Re: Torosaurus NOT Triceratops--? resent

[evelyn sobielski <koreke77@yahoo.de>]

> Body size (one genuine aspect of individual variation) is a
> big question mark. However, 
> 
> we can narrow down the issue by first pulling out the
> stratigraphic and ontogenetic signals. 
> 
> Obviously I'm biased towards strat, but this data is
> absolutely essential, and Longrich's study 
> 
> doesn't have it.


That is discouraging. 

Generally I'd make it "strat and loc[ality]", but in this case locality is 
probably resolved. Are there any pertinent specimens from outside the prairie 
belt? IIRC, no. Can they be expected, given the density of known fossil 
localities of correct age and depositional environment (e.g. in California, or 
in Appalachia)? I'd go with "no" here. All the animals in question seem to have 
been endemic to the warm-temperate to subtropical (at that time) lowlands and 
foothills of eastern Laramidia.

That leaves strat; my first thought on reading the comments was in fact "did 
they figure in the temporal range?" It's not good to hear they didn't

Using Plio-Pleisto-Holocene taxa as proxy, I'd guesstimate that for a large 
terrestrial animal, 0.1 Ma of evolutionary time is when you start getting 
morphological differences that exceed individual variation within in any 
exactly coeval population (it's definitely so for 1 Ma, and for small land 
mammals it's already so after 0.01 Ma, so it's got to be the order of magnitude 
in between).

Might try to get "exact" data from elephants; you need something with a roughly 
comparable generation time and ideally a comparable metabolism (these things 
are what affects mid/long-term rate of evolutionary change most; in the short 
term fluctuations in background mutation rate are often dominant). So 
elephants, possibly rhinos, are probably the closest we can get, and I'd think 
the quantitative value, while not terribly exact or representative, would at 
least be better than one order of magnitude, seeing that the values for *all* 
terrestrial mammals *no matter their generation time* barely span two OOM 
(meaning that the value of any one particular lineage with stereotyped 
evolutionary rate would be significantly lower).
So I think the temporal resolution necessary to resolve this question can be 
narrowed down to one order of magnitude, with an uncertainty of 0.1 OOM 
perhaps. If not better. Whether the strat sample is precisely dated and dense 
enough to fulfil this requirement I don't know, but at least we should be able 
to get a pretty good quantitative estimate for the hypodigm needed to resolve 
this.


Regards,

Eike