In a continuing biostratigraphic study of the Devonian in Virginia, more than 26,000 macrofossils have been collected, identified, and correlated. When embarking on this study, the first author was very skeptical about the Frasnian/Famennian (F/F) extinction. The so-called F/F extinction appeared to be no different than any other stage boundary in the Devonian. However, examination of the data set suggests that there was an extinction and that it had two pulses. There are 42 species of brachiopods before the extinctions. After the first event, there are only 32 species despite 5 first occurrences. After the second event, there are 22 species despite the first occurrences of 4 species. Overall, brachiopod diversity was reduced by 47%. Only some types of seafood were hurt by the extinction events: bivalves and gastropods were unaffected.

A critical factor in interpreting patterns of extinction in this data set is the validity of the correlations between stratigraphic sections. Correlations were initially based on index fossils and the order in which the fossils occurred. Graphic correlation methods (CONOP9 and GraphCor) were used to test the correlations, both including and excluding lithostratigraphic data. These analyses support both the initial correlations and the patterns of extinction they revealed. Another key factor in interpreting these data is the reliability of biostratigraphic ranges. Confidence intervals on stratigraphic ranges provide one method for gauging the reliability of range endpoints and for statistically testing extinction patterns. The results of confidence interval analyses confirm the plausibility of two pulses of extinction, which should correlate with the upper and lower Kellwasser events recognized in Europe.