RE: Microraptor also ate fish

[Mickey Mortimer <mickey_mortimer111@msn.com>]

Mike Keesey wrote-

> > As paleontologists, it seems to me we often assume a behavior was not 
> > normal without some obvious adaptation to it. But is this a realistic view 
> > of how adaptation and behavior work in living animals? I'm not talking 
> > about rare but possible behaviors like goats climbing trees, but about 
> > behaviors that living animals actually perform on a regular basis. For 
> > instance, are there any studies demonstrating species that spend a certain 
> > amount of time in trees have a certain amount of arboreal adaptations? Or 
> > is it possible/likely that say, 20% of species that spend >30% of their 
> > time in trees can't be statistically separated based on skeletal morphology 
> > from 20% of species that spend <5% of their time in trees? If the error 
> > bars and overlap are large enough, it would seem the right answer would 
> > just be to say "I don't know".
>
> Isn't this somewhat tautological? That is, if a species doesn't
> regularly perform X, don't we interpret its morphology as not adapted
> for X? And if another species does regularly perform X, don't we
> interpret its morphology as adapted for X? How do we even establish
> something as "adapted for X" without that kind of bias? (Convergence,
> maybe?)

Technically, everything is adapted to do what it does, since mutation causes 
most changes and the species has survived.  I was thinking more in the sense of 
what we as paleontologists think of as "adaptations for x" not being present in 
a significant proportion of taxa which are x.  So sure an elongate grasping pes 
is an adaptation for arboreality, but how many species which spend e.g. >30% of 
their time in trees actually have that trait, and how much can it be 
compensated for by other traits?  Or alternatively, how many species which 
spend e.g. >30% of their time in trees aren't significantly different 
osteologically from species which spend e.g. <5% of their time in trees?  How 
strict/misleading are the osteological correlates we look for, basically.

Mickey Mortimer