Holocene reef models generally emphasize the role of in-place and interlocking framework in the creation of a rigid structure that rises above its surroundings. By extension, a number of ancient biohermal deposits have been disqualified as “true reefs” owing to their lack of recognizable framework. Fifty-four cores from several eastern Caribbean sites (Fig. 1) clearly demonstrate that in-place and interlocking framework is not common in these reefs that are comprised of varying mixtures of recognizable coral (primary framework), loose sediment/rubble and secondary framework made up mostly of coralgal fragments bound together by submarine cementation and biological encrustation. Recovery of primary and secondary framework ranged from 22% (avg.) in branching-coral facies to 33% in intervals dominated by head corals. Accretion rate decreased as expected with water depth. However, the recovery of recognizable coral generally increased with water depth, inversely to presumed coral-growth rates.

This pattern reflects a spectrum in the relative importance of coral growth (primary construction), bioerosion, hydromechanical breakdown and the transport of sediment and detritus. The relative importance of each is controlled by the physical-oceanographic conditions at the site of reef development and will dictate both the architecture of the reef and the character of its internal fabric. We do not propose that framework reefs do no exist, as they most assuredly do. However, the fact that so many modern reefs are not dominated by in-place and interlocking framework suggests that its use as the primary determinant of ancient reefs may be unreasonable. We, therefore, propose the abandonment of framework-based models in favor of those that treat framework generation, physical/biological degradation, sedimentation, and encrustation as equal partners in the development of modern and ancient reefs alike.