Signor-Lipps points

[Cris Little <Cris.Little@xxxxxxxxxxxxx>]

	I'd like to raise two points with reference to the discussion on the S-L
effect.
	Firstly a semantic problem. As has already been pointed out there
are two main reasons why a catastrophic events may appear gradual - 1)
sampling error, which you can do something about and 2) Random effects,
which you can't. I've recently been picked up on by a reviewer for
refering to a case where a second field season of sampling in a section
pulled species ranges up to an event as the S-L effect. This is an example
of 1). Perhaps Jere could clear this up for everbody (including
reviewers!), does the Signor-Lipps effect include both sampling error AND
random effects or just random effects alone? A side issue here, increased
sampling effort may reduce the effect of 1) to zero but do bugger-all for
2). However, increased sampling effort will, in most cases, reduce the
size of error bars on confidence intervals, perhaps giving a handle on 2).
Ideas anybody?
	Secondly, given the S-L effect can any gradual or step-wise
extinction pattern ever be regarded with enough confidence to suggest
causes? I'm thinking here of Erle Kaufmanns groups work on the C/T in the
Western Interior where a step-wise extinction pattern has been found after
intensive sampling. Peter (Harries) presented a paper at the recent
Plymouth IGCP metting  applying confidence intervals to this data, perhaps
you could run through the results of this Peter?

Cris Little,
Geol. Dept.,
Bristol University,
UK