| [Thread Prev] | [Thread Next] | [Thread Index] | [Date Prev] | [Date Next] | [Date Index] |
Time to head off to the library and find a copy of Shaw. Sorry, Norm.
Foram overlaps: I know Norm doesn't agree, but a lot of us think that the
many of the overlaps seen only in "Zone PO" are due to reworking. But I'm
not a micropaleontologist, so I will let you other guys argue about this.
Terrestrial record: oh, it's not _that_ bad. The state of Montana is a
"locality" in some sense, but you can also look at it on a site-by-site
basis. Just the published literature on mammals (and I know there are a
ton of unpublished sites) yields 41 sites in the late Maastrichtian
(Lancian) of Montana. On top of that, we have 38 other sites in Wyoming,
Alberta, Saskatchewan, South Dakota, New Mexico, Colorado, and Utah, some
of which are quite close to the boundary (e.g., Long Fall Horizon). The
basal Paleocene is also reasonably well known from sites in New Mexico,
Wyoming, and so on.
Those who don't care about the details can skip the following, which
argues that we _do_ have enough conjunction data to show that the K-T
created a lot of turnover in mammals. For those who read it anyway and
find it dull, I apologize in advance.
There are 30 mammalian genera in the entire late Lancian record
(_including_ Bug Creek). The early Puercan (basal Paleocene) is marked by
14 and the Puercan as a whole by 61 (!) genus-level first appearances. The
Lancian record alone demonstrates 316 out of a possible 435 conjunctions
among 30 genera, or 73%. This is pretty good, considering that there are
temporal signals (turnover during a 10-million year interval) and
biogeographic signals creating disjunctions.
I extracted 9 basal Puercan ("Pu1") lists from my database using my own
correlations, which happen to agree perfectly with those of Archibald et
al. (1987) except that I include Harbicht Hill in this zone. These include
25 genera (10 carry-overs; 1 Lazarus genus reappears after Pu1) and
demonstrate 200 conjunctions, or 67% of 300 possible. This is virtually
the same as the Lancian. After _combining_ the two data sets, I get 88
lists; 45 genera; and 471 conjunctions out of a possible 990, or only 48%.
What's even more remarkable about this huge drop is that Pu1 is extremely
short, on the order of 0.39 m.y. - and the entire Puercan is only about
1.05 m.y. long (Swisher et al. 1993). If I had put in the late Puercan
(Pu2/3) record, the percentage of demonstrated conjunctions would have
been only 37% (90 genera total, 1464 conjunctions). Since the entire
Lancian is about 25 times as long as Pu1, it makes no sense to attribute
the precipitous drop in conjunctions to "background" turnover - or to
biogeography, because Pu1 is known from Montana, Wyoming, and Colorado,
and Pu2/3 is very well known all the way from New Mexico to Alberta and
Saskatchewan.
The only reasonable interpretation of the data is extremely rapid
turnover. Even if you don't buy anything I said about conjunctions, the
fact remains that 16 out of 30 (53%) Lancian genera go extinct; at least
24 of these genera are present during, and 7 (29%) go extinct after, Bug
Creek/Long Fall time; and that diversity increases from 15 genera at the
boundary to at least 24 genera within 0.4 m.y. As for the "gap," I think
it is obvious that the interval between Bug Creek and Mantua Lentil is on
the order of no more than a couple hundred thousand years. No similar
extinction/diversification event has ever occurred in the history of
mammals after the K-T, at least not in North America (there's plenty more
data to back up that claim). This to me unequivocally implies some sort of
a catastrophe.
Partial index: