Cogent interpretations of data bearing on the Cretaceous/Tertiary (K/T) extinction controversy depend on the existence of accurate chronostratigraphic models for the various K/T boundary sections. We have employed the graphic correlation technique to summarize biostratigraphic and lithostratigraphic data from 15 intensively sampled K/T boundary sections within a common chronostratigraphic model. Our results indicate that almost all of these sections, along with 13 additional boundary sections not used to construct the model, contain prolonged and in many cases multiple hiatuses. Of these 28 boundary sections, only six were found to contain a continuous record of sediment accumulation across the K/T boundary itself. These six K/T-complete sections are El Kef (Tunisia), Agost (Spain), Caravaca (Spain), and three sections along the Brazos River (Texas).
A comparative analysis of hiatus distributions among these 28 K/T boundary sections also reveals the presence of systematic differences between continental-shelf and deep-sea depositional environments. The lower Danian interval immediately following the K/T boundary, and extending into biochronozones P0 and P1a, is typically missing from the deep sea, whereas boundary sections deposited in shallower middle-neritic to upper-slope environments are in most cases complete across the K/T boundary. These shallow, neritic boundary sections, however, are in many instances disrupted by hiatuses at the P0/P1a boundary and again in the upper part of Zone P1a. These differential patterns of hiatus distribution between deep-sea and continental-shelf depositional settings appear to be linked to sea-level fluctuations. Our data suggest that the apparently sudden mass extinction of planktonic Foraminifera and anomalies in the occurrence of geochemical tracers that are characteristic of the K/T boundary in deep-sea sections may be artifacts of a temporally incomplete deep-sea stratigraphic record.