endless K-T drivel

From
Date
Reply-To
Sorry, folks. Got carried away. This is yet another reply to MacLeod.

I'm confused by your comment that we agree on the duration of the extinction  
event. My claim was that all proposed mechanisms for _terrestrial vertebrate_  
extinctions were either micro-term (<thousands of years) or macro-term (>200  
kyr), not the contrary as it appears that you are saying. I said nothing about  
the duration of the marine event other than to ask for clarification on your  
results - from others, since I have Keller & MacLeod on my desk right now (I'm  
still waiting). My point was that if 1) the terrestrial events did not occur on  
meso-time scales, 2) terrestrial and marine events were largely the result of a  
single causal factor, 3) the marine record can definitely rule out a  
macro-interval for the extinctions, and 4) the K-T impact is sufficiently well  
correlated with these events and sufficiently short-term, then the impact model  
is well-confirmed.

I gather you don't agree with 1), 3), and possibly 2) and 4). My claim 1) was  
outlined in an earlier post (on conjunctions). As for your argument on 3), I  
believe we are not talking about the same thing. My question was not about  
members of the putative survivor cohort that based on your figure 4 go well  
into Zone P1a and beyond, such as Heterohelix complanata or Globigerinelloides  
aspera. Instead, I was concerned about such species as those numbered 17 - 28  
in the Brazos chart, all but two of which (if "Globotruncanella" caravacaensis  
is the same as "Hedbergella" caravacaensis in the appendix) are ranked as  
putative survivors. If those species were to be counted as K-T victims, there  
would be a 28/34 = 82% extinction in the Brazos core. I am _not_ saying this  
interpretation is correct; I am tossing out the possibility for general  
discussion. I realize that some or many of these species may have  
well-established occurrences in other Danian sections.

Furthermore, I have no problem with there being secondary waves of extinctions  
up to 100 kyr's or even myr's after the boundary. After a physical catastrophe  
on the scale of the Yucatan impact, it stands to reason that climate,  
biogeochemical cycles, and biogeographic patterns would be highly unstable.  
This brings up another point: I realize now that so far both of us have been  
conflating the duration of a causal event and the duration of its biotic  
sequelae. My previous argument was that if the sequelae were short-term, the  
causal event must have been as well. I think this is generally reasonable.  
However, if the sequelae do turn out to be meso- or macro-term, that does not  
rule out the possibility of the causal event being micro-term. No, this is not  
an attempt to make the impact scenario untestable. This brings me to Prothero's  
"post hoc, ergo propter hoc" argument. In order to demonstrate a causal  
relationship, one _must_ show that the putative cause came before the putative  
result. But this is _not_ the only condition that must be met, and it is not  
the only thing the impact scenario has going for it.

First, the cause and effect must be spatially related. With a very large impact  
this should not be an issue; the effects are global-scale. However, it would be  
an issue for many other scenarios such as some of the strictly oceanographic  
ones (ocean anoxia, which should have no major impact on terrestrial  
organisms). Furthermore, the greater intensity of terrestrial extinctions in  
North America than elsewhere strengthens the impact scenario in this regard.

Second, the intermediate mechanisms in a causal scenario must be plausible  
based on independent evidence. I see no problem with this with respect to the  
impact; the only counter-argument I see is that there are too _many_ ways to  
create a mass extinction with a Chicxulub-sized impact on a carbonate shelf.

Third, given all of the above there still must be no more parsimonious scenario  
that accounts for the same data. A typical problem for "A causes B" arguments  
is that "B causes A" is just or even more parsimonious. This is not the case  
for the impact model, obviously: extinctions don't cause impacts. Other  
competing hypotheses ("Q") would therefore have to take the form "Q causes A  
_and_ B." This would be extraordinarily difficult because "A" is an  
extraterrestrial impact.

Based on my last point, it should be clear why non-impact models have tended to  
go after the spatial and temporal relationships of the impact and extinctions;  
the intermediate mechanisms; or the existence of the extinctions in the first  
place. But I don't see Norm strongly disagreeing on _any_ of these points, or  
offering an alternative model that accounts for _all_ of the major empirical  
observations (such as the floral extinctions, or the timing and size of the  
crater vs. its supposed lack of sequelae). Instead, he seems to be focusing on  
the _duration_ of the sequelae, trying to show that many extinctions took place  
well after the boundary - even though I see everyone agreeing that many or most  
of the extinctions took place very close to it.

Perhaps I am being uncharitable and all of us really agree that the impact (or  
impacts) did have major biotic consequences, but that other interesting things  
were going on before and after the impact that also had major consequences.  
After all, none of us has ever said that the late Maastrichtian and Danian were  
boring. So maybe we can bring this to a friendly conclusion after all.