Re: Dinosaur Diversity Analyzed

["A.P. Hazen" <a.hazen@philosophy.unimelb.edu.au>]

Interesting little paper from Mike Taylor (who gave a link for 
downloading  a pdf in his last posting), with nice graphics.
Repeatedly emphasizes that the statistics he gives should be 
approached with caution for a whole lot of reasons.

Sampling bias an obvious one: the  very large number of dinosaur taxa 
known from the  United Kingdom (about (guesing and doing mental 
arithmetic) 1/700 of the Earth's land  area), for xample, surely 
reflects the long history of geology as a science there!

Are there  any good ways to estimate sampling bias?  Paticularly: any 
ways, using available infformation, to estimate which parts of the 
Mesozoic have been better-sampled  than others?  For instance... 
Somewhere in the literature there must be a reasonably complete and 
compact list of all the dinosaur fossil sites that have been worked 
in the last century or so: are some ages much better represented than 
others?  This would give a very crude indication-- site briefly 
surveyed by a government geologist when the country it is in was a 
colony would be counted as equivalent (for sampling) to one of the 
extraordinarily rich sites that have been extensively excavated over 
decades, but I suspect getting a more refined measure of sampling out 
of the published literature would be very difficult.

   (I know attempts at estimating sampling  bias have been made in 
connection with the question of whether dinosaur diversity was 
declining in the late Cretaceous...)

--

Allen Hazen
Philosophy Department
University of Melbourne