There are still disputes about how much of this prolonged decline throughout the Permian is "real", and how much it is a distorted representation of the actual decline in species due to a poor fossil record from around this time. With the onset of the marine regression associated with this time there was a reduction in sedimentation, which makes strata from this time very rare and fossils within them even rarer.
This has meant that the time at which an organism is classified as becoming extinct is when the youngest fossil of its form is found. It is probable that there would have been later examples of the organism present which were simply not preserved. It is known that some genera have existed for long periods around this time without leaving any known fossil record by the phenomena of Lazarus taxa.
This term describes those organisms which dissapear from Permian strata, as if having become extinct, only to reappear in the Triassic with near identical morphology. It is believed that these organisms were simply not preserved during the time they are missing, or preserved in off-shore sediments as yet undiscovered. This may also be the case with many other organisms creating the illusion they are becoming extinct before they are in reality.
A better way of studying rates of complete biota extinction levels has been developed with the analysis of isotopic ratios of Carbon. When life is abundant there is almost completely carbon-12 within the geological record. This isotope is more efficiently accepted by enzymes within organisms, passing into organic matter faster which becomes lithified into rock. At times of lowered biotic activity, such as at an extinction event when a lot of life has been killed, the ratio of carbon-13 within the rocks will be higher as a higher proportion of carbon will be being fixed as carbonates inorganically. Inorganic precipitation of carbon does not differentiate between the different isotopes of carbon as life does.
By analyses of carbon isotope ratios it is then possible to see, by peaks in the carbon-13, at what times there has been a reduction of biotic activity. This is independant of whether organisms present are being preserved or not, and shows at what rates the extinction is occurring.
This carbon isotope analysis has been performed on strata from the Permo-Triassic boundary to see at what rate the levels of life are declining.
There is an obvious anomalous peak in the carbon-13 ratio at the Permo-Triassic boundary indicating that there has been a dramatic loss of biotic activity. This change in ratio value from the standard level from when life was abundant does start to occur mid-way through the last epoch of the late Permian, (Changxingian), increasing in rate towards the Permo-Triassic boundary. This suggests a far more rapid extinction than the prolonged affair which had been thought possible from study of the fossil record.
As life recovers and diversifys after the extinction event in the Triassic era the carbon isotope ratios rapidly start to return to levels similar to what they were before. There is a second smaller peak in carbon-13 early in the Triassic, which may be another small extinction event probably linked to the main one, such as a short reactivation of of some of the widespread volcanism.
Half way through the late Permian there is a small carbon-13 peak. This occurs at a time when some major species are believed to have become extinct, such as the corals and blastoids. These reef dwelling organisms are very susceptible to environmental change so may have been killed off in any climatic change before the major event which killed off the other marine species.
When the strata from different localities are studied the breadth of the carbon-13 peak does vary slightly. It is a wider, more gradual shift in localities in the Southern Alps, compared to a narrower shift in South China. This may be because of differing rates of extinction at different localities around the world, but is more likely because the Chinese sections has been condensed compared to those in the Alps.
This may help to explain the historical differences of opinion about over what period of time the extinction occurred, by geologists working at different localities. High resolution biostratigraphic correlation of the strata from different localities are starting to resolve this as analysis by age can be performed independant of how much the sections have been compressed.