https://doi.org/10.24199/j.mmv.1973.34.19 9 May 1973

PALAEOPEDOLOGY OF THE MURRAY RIVER REGION BETWEEN MILDURA AND RENMARK, AUSTRALIA

By Edmund D. Gill

Deputy Director, National Museum of Victoria

Introduction The surfaces are relict. Tn the incised river

valley is 'In dry climates, where water seldom tract, a significant area occupied by Pleistocene or intermittently leaches rocks, sapro- terrace sediments (Rufus Forma- tion). This antiquity of surfaces leads to many litization does not occur, although polygenetic profiles. the other hand, the fossil saprolites, indicating earlier On more humid conditions, may be pre- dunefields and the lunette on the E. side of Lake Victoria have preserved series of paleo- sent'. Dorothy Carroll (1970) sols. This describes well the study area with The dryness of the country, combined with which this Memoir is concerned. The sediments the general fineness of the sediments, leads to of the region (and so the substrate for the soils) consist chiefly of silica (fine sand) and 1. Considerable wind traction of sediments, clay. They have originated in the humid zone with dunes as a common landform. As of the Great Dividing Range. Silica is a stable the E-W. dune system and the lunettes primary mineral, and clay a stable secondary are relict, this process was more active in mineral. Silica results from the weathering and the past. erosion of igneous and sedimentary rocks, while 2. Wind-winnowing with airborne fine sedi- clay is a common product of that process. The ments as a result. The increase of this long traction of the sediments and the low process during droughts indicates that it dynamics of the river system in the Murray must have been a major factor during still River have resulted in only the finer fraction drier periods in the past. Thus

reaching the study area. Coarser sediments are (a) More clay and silt are found in sand common on the E. side of the Murray Basin, dunes (geomorphic highs) than would but finer sediments characterize the W. side. be otherwise expected. Water, the solvent par excellence, is active (b) Because of the accumulations of both in the provision of these sediments and mineral salts on the terrain, these like- their transport to the sedimentary plain. But wise become airborne, resulting in water is scarce in the region W. of the Dividing more accumulation in the dunes and Range, so carbonates (calcite, dolomite), sul- elsewhere than would be otherwise ex- phates (gypsum) and chlorides (halite) accu- pected (e.g. Gill, this Memoir, Fig. mulate at the surface of the terrain. For the 10). minerals involved, see Segnit, this Memoir. la the accompanying paper on the geology Thus the geologic and gcochcmical charac- and geomorphology of the region, tectonic teristics of this dry land make their unmis- movements are described that have led to the takable impress on the soils. In the humid areas sinking of the area and the preservation of seaward of the Dividing Range, these strongly paleosols by burial. On the other hand, the alkaline conditions are replaced by neutral to surface of the plain (disregarding discontinuous acid ones. Iron is there accumulated in podzolic superficial deposits) is of Pliocene/Pleistocene hardpans and as pcdalferric nodules. The mur- lacustrine clay and sand N. of the Murray ray River country is marked by the bright reds River, and Pleistocene dune sand S. of the river. and yellows of a highly oxidized terrain, but

241 242 EDMUND GILL the carbonates prevent the accumulation of a key to a number of aspects of the geologic iron. On the other hand, pyrolusite is ubiqui- history. They provide time datum planes tous. In the study area it occurs in sediments, through strongly interdigitating sediments, they on ped and crystal faces, on opal, on car- trace the palaeoclimatic change from humidity bonate nodules, on silcrete, on fossil bones, and to aridity, they mark out the evolution of so on. It occurs, often as dendrites, on Aborigi- diverging climatic patterns landward and sea- nal skeletons, including Upper Holocene ones. ward of the Dividing Range, and they measure An interesting project would be to quantify the pulse of the stable/unstable phases in the and date the accumulation of pyrolusite, so that extensive E-W. dune system. See Gill, this it could be used for relative dating. This would Memoir. be possible in the humid areas also. In Western Three Soil Types Victoria (a given climate), the accumulation of In the stratigraphic succession, three grossly maghemite on a given substrate (basaltic ash) different soil types register in the rocks major on a given slope (volcanic ash cone) is a func- changes in the environment between the Plio- tion of time; the same applies to the associated cene and the present: pyrolusite. Also with time the profile deepens and becomes more complex (uniform/duplex). 3 Soils accumulating The maghemite nodules increase in size with carbonates Quaternary time, beginning as dust and passing through 2 Soil accumulating sizes only a few millimetres in diameter to silica Plio/Pleistocene pea-size, and sometimes larger. Under the con- 1 Soil accumulating iron ditions defined, maghemite dust has formed in (lateritic) Pliocene the Holocene, while the minute nodules are Chronology is discussed in the accompanying latest Pleistocene to earliest Holocene. The paper. Each of the above soil types registers a youngest profile with pea-size buckshot gravel different climate, and each records a steady is 20,000-25,000 years old (Camperdovvn). state condition of the terrain which allowed Manganese dioxide was used as an age indi- the pedogenic processes to continue for a long cator in the dry Tule Springs country in U.S.A. time. Each of the soil types was developed over (Shutler 1968). 2 a large area (at least 250,000 km ), and the Pyrolusite accumulates in the acid humid general geomorphology has been that of a flat- areas and in the alkaline dry areas as well land. Each buried soil represents a stratigraphic because its solubility is not affected by pH. disconformity. In the dry country, a paleosol 'Under highly oxidizing conditions Mn0 is 2 represents a depression of the dune morphology. the stable phase over the whole range of pH' Most paleosols were eroded before being buried (Mason 1966, p. 171). The dry country thus under a new formation. acts as a huge fractionating column separating The present carbonate soils are not well the manganese from the iron. Manganese so understood. The early settlers came from humid accumulated on a terrain over a long period land and did not understand them at all. The could later be washed into shallow seas and methods imported from their homelands caused form the manganese-rich sediments of this type havoc to the topsoils, leading to extensive known from various parts of the world. It may erosion and widespread dust storms. This has well be that such deposits could assist palaeo- now been brought under considerable control, climatic interpretations. Tindale (pers. comm.) but it can still be questioned whether there says manganese dioxide was used by Aborigi- is not too much disturbance of the surface lead- nals as a paint. ing to erosion and loss of fine fractions plus As no pedologist could be found to prepare humus. Soil is 'one of the most important an account of the palaeopedology to include in natural resources'. The ash and charcoal of this volume, the present writer (a geologist) Aboriginal middens are probably the most sen- has written this paper because the paleosols are sitive materials to erosion on the landscape, PALAEOPEDOLOGY OF THE MURRAY RIVER 243

This yet such materials up to 2,000 yr old are exceptional features in this country) . found in the surface soil showing that the land formation is Upper Pliocene (Marshall, Gill, surface presently in a state of erosion had been this Memoir) and the paleosol is within it at stable for a couple of millenia. The youngest Yelta. The sediments lateritized are poorly soil of all is a thin carbonate crust on fans sorted sands that include grains 2 mm in dia- developed on the slopes of the W. shore of meter, which is exceptional in this region where Lake Victoria. This crust is now being rapidly the normal range is clay to fine sand, but not removed, but its age by radiocarbon dating of unusual for the Moorna Formation. is carbonized twigs is 'modern' (therefore less The part of the lateritic profile preserved than 200 yr) so this erosion has occurred the mottled zone, which consists here of white since European occupation. N/8 sand with mottles of weak red 5R 5/3 grading to red 7 5YR 4/6 with some minor Lateritic Profile reddish yellow 7 5 YR 6/6. A rhizomorph about burrow 1. Yelta, Victoria 6 mm in diameter was noted and a about 3 mm in diameter set obliquely. This The oldest buried soil found in the study horizon is strongly compacted, and forms a area is a massive mottled zone of a lateritic ledge at the foot of the cliff, although this is profile. I am indebted to Mr G. Blackburn of mostly covered with wash from the cliff. High C.S.I.R.O. Division of Soils for examining the river levels sweep parts of it clean for a while, occurrence for me and confirming that it is but at the W. end where the river impinges on lateritic. Figure 1 provides a section of the the cliff, the outcrop can always be clearly seen. Murray River cliff on the S. (Victorian) side The mottles are usually 15-25 cm in diameter. at Yelta, W. of Merbein and E. of the bridge On the day of the survey 251 m of the mottled to Curlwaa. Underneath the Blanchetown Clay, zone was exposed, but it could be seen con- and therefore older than it, is the Moorna tinuing down some distance under the water Formation, characterized by its varied sedi- of the river. It is difficult to be sure of the ments including coarse-grained ones (both underwater extension, but it was judged to be at least as deep as the section exposed. nrnrrms 0-9 mRed pedocal The sediments lateritized are of riverine fades, and apparently were part of a fluvatile sufficiently high to be drained 3 m Red 2-5YR 5/6 CHOWILLA SAND plain or terrace, in the dry season. The terrain almost certainly V?20 cm band of carbonate concretions supported a subtropical rainforest (Gill 1961 4-6 m Reddish yellow 7-5YR 6/6 CHOWILLA SAND with a, b). The climate was of the monsoon type, badlands erosion with leaching warm copious waters in the wet season, followed by a dry season when the BLANCHETOWN CLAY water table fell and air entered to oxidize the rocks. No greater contrast is possible than this " 6/3 3 m Mottled light gr.iy N/7 and light red 2-5YR rainforest palaeo-environment with abundant sand 1-1 m Light gru> N/7 deep-leaching acid waters, and the present semi- MOORNA FORMATION arid sparsely-vegetated terrain with shallow

2 iu White day 5Y 8/1 mottled brownish yellow strongly alkaline soil. 10YR 6/6 and variants MOORNA FORMATION' 2. Paringi, N.S.W. £b & &* ^ On the Sturt Highway between Euston and sb s? -£> -a Mildura, 29 km from the latter on the N. side Mottled weak red 5R 5/3 and white N/8 of the Murray River where it meets the high- profile T1MBOON PLDODKliM 1 atcritized 2 4 i m way, badlands erosion has exposed the section cliff on Pig i_Surveyed section of Murray River shown in Fig. 2. This section was surveyed For map see S. bank at Yelta, Victoria. from a lone tree on the W. side of the de- Gill on Geology of area, this Memoir. 244 EDMUND GILL

RIVER MURRAY VAtl bY NEW SOUTH WM IS

37 m Red pcdocai Carbonate fracbuK-s and cylinders

3 4 Greenish gray HE ANCHETOWN CLAY

2-8 Gray PARILLA SAND

2 I m TIMBOON PEDODERM

Gulch floor

Mottlcd_ zone of laterilizal profile

Fig. 2—Surveyed section of Murray River cliff on N. bank at Paringi (not Paringa), N.S.W. bouchement of the gulch to its floor. Beneath 3. Distribution the Blanchetown Clay is a gray clayey fine Rowan and Downes (1963, p. 17) record sand presenting vertical walls along the gulch, lateritic profiles in NW. Victoria. One locality all of which characters belong to the Parilla is Robinvale lock, where Segnit (this Memoir) Sand, so the sediments are referred to that has proved kaolinite. The writer (Gill 1966) formation. Below this a very compact mottled has recorded a lateritic profile at Ultima in N. zone occurs with nearly 2 m outcropping, but Victoria. At Tammit Station c. 14 km from continuing to a greater depth as was proved Euston, N.S.W. , I noted below the homestead by a spade hole. The background varies from a platform of ferricrete 1-1 -5 m thick extending white N8 to light gray 5Y 8/1 with mottles c. 80% of the distance across the river. The up to 18 cm diameter, although these may material was tested with a penetrometer and coalesce into larger sizes. The mottles are red found to vary from 6 kg/cm2 to hard solid 2-5YR4/8. ferricrete. Colour varied from yellowish red If the formation with the paleosol is Parilla 5YR 5/6 to very dusky red 7 5R 2/2, and the Sand or older, then the fossil soil is developed substrate from silty to sandy gravel. Discon- in an older substrate here than at Yelta. This formably over the ferricrete there was mottled need cause no surprise. Evidence of intrafor- clayey fine sand to sandy clay, which was light mational erosion is widespread. The paleosol gray 5Y 7/2 to white 5Y 8/2 with reddish occurs only as remnants, and the horizon A yellow 7 5YR 6/6 mottles 1-20 cm diameter. is nowhere preserved. At Paringi, a channel The top metre of this formation was somewhat 45 m wide is cut in the top of the mottled concretionary, forming a shelf on the shore; zone, and is c. 1 m deep. Above the 2 mottled it varied from 12 kg/cm to solid. The clay zone is a hard layer c. 18 cm thick, but it is just above the ferricrete had some rhizomorphs. not present over the channel. The hard layer A bore is needed to prove whether or not this consists of coarse sand to laminated silt. In the is a true lateritic profile. channel is a variety of sediments with current- A very extensive literature exists on lateritic bedding. profiles in Australia, but some dealing with A lateritic profile infers tectonic stability for adjacent areas are Blackburn and Leslie (1958), a considerable time, while widespread erosion Dury (1971), Faniran (1970), Firman (1967), infers terrain instability. The latter was pro- Grubb (1971),Jessup(1960a,b) andTwidale bably caused by the Kosciusko uplift. (1972). PALAEOPEDOLOGY OF THE MURRAY RIVER 245

4. Age and Nomenclature Soil Accumulating Silica

The A.O.G. oil bores show that this paleosol 1. Stratigraphic Occurrence Cheltenhamian (uppermost overlies marine The lateritic profile accumulated non-mag- the paleosol is Miocene) sediments. At Yelta netic alpha iron oxide. Higher in the sequence Formation which is Upper Plio- in the Moorna is a paleosol that accumulated silica. It formed evidence available in cene. This dating fits the common opal in clay and dolomite, while in (Gill Fisher has 5. Victoria 1971). (1958) sand it formed a range of rocks from lightly of lateritization as pointed out the importance bonded silicified sandstone to ringing silcrete. a concentrator of minerals. I think that the The Timboon Pedoderm occurs in the Parilla large concentrations of alluvial gold found in Sand and the Moorna Formation, while the Victoria were a function of deep kaolinization silica-amassing pedoderm occurs in the younger of the bedrock, followed by fluviatile sifting to Blanchetown Clay and Chowilla Sand. This remove the saprolite. statement applies to the study area, which is on The laterite constitutes an important time the down-warped side of the Pinnaroo Block, datum plane through very variable rocks. It is involving the Lake Victoria Syncline. On the the only paleosol that can be traced with con- geomorphic highs other relationships are pos- Great fidence landward and seaward of the sible. At the Chowilla dam site, the silicified Dividing Range, and indeed over the whole of zone occurs on but not in the Parilla Sand. diagrammatic ally SE. Australia. Fig. 3 shows Firman (1967, p. 170) says the siliceous cap types the contrasting series of main pedoiogic is on the eroded surface of the Parilla Sand. in section N. and S. of the Dividing Range a This is the Karoonda Surface or Terrain (Fir- from the study area to the Western District of man 1966), and so applying the nomenclatorial Victoria (Gibbons and Gill 1964). principle defined above, the paleosol would be The laterite profile is the only common de- called the Karoonda Pedoderm. nominator With respect to nomenclature, the writer 2. Distribution thinks that the close association of a paleosol This pedoderm is very widely distributed, with an ancient terrain can best be expressed and in many places influences the cliff profiles. by using the same name. In SE. Australia Both the opal and silcrete were widely used by laterite is associated with the Timboon Terrain the Aboriginals for implements. The Karoonda (Gill 1964), so applying the term of Brewer Pedoderm is present in the Chowilla dam area, et al. (1970), the nomenclature would be the along Salt Creek (PI. 32, fig. 1), at Cal Lai, Timboon Pedoderm. on Nampoo, Warrakoo, Moorna and other stations along the river, at Merbein, and along GREAT DIVIDING RANGE the river further to the SE., including Robin- vale. Indeed it is almost ubiquitous. Pedogenic BASALT PLAIN MURRAY BASIN in the dry interior of Aus- Slow Sinking silcrete is common is involved, but it suf- Climate tralia. A vast literature Mediterranean Climate Semi-arid fices for the present purpose to mention the work of a geomorphologist (Mabbutt 1965), a pedologist (Stephens 1971), and a geologist (Wopfner, H., and Twidale, C R. 1967). In Central Australia it is common for silica to be Pig 3—Diagram to show the different existing ecolo- found in various parts of the lateritic profile. gies N. and S. of the Great Dividing Range along a section from the study area to W. Much difference of opinion has been evident of soil types from Victoria. The succession as to the cause of such occurrences. It is pointed for the Pliocene to the present is shown out that in the Lake Victoria syncline there each region, indicating that in the Pliocene the two had similar ecologies, viz. mon- is an extended ( and not telescoped ) se- producing lateritic soil soonal rainforest quence. The siliceous horizon is later than the profiles. 246 EDMUND GILL

Jateritic profile, so the possibility should be Point section on Nampoo Station c. 10 km to considered that in non-depressed areas these the E. shows that a lake dried up and the soil two horizons could be telescoped, i.e. the pro- was formed on its floor. The dolomite is in- file is polygenetic. terpreted as a function of a drying up of saline waters. Rhizomorphs show that trees or shrubs 3. Age grew on the old lake floor. After the paleosol The Parilla Sand and the Mooraa Formation was developed, lacustrine conditions returned, are Pliocene in age (Gill, this the Memoir). As and the water was fresh as is shown by the age of the succeeding Blanchetown Clay is not ostracod fauna. known with accuracy, it is called Plio-Pleisto- In the cliffs near Kulcurna Station, and on cene. The Karoonda Pedoderm has not been Salt Creek, there are widespread siliceous found in the stratigraphic sequence higher than rhizomorphs that bear witness to the existence the base of the Blanchetown Clay. is It com- of a land surface, a soil, and vegetation. monly in Chowilla Sand under the clay. These In many places (e.g. in cliffs on the W. side two formations interdigitate strongly, and the of Six Mile Creek on Neilpo Station, and in paleosol is time a datum plane through them. the Bone Gulch area on Moorna Station fur- Salt Creek, On Kulcurna Station, the paleosol ther W.) deposits of gypsum characterize the can be traced Chowilla into from Sand Blanche- lowest part of the Blanchetown Clay. At Six town Clay then into and Chowilla Sand again. Mile Creek rosettes and other crystalline masses In the latter, the horizon consists of silicified of gypsum have been seen up to 5 m in sand, while in the Clay it consists of common diameter. The site was worked commercially at opal. Segnit et al. (this Memoir) have described one time. This horizon of gypsum could be due the nature of the opal in the Karoonda Pedoderm to a period of drying up that followed the at Point Sharp (N. bank of Murray River) on freshwater phase indicated by the ostracod Station, E. Nampoo of Cal Lai (PL 32, fig. 2). bands. (For stratigraphy and palaeontology, see There two opal horizons occur in this the Gill this Memoir). thickest known section of Blanchetown Clay. Pedocals A gell-like material under the lower opal proved to be highly disordered sodium montmorillonite. Three minerals characterize the terrain in the Celestite occurred between the two opal beds; study area at the present time—a carbonate barytes and sepiolite were found but not in situ (calcite), a sulphate (gypsum) and a chloride (Segnit, this Memoir). (common salt). All are present in both earthy and crystalline forms. The carbonate is par- 4. Climatic Association ticularly abundant. In the soil profiles it occurs The lateritic profile (Timboon Pedoderm) is in earthy form, as fraebules, and as a solid associated with tropical or subtropical mon- hardpan. It has been estimated that on the soonal conditions, but the Karoonda Pedoderm average there are in this country ,—- 800,000 is associated metric 2 with the incoming of the present tonnes (or tons) per km . Commonly semi-arid conditions. On the E. wall of Salt the profile at the surface is polygenetic, as suc- Creek on Kulcurna Station, where the Blanche- cessive deposits of carbonate and radiocarbon town Clay lenses out to nothing (an old lake dates establish. The persistence of all this shore) towards the south, a black clay yielded carbonate at the surface bears witness to the palynologic evidence of a flora directly persistence of semi-arid conditions over a long comparable with the present semi-arid one time preventing its solution and removal. (Churchill, this Memoir). A short distance These successive carbonates record the pulse of above these fossils in the same section is the palaeoclimatic change. silicified sandstone of the Karoonda Pedoderm. No greater pedogenic contrast could be pro- This paleosol is therefore a little younger than vided than the calcite fraebules of the highly the changeover from the previously humid cli- alkaline terrain of the study area, and the mate to the present dry conditions. The Sharp maghemite nodules (gamma iron oxide) re- PALAEOPEDOLOGY OF THE MURRAY RIVER 247

quiring an acid water-logged soil for their humid country, these would be strange to them, normal formation. Neverthess, buckshot gravel but they are certainly characteristic. X-ray is found in this region. It was noted c. 50 analysis shows the mineral to be calcite. All km W. of Cobar, N.S.W., and in Cobar itself. the nodules have amorphous centres. Some are W. of the railway station at Diapur, Victoria completely amorphous and various sizes of this (422 km from Melbourne, elevation c. 198 type up to 15 cm diameter have been sectioned. m) a cutting reveals at the top some 2 m of Such are crystalline and compact throughout. calcareous clayey fine sand with derived buck- Others have the exterior covered with fine shot gravel (maghemite fraebules). The frae- laminations of calcite, but the laminated part bules are quite out of character with the cal- is a smaller proportion of the diameter than careous terrain, and belong to some previous the amorphous part. A third type consists of period of more humid conditions. nodules that have been incorporated in a larger N. of the Murray River in the study area, nodule. The principle of super-position applies the general terrain is one of Blanchetown Clay here, and the nodules incorporated in the with surficial sand. Here the carbonate fraebules middle of the large one must be older. The are at or near the surface. S. of the Murray smaller nodules are commonly 2-3 cm in dia- River, the terrain is dominated by a system meter, while the compound ones occur up to of E-W. dunes with two or more paleosols 20 cm in diameter. Plate 33, fig. 1 shows a marked by aggregations of earthy carbonate. section through such a nodule. A comparatively Beneath the dunes are hard calcitic fraebules rare type is the septarian nodule, and is charac- as are found at or near the surface N. of the terized by cracks over its surface that have river. The geomorphology and stratigraphy of been infilled since formation. The cracks are these paleosols is given in an accompanying pronounced on the upper surface but fade paper (Gill, this Memoir). In that paper the away on the lower surface. They appear to a of evidence for climatic oscillation is given. Alter- be function long exposure. One of these nate phases of terrain stability and instability was black, but the colour disappeared on firing existed. The paleosols are associated with the in a crucible so the black material was ap- stable phases, when the climate was more parently organic. Other black nodules owe their humid, because the carbonates were carried colour to being used by the Aboriginals as oven down to the B horizon instead of accumulating stones. at the surface. Southward of the Dividing 2. Ages of Carbonates Range, during the drier phases, carbonates The oldest soil at the present surface is rep- developed in the B horizons in areas where no resented by the top 6 m of the terrain on the carbonates accumulate at present (Gill 1973). N. side of the Murray River, where it is Where carbonate deposition is well developed, developed mostly in Blanchetown Clay. It is the amount of mineral present is far than more illustrated by the cover picture of this Memoir. could possibly derived from the rocks pre- be Superimposed on this is the complex of calcrete- sent, so have transported in. the must been On bearing soils, which extend down over eroded highs at least, the transport must geomorphic surfaces to within about 6 m of the river level, aeolian further evidence of terrain have been — e.g. at Moorna E. oxbow lake. Smaller, less instability. Firman (1969) has referred to this regular fraebules occur on the Rufus Forma- pedocalcic complex as the Bakara Soil, which, tion, a ?Mid-Pleistocene terrace in the present nomenclature of Brewer et al. following the valley tract. The E-W. dune system, best de- the Bakara Pedoderm. (1970), would become veloped in the study area S. of the Murray R., is characterized by a series of pedocals that 1. Carbonate Glaebules have amassed earthy carbonate, which also is Blandowski and many other early explorers calcite. The large lunette on the E. side of Lake comment on the 'numerous nodules of lime- Victoria has red paleosols, and the amount of stone' in the dry country. As they came from a carbonate amassed is small. The lunette onlaps 248 EDMUND GILL

relict 'islands' of Rufus Formation, and so is illustrated in PI. 32, fig. 1, gave a radiocarbon younger than that formation. age of 27,800 ± 1,900 yr B.P. (GaK-1727) As the paleosols are of considerable im- for the outermost 5 mm of the laminated portance for stratigraphy, they are discussed cortex, 28,000 ± 1,800 yr (GaK-1728a) for under that section in the accompanying paper, 5-10 mm from the surface, and > 31,700 yr as also is the chronology. The youngest soil (GaK-1728b) for the small included fraebules dated gave a modern age on charcoal, and this forming the centre of the nodule. The specimen shows that the process of carbonate deposition came from the inland end of the Moorna E. is continuing at present. Paleosols of the order oxbow lake on Moorna Station, W. of Went- of 16,000 yr and 28,000 yr B.P. are recorded worth, N.S.W., where the nodules are grubbed from the E-W. dune system, these dates accord- from the ground for road-making. Solid pans ing well with those of other workers. Under of calcrete appear in places. Dr T. A. Rafter the dune at Berribee Station was a paleosol of the New Zealand Institute of Nuclear exceeding the range of C14 (> 34,300 yr, Sciences expressed interest in these nodules so N.Z. R2729/3). some were forwarded to him. He analysed one A complex calcite fraebule, such as that with the following results:

R2182/2A Outer layer 8 ,;1 CwrtPDB— 5-2% 18,800 ± 380 yr B.P. R2182/2B Intermediate layer 34,400 ± 100 yr B.P. R2182/2C Inner layer a l3c — 3-1% 33,500 ± 900 yr B.P.

There is not necessarily any contradiction in of the carbonate cycle, and the surprises that the results from Japan and New Zealand. Dif- have come from U/Th check on the C14 ages ferent nodules were assayed, although from the of marine carbonate, I have some doubts about same site. Both obtained younger dates on the the older calcrete datings. outside and older ones in the middle, which There is quite an amount of variation in the super-position demands. The soil is polygenctic, chemistry of these large nodules because they and so evidences of soil formation at various have incorporated different portions of the sub- periods are to be expected. In this region, soil strate, quite apart from any variation in the formation occurred at roughly 16,000 yr, secondary carbonate. Professor K. Kigoshi has 28,000 yr and 34,000 yr. Each assay picked kindly permitted me to quote the following up two of these three phases. The matter re- analyses by H. Ito and himself of the nodule quires further investigation. In view of the long he assayed. The first three parts numbered in time gap prior to the 34,000 yr paleosol, the the table below, are the three portions used comparatively small fraebules on the Rufus for dating, while the fourth is from the very Formation with extinct marsupials that is prob- centre of the nodule. The figures are percent- ably mid-Pleistocene, the lack of understanding ages:

FeoO-s A1 2 3 CaO MgO Si02 Na2 co2 H 2 Total 1. 103 2-60 430 210 1205 0-257 320 3 10 96 17 2. 0-322 933 48-5 181 6-21 0-214 32-9 3-51 94-40 3. 682 1-24 43 6 1-28 15-48 328 31-36 1-80 95-82 4. 0-951 1 98 391 1-41 22-3 0338 27-40 2-25 95-74

The three nodule fractions assayed by Rafter, derm, but below it (and therefore earlier) were 2A-C respectively, had 35 8, 35 -5 and 34-2% found the sediments which yielded palynologic of CO.. evidence of the incoming of arid conditions The Moorna Formation (Gill, this Memoir) similar to the present. That time is 1-5—2m.y. is Pliocene in age, and over it is the Blanche- ago, on present understanding, and more car- town Clay-Chowilla Sand complex. Low in this bonate paleosols than exist would be expected stratigraphic horizon is the Karoonda Pedo- for that period. It is possible that they have )

PALAEOPEDOLOGY OF THE MURRAY RIVER 249 been formed and then leached or eroded away, of the lake, and this encountered such nodules during humid intervals. in place: For discussion of the dating of these pedo- 15 m Loose red sand genic carbonates, see Rightmire (1967), Wil- 0-25 Closely packed carbonate nodules liams and Polach (1969), Bowler and Polach with complex ones as in PL 33, and Williams and (1971), Polach (1971). fig. 1 at the top, merging into 76 Scattered nodules in red sandy 3. Vectors of Carbonate Deposition Blanchetown Clay, merging into Difference of opinion exists on how carbonate 15+ Light gray weathered Blanchetown is deposited in the soil profiles of dry lands: Clay with five layers of softer rather

(a) That the carbonate is carried down gravi- irregular ( but generally platey tationally by rain water, i.e. vector normal carbonate nodules, plus some rhi- to ground surface and sense downward. zomorphs. (b) That the carbonate is carried upwards in Nearby are areas where a solid harpan covers solutions rising by capillary attraction. the calcite spherical nodules (PL 33, fig. 2). 'Caliches seem to form per ascendum by This locality is at the end of long low slope, capillarity' (Termier and Termier 1963, and it has been noted that other areas of maxi- p. 402), i.e. same vector but sense up- mal development of calcrete are so placed. It ward. would seem that the heavy rain that occasionally The above two views have been placed in appo- falls soaks through the loose surficial sand, sition, but this may be a false antithesis. Special then accumulates at the foot of the slope where conditions are required for capillary action, its load of carbonate is deposited on drying out. but when they are present, the rise of solutions This also appears to be the process by which may indeed occur. On the other hand, calcrete the 'indurated layer' on the alluvial fans on has been deposited in coarse gravels, where the W. side of Lake Victoria (C14 date capillary action probably could not occur. 'modern') are formed. Nevertheless, the small The author's observations suggest that de- spherical nodules occur over the whole terrain. position can take place in any vector and in These nodules are quarried for the roads, and either sense, but that the downward transport some of these quarries are on geomorphic of carbonate in solution is the commonest pro- highs, e.g. the road up the E. side of Lake

cess. That it was in solution is shown by Victoria. They are a normal pedogenic product radiocarbon dating. Calcite dust from Tertiary and not evaporites as some have suggested. limestones, older paleosols, and such, all Hawley et al. (1968), working in the Rio combine to form a soil that gives consistent Grande valley, noted that 'There is a direct radiocarbon dates over thousands of square correlation between geomorphic surface age kilometres. and thickness and morphological complexity Deposition from all directions is shown by of associated caliche'. 'Ground water and the millions of spherical calcrete nodules with capillary-fringe processes did not play a sig- a cortex of regular laminations (PI. 33). Down- nificant role in caliche genesis. The C14 of ward movement is shown by PL 33, fig. 3, analyses indicate that the age the caliche where the large nodules are aggregations of generally increases with depth. There is strong smaller ones united by enveloping laminate cal- evidence for aeolian origin of much of the caliche.' cite, as shown in the section PI. 33, fig. 1. The carbonate in the These comments small nodules protrude through the bottom, but apply well to our study area. are thickly covered on top (PL 33, fig. 1). The In Western Victoria, a thick hardpan of in nodules in PL 33, fig. 3 are in place, having calcrete is covered some areas by a com- been exposed by deflation at the edge of the paratively thin layer of laminated mammillary Moorna E. oxbow lake. An excavation was calcite (Gill 1973). The hardpan is Last carried out on the flat ground at the NW. end Glacial (^ 20,000 yr B.P.) while the lamina-

Kl 1

250 EDMUND GILL ted calcitc is Postglacial (— 8,700 yr). As the for dunes (see grain size analyses, Gill hardpan is more or less impenetrable, the this Memoir) due to the general fineness younger layer must have been deposited from of the sediments in this river regime of above and not per ascendum. Moreover, the unusually low dynamics, hardpan has developed in aeolianite as the B (b) The stratification is mostly subhorizontal. horizon of a terra rossa. The highly calcareous The strong W. winds apparently kept this matrix has been leached almost free of car- fine sand in a state of constant blowout. bonate which has been deposited downwards As a result the paleosols are mostly subhori- into the hardpan. zontal. Figure 4 shows diagrammatically the In both Western Victoria and the study area basic pattern of paleosol occurrence. However, there are some well developed rhizomorphs. As the red soils frequently bifurcate. The largest these root casts thin downwards, it is difficult number seen was when in drought conditions, to see how the root cavity (and no more) the S. wall of a gulch showed five red paleosols could be filled with secondary carbonate if it plus the columnar gray soil at the present were ascending by capillary attraction. In the surface (Gill, his Memoir, Fig. 42). Aboriginal study area (and many other sites in the semi- middens are associated with the paleosols, arid country) cylinders of carbonate have been making possible the radiocarbon dates shown collected loose on the surface, and in situ. Some in Fig. 4. This series of paleosols needs study of these are thought to be rhizomorphs that by a pedologist. They reflect more humid con- have broken into sections, while others almost ditions in that they are not full of carbonates certainly are infilled burrows. Such were studied as is the present terrain. at the N. end of Lake Wcthercll in the Menin- A comparable suite of fossil soils occurs in dee district. Some were about 25 cm in the E-W. dune system (Woorinen Formation) diameter with irregular sides, and a tendency of this region. The lowest soil shown in Fig. 4 to a spiral arrangement (cf. Gill, this Memoir, is of the same age as the youngest found in the PI. 7, fig. 5). Some carbonate cylinders from E-W. system in the study area. Firman (1969 apparent burrows were tested for phosphate, and references) has applied the name Loveday but the results were negative. Soil to the paleosols of the Woorinen Forma- tion. 4. Lunette Pcdeosots Paleosol Succession On the E. side of Lake Victoria is a large lunette some 14 km long. It is remarkable for The sequence of fossil soils described in two features: tli is paper is thus:

(a) The sand is a grade finer than is usual Holocene Unnamed

LAKE VlCji >i:l \ swns I UkM w u >\ (in? uilumiMi vnl

I II I II ! 1 1 1 M 1 I 1 1 1 I rTTTTTTTTTTTTTTTT

1 \i OARm 5 WD Ml Mill R \hnri« n.il Miilcie ' '.'.'. l i i I /i i i i i i i i i 1 rU L" .'. \ n n m r i i i i i ix

IM-' "1 ' Rwl \ r VO Slli II f] IMKi Fossil Giikk J CT " -"-'.• Hon.: U IfilUr '. Old H.M.I Site

1"1 I I 1 I I I 1 I'l I'l TTT 1 I'l I I I I ^1 I ITTTTTI II II I I I') I 1 I 1 I I'll I'M I 'I I ipH I III rf II II iTl II MTTTT-

MiJI |fi, Mm yj V .K.-.l IM.m-h SlKll - I 'i ,,,.. ,1 I K> ;

M1IAM1I \ S\M) Ml \1Ul K

Fig. 4—Diagram of the basic paleosol occurrence in the Lake Victoria Sand (lunette, E. side Lake Victoria, N.S.W.) with chronology. The paleosols are of the order of 6,000 and 17,000 yr old. The latter is the date of the youngest paleosol in the E-W. dune system (Woorinen Formation). In places in the lunette the paleosols divide to give a complex sequence. —

PALAEOPEDOLOGY OF THE MURRAY RIVER 251

Search 2: 32. Pleistocene Loveday Pedoderm (pedocal , 1971. Laterite chronology. rhizomorphs —, 1973. Calcrete hardpans and series) in Western Victoria, Australia. Pacific Geol. In Bakara Pedoderm (pedocal press. complex) Grubb, P. L. C, 1971. Genesis of bauxite deposits in the Boolarra-Mirboo area of Gippsland, Vic- Plio- toria. J. geol. Soc. Aust. 18: 107-113. Pleistocene Karoonda Pedoderm (silcrete/ Hawley, J. W., Gile, L. H., and Grossman, K. B., Caliche development related to the geo- opal) 1968. morphic evolution of the Rio Grande Valley. Pliocene Timboon Pedoderm (lateritic). Geol. Soc. Am. Ahstr. Ann. Mtg. 1968, p. 130. Jessup, R. W., 1960a. The lateritic soils of the SE- Comment portion of the Australian Arid Zone. /. Soil Sci. 11: 106-113. These notes have been prepared in the hope , 1960b. Introduction to the soils of the SE. portion of the Australian Arid Zone. /. Soil Sci. that they will encourage a full study of the 11: 92-105. soils and paleosols of this area. Geologically Mabbutt, J. A., 1965. The weathered land surface and palaeoclimatologically they have proved to in Central Australia. Xeit. Geomorph. 9 (1): 82-114. significant, and no doubt they will be very Mason, B., 1966. Principles of geochemistry (3rd prove to be as rewarding pedologically. Ed.) New York. Rightmire, C. T., 1967. A radiocarbon study of the age and origin of caliche. Geol. Soc. Amer. Spec. References Pap. 115: 184 (Abstr.) Shutler, R., 1968. Early man in Western North charac- Blackburn, G., and Leslie, T. I., 1958. The America (Symposium). E. New Mexico Univ. of soils in the Coleraine teristics and origins Contrih. in Anthrop. 1 (4). district, Victoria. CSIRO Soil Publ. 12. Stephens, C. G., 1971. Laterite and silcrete in Aus- 1971. Radio- Bowler, J. M., and Polach, H. A., tralia: a study of the genetic relationships of carbonates: an evaluation carbon analyses of soil laterite and silcrete and their companion materi- Southeastern Australia. In from paleosols in als, and their collective significance in the for- Yaalon, D. H. Paleopedology, Jerusalem, pp. mation of the weathered mantle, soils, relief 97-108. and drainage of the Australian continent. Geo- Speight, J. G., Brewer, R., Crook, K. A. W., and derma 5: 5-52. the 1970. Proposal for soil-stratigraphic units in Tkrmier, H., and Termier, G-, 1963. Erosion and Australian Stratigraphic Code. J. geol, Soc. Aust. sedimentation London. 103-111. 17: A , . Twidale, C. R., 1972. Landform development in the Dury, G. H., 1971. Relict deep weathering and dun- Lake Eyre region, Australia. Geogr. Rev. 62; crusting in relation to the palaeoenvironments of 40-70. middle latitudes. Geogr. J. 137: 511-522. Williams, G. E., and Polach, H. A., 1969. The Their Faniran, A„ 1970. The Sydney duricrusts: evaluation of 14 C ages for soil carbonate from Sci. J., terminology and nomenclature. Earth the arid zone. Earth plan. Sci. Lett. 1: 240-242. 117-128. New Zealand. 4 (2): , 1 97 1 . Radiocarbon dating of arid-zone the Chowilla FreMAN, J. B., 1966. Stratigraphy of calcareous paleosols. Bull. geol. Soc. Am. 82: area in the Murray Basin. Geol. quart. Notes, 3069-3086. Geol. Surv. S.A. 20: 3-7. Wopfner, H., and Twidale, C. R., 1967. Geomor- late Cainozoic de- ., 1967. Stratigraphy of phical history of the Lake Eyre Basin. Landform Soc. S. posits in South Australia. Trans, R. st {ulies from Australia and New Guinea (Ed. Aust. 91: 165-178. J, N. Jennings and J. A. Mabbutt). Canberra. analysis of soils near -, 1969. Stratigraphic Pp. 119-143. Adelaide, South Australia. Trans. R. Soc. S. Aust. 93: 39-54. , ... and Explanation of Plates 32-33 Fisher, N. H., 1958. Notes on latentization L. mineral deposits. Aust. Inst. Min. Metall. F. Plate 32 Stillwell Aniv. Vol. pp. 133-142. R. at Sharp Point. and Fi e . i n. bank of Murray Nam- Gibbons, F. R., and Gill, E. D., 1964. Terrains Vic- poo Stn., E. of Cal Lai and W. of Went- soils of the basaltic plains of far Western worth, N.S.W. (Site 4) showing at base white toria. Proc. R. Soc. Vict. 11: 387-395. Gondwanaland dolomite (partly replaced by opal) in Blanche- Gili E D., 1961a. The climates of Palaeo- town Clay. Chowilla Sand overlies this, and in Kainozoic time. Ch. 14 of Descriptive the sand in turn by another layer of dolomite climatology (Ed. Nairn) New York and opal. When Lake Bungunnia dried up, it 1961b. Cainozoic climates of Australia. ceased depositing Blanchetown Clay and de- Ann! New York Acad. Sci. 95 (1): 461-464. with the basaltic posited dolomite. Pedogenic processes on the -, 1964. Rocks contiguous Soc. Vict. dry lake bed resulted in deposition of silica cuirass of Western Victoria. Proc. R. and formation of rhizomorphs. The Chowilla 11: 331-355...... Sand is a channel deposit. During the re- 1966 Geochronology of Victorian Mallee establishment of Lake Bungunnia, another ridges. Proc. R. Soc. Vict. 79: 555-559. R2 252 EDMUND GILL

dolomite layer was deposited, and the process bonate fraebules enclosed in the laminated of silica deposition repeated. calcite envelope. The coin is 1 -7 cm diameter. Fig. 2—Rxhumed Karoonda Terrain forms a platform Site 13. on the E. bank of Salt Creek, (as it does at Fig. 2—Same site. Carbonate fraebules as they occur Cal Lai and elsewhere), Kulcurna Sin., N.S.W. on the ground surface where it has been Opal was deposited in clay or dolomite, while eroded by wind and rain at the edge of the sands were silicified to various degrees. This oxbow cliff. The coin is 1-9 cm in diameter. horizon is the Karoonda Pedodcrm. Fig. 3 —Same site. Large nodules as in Fig. 1 in situ, Fig. 3 — Mottled zone of a lateritic profile at the base but exposed by erosion. The upper surface of the cliff forming the S. bank of the Murray has always a thick layer of secondary car- R. at Yelta, N. Victoria. This is the Timboon bonate, but the enclosed fraebules project Pedodcrm. out of the thin bottom layer, proving a pre- dominantly downward movement of the car- Plate 33 bonate, but the enclosed fraebules project Fig. 1 —Polished section through one of the com- out of the thin bottom layer, proving a pre- pound nodules at the NW. corner of the E. dominantly downward movement of the Moorna Stn. oxbow W. of the Darling Ana- carbonate. The coin is 2-8 cm in diameter. branch. Note the older generation of car- —All photos by the author. MEM. NAT. MUS. VICT. 34 PLATE 32 MEM. NAT. MUS. VICT. 34 PLATE 33