Have you ever had one of those moments, where you read a paper, and you realize the paper is saying...pretty much everything you have been working on for the past few seasons? I just had one of these precious moments. Fortunately, I don't think negates my efforts though. In this case, the paper's location is on the opposite side of the world and approximately 190 Ma earlier. So I think I am safe. *Phew*
The paper in question is Gastaldo et al, 2009 (availble here). This was published in the March issue of Geology, which means you will have to meander over to your local library/university computer to gain access, or just join the GSA.
The deposits spanning the terrestrial Permo-Triassic (P-Tr) boundary are fluvio-lacustrine, with the placement of the boundary based on the position of a succession of laminated beds. Gastaldo et al. seek to determine the legitimacy of using lithostratigraphic means to evaluate the position of the P-Tr boundary. For those not in the know, lithostratigraphy is a subset of stratigraphy which focuses only on characteristics of the rocks themselves to define stratigraphic relationships. In other words, sandstones correlate with sandstones, mudstones with mudstones, etc. Lithostratigraphy is a fairly useful tool for some purposes, but it is inadequate for other purposes. So if you are using lithostratigraphy to help examine some aspect of the geologic record, you better make certain that lithostratigraphy can do what you think it can do.
In the case of fluvial (rivers) deposits and lacustrine (lake) deposits, lithostratigraphy is rife with internal discontinuities. To put it another way, rivers move and lakes can dry up. The upshot of this is that moving laterally in these depositional environments might mean that you are crossing timelines as well. If workers don't take this into account when they are recording their observations, they will have transcribed errors into their data set. For some studies, this is a trivial problem. However, when workers are examining an extinction event, like the one that happened at the P-Tr boundary, these subtle internal discontinuities can lead to erroneous interpretations.
And, in point of fact, the authors determine that the P-Tr boundary (the laminated beds) are diachronous. Which means that these beds were not laid down at the same time, and should not be used to interpret a basin-wide event.
The only problem I have with this paper is how they established their stratigraphic framework. The authors set out to evaluate the utility of lithostratigraphy as a proxy for basin-wide events. Yet, when they set up their stratigraphic framework, they based it on.... lithostratigraphy. To me, this seems akin to using a word to define itself. It doesn't invalidate the conclusions, but I think they could have made a better point by evaluating the lithostratigraphy with some other form of stratigraphy. For example, this paper would have been a much stronger argument if they set up the stratigraphic framework based on some chronostratigraphic criteria.
Chronostratigraphy is essentially "Time Stratigraphy". By looking at deposits that could only have been deposited at approximately the same time (or, in the case of ash deposits, at the same time), it is possible to evaluate the timing of depositional events throughout a basin. A commonly used chronostratigraphic marker, in a terrestrial setting, is palynology (looking at pollen preserved in the rocks). This will provide some "fuzziness" in the data set, because the pollen can be reworked, but this is a better chronostratigraphic framework than nothing. The best chronostratigraphic markers are event beds, like an ash deposit. These can be radiometrically dated, and best of all, they can't be reworked without diluting the ash to the point where it is unrecognizable (it turns out ash deposits have a VERY low preservation potential in most terrestrial settings).
That critique aside, this paper still makes a valid point. Lithostratigraphy, on its own, is an inadequate tool to evaluate the timing of non-localized events (events limited to ~ 1km radius) in terrestrial settings.
Gastaldo, R., Neveling, J., Clark, C., & Newbury, S. (2009). The terrestrial Permian-Triassic boundary event bed is a nonevent Geology, 37 (3), 199-202 DOI: 10.1130/G25255A.1