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

RE: Paravian claw studies

I've been reading more claw studies today.

Different authors have used disparate methods for measuring claw angles and 
radii. The differences can make it difficult to compare results directly. 

There is Feduccia's 1993 method which measures the inner (ventral) curvature, 
following it down to the plantar surface of the fleshy toe pad.. As far as I 
can find, he never published his data set. He divided his samples into three 
behavioral categories: ground, perching, and trunk climbing birds, without 
defining the criteria used to distinguish between them. He did not investigate 
whether claw curvature scales with body mass. He provided no statistical 
evidence for the strength of his signal either. Later authors faulted the use 
of the inner curvature, fit to the best fitting circle, as the inner curvature 
of many birds is not circular.

Glen and Bennett 2007 use the outer (dorsal) curvature of the ungual and 
keratinous sheath as two separate measurements. They divide their samples into 
four categories which divide a continuous spectrum defined by the % of time 
spent foraging on the ground vs. in trees. They did publish their statistical 
confidence values (P values). They critiqued Feduccia's method, as the inner 
surface of a claw can have a highly variable (non - circular) shape, and they 
noted that it also included the joint and toe pad, increasing the apparent 
curvature. I wrote to Chris Glen who was very kind to inform me that he is 
still working on his gigantic dataset and hopes to publish it in the next year 
or so.

Fowler et al. 2009 state that they use the method of Pike and Maitland, 2004. 
The latter demonstrated that claw curvature scales with body mass. Fowler et 
al. measured inner AND outer curvatures. Their data do not seem to draw strong 
conclusions about cursorial vs. perching behaviors.

Burnham and Martin 2010 followed Feduccia's method, and failed to address Glen 
and Bennet's criticisms of it. They found curvature values for Microraptor much 
greater than Glen and Bnennett did. They also make qualitati
essments of claws - that they have pin like points or "a thicker base". In this 
method they cite Yalden. Yet no measurements are taken, and the possibility of 
diagenetic lateral compression of the claw shapes is not excluded.

Dececchi and Larsson 2011 grouped large numbers of limb and foot measurements 
together to form ratios. They did provide their data sets in supplemental info. 
They seem to have measured claws only by sorting them into categories of 
straight, recurved, and highly recurved.

The literature in this field shows a state of confusion and disagreement. Any 
researcher who attempts to investigate claw curvature correlations would do 
well to begin with a systematic review of methodologies, and then provide a 
very explicit method, and publish their entire data set with the paper, to 
allow the experiment to be reproduced by others.

There may be a consensus that there is some correlation between increased 
curvature and tree climbing behaviors, but no clear measurements of the 
strength of this correlation, and little or no consensus beyond that.

Lastly I would just state again that distortion is an important matter to 
address. Photographs can unintentionally distort curvatures. A perfect circle 
viewed from just a slight angle is distorted into an ellipse. If a camera is 
held even slightly out of perfect perpendicular in both degrees of freedom to a 
perfectly flat fossil there will be distortion. Moreover, fossils of basal 
paravians are often preserved as flattened slabs or carbonized films, and the 
bones are often crushed. lateral crushing of claws cannot be easily ruled out.
From: Jason Brougham
Sent: Monday, May 28, 2012 10:49 AM
To: tijawi@gmail.com; dinosaur@usc.edu
Subject: Paravian claw studies

Claw studies are frought with peril. Please, please, whatever we do, let's be 
aware that claw curvature studies must fully respond to the literature. And 
let's reconsider using photographs as the only evidence in any claw curvature 

In 2010 Martin and Burnham
hird toe claw curvature of 138 degrees, and Archaeopteryx had 137. They ruled 
this to be arboreal. They did not even acknowledge Glen and Bennett 2007, or 
Hopson 2001.

Burnham and Martin used Feduccia's method with minor adjustments. He ruled 
Archaeopteryx to be arboreal. Feduccia, of course, never published his data set 
for his original 1993 Science paper.

In 2007 Glen and Bennett developed a more detailed and explicit system for 
measuring toe claw curvature, both with the ungual core and preserved sheaths. 
They got a range of results from different specimens, from about  90 degrees in 
Microraptor to about 70 in Archaeopteryx. their methodology also recognized 
that bahevioral categories are a continuous spectrum, rather than discrete 
cohorts of ground vs. tree birds. They found both animals to lie near the 
overlap between strict ground birds and predominantly ground birds. I'm not 
sure they ever published their data set either.

Larsson and DeCecchi in 2011 did a multivariate analysis and again found basal 
paravians to cluster within the range for ground birds. Their methods were a 
bit harder to interpret, as they used indices that grouped many measurements 
together and expressed most measurements as ratios.

So, to be of clear value, any new study on this subject must respond explicitly 
to all these previous studies and account fully for all methodological choices 
and for any differences with the measurements in past studies.

Lastly, let me mention that holding a camera at anything other than absolute 
perpendicular in both free dimensions to an absolutely flat bedding plane can 
distort the apparent curvature in the photograph. When we draw a teacup resting 
on a table the top is an ellipse, not a circle. Therefore, anyone who proposes 
to use a method of measuring curvature from a  photograph should calibrate 
their photographic procedures first against subjects of known curvature. I 
don't think that using published photos without an explicit photographic method 
will really be valid at all.

Therefore it may be
e in basal paravians is obsolete. Even after publication of such work problems 
with or limitations of the methodology may be discovered.

From: owner-DINOSAUR@usc.edu [owner-DINOSAUR@usc.edu] on behalf of Tim Williams 
Sent: Sunday, May 27, 2012 11:00 PM
To: dinosaur@usc.edu
Subject: Re: Microraptor hanqingi, new species from China.

GSPaul <GSP1954@aol.com> wrote:

> I have measured central toe claw curvature in a large number of specimens
> with complete keratin sheaths via high res images and they are strongly
> curved and in the arboreal range. No bird that has such hook toe claws spends
> much time on the ground. This will effectively close the case on this, and 
> past
> publications on this are errant becaus they did not have access to quality
> images of a lot of specimens. Can't discuss this openly because some
> journals won't then consider the paper.

Thanks for the info.  I look forward to the publication and to the
mooted data.  As a way of making the manuscript as strong as possible,
and just in case you haven't adopted this approach already: when
discussing claw curvature it's a good idea to differentiate
"trunk-climbing" and "perching" in the context of arboreality.  Also
helps to differentiate claw curvature in predatory birds (many of
which don't spend much time on the ground).  I'm sure these are things
that have occurred to you already, but I guess it can't hurt to
mention them.

Anthony Docimo <keenir@hotmail.com> wrote:

>  Given that those modern bird you name, are all *groups* of birds, I 
> completely expect that they'd be a bad comparison with *A. lithographica*. (If
> you were to use that, as it, you wouldn't be comparing apples with lemons, 
> you're comparing apples with citruses (or apples with fruits))
> So perhaps the question should be, of those pigeons, which are in the same 
> morphospace as *A. lithographica*? (and the same for the birds of
> prey, trogons, galliforms)

No, I don't think it makes a difference.  The differenc
locomotory styles (hip-based vs knee-based) means that using hindlimb
proportions to compare _Archaeopteryx_ to *any* modern avian has
profound limitations.