Can ancient shelf sand ridges be mistaken for Gilbert-type deltas? Examples from the Vryheid Formation, Ecca Group, KwaZulu-Natal, South Africa

Medium scale clinoforms within sandstones of the Vryheid Formation, northern KwaZulu-Natal, South Africa have in the past been interpreted as the products of Gilbert-type or river-dominated deltaic deposition. Detailed sedimentary logging of these rocks reveals a tempestite association, overlain by a series of moderate to very large scale climbing bedforms. These are intersected at various stratigraphic levels by massive sandstone filled channels, some with subordinate rippled bases, which are orthogonally orientated to the largest scale bedding surfaces (clinoforms) of the climbing bedform association. The sedimentary architecture is dominated by the south-easterly, south-westerly and north-westerly dipping clinoforms, all orientated oblique to the east–west trending palaeo-coastline. Local dip direction reversals are commonplace. This is at odds with Gilbert-type or river-dominated deltaic settings where: (1) sediment progrades normal to the shoreline; (2) distributary channels do not intersect the underlying deltaic foresets orthogonally; and (3) wave action is typically limited. We subsequently interpret the succession as a shelf sand ridge that evolved on a subaqueous delta, fed by sediment exchange from the upper to lower shoreface via a series of rippled scour depressions. To foster such thick (∼30 m) shelf ridge development, it is likely that these remained attached to the shoreface for a significant period of time. In light of these findings, caution should be adopted in applying a blanket deltaic interpretation of shallow marine deposits comprising similar medium scale clinoforms.