Re: Geographic Distribution of Deinonychosauria

[David Marjanovic <david.marjanovic@gmx.at>]

>  True, but the stages don't last so long that one cannot infer
>  distributions of nonvolant taxa fairly well from knowledge of the
>  topographical layout (particularly if taking into account
>  phylogeny).

Maastrichtian: 5 Ma, Campanian: 13 Ma...

Many stages in the Jurassic are estimated to be about 4 Ma long, which 
is also the confidence interval on their start and end dates.

>  The tricky thing is shelf/epicontinental seas; sea level fluctuates
>  more rapidly than can be reliably discerned from the geological
>  record (even in warm ages, we have glacial/interglacial oscillations,
>  and these take only a few dozen Ka).

The marine geological record captures those just fine; keyword: 
cyclostratigraphy. The problem is how to correlate continental sediments 
to that.

>  Apart from that, one can assume that open seas/oceans are a barrier
>  to dispersal until proven otherwise, and simply deal with the major
>  land areas (i.e. as isolated by open sea and high mountains).

That's where uncertainties in paleogeographic reconstructions come in. 
How much connection was there between South America and Antarctica and 
between Antarctica and India in the Cretaceous? For how long? What about 
Antarctica and Australia or Madagascar and India? Or India and Africa? 
Etc., blah, blah.

>  What might be informative is the (fossil and Recent) distribution of
>  Crocodylomorpha.

Ecologically speaking, that sounds like a bad idea.