Ediacaran Life on Land

Australia: The Land Where Time Began

Ediacaran Life on Land

For some time the general view of the Ediacaran fauna, 635-542 Ma, has been that they were the ancestral forms of the marine invertebrate phyla¹ that gave rise to the forms associated with the Cambrian Explosion, the rapid diversification of marine animals that occurred in the Cambrian, including giant marine protists², and as lichenised fungi³. Palaeosols in the Ediacaran Member of the Rawnsley Quartzite in South Australia⁴ have now been documented to confirm earlier interpretations of deposition of lagoonal-aeolian origin based on synsedimentary ferruginisation and loessic texture^5,6. Non-marine facies have also provided evidence for palaeosols, as published in the Nature paper by the author¹, dilation cracks, soil nodules, sand crystals, stable isotope data and mass balance geochemistry⁴. A variety of fossils are present on the surface of the palaeosols that are in growth position, which includes Charniodiscus, Dickinsonia, Hallidaya, Parvancorina, Phyllozoon, Praecambridium, Rugoconites, Tribrachidium and 'old elephant skin' ichnogenus Rivularites⁷. As with plant fossils⁸, these fossils were preserved as ferruginous impressions, and biological soil crusts^9,10 from the sandy palaeosols from the Phanerozoic. Shallow within the palaeosols, sand crystals after gypsum¹¹ and carbonate nodules¹², even following correction for burial compaction¹³. A dry, cold temperate palaeoclimate is indicated for South Australia⁴ during the Ediacaran, based on periglacial involutions and a modest degree of geochemical differentiation of the palaeosols. The new interpretation of some fossils from the Ediacaran, as presented by the author¹, with large, sessile organisms living on cool, dry soils is consistent with observations that the appearance of some Ediacaran fossils, and their preservation, were similar to those of lichens and other microbial colonies present in biological soil crusts³, instead of their usual interpretation as marine animals¹ or protists².

The author¹ suggests the new evidence from the Rawnsley Quartzite⁴ indicates a re-evaluation of the Ediacaran fossils, that are widely believed to be the ancestral forms of the marine animal phyla involved in the Cambrian explosion¹, is in order. Douglas Mawson and Ralph Segnit⁵ originally suggested the red beds in South Australia, of Ediacaran age, were of terrestrial origin. Reginald Sprigg, a student of Mawson, who discovered the Ediacaran fossils in South Australia, interpreted the fossils as marine jellyfish¹⁴, this apparent contradiction of the palaeoenvironmental conditions under which the fossils were preserved in the Ediacaran was eventually resolved by the compromise that they had been washed up onto tidal flats during storms¹⁵. Fossils from the Ediacaran have now been found at a variety of sedimentary facies¹⁶ around the world that have generally been interpreted as marine deposits, from shallow to deep, following on from Sprigg's¹⁴ interpretation of them, based on comparisons with marine animals, though it is becoming increasingly likely that such comparisons are doubtful. Most of the Ediacaran fossils appear to have no clear relationship with animals of the present^2,3,16,17. The author¹ suggests putative 'embryos' from the Neoproterozoic were more likely to have been protists¹⁸. He also suggests that putative metazoans that had been permineralised may actually been vughs¹⁹ (small holes, cracks inside a rock) that were crystal lined, and other fossils of Ediacaran age that had been permineralised (process of fossilisation involving mineral deposits forming internal casts of the organism) were actually red algae or glomeromycotan lichens²⁰. The shallow trails from the Cambrian may have actually been made by slime moulds in their slug aggregation phase and so not by worms²¹. Ediacaran fossils have been suggested to have been giant protists², such as xenophyophores, or fungi such as a variety that forms an association with an alga to form lichen³.

According to the author¹ as a result of palaeosols being less strongly developed in the Ediacara Member than palaeosols at Precambrian unconformities, or formed beneath Devonian forests, as well as from later periods⁴, they have been overlooked in the past. In the field, as well as obvious soil structures (platy peds) and horizons (A-Bk and A-By), disruption of the beds on scales that are predominantly microscopic contributes to the massive appearance of palaeosols compared to that of sedimentary rocks. There are 4 distinct episodes of soft sediment deformation, which is followed by 'successive weathering and bedding disruption of previous episodes, comparable with successive periglacial soil involutions, rather than seismic or load deformation⁴'¹. There is more evidence for palaeosols from the Ediacaran that has been detailed elsewhere:

geochemical mass-balance negative strain and cation mass transfer;
loessic grain-size distribution and texture;
compositions of oxygen and carbon isotopes that are unusually light showing linear covariance; and
gypsum sand crystals and micritic replacive nodules that have a consistent depth from the tops of the beds.

In Brachina Gorge there are 47 different stratigraphic levels that show pedogenic features in the form of 5 distinct palaeosol types that are repetitive, the pedotyp names deriving from the indigenous Adnamatna language.

Deep weathering in the Cretaceous or later has been believed to be the cause of the red colour and weathering of the rocks in the Flinders Ranges, but according to the author¹ several observations have falsified this view. In the Ediacara Hills drill cores have shown grey shales and limestones to overlay red beds of the Ediacara Member, the cores showing the same unusual, distinctive carbonate carbon isotope composition and the composition of major elements as have been found in outcrops in Brachina Gorge⁴. In the Ediacara Member palaeochannels of grey sandstone contain deposits that include red clasts, that were redeposited from the Ediacara Member and Bonney Sandstone, together with grey calcareous clasts derived from the Wonoka formation, though if they had been weathered downwards from the land surface of the present they should all be red. The outcrop of the Ediacara Member has been analysed by a variety of crystallinity indices that includes X-ray diffraction and microprobe analyses with the results that the Ediacara Member outcrop and core is illite-chlorite that has been heated to low within the greenschist metamorphic facies⁴. The sandstone of the Ediacaran Member have been found to contain abundant feldspar, and carbonate nodules are present within the red siltstones. Neither feldspar nor carbonate are present in deep weathering profiles, and have kaolinite clays and their chemical composition is strongly depleted in alkali and alkaline earth elements. As both palaeosols and sediments can be derived from soils in source terrains palaeosols cannot be distinguished from sediments based on their colour and degree of weathering, but the relative timing of weathering, and weathering is crucial in an aeolian-terrestrial interpretation⁵.

A surface texture termed 'old elephant skin', that is preserved best beneath sandstone beds, distinguishes palaeosols and fossils in the Ediacaran Member⁷. The ichnospecies Rivularites repertus is suggested by the author¹ to be validly named for comparable surfaces that are cracked and pustulose⁷. Cracks that have healed, irregular fine ridges (cracks in cover sandstone) and a relief of intergrown radial growth centres that give it a pustulose appearance, all combine to make it look old. There are a variety of features of these surfaces, that are distinctively cracked and pustulose, that give it an appearance that is more like the biological soil crusts that are found on the surface in deserts and tundra of the present than the hydrated microbial mats found on intertidal flats and shallow seas that are parallel-wrinkled and undulose. Aquatic microbial mats are laminated and detachable from substrate as flakes, skeins and roll-ups, structures that are not found in the Ediacara Member. Biological soil crusts and their soils differ in that the crusts and their soils have organisms that are oriented vertically, and are admixed intimately with minerals of the soil. Even in sandstone that is clay-pore, soil crusts have irregular relief, desiccation cracks that have healed and pressure ridges, while in microbial mats there are flexuous striated domes and tufts of a form that are not seen in the Ediacaran Member. Microbial mats form caps above sedimentary layers that not weathered, and are chemically reduced. Soil crusts are the upper parts of deeper soil profiles in which oxidation is varied downward, the abundance of clay and replacive nodular subsurface horizons like palaeosols found in the Ediacaran Member. Over time soil profiles that are increasingly differentiated are developed by soil crusts, microbial mats form laminar to domed (stromatolitic) increments by building upwards. Sandy palaeosols dating to the Ordovician⁹ and Cretaceous¹⁰ are known that contain fossils in the form of impressions with a lichen-like appearance. In red sandy palaeosols from the Cretaceous⁸ comparable preservation of terrestrial vascular plants is well k known.

Fossils from the Ediacaran have been found as impressions in old elephant skin sandstones above 4 of the 5 different types of palaeosol from 10 different horizons in 4 of the classic locations. The overhang at 39.7 m in Brachina Gorge is the best place to see Ediacaran fossils in place above a palaeosol, 2 good specimens that are still present, Dickinsonia costata and Pseudorhizostomites howchini. In life the body fossils were attached firmly of embedded in these soil surfaces, as is shown by the growth series, specimens not being found overlapping, and the thickening of specimens that are adjacent which is comparable with competitive reaction. Some taxa are embedded in the surface layer as do window lichens and rhizines (bundles of hyphae) in desert crusts³, e.g. Phyllozoon hanseni and Aulozoon sp. Branching tubular structures are seen extending deep into the palaeosols in petrographic thin sections, as occurs with cyanophyte ropes, fungal hyphae and lichen rhizines of desert crust soils of the present⁷. According to the author¹ when thin sections of the lower half of the impression of a fossil of Dickinsonia is studied petrographically comparable bedding disruption by irregular tubular structures is shown. Upper impression fossils, that are the most common type found, show that only bedded sandstone is present above the fossil. In surface horizons of the palaeosols comparable bifacial fossil features were found that had smooth and finished upper surfaces and ragged lower surfaces. As the definition of Ediacaran fossils has been by shape rather than thin section appearance it is uncertain which specific Ediacaran fossils are represented by these thin sections, the author¹ suggesting the images of complex chambered structures that have basal tubules can be viewed as preliminary indications of their appearance in thin sections that had been prepared for a detailed study in progress. When viewed in petrographic thin sections the preservation style of Ediacaran fossils is comparable with that of plants and lichens that have been recovered from palaeosols dating from the Phanerozoic^8-10.

A number of body fossils have so far been recovered from palaeosols, including 'Aulozoon' sp indet., Charniodiscus arboreus, 'Cyclomedusa davidi' Dickinsonia costata, D. elongata, D. rex, Hallidaya brueri, cf 'Kimberella sp. indet., 'Medusinites asteroides', Parvancorina minchami, Phyllozoon hanseni, Precambridium sigillum, Rugoconites enigmaticus, Sprigging floundersi and Tribrachidium heraldicum. In the palaeosols trace fossils have also been found that include Archaeonassa sp. indet. Pseudorhizostomites howchini, cf. 'Radulichnus' sp. indet. and Rivularites repertus. There is a diverse assemblage of fossils in some palaeosols, such as Muru and Wadni pedotypes, that is dominated by Dickinsonia, while in pedotypes, such as Yaldati and Inga, the fossil assemblage is of low density, being comprised mostly of discoid fossils that have been variously attributed to microbial colonies, medusoids or holdfasts, of taxa such as Medusinites and Cyclomedusa. Deformation that has been interpreted as periglacial involutions⁴, the Warrutu pedotype, have not yielded fossil specimens in outcrops, though it has been shown that they may have Ediacaran fossils in thin sections. The author¹ suggests these diversity differences may be evidence that in the Ediacaran organisms preferred soils that were not frozen, of low salinity, and nutrient-rich, as do most terrestrial organisms of the present.

It is suggested by the author¹ that as palaeosols developed there may have been an increase in the size of fossils as well as their diversity. In the Sinai and Negev Deserts in Israel the relationship between the abundance of gypsum (G in area(%)) and the geological age (A in kyr) is given by the following equation:

A = 3.987G + 5.774

There is a standard error of ± 15 kyr for this relationship. In the main collecting gully in the Ediacaran Hills the largest Dickinsonia recovered from the Muru palaeosol that has 10 % gypsum, D. rex¹², measured 32 cm long, indicating that, according to the above equation it was 53.6 ± 15kyr in the making. The long-term growth for this D. rex (0.006 ± 0.002 mm/yr) would have been more similar to lichens of the present, such as Rhizocarpon geographicum from the Brooks Range in northern Alaska, that grows at 0.04 mm/yr, than lichens from southern alaska²⁵ that have growth rates of 0.1 mm/yr, or in southern Norway²⁶ 0.5-0.7 mm/yr. It is suggested by the author¹ that between the largest Dickinsonia found in a particular collection and the enrichment of gypsum in the relevant palaeosol there is a correlation, evidence that in Dickinsonia there was slow growth rates that were non-linear. In the Ediacaran Hills is one of the most diverse assemblages known, indicating that organisms may have diversified as well as grown with the age of the palaeosol, comparable with terrestrial communities of the present.

Comparison of palaeosols found in the Ediacara Member with soils of the present give some indication of the palaeoclimate in the Ediacaran. According to the author¹ the closest modern analogue to the gypsic palaeosols of the Ediacaran appear to be soilscapes found on the coastal plain of the Caspian Sea near Atyrau, Kazakhstan, where the mean annual temperature is 8^oC and the mean annual precipitation is 160 mm. In the nearby Emba River floodplain there are calcic floodplains. Evidence for a cold temperate palaeoclimate has been found in the form of dropstones in the coeval Billy Springs Formation, and this has been supported by periglacial involutions in one palaeosol⁴. Further evidence of a cool dry climate has been included in supplementary information for this article in Nature. The Flinders Ranges are near the gypsic-calcite soil ecotone, which is the zone of best preservation of large Ediacaran fossils, and in deserts of the present it is an important biotic boundary such as seen in the Atacama Desert²⁹ and the Negev Desert²³.

The author¹ points points out that the discovery in the surface horizons of palaeosols of some Ediacaran fossils does not imply that all Ediacaran fossils everywhere were terrestrial. In the Ediacaran Member⁴ intertidal to shallow marine facies have been recognised, though at the time of writing they have not produced any fossils, an exception may be Palaeopascichnus that has been found in both old elephant skin surfaces of the Ediacaran Member, that are here interpreted as terrestrial, as well as in pyritic black shales of the Wonoka Formation that have been interpreted as of marine origin. Such multisegmented fossils as Dickinsonia, Charnia, Praecambridium and Spriggina, recovered from palaeosols of the Ediacaran Member are more likely, according to the author¹, to be either lichens or some other consortia of microbes³ than marine invertebrates¹ or giant protists³. Discoid fossils, such as Cyclomedusa, Medusinites and Rugosoconites[?] could have been microbial colonies³⁰, though they could not be jellyfish in such dry soils. If small fossils such as Parvancorina or Tribrachidium could not have been pre-trilobites or proto-[sea?]-stars if they were terrestrial, if they liven on dry land they could possibly have been the fruiting bodies³ of fungi. It has been suggested that metazoan slugs or worms could have created trace fossils such as Archaeonassa after rainstorms on land, but there is also the possibility that in terrestrial habitats these trails were created by the slug-aggregating phase of slime moulds²¹. It has also been suggested that impressions recovered from the Ediacaran Member, such as 'Radulichnus', are too straight and sharp to have been scratches made by the radula of molluscs, and if terrestrial, may have been casts of needle ice. When considered in a marine setting¹⁶ Pseudorhizostomites has been interpreted as gas-escape structure, though it is most like a flanged pedestal of a biological soil crust⁷. The recognition of palaeosols and their comparison with soils and soil processes of the present has led to surprising alternative terrestrial hypotheses for habitats and affinities of these enigmatic fossils. The author¹ suggests these unconventional ideas and comparisons are still to be tested for the various Ediacaran fossils in South Australia, as well as Ediacaran fossil assemblages in other parts of the world.