be validated by other methods of estimation, it 
cannot be accounted for through lightening of the skeleton alone (or even 
primarily).

> To arrive at a more certain conclusion, one would need to consider, perhaps, 
> Diomedeidae (which are closest in gross morphology) and Anhimidae (which 
> *might* be closely related to Pelagornithidae, and have a very low 
> volume/mass ratio). With the material now at hand, there should be some 
> rather uninformative specimens available for thin-sectioning, which should 
> provide satisfying answers to the question of how heavy/strong their bones 
> were (http://onlinelibrary.wiley.com/doi/10.1002/jmor.10029/abstract and 
> doi:10.1006/cres.1997.0102 give some ideas on how to tackle this).

Diomedeidae actually have reasonably thick-walled limb bones (still pretty 
hollow by mammal standards, for example, but nothing like Pelagornis).  
Calculating the strength of the bones in torsion and bending is actually pretty 
easy, and does not require thin sectioning, necessarily - it can be done using 
radiographic methods, or using external measurements and the assumption that 
the new material has similar relative bone wall thickness to other known 
pseudodontorns (which the authors indicate is true).  Having done this back of 
the envelope version already, I can say that it appears the hind limbs are not 
all that much weaker, relative to the forelimb, than in living albatrosses (a 
bit, but nothing tremendous).

> I would not discount the "automatic takeoff" technique (also note that this 
> only applies to the gigantic spp.), as long as we don't know where these guys 
> nested; for all we know they might only ever have alighted on high/volcanic 
> islands to breed 
> (http://si-pddr.si.edu/dspace/bitstream/10088/2006/2/SCtP-0090-Lo_res.pdf 
> p.260 seems to imply that at least one species nested in the Appalachians or 
> on low-lying islands that today are hills in the Carolinan Piedmont).

I am extremely skeptical of the proposed launch mechanisms because it will not 
work in essentially any nesting location.  It's not a matter of having the 
right conditions so much as first principles of aerodynamics.  Steady winds do 
not provide energy, but gusts do.  When a bird (or aircraft, for that matter) 
hits a headwind, it lifts up temporarily because it feels the equivalent of a 
gust.  However, unless the animal/aircraft then powers up and produces 
sufficient thrust on its own, drag will quickly decelerate the animal/aircraft 
back to the original groundspeed - i.e. they will come back down quite quickly. 
 Given the size of the giant species in question, their exceptionally large 
spans, and substantial wing loadings (even if the light mass is correct, 
especially if it is not), it is very improbable that a giant pseudodontorn 
could then build up circulation on the wing and power into flight through 
flapping quickly enough (and with sufficient power) to stay airborne after the 
"automatic takeoff".  It is much more probable that some degree of additional 
clearance and forward speed was required to launch, especially for a heavily 
loaded, narrow-winged seabird.  Modern birds all generate this required speed 
and power through a hind limb impulse (as do ground-launching bats and most 
insects).  In some cases (such as albatrosses) gusts are regularly used as 
augmentation.  For small taxa (hummingbirds, insects) that initial impulse is 
likely important mostly for the sake of clearance and timing for efficient 
circulation generation.  For large species the additional speed is critical 
because they cannot generate enough fluid force over the wings to launch 
without it.

> Albatrosses can get away with their extremely clumsy takeoff/landing 
> techniques because their way of life minimizes the total time spent in 
> either, and of course by choosing predator-free islands to breed; they 
> regularly seem to spend 1-2 years without ever landing. (In that regard, is 
> there any takeoff footage of _Phoebastria albatrus_ from Torishima?)

Whether or not we choose to consider their launch clumsy is a matter of 
opinion.  Personally, I would consider the albatross launch system to be 
reasonably efficient in that it works on both land and water, and does not 
appear to be exceptionally power hungry (given they have modest muscle mass).  
Furthermore, note that albatrosses launch in the manner that they do because of 
a very specific morphology adapted to very high speed gliding - it is not, as 
often suggested, a matter of total body size (we know this because 1) small 
albatrosses still run to take off in the same manner as large ones, and still 
prefer gust assistance, and 2) much larger birds with a differing morphology 
launch from a standstill).

> Pelagornithidae were such a unique, extreme and specialized lineage - and 
> highly successful at the same time - that they do not compare well with e.g. 
> large pterosaurs (which were also extreme and specialized in their own 
> regard, though not as unique), except in the loosest terms of crude physics; 
> i.e., both must have some sort of vaguely similar mechanism to overcome 
> gravity. But even so they had to approach the problem from different angles. 
> Ditto _Argentavis_ - its soaring technique must have differed from that of 
> Pelagornithidae, if it did not travel all the way to the ocean to forage 
> there (and there is really no good reason to think it did).

The basic rules of aerodynamics are strict enough that the differences are not 
as great as you might expect, and the physics involved aren't quite a crude as 
one might expect, either.  But in general, yes, I agree - directly comparing 
birds and pterosaurs is often not functional.  This is why I did not make any 
comparisons between pseudodontorn function and pterosaur function in my initial 
post.  Personally, I would consider pterosaurs to be more unique, in comparison 
to living taxa, than pelagornithids, but that's besides the point.

> Any pterosaur researchers and flight-interested biophysicists who consider 
> the question relevant can at this point do much worse than to contact Mayr, 
> and keep track of his research, to ultimately cooperate in a study outlining 
> the specific adaptations to the problem found in either group, and in which 
> respects they differed and in which ones not. (One would also need to 
> consider teratorns, because Pelagornithidae alone are too singular to make a 
> good comparison with anything)

I'll drink to that.

Cheers,

--Mike H.

> 

Michael Habib
Assistant Professor of Biology
Chatham University
Woodland Road, Pittsburgh PA  15232
Buhl Hall, Room 226A
mhabib@chatham.edu
(443) 280-0181