Apparently homogenous "reef"-limestones built by high-frequency cycles: Upper Jurassic, SW-Germany

Central European Upper Jurassic sponge-microbial "reefs" are commonly regarded as compact, massive and homogenous facies bodies. A continuous study of borehole cores from SW Germany clearly shows a very systematic architecture of these "reefs" built by sedimentary cycles of various hierarchical levels. Altogether five orders of cycles can be distinguished: decimeter-scale microcycles building meter-scale mesocycles, which in turn are stacked to form several-meter thick macrocycles. These high-frequency cycles in turn are parts of basin-wide medium-scale and large-scale cycles reported in a previous paper. This paper documents the high-frequency micro-, meso- and macrocycles. Microcycles (0.1-0.8 m thick) usually have a sharp erosional base covered by reworked sponges, shells and intraclasts, interpreted as storm events. These bioclastic layers are colonised by a succession of thrombolitic and stromatolitic microbialites. Successive microcycles form distinct stacks that record gradually increasing and decreasing depositional energy, thus defining mesocycles (0.5-4 m thick). Several mesocycles construct macrocycles (3-25 m). The cycle stacking patterns were analysed by means of facies-proportion-diagrams. The dominant microscale cyclicity records changes in water agitation, possibly caused by fluctuating storm wave base. The sponge-microbial mounds could probably build up just to average storm wave base. Within the cycles of larger hierarchical levels the successive smaller-scale cycles show systematic trends in their symmetry and in thickness. This may be caused by variations in the ratio of accomodation space (and associated energy level) to carbonate production. External reef geometries (progradation, aggradation, retrogradation, condensation) can be related to internal cycle stacking patterns. To approximate the duration of the high-frequency cycles two different approaches were used. Firstly, the duration of cycles was determined with the help of published chronostratigraphic time scales. Secondly, cycle duration was estimated using published microbial reef growth rates. Both approximations result in ca. 0.4 Ma for the medium-scale cycles and 0.1 Ma for the macrocycles. Both are regionally correlatable, thus appear to be allocyclic in nature and probably are Milankovitch-driven. In contrast, the microcycles are only locally correlatable and thus indicate an autocyclic (storm-induced) nature. This study poses the question if not more of the apparently homogenous "reef bodies" of other settings and geological ages may possibly also show internal high-frequency cycles. (C) 2002 Elsevier Science B.V. All rights reserved.