Southern Australia is a divergent, passive continental margin that formed during the protracted period of extension and rifting that led to the separation of Australia and Antarctica in the Cretaceous, and evolved during the subsequent northward drift of the Australian continent. Continental extension began in the Jurassic, followed by breakup in the middle Cretaceous (96 Ma) and slow spreading until the middle Eocene (49 Ma). Spreading accelerated in the middle Eocene (until 44.5 Ma), followed by faster spreading to the present (Veevers et al., 1990).
The initial extension phase prior to breakup, together with the following period of slow spreading, resulted in deep continental margin basins filled with up to 12 km of mainly terrigenous clastic sediment (Willcox et al., 1988; H. L. Davies et al., 1989). These basins broadly correspond to the sites of modern upper slope terraces (e.g., the Eyre Terrace at 400-1600 m depth in the western GAB; Figure 1). Cenozoic sedimentation resulted in an extensive, relatively thin (up to 800 m thick) Eucla basin succession deposited in a predominantly platform-sag to platform-edge tectonic regime (Stagg et al., 1990). Throughout the Cenozoic, the western GAB has been particularly stable, with geohistory analysis of exploration well Jerboa 1 indicating minimal Tertiary subsidence (Hegarty et al., 1988). Slight regional tilting (<1°) in the late middle Miocene resulted in uplift and exposure of the immense, arid, essentially featureless Nullarbor Plain adjacent to the GAB (Figure 1), and restriction of Neogene sedimentation to the modern outer shelf and upper slope.
Onset of faster sea-floor spreading in the middle Eocene also corresponded to establishment of fully marine conditions and initiation of carbonate sedimentation in the widening gulf between Australia and Antarctica. Carbonate sedimentation continued throughout the Cenozoic as the gulf evolved into a broad, open seaway, and then into the modern Southern Ocean. Oceanographic conditions in the GAB after the middle Eocene are likely to have been similar to modern conditions. The modern shelf is a broad (up to 220 km wide), gently southward-sloping surface continuously swept by long period swells (James et al., 1994). Much of the shallow shelf (<70 m deep) is bare Cenozoic limestone supporting active carbonate production (coralline algae, bryozoans, foraminifers, bivalves), but with minimal sediment accumulation; farther seaward, the outer shelf is mantled by fine, microbioclastic muddy sand (Feary et al., 1993b; James et al., 1994).
The swell- and storm-dominated oceanographic regime is the predominant physical factor presently controlling shelf sedimentation, although episodic influxes of warm, less saline water also affect carbonate production. These incursions reflect activity of the Leeuwin Current, which carries low-latitude, Indian Ocean waters south along the coast of Western Australia and then eastward into the GAB (Cresswell and Golding, 1980; Rochford, 1986). Modern records indicate that these incursions are of variable annual and interannual intensity. Cores from Australia’s Northwest Shelf and the GAB indicate that this current was inactive during glacial times (Almond et al., 1993; Wells and Wells, 1994). Warmer water faunas recovered from middle Eocene and middle-late Oligocene samples from the central and eastern GAB have been attributed to activity of a proto-Leeuwin Current (Shafik, 1990). However, the very limited sampling of this succession makes it difficult to reliably differentiate between local current activity and global paleo-oceanographic effects.
One of the most unexpected findings of the seismic stratigraphic analysis of the western Eucla basin Cenozoic succession was the identification of numerous carbonate buildups in the subsurface (Feary and James, 1995). The seismic attributes of these buildups can be used to differentiate between reefs and biogenic mounds (cf. James and Bourque, 1992).
Reefs are coral-algal structures similar to those growing today in warm-water, warm subtropical, or tropical settings. The buildups in the Eucla basin succession that we interpret as reefs have distinctive seismic characteristics: (1) they have a large height:breadth ratio, typically 1:10-1:15, and with one very large example of 1:7 (Figure 3); (2) they are clearly delineated by high-amplitude reflections resulting from high-impedance contrast between the buildup surface and enveloping sediment. In addition, disruption and low amplitude of reflections beneath these buildups and minor velocity pullup indicate that these high-amplitude reflections result in diminished transfer of seismic energy into underlying strata; and (3) on many lines, stacked buildups show that later reefs developed on top of earlier reefs. These stacked reefs merge together vertically and horizontally to form extensive complexes up to 6.5 km across and extending over 200-250 m of section. This seismic pattern is similar to that displayed by Pleistocene stacked reefs within the outer Great Barrier Reef of northeast Australia, where individual reefs represent periods of high sea level, and other phases of the sea level cycle are represented by eroded karst surfaces separating the stacked reefs (Davies, 1983; Symonds et al., 1983).
Biogenic mounds consist of small skeletons and large amounts of fine-grained sediment; these mounds grow in relatively low-energy inner shelf or slope environments, typically forming on ramps, in cooler waters (warm temperate or cool subtropical settings) compared with their reefal equivalents. The attributes of the Eucla basin biogenic mounds are (1) broad, low-relief geometry with characteristic height:breadth ratios ranging from 1:12 to 1:20 (and up to 1:30); (2) slight to moderate impedance contrast between mound margins and surrounding sediment, so that the amplitude of reflections delineating these mounds is low to moderate compared with commonly higher amplitude, continuous or semicontinuous reflections in adjacent sediments (Figure 4); (3) internal reflections within mounds are also of low to moderate amplitude, and generally mimic the mound shape of the upper mound surface, but with lower relief; and (4) although individual, isolated mounds are present in all carbonate sequences, composite mound complexes composed of numerous, apparently coalesced mounds are found in the youngest (late Neogene) sequences. Such mound complexes predominantly occur toward the landward margins of these sequences, and sequence geometry indicates that these complexes formed on the shelf, either at the shelf edge or on the inner shelf.