Nature template - PC Word 97 · sedimentary facies and palaeosols (Fig. S2, Tables S1- 2) supports...

Geological background

Fig. S1. Map and general section of the Flinders Ranges, South Australia. a) isopach map revealing palaeovalleys at the base of the Ediacara Member and key localities for Ediacaran fossils1; b) geological section of alternating red and grey beds1.

WWW.NATURE.COM/NATURE | 1

SUPPLEMENTARY INFORMATIONdoi:10.1038/nature11777

Fig. S2. Conjectural palaeoenvironmental reconstruction of palaeosols and sedimentary facies of the Ediacara Member, Rawnsley Quartzite, Flinders Ranges, South Australia.


SUPPLEMENTARY INFORMATIONRESEARCHdoi:10.1038/nature11777

The Ediacara Member of the Rawnsley Quartzite is red siltstone and sandstone within palaeovalleys1 in a thick Neoproterozoic-Cambrian sedimentary sequence in the Flinders Ranges of South Australia (Fig. S1). The Ediacara Member is famous for exceptional preservation of unskeletonized large fossil impressions2. It is generally a recessive-weathering unit within strike ridges of thick folded Rawnsley Quartzite, which forms the most impressive peaks of the Flinders Ranges. Recent reassessment of its sedimentary facies and palaeosols (Fig. S2, Tables S1- 2) supports earlier interpretations of the Ediacara Member as coastal floodplain and lagoonal deposits3-6, rather than later interpretations as submarine canyon and fan deposits7-8. This research reports new observations of Ediacaran fossils in the surface horizons of palaeosols, as well as palaeoclimatic and other palaeobiological reinterpretations arising from discovery of palaeosols in the Ediacara Member1. Table S1. Summary of Ediacara Member palaeosol definition and classification Pedotype Aboriginal

meaning Diagnosis US taxonomy FAO map unit

Inga Hard Grey sandstone capped with algal wrinkling and expansion cracks, with shallow and sparse red mottles (A horizon), over deeper sand-crystal rosettes (By horizon)

Haplogypsid Gleyic Solonchak

Muru Light-coloured rock

Red massive siltstone to sandstone surface (A horizon), over a shallow horizon of hard, ellipsoidal, non-calcareous nodules with enclosed crystal pseudomorphs (By horizon)

Haplogypsid Gypsic Yermosol

Wadni Small Intensely red ferruginized surface (A horizon) with expansion cracks and grey mottles, within bedded siltstone (C horizon),

Fluvent Eutric Fluvisol

Warrutu Undulating ground

This red, massive siltstone with pedoturbation of surface anticlines (A horizon) and subsurface load structures (Bw horizon), above deep horizon of hard, ellipsoidal, non-calcareous nodules with enclosed crystal pseudomorphs (Bw horizons)

Gypsic Anhyturbel

Gelic Cambisol

Yaldati Red Red siltstone with intensely reddened and expansion-cracked surface (A horizon) and some relict bedding, above ellipsoidal calcareous mottles (Bk horizon).

Haplocalcid Calcic Xerosol

Table S2. Summary of Ediacara Member palaeosol interpretation Pedotype Climate Organisms Topography Parent material Soil durationInga Arid (MAP 139±17 mm

to 257±129 mm) “Cyclomedusa davidi” Near stream

floodplain Waterlain quartz sand

50-1000 years

Muru Cold temperate (6.47±0.67oC), arid (MAP 204±129 to 267±129 mm)

Dickinsonia costata, D. elongata, D. rex, Charniodiscus arboreus, Tribrachidium heraldicum, Parvancorina minchami, Kimberella sp. indet., Spriggina floundersi, Praecambridium sigillum, Archaeonassa sp. indet.

Well drained floodplain

Aeolian silt-sand, quartzofeldspathic

500-2000 years

Wadni Not diagnostic for climate

“Medusinites asteroides”, Pseudorhizostomites howchini, Paleopascichnus sp. indet., Dickinsonia costata, Parvancorina minchami,

Near stream floodplain

Aeolian silt, quartzofeldspathic

5-100 years

Warrutu Cold temperate (MAT 7.62±0.67oC), semi-arid (579±129 mm)

None found Well drained floodplain

Aeolian silt-sand, quartzofeldspathic

500-1000 years

Yaldati Cold temperate (7.95±0.69oC), semiarid (344±147 to 407±147mm)

“Cyclomedusa davidi” Well drained floodplain

Aeolian silt, quartzofeldspathic

100-2000 years



Localities and collections of Ediacaran fossils examined Ediacaran fossils have been found in a variety of localities within the Flinders Ranges (Fig. S1). At the following localities, specific observations were made for this study of Ediacaran fossils in palaeosol profiles. Measurements of the longest Dickinsonia fossil and key palaeosol variables at these sites are presented in Table S3. 1. Crisp Gorge (1 level). Helicopter recovery in 1992 and 1994 by Jim Gehling from

Crisp Gorge, 6 km south of Parachilna Gorge (S31.176572o E138.528533o)9-13 included a large slab with reddened and textured surfaces (“old elephant skin” or trace fossil Rivularites repertus1), from above a Wadni palaeosol. This slab mounted in the South Australian Museum, has 148 small Dickinsonia costata, as well as Hallidaya brueri, Parvancorina minchami, Rugoconites enigmaticus, Spriggina floundersi and Tribrachidium heraldicum14.

2. Bathtub Gorge (1 level). In Bathtub Gorge (S31.24711o E138.53655o), a Muru palaeosol was examined beneath remnants of the large slab collected by Jim Gehling and Bruce Runnegar in 1992 and 1994, now on display in the South Australian Museum, with Dickinsonia costata, Phyllozoon hanseni, Pseudorhizostomites howchini, and “Aulozoon” 10, 15-16.

3. Brachina Gorge (7 levels). Brachina Gorge is a publicly accessible and protected location beside the main access road to Flinders Ranges National Park. Collection of Ediacaran fossils in Brachina Gorge (S31.34422o E138.55763o) recovered Dickinsonia costata, and Rugoconites enigmaticus (=“Tribrachidium heraldicum” of ref. 17) from atop a Muru palaeosol (42.6 m in section of ref. 1), Dickinsonia costata from atop another Muru palaeosol (41.7 m), and “Medusinites asteroides” (junior synonym of Aspidella terranovica18) from the base of ripple marked sandstone above a Wadni palaeosol (34.4 m). Fossils observed in place in Brachina Gorge in 2006-7 included Dickinsonia costata and “Pseudorhizostomites howchini” in the sole of the deformed layer capping a Wadni palaeosol (39.7 m); Charniodiscus arboreus, Dickinsonia elongata, Parvancorina minchami, and cf “Kimberella” sp. indet. (unusually large and effaced) from atop Muru palaeosol (40.5 m); and “Cyclomedusa davidi” (also Aspidella terranovica18) from atop Inga and Yaldati palaeosols (0.2 and 13.2 m respectively).

4. Ediacara Hills (1 level). Fossils observed in the Ediacara Hills in 2007 (“Spriggina Gully” at S30.79423o E138.13276o and “Praecambridium spur” at S30.79318o E138.13579o) were in loose slabs with ice-needles (“Radulichnus”) and old elephant skin (Rivularites repertus) above sandstones with “peachstone” nodules of Muru palaeosols, and included Dickinsonia costata, D. rex., Spriggina floundersi, Praecambridium sigillum, Tribrachidium heraldicum, Parvancorina minchami, “Pseudorhizostomites howchini”, and Archaeonassa sp. indet.

Table S3. Size of longest Dickinsonia in different palaeosols.

Location Species Number of specimens

Length (cm) Palaeosol Depth to gypsum

(cm)

Gypsum %

Crisp Gorge D. costata 148 3 Wadni n/a 0 Brachina 39.7 m D. costata 2 6 Wadni n/a 0 Brachina 40.5 m D. elongata 3 6 Muru 25 6 Bathtub Gorge D. costata 4 9 Muru 23 5

Brachina 41.7 m D. costata 1 16 Muru 29 8 Brachina 42.3 m D. costata 4 19 Muru 25 7 Spriggina Gully D. rex 97 32 Muru 28 10



Additional notes on clay minerals and their crystallinity in the Ediacara Member The notion that red colour and other pedogenic features of the Ediacara Member are due to Cretaceous or later deep weathering8 was addressed with a variety of studies. Deep weathering is falsified by field observations of claystone breccias with both red clasts and green clasts in a grey matrix, and alternating grey and red beds in outcrop and deep boreholes1,3,5,19. Deep weathering is also falsified by little weathered chemical composition of Ediacara Member palaeosols, which include micritic carbonate nodules and much feldspar in both outcrop and in deep boreholes1. Finally the idea is falsified by the nature of clays (Fig S3) and their crystallinity index (Table S4), as previously reported for South Australian Cambrian palaeosols19. There is no kaolinite in these rocks, as would be expected in a deep weathering profile. The high degree of crystallinity progressively higher in more deeply buried rocks (Fig. S4), corroborated by correlated alternate crystallinity indices (Fig. S5), are evidence of alteration expected for such deeply buried rocks of lower greenschist facies metamorphism. Weaver indices of illite crystallinity (10Å/10.5Å peak height) increase down section (Fig.S4) beyond 2.3 into greenschist facies regional metamorphism20. Low grade metamorphic alteration also is indicated by Kübler indices (width at half height of 10Å peak) less than 0.42 and Weber indices (width at half height of 10Å peak standardized to quartz 100 peak) less than 181 (Table S4). Chemical composition of clays by electron microprobe confirmed that these clays are illitic1, and reveal distinct iron and magnesium varieties, presumably from deep burial alteration of smectites. Bulk chemical composition of palaeosol samples in both core and outcrop also reveal illitization trends (Fig. S6). Table S4. Clay crystallinity indices for Cambrian and Ediacaran palaeosols Age Ma burial depth km m level sample Weaver index Kübler index Weber index 483.10 0.993 5607 F45 Matarra 2.2 1.1 275 483.90 1.058 5542 W26 Adla 2.3 1 250 504.16 2.705 3895 F34 Warru 2.9 0.7 233 507.75 2.997 3603 F15 Irkili 3.0 1.05 210 507.77 2.998 3602 F2 Natala 3.1 0.75 214 507.78 2.999 3601 F10 Viparri 3.6 0.45 180 522.50 4.196 2404 W12 Warru 3.1 0.7 200 538.72 5.515 1085 B108 Madla 3.4 0.8 200 538.81 5.522 1078 B86 Mata 3.1 0.9 180 538.84 5.524 1076 B94 Vidnapa 4.0 0.6 150 538.90 5.529 1071 B99 Watuna 4.0 0.5 167 539.13 5.548 1052 B80 Arrari 4.5 0.5 143 552.34 6.956 4653 B14 Muru 3.1 0.55 138 552.36 6.957 4652 B20 sediment 4.4 0.6 171 552.37 6.958 4651 B24 Warrutu 3.4 0.4 160 552.51 6.966 4643 B5 Yaldati 3.3 0.65 163 554.52 7.084 4525 B45 Muru 4.4 0.55 157 554.86 7.104 4505 F38 Yaldati 4.2 0.4 133 555.11 7.119 4490 B34 Inga 3.1 0.5 143



Fig. S3. X-ray diffractogram traces for palaeosols and sediments of the Ediacara Member and Bonney Formation, showing chlorite and illite, but no kaolinite or smectite. Stratigraphic levels for Ediacara Member are as in Fig. 2, and for Bonney Sandstone are meters below the Chace Quartzite along the south side of Brachina Gorge (S31.34374o E138.56279o).

Fig. S4. Increased clay crystallinity of selected Cambrian and Ediacaran Formations in the Flinders Ranges of South Australia, showing greenschist facies metamorphism of the lower part of the sequence.



Fig. S5. Correlation of different clay crystallinity indices from Cambrian and Ediacaran palaeosols and sediments of South Australia.

Fig. S6. Illitization trends in chemical analyses of Ediacaran palaeosols in outcrop and core of the Ediacara Member of South Australia.



Additional notes on age model and times of Ediacaran palaeosol formation

Two separate proxies for duration of soil formation are based on size and abundance of gypsum and carbonate nodules. First, the diameter of carbonate nodules (S) is related to soil age (A in kyrs) by equation S1 (R2=0.57; S.E.± 1.8 kyrs, ref. 21). A = 3.92S0.34 — equation S1

Second, the relationship between gypsum abundance (G in area %) and geological age (A in kyrs) in the Sinai and Negev Deserts of Israel22-23 is given by equation 1. Burial compaction has a minor flattening effect on carbonate nodules and gypsum sand crystals, but cannot alter their horizontal diameter because of lithostatic pressure from the side24. Such nodules are characteristic of desert soils of low vascular plant productivity, where nodule precipitation is regulated by soil microbes25. These proxies are thus applicable to palaeosols predating the Siluro-Devonian evolution of vascular land plants26-27.

These chronofunctions for modern gypsic and calcic soils can be used to calculate individual palaeosol durations (Tables S5-S6) and also changing rates of rock accumulation during deposition of the Ediacara Member. Intervening floods and other sedimentation events occupied insignificant lengths of time (days to weeks) compared with soil durations (102-105 years), so can be ignored in this kind of analysis28-29. Adjusting the time scale of Ediacara member deposition for palaeosol durations, the calcic palaeosol interval can thus be seen as a brief (ca. 34 kyrs) semiarid episode interrupting persistently arid conditions (Fig. S7).

Long-term rock-accumulation rates can also be estimated. The 16 gypsic palaeosols seen in the measured section in Brachina Gorge (Fig. 1) to 9 m above the top of the Ediacara Member averaged a duration of 33.7±3.3 kyrs, whereas the 8 calcic palaeosols within the Ediacara Member averaged a duration of 5.3±5.6 kyrs. Total duration of Ediacara Member palaeosols in that measured section is 0.412 myrs for a long-term rock accumulation rate of 0.11 mm.yr-1. This is faster than estimates gained by taking the top of the Rawnsley Quartzite in Brachina Gorge as the Cambrian-Precambrian boundary at 542 Ma and the base of the Nuccaleena Dolomite in nearby Enorama Creek as Marinoan cap carbonate30, dated in China at 635 Ma (ref.31), and local thicknesses measured by Mawson32. Long term Ediacaran rock accumulation rate calculated in this way was 0.034 mm.yr-1, half as fast as local Cambrian rates (0.076 mm.yr-1 for basal Cambrian to 509.5 Ma Onaraspis rubra zone, and 0.069 mm.yr-1 from O. rubra to 504.5 Ma Leiopyge laevigata zone in Ten Mile Creek, eastern Flinders Ranges19). Duration of the Ediacara Member with such assumptions of long-term steady Table S5. Ediacaran pedogenic carbonate, palaeoprecipitation, and duration. Stratigraphic

Level (m) Bk depth

(cm) Burial depth

(km) Bk depth

decompacted (cm)Predicted

MAP (mm) Nodule size

(cm) Predicted duration

(kyrs) 13.2 24 5.621 44.18 397 0.3 1.7 14.1 25 5.620 46.02 407 0.4 2.2 16.5 24 5.618 44.17 397 1.1 6.1 23.7 27 5.611 49.69 426 3 16.5 30.9 26 5.604 47.84 416 2 11.0 31.3 28 5.603 51.52 435 0.3 1.7 31.7 32 5.603 58.88 472 0.2 1.1 38.3 24 5.596 44.15 397 0.4 2.2



Table S6. Ediacaran pedogenic gypsum, palaeoprecipitation, and duration. Stratigraphic

Level (m) By depth

(cm) Burial depth

(km) By depth

decompacted (cm)

Predicted MAP (mm)

Gypsum (%) Predicted duration (kyrs)

0 29 5.726 53.51 268 7 33.69 1 12 5.725 22.14 139 7 33.69

2.5 20 5.724 36.90 189 8 37.67 3.3 13 5.723 23.99 145 9 41.66

39.3 49 5.687 90.34 579 6 29.70 40.5 27 5.686 49.78 248 8 37.67 41.6 29 5.684 53.46 268 7 33.69 42.3 27 5.684 49.77 248 6 29.70 42.9 22 5.683 40.56 204 7 33.69 44 16 5.682 29.49 162 7 33.69

44.6 13 5.681 23.96 145 7 33.69 45.4 12 5.681 22.12 139 6 29.70 49.5 14 5.677 25.80 150 6 29.70 50.3 15 5.676 27.65 156 7 33.69 52.3 17 5.674 31.33 169 7 33.69 54.3 14 5.672 25.41 150 7 33.69

Spriggina Gully

28 5.686 51.21 255 10 45.65

rock accumulation rate would be 1.3 m.yr. The threefold time discrepancy may be lost in erosional disconformities abundant in palaeosol-palaeochannel sequences28-29, but most of this time is lost in major palaeovalley incisions (Fig. S1b). Additional notes on Ediacaran palaeosols and palaeoclimate

Palaeosols of the Ediacara Member also furnish evidence for palaeoclimate and its temporal variation during late Ediacaran time (Fig. S7, Table S5-S6) using a variety of modern climofunctions (equations S2-S4), and some of these results are shown in Tables S4-S5. A variety of other palaeosol features also point to semi-arid to arid, cold temperate palaeoclimate.

One example is the Warrutu palaeosol of the Ediacara Member, recognized as an ancient soil by deformation onlapped by overlying sediments and downward diminution of successive generations of deformation (Fig. 1b). Successive deformations followed by soil formation distinguish this as soil deformation rather than a seismite or load cast1. Its surface microrelief is comparable with gilgai of self-mulching soils (Vertisols), but the deformation is plastic, not sheared, and lacks the slickensided lentil peds and high clay content of Vertisols33. Anticline crests and overlying sandstone load casts are elongate but not quite parallel, and one load cast is curved and orthogonal to the others1. This indicates curved local slumps, like solifluction lobes, mostly wrinkling downhill along contours of palaeorelief, as determined from isopachs (Fig. S1a), and palaeocurrents1. This orientation is unlike linear gilgai microrelief, which is straight and runs perpendicular to contours33. This unusual combination of successive liquefactions of siltstone are most like Pleistocene periglacial involutions34-37, and also comparable with palaeoperiglacial deformation in the Neoproterozoic Port Askaig Formation of Scotland38-39 and Whyalla Sandstone of South Australia40. This horizon of cryoturbation was seen at comparable levels within the Ediacara Member at most sites examined, spanning 225 km along strike of current outcrops (Fig. S1a). This ancient microrelief is thus interpreted as evidence of periodic thaw within a widespread frozen (though not necessarily near-glacial) soil of varied palaeotopographic setting. In Canada today, the southern limit of discontinuous permafrost is at mean annual temperature (MAT) -1oC, which is at 58oN in Alberta and 55oN in Saskatchewan41. At least four episodes of soil



freezing and thaw are indicated for the Warrutu palaeosol, which lacks large (>1 m) ice wedges or sand wedges of other Proterozoic periglacial palaeosols of South Australia40 and Scotland39.

Other palaeoclimatically significant sedimentary structures in the Ediacara Member are expansion cracks1, previously interpreted as petee structures from microbial mat degassing or salt crystallization in supratidal sandflats8, but regarded here as evidence of frost heave (thufur mounds) of Inga palaeosol surfaces1. Key observations supporting this interpretation are that in all known cases of surficial expansion cracks are at the bed surfaces whereas sand crystals like those of very arid soils are 25-30 cm deep within the bed. There is no evaporite mineral or pseudomorph in the expansion cracks, which are interpreted as expanded by freezing soil water1. Such periglacial structures in modern soils are called earth hummocks or thufur mounds, and indicate mean annual temperature less than 6oC (ref. 41), though more common in mean annual temperature less than 3oC, in climates with a freezing index of 700-3800oC days.yr-1 and thawing index of 500-3000oC days.yr-1 (ref. 42).

Also common in association with Ediacaran fossils are moulds of sharp needle-like objects, either solitary or in radiating groups, up to 15 cm long by 2 mm wide9-10,44, or in arcuate, concentric palisades, up to 5 cm long and 2 cm wide16,45. These narrow casts are similar to needle ice crystals extruded from pores and cracks in the surface of frozen soils in frigid to cold temperate climates 42,46. These palisade-like marks in South Australian Ediacaran rocks have been attributed to Radulichnus, an ichnogenus for putative mollusc scratching16,45, but they are straight, sharp ended, and parallel within palisades, unlike subparallel, curved, and round-ended Radulichnus associated with genuine molluscs in Cambrian rocks47 or formed by modern molluscs48. South Australian “Radulichnus” was considered problematic16,45, because the scratches remain as if on a hard substrate, unlike geologically younger mollusc scratching into microbial mats obscured by subsequent microbial mat growth and mollusc forward motion. This objection was countered by proposing10 uniquely Ediacaran mat toughness and molluscan feeding backwards with raking motions of a radula at the tip of a long proboscis of the putative Ediacaran mollusc Kimberella. Russian Kimberella have high relief, central invaginations, and marginal wrinkles49 not seen in South Australian “Kimberella” near “Radulichnus”16,45. Low-relief oval mounds near South Australian “Radulichnus” could equally be a biological soil-crust pedestal or tuft in which water froze, expanded, and protruded radially as ice needles or sheets. A similar situation may be represented by Coronacollina acula, which has been interpreted as a sponge with 4-9 spicules extending radially up to 17 body-lengths beyond its body44. These may also be ice needles extruded from frozen mounds, though some of these radial ridges (grooves in casting sandstone) may also have been flanges of biological soil crust pedestals, as here envisaged for the comparable Pseudorhizostomites howchini2. Comparable ice needles are well known from high latitude50 and high altitude51 modern biological soil crusts, in which focused frost heave is considered an important cause of rugose and pinnacled, as opposed to flat and rolling, biological soil crusts50. Ice needles commonly form at night and melt by day as an important source of moisture during intervals of photosynthetic activity51. Needle impressions are better preserved by soil cyanobacteria and their extracellular polysaccharides than by lichens and mosses, which today dominate frost-heaved biological soil crusts50-51. Mosses are unknown before the Ordovician52, but Glomeromycotan lichens are well preserved in Ediacaran rocks of China53. Horizontal needle ice can also be a predecessor of frazil ice in vernal ponds54-55, and grows well



from remnants of vegetation56. Freezing of very shallow vernal ponds is an appealing explanation for subhorizontal needles of “Coronacollina” and “Radulichnus”, which are regionally rare, but locally common on Ediacaran elephant skin (Rivularites) slabs16,44-45.

Because both Coronacollina and Pseudorhizostomites are fossils based on holotypes from South Australian palaeosols, the interpretations offered here imply that they were microbial mounds and pedestals comparable with stromatolites, not animal nor body fossils. The trace fossil name Radulichnus was proposed for genuine molluscan scratching of Jurassic to Pliocene mollusc shells57, and should thus not be applied to the straighter and sharper South Australian crystal casts9-10, 16, 44-45. Whether Radulichnus is appropriate for Ediacaran linear grooves from other regions such as the Russian White Sea2,16,45 is considered doubtful, but will require re-evaluation of that material from the new perspective of needle ice and other data on palaeoclimatology.

These general comparisons of cold, dry palaeoclimate could be quantified, if modern and Edicaran soil formation can be assumed to have been comparable. Depth of gypsum nodules (Dg in cm) is known to correlate with mean annual precipitation (P in mm) in modern soils, according to the following relationship58.

P = 87.593e0.0209Dg (equation S2)

This relationship (R2 = 0.63) has standard error ± 129 mm. Gypsic soils used to establish this relationship in arid to hyperarid regions have little vascular plant cover22-

23, and so may be applicable to palaeosols of Ediacaran age. This relationship can be used to infer precipitation from palaeosols once depths are corrected for burial compaction24 by some 5.8 km of latest Precambrian and Cambrian overburden1. Three modes in Ediacaran palaeosol depth to gypsum of 14, 27, and 49 cm, correspond with mean annual precipitation of 150 ± 129, 248 ± 129, and 579 ± 129 mm, respectively (Fig. S7). This last high result for a single gypsic palaeosol (Warrutu palaeosol) is supported by associated calcic palaeosols (Muru and Yaldati pedotype) using a climofunction for depth of carbonate nodules (Dc) in soils worldwide, according to the following relationship21.

P = 137.24 + 6.45Dc + 0.013Dc2 (equation S3)

This relationship (R2 = 0.52) has standard error ± 147 mm, and gives a mean value for 8 palaeosols near the deep gypsic palaeosol of 418 ± 147 mm. Geochemical indicators of palaeoclimate using alkali elements are unreliable for palaeosols with pseudomorphs leached of original evaporites after burial, evidence of burial illitization, and greenschist facies metamorphism. Nevertheless, an alumina/silica palaeothermometer59 based on North American weakly developed soils (Inceptisols comparable with Ediacaran palaeosols1) has the advantage of using metamorphically immobile elements, and gives mean annual palaeotemperature (T in oC) from the ratio of silica to alumina (R mole per mole) according to the following formula.

T = 46.94R + 3.99 (equation S4)

This relationship (R2 = 0.96) has standard error of± 0.6 oC, which is small enough that analytical error for silica (± 2.703 %) and alumina (± 0.875 %) by x-ray fluorescence should also be included by Gaussian error addition in quadrature. For Ediacaran palaeosols, such calculations yield cold-temperate mean annual palaeotemperatures: 6.62 ± 0.67oC for the Warrutu palaeosol, 6.47 ± 0.67oC and 7.95 ± 0.69oC for analyzed4 Muru and Yaldati palaeosols respectively. Comparable results were obtained from chemical analysis of Warrutu palaeosols in Ediacara core 3 and 4 (8.45 ±



Fig. S7. Palaeoclimatic and sedimentation variations during accumulation of the Ediacara Member of the Rawnsley Quartzite in Brachina Gorge, South Australia, on a time scale adjusted for relative palaeosol development1. 0.7o and 6.11 ± 0.67oC), and for separate Yaldati paleosols in Ediacara core 3 (5.39 ± 0.67o and 6.46 ± 0.72oC). These are reasonable values for cold-temperate palaeosols, and do not falsify assumptions that Ediacaran and modern weathering were comparable.

These various indications of cold to frigid temperatures are surprising for a palaeolatitude of less than 20o, based on palaeomagnetic inclination and plate reconstructions for South Australia during the Neoproterozoic and Cambrian60-61. Comparably low palaeolatitudes of Marinoan glacial deposits, underlying Ediacaran sedimentary rocks in the Flinders Ranges, was the initial stimulus for the Snowball Earth hypothesis62-63. Other indications that cool low-latitude palaeoclimates persisted from Marinoan (635 Ma) and Gaskiers (580 Ma) ice ages64-65 into the late Ediacaran (550 Ma) include tillites in the Billy Springs Formation66-67(equivalent to Bonney Sandstone and Ediacara Member of Fig. S1b), glacio-eustatic palaeovalleys in the upper Wonoka Formation68, and ubiquitous loess-like grain-size and microtexture in Neoproterozoic red beds of South Australia3.



References 1. Retallack, G.J., Were Ediacaran siliciclastics of South Australia coastal or deep

marine? Sedimentology 59, 1208-1236 (2012). 2. Fedonkin, M.A., Gehling, J.G., Grey, K., Narbonne, G.M. & Vickers-Rich, P. (eds.),

The Rise of Animals : Evolution and Diversification of the Kingdom Animalia (Johns Hopkins University Press, Baltimore, 2008).

3. Glaessner, M.F., Precambrian animals. Scientific American 204, 72-78 (1961). 4. Mawson, D. & Segnit, E.R., Purple slates of the Adelaide System. Roy. Soc. South

Australia Trans. 72, 276-280 (1949). 5. Goldring, R., & Curnow, C.N., The stratigraphy and facies of the Late Precambrian

at Ediacara, South Australia. Geol. Soc. Austral. J. 14, 195-214 (1967). 6. Jenkins, R.J.F., Ford, C.H. & Gehling, J.G., The Ediacara Member of the Rawnsley

Quartzite: the context of the Ediacara assemblage (late Precambrian, Flinders Ranges). Geol. Soc. Austral. J., 30, 101-119 (1983).

7. Mount, J.F., Re-evaluation of unconformities separating the 'Ediacaran' and Cambrian systems, South Australia. Palaios 4, 366-373 (1989).

8. Gehling, J.G., Environmental interpretation and a sequence stratigraphic framework for the terminal Proterozoic Ediacara Member within the Rawnsley Quartzite, South Australia. Precambrian Res. 100, 65-95 (2000).

9. Droser, M.L., Gehling, J.G. & Jensen, S.R., in Evolving Form and Function: Fossils and Development (D.E.G. Briggs, ed.), 125-138 (Peabody Museum of Natural History, New Haven, 2005).

10. Gehling, J.G., Droser, M.L., Jensen, S.R. & Runnegar, B.N., in Evolving Form and Function: Fossils and Development (D.E.G. Briggs, ed.), 45-56 (Yale Peabody Museum, New Haven, 2005).

11. Droser, M.L., Gehling, J.G. & Jensen, S.R. Assemblage palaeoecology of the Ediacara biota: the unabridged edition? Palaeogeogr. Palaeoclim. Palaeoec. 232, 131-147 (2006).

12. Clites, E.C., Droser, E.L., & Gehling, J.G., The advent of hard-part structural support among the Ediacara biota: Ediacaran harbinger of a Cambrian mode of body construction. Geology 40, 307-310 (2012).

13. Droser, M.L., & Gehling, J.G., Synchronous aggregate growth in an abundant new Ediacaran tubular organism. Science 319, 1660-1662 (2008).

14. Retallack, G.J., in Microbial Mats in Siliciclastic Sediments (N. Noffke, & H. Chafetz, eds.), 136-152 (Soc. Econ. Paleont. Mineral. Spec. Pap. 101, 2012).

15. Retallack, G.J., Growth, decay and burial compaction of Dickinsonia, an iconic Ediacaran fossil. Alcheringa 31, 215-240 (2007).

16. Seilacher, A., Trace Fossil Analysis (Springer, Berlin, 2007). 17. Retallack, G.J., Were the Ediacaran fossils lichens? Paleobiology 20, 523-544 (1994). 18. Gehling, J.G., Narbonne, G.M., & Anderson, M.M., The first named Ediacaran body

fossil; Aspidella terranovica. Palaeontology 43, 427-456 (2000). 19. Retallack, G.J., Cambrian palaeosols and landscapes of South Australia. Australian J.

Earth Sci. 55, 1083-1106 (2008). 20. Frey, M. 1987. In Low Temperature Metamorphism, (M. Frey. Ed.) 9-58 ( Blackie,

Glasgow, 1987). 21. Retallack, G.J., Pedogenic carbonate proxies for amount and seasonality of

precipitation in paleosols. Geology 33, 333-336 (2005).



22. Dan, J., Moshe, R. & Alperovich, N., The soils of Sede Zin. Israel J. Earth Sci. 22, 211-227 (1973).

23. Yaalon, D.H., Moshe, R. & Nissim, S., Evolution of reg soils in southern Israel and Sinai. Geoderma 28, 173-202 (1982).

24. Sheldon, N.D. & Retallack, G.J., Equation for compaction of paleosols due to burial. Geology 29, 247-250 (2001).

25. Monger, H.C., Daugherty, L.A., Lindemann, W.C., & Liddell, C.M., Microbial precipitation of pedogenic calcite. Geology 19, 997-1000.

26. Driese, S.G., & Mora, C.I., in Plants Invade the Land: Evolutionary and Environmental Perspectives (P.G. Gensel & D. Edwards, eds.), 237-253 (Columbia Univ. Press, New York, 2001).

27. Retallack G.J. and Huang, C.-M., 2011, Ecology and evolution of Devonian trees in New York, USA. Palaeogeogr. Palaeoclim. Palaeoec. 299, 110-128.

28. Retallack, G.J., Fossil soils as grounds for interpreting long term controls on ancient rivers. J. Sediment. Petrol. 56, 1-18 (1986).

29. Retallack, G.J., in The Adequacy of the Fossil Record (S.K. Donovan & C.R.C. Paul, eds.), 131-162 (Wiley, Chichester, 1998).

30. Retallack, G.J., Neoproterozoic glacial loess and limits to snowball Earth. Geol. Soc. London J. 168, 1-19 (2011).

31. Condon, D., Zhu M., Bowring, S., Wang, W., Yang, A., & Jin, Y., U-Pb ages from the Neoproterozoic Doushantuo Formation, China. Science 308, 95-98 (2005).

32. Mawson, D., The Cambrian sequence in the Wirrealpa Basin. Roy. Soc. S. Australia Trans. 63, 331-347 (1939).

33. Paton, T.R., Origin and terminology for gilgai in Australia. Geoderma 11, 221-242 (1974).

34. Johnsson, G., Periglacial phenomena in southern Sweden. Lund Stud. Geogr. A21, 378-403 (1963).

35. Hamilton, T.D. & Ashley, G.M., Epiguruk: a late Quaternary environmental record from northwestern Alaska. Geol. Soc. Amer. Bull. 105, 583-602 (1993).

36. French, H.M., The Periglacial Environment (Longman, London, 1996) 37. Davis, N., Permafrost: a Guide to Frozen Ground in Transition (Univ. Alaska Press,

Fairbanks, 2001). 38. Spencer, A.M., Late Precambrian glaciation in Scotland. Geol. Soc. London Mem. 6,

1-100 (1971). 39. Benn, D.I. & Prave, A.R., Subglacial and proglacial glacitectonic deformation in the

Neoproterozoic Port Askaig Formation, Scotland. Geomorphology 75, 266-280 (2006).

40. Williams, G.E. & Tonkin, D.G., Periglacial structures and palaeoclimate in the Cattle Grid Copper Mine, Mount Gunson, South Australia. Austral. J. Earth Sci. 32, 287-300 (1985).

41. Smith, M.W. & Riseborough, D.W., in Landscapes of Transition (M.-L. Byrne & G. Young, eds), 121-139 (Kluwer, Dordrecht, 2002).

42. Washburn, A.L., Geocryology. (Wiley, New York, 1980). 43. Karte, J., Periglacial phenomena and their significance as climatic and edaphic

indicators. Geojournal 7, 329-340 (1983). 44. Clites, E.C., Droser, E.L. & Gehling, J.G., The advent of hard-part structural support

among the Ediacara biota: Ediacaran harbinger of a Cambrian mode of body construction. Geology 40, 307-310 (2012).



45. Seilacher, A., Biomat-related lifestyles in the Precambrian. Palaios 14, 86-93 (1999).

46. Jennings, J.N., Needle ice raked ground. Austral. Geogr. 15, 328-330 (1983). 47. Dornbos, S.Q., Bottjer, D.J. & Chen, J.-Y., Evidence for seafloor microbial mats and

associated metazoan lifestyles in Lower Cambrian phosphorites of southwest China. Lethaia 37, 127-137 (2004).

48. Koy, K. & Plotnik, R.E., in Trace Fossils: Concepts, Problems, Prospects (W. Miller, ed.), 428-441 (Elsevier, Amsterdam, 2007).

49. Fedonkin, M.A., Simonetta. A., & Ivantsov, A.Y. in The Rise and Fall of the Ediacaran Biota (P. Vickers-Rich & P. Komarower, eds.), 157-179 (Geol. Soc. London Spec. Pap. 286, 2007).

50. Belnap, J., The potential roles of biological soil crusts in dryland hydrologic cycles. Hydrol. Proc. 20, 3159-3178 (2006)

51. Pérez, F.J., Microbiotic crusts in the high equatorial Andes and their influence on paramo soils. Catena 31, 173-198 (1997).

52. Lenton, T.M., Crouch, M., Johnson, M., Pires, N., & Dolan, L., First plants cooled the Ordovician. Nature Geoscience 5, 86–89 (2012).

53. Yuan, X., Xiao, S. & Taylor, T.N., Lichen-like symbiosis 600 million years ago. Science 308, 1017-1020 (2005).

54. Martin, S., Frazil ice in rivers and oceans. Ann. Rev. Fluid Mech. 13, 379-397 (1981).

55. Purer, E.A., Ecological study of vernal pools, San Diego County. Ecology 20, 217-229 (1939).

56. Matthews, V., Origin of horizontal needle ice at Charit Creek, Tennessee. Permafrost Periglacial Proc. 10, 205-207 (1999)

57. Voigt, E., in Trace fossils 2 (T.P. Crimes & J.C. Harper, eds.), 335-346 (Seel House Press, Liverpool, 1977).

58. Retallack, G.J., & Huang, C.-M., Depth to gypsic horizon as a proxy for paleoprecipitation in paleosols of sedimentary environments. Geology 38, 403-406 (2010).

59. Sheldon, N.D. & Tabor, N.J., Quantitative paleoenvironmental and paleoclimatic reconstructions using paleosols. Earth Sci. Rev. 95, 1-52 (2006).

60. Schmidt, P.W., Embleton, B.J.J. & Williams, G.E., Early timing of remnance in haematite of the Neoproterozoic Elatina Formation, Adelaide Geosyncline: confirmation of the low latitude of Neoproterozoic glaciation. Explor. Geophys. 24, 239-242 (1993).

61. Sohl, L.E., Christie-Blick, N. & Kent, D.V., Paleomagnetic polarity reversals in Marinoan (ca. 600 Ma) glacial deposits of Australia: implications of low latitude glaciation in Neoproterozoic time. Geol. Soc. Amer. Bull. 111, 1120-1139 (1999).

62. Kirschvink, J.L., in The Proterozoic Biosphere: a Multidisciplinary Study (J.W. Schopf & C. Klein, eds.), 51-52 (Cambridge Univ. Press, Cambridge, 1991)

63. Hoffman, P.F. & Schrag, D.P. The snowball Earth hypothesis: testing the limits of global change. Terra Nova 14, 129-155 (2002).

64. Carto, S.L. & Eyles. N., in The Geological Record of Neoproterozoic Glaciations (E. Arnaud, G.P. Halverson & G. Shields-Zhou, eds.), 467-473 (Geol. Soc. London Mem. 36, 2011).

65. Halvorsen, G.P., in Neoproterozoic Geobiology and Paleobiology (S.-H. Xiao & A.J. Kaufman, eds.), 231-271 (Springer, Berlin, 2006).



66. DiBona, P.A., A previously unrecognized Late Proterozoic succession: upper Wilpena Group, northern Flinders Ranges, South Australia. Geol. Surv. S. Australia Qu. Notes 117, 2-9 (1991).

67. Jenkins, R.J.F., in The Geological Record of Neoproterozoic Glaciations (E. Arnaud, G.P. Halverson, & G. Shields-Zhou, eds.), 693-699 (Geol. Soc. London Mem. 36, 2011).

68. Eickoff, K.-H, von der Borch, C.C. & Grady, A.E., Proterozoic canyons of the Flinders Ranges (South Australia): submarine canyons or drowned river valleys? Sediment. Geol. 58, 217-135 (1988).



Date post:	10-Apr-2020
Category:	Documents
Upload:	others
View:	4 times
Download:	0 times

Nature template - PC Word 97 · sedimentary facies and palaeosols (Fig. S2, Tables S1- 2) supports...

Documents