In 1996, we expanded our research efforts to the northeast sector of the Gulf of Mexico offshore west Florida and Alabama. In contrast to much of the Texas and Louisiana coast, this area is characterized by numerous rivers with high bedload and low to moderate discharge. From west to east, the major rivers in the area are the Tombigbee, Mobile, Perdido, Escambia, Blackwater and Yellow rivers. The near-surface stratigraphy is undisturbed by faulting and salt tectonics. These factors taken together represent an end-member setting rather different from our study areas on the Texas and Louisiana shelves.

Our 800 km of high-resolution single-channel seismic data allow us to connect the previously acquired 1300 km of seismic data to age and lithologic information at a well site in offshore block Main Pass 303 Figure 1. In addition, 15 piston cores and grab samples have been acquired from key areas on the west Florida shelf.

An initial objective of this regional study was to understand how shelf-derived sands might be transported to the head of DeSoto Canyon. In contrast to Mississippi Canyon, DeSoto Canyon does not dissect the shelf edge, but is confined to the slope (Figure 2). In fact, it occurs at a water depth of approximately 450 m. Up-slope from the canyon head, there are major indentations and bulges in the trend of the shelf edge (Figure 2). We first surmised that the broad indentations might represent upper-slope canyons. However, our subsurface mapping and correlations from the seismic grids has shown that these shelf-edge features are actually lows between large shelf margin delta lobes. Nonetheless, inter-deta lobe lows may have acted as zones of sediment bypass during eustatic lowstands.

The shelf-edge morphology has varied during the last few glacial eustatic cycles. With the creation of each new delta lobe, the locations of inter-lobe depressions (potential sediment conduit to the slope) changed during each depositional cycle.

Stratigraphic analysis shows that there are seven distinct depositional episodes. Based on correlation with a top of Pliocene marker, Unconfomity A of Mitchum (1978), we infer that the seven depositional episodes represent the middle Pleistocene to present (Figure 3). During each depositional epidsode, a large lobe was deposited and most sediment was sequestered on the shelf/margin. These large lobes are distinct in their location and in their stratigraphic position. For example, the early transgressive prograding wedge strata on profile 48 (Figure 3) is located offshore Florida, whereas similar shelf margin stratigraphy at profile 37 is restricted to offshore Alabama and was deposited durng the previous highstand. Lobes are themselves composites of sub-lobes deposited during what we interpret as highstand to lowstand falling limbs of sea-level cycles. Externally, lobes exhibit lens geometry with high-angle forsets truncated by a major toplap truncation surface. Bottomset strata are condensed, but can be regionally extensive. In some cases, prograding foreses exhibit a true downlap limit.

Foreset slopes very often show major erosional relief, but not at the scale of canyon heads. At some stratigraphic levels, a base of foreset slope channel cuts into the delta lobe's clinoforms and aggrading bottomsets/onlapping wedge strata. We infer the ravinement process by which this erosional stripping of the foreset spope occurs is always active; most active at the crest and up-current flank of these delta lobes which created bulges in the trend of slope.

Seismically, sub-lobes bear resemblance to what an outcrop or well-log geologist might call a parasequence or parasequence set. In that intermediate toplap truncation surfaces within a lobe are often downlapped by foresets of the next sub-lobe. The seismic stratigraphic configuration of toplap turncation downlapped by prograding foresets (Figure 3) indicates a sizable increase in accommodation (up to 20 m between sub-lobes). Based on the distribution of toplap truncation and offlap break geometry exhibited at the culmination of each depositional lobe, we infer that most of the continental shelf was exposed during eustatic lowstands.

Local amplitude build-ups along toplap truncation surfaces are interpreted as relatively wide and disorganized shallow braided fluvial feeder systems (Figure 4). In spite of this fluvial pattern, the deltas are distinctly lobate. Disorganization and braiding of the fluvial system is believed to be a result of the relatively high-gradient and sandy nature of the lowstand exposed shelf.

Our analysis shows that sandy-shelf sediments remain essentially intact on the shelf during eustatic lowstands. Detectable shelf-edge/upper slope canyon topography was not observed in seismic profiles. It does not appear likely that the shelf margin suffered major slope failure during the middle Pleistocene to recent eustatic cycles. The apparent lack of canyons cutting into the shelf edge may be related to the sandy nature of the shelf, which might have inhibited organized drainage on and off the shelf during eustatic lowstands. Foreset slope ravinement seems to have been the most likely mechanism by which sediment escaped the shelf-margin. Base of foreset slope channels probably scoured into silts and fine sands of the prodelta. The sediments generated from this scouring were deposited on the upper slope as strike-aligned onlapping wedges.

For a more detailed discussion see Bart and Anderson (2004).

Highstand Systems Tract

Lowstand Systems Tract

Transgressive Systems Tract

Gulf Bottom Imagemap

Comments, questions? Contact us at