The Tumblagooda sandstone is a geological formation deposited during the Silurian or Ordovician periods, around four to five hundred million years ago, and is now exposed on the west coast of Australia, straddling the boundary of the Carnarvon and Perth basins. Visible trackways may prove to be the earliest evidence of terrestrial animals.
The Tumblagooda is over 1,400 m (4,500 ft) deep: the bottom has never been found, but seismic data suggests it unconformably overlies a Proterozoic basement. The formation is divided into four facies associations (FAs), numbered stratigraphically, that is from bottom to top. These lithified sediments portray an environment dominated by high-energy braided streams, flowing into the sea in places; ephemeral pools were also common.
Facies Association 1
The lowest facies association in the unit is dominated by trough cross-stratification, deposited by broad, high-energy braided rivers, which formed the outwash plain of an alluvial system. Trace fossils are virtually absent, because the high depositional energy meant burrowing organisms could not survive. The downslope flow was to the north west.
Facies Association 2
These facies reflect a quieter environment; the unit is occasionally interrupted by lenses of FA1 sediments. Beds are on the whole thin, planar and well sorted. Beds about 5 centimeters (2.0 in) thick form 2 meters (6.6 ft) units of "bedded sandsheets"—layers of sand blown by the wind—which form the dominant lithology of this facies.
Low angle (<20°), cross-stratified sandstones form units up to 50 centimeters (20 in) thick, sometimes reaching thicknesses as much as 2 meters (6.6 ft). The current's direction here is to the south east - up slope - and reinforces their interpretation as aeolian dunes. A further suite of layers contain a dense trace fossil assemblage; other layers bear current ripple marks, which likely formed in shallow streams, with flooded hollows probably hosting the creators of the trace fossils. Cyclicity is absent, suggesting that, rather than being seasonal events, the occasional inundation was based on unpredictable events such as storms, a varying water table, and changing stream courses.
Facies Association 3
This facies is much like FA1, with an increased supply of clastic material represented in the sedimentary record by coarse-grained, poorly sorted, upwards-fining (i.e. largest grains at the bottom of the unit, becoming progressively finer towards the top), pebbly trough cross-bedded units up to four metres thick. Trace fossils are rare. Sheet-like braided rivers are inferred as the dominant control on sedimentation in these facies.
Facies Association 4
The uppermost facies association appears to reflect an environment on the fringes of the sea. Fining-up is observed on 0.5 meters (1.6 ft) to 2 meters (6.6 ft) scales, with trough cross bedding at the bases of units overlain by current ripples. Fine sandstones and green shales are also present. The upper units are strongly bioturbated, with an abundance of Skolithos - a fossil typically found in marine environments.
It has been interpreted as an inter-distributary bay, or alternatively as a sandy coastline featuring wave-generated bars, perhaps with tidal influence; braided fluvial streams often reworked the sediments.
Since the Tumblagooda sandstone comprises a sedimentary succession with no volcanic layers (which could be dated radiometrically) and with virtually no body fossils, its age is very difficult to constrain. It was first thought to have formed around , during the Cretaceous period, on the basis of stratigraphy; current estimates place it far earlier, in the lower Ordovician, . It is hoped that a new technique based on the uranium-thorium dating of diagenetic monazite crystals may produce a more precise estimate of the age, but initial attempts have failed to extract sufficient monazite from the unit.
Such a method would be of great value, as previous attempts to date the unit have been rather inconsistent. The initial Cretaceous estimate was soon reviewed with a "mid-Cambrian to early Ordovician" (~ and was shortly afterwards replaced by a mid-Silurian age based on spores and acritarchs. This was apparently confirmed by the identification in the overlying beds of a conodont fauna with a recognizably Silurian character, but when the palæomagnetism of the area was studied, an early Ordovician age was deduced. A single conodont element, again drawn from overlying sediments, was taken to support a late Cambrian to early Ordovician age, but this was refuted by the observation that the trace fossil assemblages bore great similarity to well constrained lower Silurian assemblages from Antarctica. The current early Ordovician age estimation is based on a much more diverse and numerous—therefore more securely dated—assemblage of conodonts, again from overlying sediments.) estimate based on trace fossils,
In common with most of the Australian cratonic rocks, the Tumblagooda has undergone minimal tectonic activity since its formation. Therefore faulting is rare, and units continue laterally for great distances. Jointing is the dominant control of the landscape, with meandering rivers enhancing joint locations. Miocene uplift has resulted in the formation of deep (approximately 60m/200 ft) gorges exposing large cliff sections; with the sparse vegetation characteristic of arid Western Australia, this means the unit is exceptionally exposed, making detailed study easy. Despite this, the unit was not studied until 1948, due to its inaccessibility - the 600 miles (1000 km) from Perth were mainly dirt tracks until the 1970s.
Only one body fossil, Kalbarria (an early euthycarcinoidic arthropod) has been found in the Tumblagooda, mainly due to the large clast size and the abundance of predatory and burrowing organisms. (This meant that oxygen could penetrate to good depths in the sediment, permitting decomposing organisms to decay anything that burrowing animals had not eaten too rapidly for fossils to form.) Since Kalbarria had 11 pairs of legs, it can be tentatively matched to some Protoichnites arthropod trackways of the same size. Protoichnites is abundant in subaerial facies in FA2-4. Marks which can only have been made on exposed wet sand are seen: for example "splurges" where the legs of the organism flipped sand out behind them. The marks vary in crispness and character according to the wetability of the underlying sediment; this is particularly marked where the traces cross ripples, with the lee slopes recording a trace markedly different in appearance to those in the troughs, and the stoss slopes recording no trace at all. Behaviour can be inferred from these traces; in places, they parallel features which modern observation notes forming at the edge of a wind-blown pond, just on the landward side of the shore. This behaviour has been interpreted as a feeding trace; presumably the trace-maker dined on organic matter blown out of the pool, or detritus left as the pool had shrunk. Further tracks can be traced across dunes; a slow walk up turns into a skid as the organism slid down the lee slope and into the pool on the other side. Another instance shows the trackways of two organisms converging, then becoming one trackway, before one individual swerves away to the left, leaving the other to walk onwards. These trackways are the earliest evidence of terrestrial animals. Due to the poor dating of the unit, it is currently impossible to speculate whether the plants, which colonised the land in the mid-Ordovician, got there first.
Aquatic trace fossils are also abundant. Two major ichnofacies are observed, bearing close resemblance to assemblages found in Antarctica and demonstrating proximity of western Australia and Antarctica at the time of deposition. One is dominated by Skolithos, suggesting marine deposition. The fabric of the other is dominated by Heimdallia, a strange planar trace that does not have a circular cross-section; it is interrupted in places by Beaconella, a wide trace thought to be constructed by a burrowing arthropod ploughing through the sediment for food, leaving a mound of piled sediment at the end of each trace. This may suggest the Beaconella organism feeding on Heimdallia. Daedalus, a trace fossil resembling a giant garlic, is also present in this assemblage.
- Trewin, N.H.; McNamara, K.J. (1995). "Arthropods invade the land: trace fossils and palaeoenvironments of the Tumblagooda Sandstone (? late Silurian) of Kalbarri, Western Australia". Transactions of the Royal Society of Edinburgh: Earth Sciences 85: 177–210. doi:10.1017/s026359330000359x.
- Clarke, E. de C. & Teichert, C. (1948). "Cretaceous stratigraphy of the lower Murchison River area, Western Australia". Journal of the Royal Society of Western Australia 32: 19–48.
- Iasky, R.P.; Mory, A.J. (1999). Geology and Petroleum Potential of the Gascoyne Platform, Southern Carnarvon Basin, Western Australia. Geological Survey of Western Australia. ISBN 0-7309-6634-8.
- Opik 1959
- Balme 1962
- Hocking et al. 1987
- Schmidt and Hamilton 1990
- Gorto et al.. 1994
- McNamara, K.J.; Trewin, N.H. (1993). "A euthycarcinoid arthropod from the Silurian of Western Australia". Palaeontology 36: 319–335.
- Gray, J.; Chaloner, W. G.; Westoll, T. S. (1985). "The Microfossil Record of Early Land Plants: Advances in Understanding of Early Terrestrialization, 1970-1984". Philosophical Transactions of the Royal Society of London. Series B, Biological Sciences 309 (1138): 167–195. Bibcode:1985RSPTB.309..167G. doi:10.1098/rstb.1985.0077. JSTOR 2396358.