Ancient Soils and Ancient Climates
FRANKLYN B.VAN HOUTEN
The essence of this paper is a consideration of how ancient soils may reflect long-term changes in climate and in the composition of the atmosphere. Detailed field studies and laboratory analyses of ancient nonmarine deposits commonly reveal remnants of fossil soils. These paleosols, together with transported clay- and iron-rich sediments derived directly from them, provide clues for reconstructing ancient climates, because they were formed under a limited range of temperature and humidity.
Basic inquiries that direct such an investigation focus on (1) the relation of climate to a global temperature gradient and (2) patterns of precipitation controlled by the distribution of landmasses and their mountain belts, as well as by (3) the global distribution of drifting continents and (4) the modifying effect of the stand of global sea level and flooding of the continents. Paleosols amplify other geologic information about trends in climate change through Earth history and about those conditions that produced widespread coal swamps and continental glaciation. The data also reflect the role of major regressions of the seas and impose critical constraints on speculation about Precambrian climate and atmosphere.
The discipline of paleopedology (Yaalon, 1971) is founded on the well-documented proposition that widespread soils in the geologic past, as in the present (White, 1979), reflect atmospheric temperature and humidity, the former varying mostly with latitude, the latter asymmetrically east to west. These, in turn, are conditioned and augmented by effects of local relief and biota, of stability of the landmass (time factor), and of the distribution of land and sea.
Analysis of an ancient weathered profile necessary to establish its particular pedologic character (Table 11.1) requires data concerning the petrography of the parent rock, mineralogic and chemical composition of the profile, and possible diagenetic alteration. Such a study deals with those downward-mobilizing processes that oxidize plant debris, lower soil-water pH, dissolve minerals and remove soluble cations, produce new clay minerals (Singer, 1980), and concentrate insoluble ferric and aluminum oxides in the upper horizon of the soil profile. Current procedures and modes of interpretation involved in paleosol analyses are well illustrated in studies by McPherson (1979) and Retallack (1976, 1977).
TABLE 11.1 Criteria Diagnostic of Pedogenic Profiles
Paleosols comprise those profiles of weathering developed on rocks or sediments that were exposed long enough to be modified by soil-forming processes. For the present review, potentially meaningful ones are limited largely to lateritic products of hot and humid weathering (Thomas, 1974) composed essentially of ferric and aluminous oxides (sequioxides) and high-aluminum kaolinite clay and to calcretes produced by less intense dry-climate weathering and composed of ferric oxides, clay minerals, and calcium carbonate. Today laterite and calcrete predominate on relatively stable landmasses within the intertropical climate zone (Figure 11.1; Van Houten, 1961, 1973). Here the temperature is less important than differences in amount and seasonal distribution of rainfall in determining the kind of soil produced.
Lateritic soils or latosols (Millot, 1970; Paton and Williams, 1972), both aluminous bauxite and ferruginous laterite, form under conditions of prolonged stability and intense weathering, accompanied by essentially no erosion or aggradation. They are favored by relatively uniform maritime conditions on windward sides of continents within the humid intertropical, or tropical forest, zone. Sequioxides and kaolinitic products of laterization can be transported to local depressions, as in a karst terrane, or to depositional basins where distinctive aluminum-rich claystones and oolitic ironstone may accumulate (Millot, 1970).
Carbonate-enriched calcrete (caliche) and calcareous red-earth soils reflect warm, seasonally dry climate and commonly develop on drier leeward sides of continents. In as much as calcrete requires less-intense weathering than laterite, it can form in periods of thousands of years. Estimates suggest that calcrete profiles develop at rates of 10–50 mm/103 yr with a maximum of about 1000 mm/103 yr. In the geologic record calcretes are commonly associated with reddish-brown detrital deposits containing more silica-rich clay minerals such as illite, montmorillonite, and chlorite. Used prudently these redbeds (Millot, 1970) can provide suggestive evidence of warm, dry climate.
Ancient soils as clues to ancient climate have limitations. Even though laterites and calcretes commonly developed on uplands and well-drained nonmarine deposits in the past, terranes of this sort constitute only a minor part of the geologic record. Moreover, to provide useful information, ancient soils, wherever they formed, must have been buried relatively rapidly to prevent subsequent chemical modification or loss by erosion. Conversely, soils could develop on nonmarine deposits only where accumulation was interrupted or was slow enough to permit development of a distinctive profile. Yet, commonly in such a situation pedological processes were arrested before completion, or the upper diagnostic part of the profile was eroded. In addition, many ancient soils cannot be dated accurately as to time of origin, and after burial some were altered by diagenetic effects that obscure primary climatic indicators. The use of Recent laterite and calcrete as a model is limited, particularly because modern soils have developed during a time of considerable tectonic activity, a low stand of sea level, and assembled continental blocks (Figure 11.1). These factors reduce the global extent of maritime conditions and cratonic stability that favored the formation of laterite in the past.
The available data plotted on currently reasonable paleocontinental drift reconstructions (Smith and Briden, 1977; Scotese et al., 1979) and transferred to a time-latitude chart (Figure 11.1) provide a basis for evaluating paleoclimatic information supplied by ancient laterite and calcrete. The record of their temporal and geographic distribution reflects only large-scale motion and change in the atmosphere, hydrosphere, and lithosphere. During the Phanerozoic interplay of global climate zones, latitudinal position of wandering continents, and the longitudinal extent of orogenic belts, and of assembled cratonic blocks, the role of global climate change probably was minimal.
As today, laterites in the past lay mostly within the intertropical zones in east-coast locations (Valeton, 1972; Bardossy, 1973). They were widespread during siderolithic periods (Millot, 1970) when extensive, relatively stable areas of the continents were subjected to prolonged weathering in hot and humid climate, producing ferric oxides, kaolinitic clay, beds of lignite and coal, and quartz-rich sand. Commonly, development of laterite was accompanied by the nearby accumulation of transported bauxite or kaolinite and oolitic ironstones.
A general northward drift of most of the major landmasses during Phanerozoic time (Figure 11.1) produced a northward shift of the broad belt of laterite on Laurasian continents and
led to low-latitude development on northern blocks of Gondwana in Mesozoic and Cenozoic time. This trend was modified by a late Mesozoic-Cenozoic southward expansion of the laterite belt when Australia broke away from Antarctica. In middle and late Paleozoic time favorable lateritic conditions were widespread 011 dispersed Laurasian blocks in middle and high latitudes but occurred on assembled Laurasian blocks only in low latitudes. Similarly, in Mesozoic and early Cenozoic time laterites developed in high northern latitudes toward leeward coasts (Figure 11.2) when the open Atlantic, Tethys, and Caribbean permitted latitudinal oceanic circulation among continental fragments.
The limited resolution of the laterite data is illustrated by the fact that a reconstructed 10° northward drift of Africa and Australia between Eocene and Miocene time is not reflected in the known distribution pattern of laterite. Early Permian to Cenozoic laterite in southern latitudes higher than 30° S are known only in southeastern Australia (Figures 11.1–11.3). The Permian and Triassic paleosols have been interpreted to be products of humid tropical to subtropical intervals within and following the extensive Gondwana glaciation (Loughnan et al., 1974; Loughnan, 1975). Development of oolitic ironstone in Australia in late Permian and early Jurassic time also suggests periods of warm, moist climate. Nevertheless, Retallack
(1977) claims that the early Triassic profiles are podzols and that “the climate in southeastern Australia seems to have remained cool temperate from the later Permian to the later Triassic.” Late Paleozoic and early Mesozoic laterite in high northern latitudes accumulated in Russia and northern Asia-Siberia when they were the principal landmasses in the far north (Figures 11.1 and 11.3). According to Strakhov (1969, p. 214, Figure 94), these paleosols developed in moist, tropical to subtropical climate.
Almost all of the ancient calcretes recorded formed on intermittently aggraded redbeds in unstable (mobile) belts, commonly within the intertropical zone and on the leeward side of a landmass. All but one lay between latitudes 45° N and 45° S, and most of these are preserved on Laurasian continental blocks. During late Permian and Triassic time, longitudinal assemblage of Laurasia and Gondwana prevented latitudinal oceanic circulation and led to a preponderance of calcrete over laterite in the intertropical zone of North America and Europe (Figures 11.1 and 11.3). This reconstructed low-latitude belt of dryness was rapidly succeeded by relatively humid Early Jurassic climate and the widespread development of laterite and oolitic ironstone in Eurasia and northeastern Gondwana.
A Proterozoic record of laterite and oolitic ironstone, calcrete, and nonmarine redbeds (Chandler, 1980) and aluminous sandstone (Young, 1973) on the Canadian Shield is augmented by redbeds as old as 2000 million years (Ma) in southern Africa (Button and Tyler, 1979) and 2380 Ma in India (Windley, 1977). These data imply an oxygen-deficient atmosphere since about 2300 Ma, and a warm, humid to semiarid climate when the ferric oxide-bearing sediments accumulated. Detailed analysis of this record is complicated by the fact that Canada apparently drifted widely into high and low latitudes during Proterozoic time (Irving, 1979). Nevertheless, both the extensive iron formations and various redbeds that accumulated between 2300 and 1800 Ma lay in relatively low paleolatitudes (Donaldson et al., 1973).
Accumulation of Proterozoic products of oxidative weathering overlapped the widespread blooming of cherty and banded-iron formations by as much as 500 m.y. (Figure 11.4). A specific association of the two, about 2000 Ma, comprises a redbed with weathered andesite pebbles that lies below the Labrador cherty iron formation (Dimroth, 1976). Clearly,
references to an oxygen-deficient atmosphere after 2000–2300 Ma (Button and Tyler, 1979; Chapter 5) are speculations that ignore the facts of the geologic record. In contrast, a marked abundance of CO2 in the atmosphere and hydrosphere during this interval (Strakhov, 1969) may have been an important factor in the genesis of the unique Proterozoic iron formations.
Evidence of chemical weathering older than 2300 Ma in Canada (2200 Ma in southern Africa; Button, 1979; Button and Tyler, 1979) includes metamorphosed aluminum-rich weathering crusts as old as 3500 Ma (Serdyuchenko, 1968), Commonly, iron has been leached from these profiles, indicating at least local reducing conditions. The evidence does suggest an atmosphere with a limited supply of oxygen (Pienaar, 1963). Nevertheless, one of the gray paleosols is virtually identical to some modern soils formed in temperate, humid climate (Gay and Grandstaff, 1979). Others may be products of oxidative weathering that were altered diagenetically by reducing paleogroundwater. In lieu of substantial evidence, gray profiles alone afford no compelling indication that the early Proterozoic atmosphere was significantly different from that of later time.
Pedologic effects of ancient climate are more boldly stamped on nonmarine sedimentary rocks than generally acknowledged. Recent detailed studies, such as those by Allen (1974), McPherson (1979), and Retallack (1976, 1977), reveal the amount of useful information that can be gleaned. With careful work the amount of available data should be increased significantly.
There is need for a wider understanding of the criteria for identifying ancient soils (Table 11.1), and for better-trained students who can use the techniques of modern soil science in detailed analyses of nonmarine deposits. Problems involved in this practice include identifying a soil type from a partially eroded profile or recognizing ferruginous and kaolinitic profiles produced by processes other than laterization and distinguishing between the effect of intense warm-climate weathering and of prolonged weathering under more moderate conditions and regional stability.
Generally, the available record of paleosols is more directly related to episodes of widespread nonmarine sedimentation than it is to a global control of climate. This situation, in turn, was closely controlled by tectonic development of appropriate basins. In addition, development and distribution of laterite and calcrete reviewed here imply that the paleolatitudinal position and interrelation of landmasses, and the latitudinal and longitudinal oceanic circulation through open gateways (Chapter 12) during Phanerozoic history of drifting continents, were the principal forcing factors in long-range climate change. These conditions were augmented by effects of the stand of global sea level and extent of flooding of the continents, as well as by the development of continental glaciation or of ice-free polar regions. In contrast, ancient soils contribute little to an understanding of short-term climate-controlled processes or periodicities like those exhibited in the Pleistocene record.
Throughout Phanerozoic time the intertropical belt apparently persisted in essentially steady state but shifted somewhat northward, except during Permo-Triassic time when a meridional assemblage of major landmasses (Pangaea) was a barrier to latitudinal oceanic circulation. During episodes of widespread regression, as in Early Cretaceous time, nonmarine sequences of stable landmasses commonly preserved a record of hot and humid climate in successive ferruginized profiles, as well as in deposits of transported laterites and of kaolinite associated with oolitic ironstones.
Laterite and oolitic ironstones in the early Paleozoic and Proterozoic record demonstrate that development of these products of intense weathering were not uniquely dependent on the presence of vascular plants. In fact, an absence of abundant plant debris may have fostered the development of oxidized soils in an atmosphere with a somewhat diminished oxygen content.
Data and interpretations presented here suggest that both the Earth’s atmosphere and its global climate have varied but little during the last 2000 million years. As in Phanerozoic time, Proterozoic paleosols and redbeds reflect conditions favoring preservation of nonmarine deposits rather than variation in oxygen content of the atmosphere. Correlative banded-iron formations may reflect some aspect of a progressive change in the atmosphere. Speculation about this condition is constrained, however, by evidence of an oxidizing atmosphere since about 2300 Ma. In fact, accumulation of banded iron for-
mations may not have been directly related to the oxygen content of the atmosphere and oceans. A significant role of a reducing atmosphere has not been demonstrated (Dimroth and Kimberly, 1976). At present, hypotheses proposing an oxygen-deficient atmosphere in early Proterozoic time are not rigorously convincing.
Research for this report was supported by National Science Foundation grant EAR 77–06007 and funds provided by the Department of Geological and Geophysical Sciences, Princeton University. J.F.Hubert, J.C.Lorenze, J.G.McPherson, and D.L.Woodrow supplied useful information and helpful suggestions.
Allen, J.R.L. (1974). Studies in fluviatile sedimentation: Implications of pedogenic carbonate units, Lower Old Red Sandstone, Anglo-Welch outcrops, Geol. J. 9, 181–208.
Bardossy, G. (1973). Bauxite formation and plate tectonics, Acta Geol. Acad. Sci. Hung. 3, 141–154.
Button, A. (1979). Early Proterozoic weathering profile 011 the 2200 m.y. old Hekepoort Basalt, Pretoria Group, South Africa, U. Wits. Econ. Geol. Res. Unit Info. Circ. 133, 19 pp.
Button, A., and N.Tyler (1979). Precambrian paleoweathering and erosion surfaces in Southern Africa: Review of their character and economic significance, U. Wits. Eco. Geol. Res. Unit Info. Circ. 135, 37 pp.
Chandler, F.W. (1980). Proterozoic redbed sequences of Canada, Geol. Surv. Canada 311, 53 pp.
Dimroth, E. (1976). Aspects of sedimentary petrology of cherty iron-formations, in Handbook of Stratabound and Stratiform Ore Deposits, Vol. 7, Elsevier, Amsterdam, pp. 203–254.
Dimroth, E., and M.M.Kimberley (1976). Precambrian atmospheric oxygen: Evidence in sedimentary distribution of carbon, sulfur, uranium and iron, Can. J. Earth Sci. 13, 1161–1185.
Donaldson, J.A., E.Irving, J.C.McGlynn, and J.K.Park (1973). Drift of the Canadian Shield, in Implications of Continental Drift to the Earth Sciences, Vol. 1, Academic, London, pp. 3–17.
Gay, A.L., and D.E.Grandstaff (1979). Precambrian paleosols at Elliot Lake, Ontario, Canada (abstract), Geol. Assoc. Canada— Mineral. Assoc. Canada, Joint Ann. Mtg., Abstracts, 52.
Irving, E. (1979). Paleopoles and paleolatitudes of North America and speculations about displaced terrains, Can. J. Earth Sci. 16, 669– 694.
Loughnan, F.C. (1975). Laterites and flint clays in the early Permian of the Sydney Basin, Australia, and their paleoclimatic implications, J. Sed. Petrol. 45, 591–598.
Loughnan, F.C., R. Goldberg, and W.N. Holland (1974). Kaolinitic clay rocks in the Triassic Banks wall sandstone of the western Blue Mountains, New South Wales, J. Geol. Soc. Australia 21, 392–402.
McPherson, J.G. (1979). Calcrete (caliche) paleosols in fluvial red-beds of the Aztec Siltstone (Upper Devonian), southern Victoria Land, Antarctica, Sed. Geol. 22, 267–285.
Millot, G. (1970). Geology of Clays, Springer-Verlag, New York, 429 pp.
Paton, T.R., and M.A.J.Williams (1972). The concept of laterite, Ann. Assoc. Am. Geogr. 62, 42–56.
Pienaar, P.J. (1963). Stratigraphy, petrology, and genesis of the Elliot Lake Group, Blind River, Ontario, including the uraniferous conglomerate, Geol. Surv. Canada 83, 140 pp.
Retallack, G.J. (1976). Triassic paleosols in the Upper Narrabeen Group of New South Wales. Pt. I. Features of the paleosols, J. Geol. Soc. Australia 23, 383–400.
Retallack, G.J. (1977). Triassic paleosols in the Upper Narrabeen Group of New South Wales. Pt. II. Classification and reconstruction, J. Geol. Soc. Australia 24, 19–36.
Scotese, C.R., R.K.Bambach, C.Barton, R.Van der Voo, and A.M. Ziegler (1979). Paleozoic base maps, J. Geol. 87, 217–277.
Serdyuchenko, D.P. (1968). Metamorposed weathering crusts of the Precambrian, their metallogenic and petrographic features, XXIII Internat. Geol. Congr. 4, 37–42.
Singer, A. (1980). The paleoclimatic interpretation of clay minerals in soils and weathering profiles, Earth Sci. Rev. 15, 303–326.
Smith, A.G., and J.C.Briden (1977). Mesozoic and Cenozoic Paleocontinental Maps, Cambridge U. Press, Cambridge, 63 pp.
Strakhov, N.M. (1969). Principles of Lithologenesis, Vol. 2, Oliver and Boyd, Edinburgh, 609 pp.
Thomas, M.F. (1974). Tropical Geomorphology, Wiley, New York, 332 pp.
Vail, P.R., R.M.Mitchum, Jr., and S.Thompson III (1977). Seismic stratigraphy and global changes of sea level, pt. 4, in Seismic Stratigraphy, Am. Assoc. Petroleum Geol. Mem. 26, pp. 83–97.
Valeton, I. (1972). Bauxites, Elsevier, Amsterdam, 226 pp.
Van Houten, F.B. (1961). Climatic significance of redbeds, in Descriptive Paleoclimatology, Interscience, New York, pp. 89–139.
Van Houten, F.B. (1973). The origin of redbeds: A review—1961– 1972, Ann. Rev. Earth Planet. Sci. 1, 39–61.
White, R.E. (1979). Introduction to the Principles and Practice of Soil Science, Wiley, New York, 198 pp.
Windley, B.F. (1977). The Evolving Continents, Wiley, London, 385 pp.
Yaalon, D.H., ed. (1971). Paleopedology—Origin, Nature, and Dating of Paleosols, Israel U. Press, Jerusalem, 350 pp.
Young, G.M. (1973). Tillites and aluminous quartzites as possible time markers for Middle Precambrian (Aphebian) rocks of North America, Geol. Assoc. Canada 12, 97–127.