Earth's glacial record and its tectonic setting

Earth's glacial record and its tectonic setting

Earth-Science Reviews, 35 (1993) 1 - 2 4 8 Elsevier Science Publishers B.V., A m s t e r d a m Earth's glacial record and its tectonic setting * N. ...

22MB Sizes 0 Downloads 48 Views

Earth-Science Reviews, 35 (1993) 1 - 2 4 8 Elsevier Science Publishers B.V., A m s t e r d a m

Earth's glacial record and its tectonic setting * N.

Eyles

Glaciated Basin Research Group, Department of Geology, Scarborough Campus, University of Toronto, 1265 Military Trail, Scarborough, Ontario M1C 1A4, Canada (Revised and accepted December 23, 1992)

ABSTRACT Eyles, N., 1993. Earth's glacial record and its tectonic setting. Earth-Sci. Rev., 35: 1-248. Glaciations have occurred episodically at different time intervals and for different durations in Earth's history. Ice covers have formed in a wide range of plate tectonic and structural settings but the bulk of Earth's glacial record can be shown to have been deposited and preserved in basins within extensional settings. In such basins, source area uplift and basin subsidence fulfill the tectonic preconditions for the initiation of glaciation and the accomodation and preservation of glaciclastic sediments. Tectonic setting, in particular subsidence rates, also dictates the type of glaciclastic facies and facies successions that are deposited. Many pre-Pleistocene glaciated basins commonly contain well-defined tectonostratigraphic successions recording the interplay of tectonics and sedimentation; traditional climatostratigraphic approaches involving interpretation in terms of either ice advance/retreat cycles or glacio-eustatic sea-level change require revision. The direct record of continental glaciation in Earth history, in the form of classically-recognised continental glacial landforms and "tillites", is meagre; it is probable that more than 95% of the volume of preserved "glacial" strata are glacially-influenced marine deposits that record delivery of large amounts of glaciclastic sediment to offshore basins. This flux has been partially or completely reworked by "normal" sedimentary processes such that the record of glaciation and climate change is recorded in marine successions and is difficult to decipher. The dominant "glacial" facies in the rock record are subaqueous debris flow diamictites and turbidites recording the selective preservation of poorly-sorted glaciclastic sediment deposited in deep water basins by sediment gravity flows. However, these facies are also typical of many non-glacial settings, especially volcanically-influenced environments; numerous Archean and Proterozoic diamictites, described in the older literature as tillites, have no clearly established glacial parentage. The same remarks apply to many successions of laminated and thin-bedded facies interpreted as "varvites". Despite suggestions of much lower values of solar luminosity (the weak young sun hypothesis), the stratigraphic record of Archean glaciations is not extensive and may be the result of non-preservation. However, the effects of very different Archean global tectonic regimes and much higher geothermal heat flows, combined with a Venus-like atmosphere warmed by elevated levels of CO 2, cannot be ruled out. The oldest unambiguous glacial succession in Earth history appears to be the Early Proterozoic Gowganda Formation of the Huronian Supergroup in Ontario; the age of this event is not well-constrained but glaciation coincided with regional rifting, and may be causally related to, oxygenation of Earth's atmosphere just after 2300 Ma. New evidence that oxygenation is tectonically, not biologically driven, stresses the intimate relationship between plate tectonics, evolution of the atmosphere and glaciation. Global geochemical controls, such as elevated atmospheric CO 2 levels, may be responsible for a long mid-Proterozoic non-glacial interval after 2000 Ma that was terminated by the Late Proterozoic glaciations just after 800 Ma. A persistent theme in both Late Proterozoic and Phanerozoic glaciations is the adiabatic effect of tectonic uplift, either along collisional margins or as a result of passive margin uplifts in areas of extended crust, as the trigger for glaciation; the process is reinforced by global geochemical feedback, principally the drawdown of atmospheric CO 2 and Milankovitch "astronomical" forcing but these are unlikely, by themselves, to inititiate glaciation. The same remarks apply to late Cenozoic glaciations. Late Proterozoic glacially-influenced strata occur on all seven continents and fall into two tectonostratigraphic types. In the first category are thick sucessions of turbidites and mass flows deposited along active, compressional plate margins recording a protracted and complex phase of supercontinent assembly between 800 and 550 Ma. Local cordilleran glaciations of volcanic peaks is indicated. Many deposits are preserved within mobile belts that record the subduction of interior oceans now preserved as "welds" between different cratons. Discrimination between glacially-influenced and non-glacial, volcaniclastic mass flow successions continues to be problematic.

* A c o n t r i b u t i o n to I G C P Project 260. 0 0 1 2 - 8 2 5 2 / 9 3 / $ 2 4 . 0 0 © 1993 - Elsevier Science Publishers B.V. All rights reserved

2

N. EYLES The second tectonostratigraphic category of Late Proterozoic glacial strata includes successions of glacially-influenced, mostly marine strata deposited along rifted, extensionalplate margins. The oldest (Sturtian) glaciclastic sediments result from the break-out of Laurentia from the Late Proterozoic supercontinent starting around 750 Ma along its "palaeo-Pacific" margin with a later (Marinoan) phase of rifting at about 650 Ma. "Passive margin" uplifts and the generation of "adiabatic" ice covers on uplifted crustal blocks triggered widespread glaciation along the "palaeo-Pacific" margin of North America and in Australia. A major phase of rifting along the opposite ("palaeo-Atlantic") margin of Laurentia occurred after 650 Ma and is similarly recorded by glaciclastic strata in basins preserved around the margins of the present day North Atlantic Ocean. Glaciation of the west African platform after 650 Ma is closely related to collision of the West African and Guyanan cratons and uplift of the orogenic belt; the same process, involving uplift around the northern and western margins of the Afro-Arabian platform subsequently triggered Late Ordovician glaciation at about 440 Ma when the south polar region lay over North Africa. Early Silurian glaciation in Bolivia and Brazil was followed by a non-glacial episode and renewed Late Devonian glaciation of northern Brazil and Bolivia. The latter event may have resulted from rotation of Gondwana under the South Pole combined with active orogenesis along the western margin of the supercontinent. Hercynian uplift along the western margin of South America caused by the collision and docking of "Chilinia" at about 350 Ma (Late Tournasian-Early Visean) was the starting point of a long Late Palaeozoic glacial record that terminated at about 255 Ma (Kungurian-Kazanian) in western Australia. The arrival of large landmasses at high latitude may have been an important precondition for ice growth. Strong Namurian uplift around virtually the entire palaeo-Pacific rim of Gondwana culminated in glaciation of the interior of the supercontinent during the latest Westphalian (c. 300 Ma). There is a clear picture of plate margin compression and propagation of "far field" stresses to the plate interior allowing preservation of glacially-influenced strata in newly-rifted intracratonic basins. Many basins show a "steer's head" style of infill architecture recording successive phases of subsidence and overstepping of younger strata during basin subsidence and expansion. Exploration for oil and gas in Gondwanan glaciated basins is currently a major stimulus to understanding the relationship between tectonics and sedimentation. Warm Mesozoic palaeoclimates do not rule out the existence of restricted ice covers in the interiors of continental landmasses at high palaeolatitudes (e.g. Siberia, Antarctica) but there is as yet, no direct geological record of their existence. The most likely record of glaciers is contained in Late Jurassic and early Cretaceous strata. In any event, these ice masses are unlikely to have had any marked effect on global sea levels and alternative explanations should perhaps be sought for 4th order, so-called "glacio-eustatic" changes in sea level, inferred from Triassic, Jurassic and Cretaceous strata. The growth of extensive Northern Hemisphere ice sheets in Plio-Pleistocene time (c. 2.5 Ma) was the culmination of a long global climatic deterioration that began sometime after 60 Ma during the late Tertiary. Tectonic uplift of areas such as the Tibetan Plateau and plate tectonic reorganizations have been identified as first-order controls. Initiation of the East Antarctic ice sheet, at about 36 Ma, is the result of the progressive thermal isolation of the continent combined with uplift along the Transantarctic Mountains. In the Northern Hemisphere, the upwarping of extensive passive margin plateaux around the margins of the newly-rifted North Atlantic may have amplified global climatic changes and set the scene for the growth of continental ice sheets after 2.5 Ma. Ice sheet growth and decay was driven by complexly interrelated changes in ocean circulation, Milankovitch orbital forcing and global geochemical cycles. It is arguable whether continental glaciations of the Northern Hemisphere, and the evolution of hominids, would have occurred without the necessary precondition of tectonic uplift.

Nicholas Eyles was born in London, England in 1952. He graduated from the University of Leicester in 1974 before completing graduate work on modern Canadian, French and Ice-

landic glaciers and Pleistocene glacial deposits, at Memorial University of Newfoundland, Canada and the University of East Anglia in Great Britain. He is Professor of Geology at the University of Toronto and was recently awarded the degree of D.Sc. from the University of Leicester. His research includes detailed facies investigations of modern and ancient glacial deposits and the identification of their basinal setting. His interests include Palaeozoic glaciated basins containing oil and gas, and urban geological studies in Pleistocene glaciated terrains.

3

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T T I N G

PREFACE

This paper grew out of a graduate glacial sedimentology course at Toronto which identified a need for a critical overview of the relationship between plate tectonic setting and glaciation in Earth history. This had last been attempted in the mid-1970's by L.M.G. Schermerhorn but met with strong resistance because of his central thesis that most ancient tillites are in fact, mass flow deposits related to active tectonism; unfortunately, the inclusion of strata that were indisputably of a glacial origin weakened the essential correctness of Schermerhorn's argument. His review has nevertheless, been a stimulant to closer examination of ancient glaciated basins and more sophisticated interpretations of strata uncritically labelled as "glacial". There is now a renewed appreciation of plate tectonic and structural settings as a major control on the timing of glaciation in Earth's history, the style of glacial deposition and the character and preservation potential of different sedimentary facies. Unfortunately, this information is scattered across a diverse and international literature which has retarded cross-fertilization. One of the objectives of this paper is to accelerate a unified approach to glacial sedimentology by emphasising the common (and uncommon) geodynamic controls on Earth's glacial record from the Archean to the Pleistocene. Past practice has been to divide this record and for different groups of workers to concentrate on a particular time slice without much referencing from one episode of glaciation to another. This has been a major obstacle to the formulation of depositional facies models that can aspire to some degree of universality (e.g. Walker, 1979, 1992). "All purpose" depositional models may not exist of course (Anderton, 1985), but this needs to be tested by a catholic approach to the rock record. This paper does not claim to be a comprehensive treatment. Many aspects of Earth's glacial history are controversial and enjoy fast moving debate; many more data need to

be gathered by multidisciplinary research groups involving sedimentologists, geochronologists, biostratigraphers, palaeoecologists, climate modeUers and structural geologists. Subject to these limitations, the present paper should be viewed not as a catalogue of Earth's long glacial heritage but as a consumer's guide. 1. PAST P R O G R E S S GLACIATED BASINS

IN

THE

STUDY

OF

" . . . a strict search for glacial indications among all deposits, primary, secondary, and tertiary, would be one of the most valuable pieces of scientific work possible" (Ball, 1891, p. 149). By the mid-nineteenth century, most geologists were aware of the argument of Agassiz (1840) that existing glaciers had been much more extensive in the recent past. The growth of ice sheets to continental dimensions, whilst not universally accepted by the scientific community, fitted well with the then dominant geophysical theory of a progressivelycooling planet. The effects of past glacio-eustatic changes in sea level caused by ice sheet growth were first recognised by Maclaren in 1842; by 1865 evidence of glacio-isostatic crustal downwarping under large ice sheets had been identified by Jamieson. By this time, the first convincing reports of PrePleistocene glacial deposits had been made from the Talchirs of central India (Blandford and Blandford, 1859) and from the Inman Valley of south Australia (Selwyn, 1859). Ramsay (1855) had earlier argued, incorrectly as it transpired, for glaciation in the Permian of Great Britain. Other ancient glacial deposits were found in quick succession in Scotland (Thompson, 1871, 1877), Canada (Dawson, 1872), South Africa (Sutherland, 1870) and in Brazil (Derby, 1888). The finding of ancient glacial strata, implying glaciations on a scale equal to Pleistocene Ice Ages, dispelled ideas of a progressively cooling planet which had so upset late nineteenth century intelligentsia; "this cheerless

4

vision of the end of all things" (Coleman, 1926, p. xix). Coleman's (1907, 1908) reports of the early Proterozoic Gowganda tillites in Ontario was revolutionary since they demonstrated the existence of cold conditions very early in the history of planet Earth and stimulated intense debate on the timing and origin of past ice ages. Coleman (1926) reviewed what was known of ice ages recent and ancient and referred to many reports made in the last few decades of the nineteenth century when glaciation provided a ready explanation for poorlysorted sediments in the rock record; the classic work of Geikie (1896) on continental Pleistocene glacial deposits is often cited. Geikie's work was particularly important because it marshalled geological evidence from North America and Europe showing that "glacial drift" deposits recorded multiple advances and retreats of ice sheets. This fitted well with newly developing theories of extraterrestrial forcing of climate change. "Astronomical" theories of glaciation, those which argue that Pleistocene ice ages were the result of periodic changes in the Earth's orbit or fluctuations in solar luminosity, have a lengthy pedigree starting with the qualitative work of Adhemar (1842), Croll (1875) and Ball (1891). Milankovitch's classic quantitative treatise "On the problem of the Astronomical Theory of the Ice Ages" (1914) was first printed in Serbian and was more widely publicized by K6ppen and Wegener's "Climates of the Geological Past" issued in 1924. These works initiated feverish geological activity in the 1930's to ground truth the Milankovitch model. Initial results were favourable (Simpson, 1930) but with the advent of radiometric dating in the 1950's it became apparent that the terrestrial record of Late Pleistocene ice sheet growth and decay was more complex than that modelled by Milankovitch; the astronomical theory fell into disfavour. Detailed isotope and paleomagnetic stratigraphies from the deep oceans subsequently confirmed the essential "pacemaker" role of astronomical variables on the global climate changes that dictated late

N. EYLES

Pleistocene glacial and interglacial cycles (Shackleton and Opdyke, 1973; Imbrie and Imbrie, 1979). At the present time, the identification of "Milankovitch" controls on both pre-Pleistocene glacial and non-glacial deposition is a rapidly growing industry (Fischer, 1986; Ginsburg and Beaudoin, 1990; Franseen et al., 1991) though the application of chaos theory to behavior of the solar system (Laskar, 1989) threatens to make many workers redundant. Explaining the timing and climatic origin of pre-Pleistocene ice ages in Earth history is one of the longest standing problems in geology. Recent reviews of a very large literature have been made by Harland and Herod (1975), Frakes (1979), Chumakov (1992) and Frakes et al. (1993). The first thorough inventory of the pre-Pleistocene rock record was assembled by Hambrey and Harland (1981). Since then, several authors such as Anderson (1983) and C.H. Eyles et al. (1985) have emphasised the importance of glaciomarine deposits in the rock record. It is evident that the primary data base of past glaciations is the rock record and models of past ice ages are only as good as our ability to describe and interpret that record in the field. Particularly influential depositional models, based on descriptions of modern processes and environments, have been developed by Carey and Ahmad (1960), Boulton (1972, 1975), Powell (1981, 1990) and increasingly, Alley et al. (1987) and Hughes (1987). Radically new concepts of glacial behaviour and deposition proposed by Boulton (1987), Boulton and Hindmarsh (1987) and Alley et al. (1987) have been major stimulants to field investigations by other workers (MacAyeal, 1992). J. Crowell and L. Frakes (e.g. Crowell and Frakes, 1970, 1971a, b; Crowell, 1978, 1982, 1983a,b; Frakes, 1979) introduced a global, intercontinental view of ancient glaciated basins and recognised the predominance of glacially-influenced marine deposits. Reading and Walker (1966) and Spencer (1971) set new standards in the description and interpretation of ancient glacial strata; the

EARTH'S GLACIAL RECORD AND ITS TECTONIC S E T l l N G

work of Beuf et al. (1971), Eisbacher (1981), Rochas-Campos and dos Santos (1981), Young and Nesbitt (1985), Nystuen (1976a), Edwards (1984), Deynoux (1985a, b), Visser (1989), Vaslet (1990) and Franca and Potter (1991) provides new insights into glacial deposition by integrating data across entire sedimentary basins. Arguably, the most important and provocative paper on ancient glacial deposits to have been published in the last twenty years is that of Schermerhorn (1974) who argued that most "glacial" strata preserved in the rock record are mass flow deposits. The increasing economic significance of glaciated basins and the availability of detailed subsurface drill core and various geophysical data (e.g. Levell et al., 1988; Franca and Potter, 1991; C.H. Eyles et al., 1993) will accelerate understanding of glaciated basins by allowing the application of sequence stratigraphic concepts (e.g. Proust et al., 1990; Deynoux et al., 1991). Important constraints on ice age models will be established by theoretical palaeoclimate modelling that can isolate the role of global geochemical controls and feedback mechanisms (e.g. Fischer, 1984; Veevers, 1990; Crowley et al., 1991; Socci, 1992). Such modelling of past glaciations is dependent on detailed palaeoglaciological studies of Pleistocene ice sheets and their relationship with global climate change (Shackleton and Opdyke, 1973; Imbrie and Imbrie, 1979; Andrews, 1987; Broecker and Denton, 1989; Ruddiman and Kutzbach, 1989; Bond et al., 1992; Raymo and Ruddiman, 1992). The next phase in the study of Earth's glacial record will increasingly focus on the relationship of plate tectonic processes and basin forming processes to the initiation, preservation and character of the glacial record in sedimentary basins. Discrimination of regional tectonic and global controls is an important aim of such work. 2. TERMINOLOGY

"We must be very cautious about tillites; many are pseudo-glacial conglomerates and

5

are not dependable evidence of climate" (Wegener, 1929, pp. 132-137). Examination of Earth's glacial record must be based on a thorough understanding of glacial and non-glacial sedimentary processes, the associated depositional systems and their sedimentary products. This is because information regarding tectonics setting, climate and water depth changes is complexly recorded in the stratigraphic record. A rigorous sedimentological approach is essential because major uncertainty still exists regarding the precise origin and real climatic significance of most ancient strata alleged to be glacial. An important requirement of such work is the need for careful definition and use of terms relating to poorly-sorted sediments. Much of the older literature describing Earth's glacial record is based, either explicitly or implicitly, on the assumption that any coarse-textured deposit is of glacial origin. Commonly, these sediments were first described when the presence of poorly-sorted, conglomeratic horizons was sufficient evidence for a glacial origin. Many of these sediments have subsequently been shown to be non-glacial deposits formed by a wide range of depositional processes in terrestrial, marine and volcanic environments (e.g. Crandell and Waldron, 1956; Crowell, 1957; Newell, 1957; Van Houten, 1957; Schwarzbach, 1958; Dott, 1961; Winterer, 1963; Mather and Wengerd, 1965; Schermerhorn and Stanton, 1966; Borns and Hall, 1969; Schermerhorn, 1974; Plafker et al., 1977; Martin et al., 1985; Charrier, 1986; N. Eyles et al., 1988a; Laznicka, 1988; Stanistreet et al., 1988a; C.H. Eyles and N. Eyles, 1989; Collinson et al., 1989; Martin et al., 1989; Bailey et al., 1990; N. Eyles, 1990; Long, 1991; Hefferan et al., 1992). New work also suggests that coarse-grained, poorly-sorted, ejecta deposits of impact craters are surprisingly common in the rock record and have been widely mistaken for glacial deposits (Aggarval and Oberbeck, 1992; Marshall and

6 Oberbeck, 1992; Oberbeck et al., 1993). These ideas may be considered as extreme but they can only promote the closer scrutiny of diamictites in the rock record. Even where a glacial setting can be demonstrated from other contextual evidence, the role of glaciers has in most instances, been simply to supply large volumes of glaciclastic sediment to surrounding environments where it is then redistributed by non-glacial processes sometimes long after the glaciers vanished; rare striated and glacially-shaped clasts may be the only evidence of a glaciclastic contribution to the sedimentary basin. This is easily demonstrated even for relatively recent Pleistocene glaciations. An impressive array of glacial landforms and sediments remain on the continental surfaces in mid- to high-latitudes but the long term fate of such sediment is to be reworked to the continental margin and offshore basins which are the ultimate repository of any glacial record (Section 5). Geologists have described poorly-sorted sediments, consisting of glacially-mixed clasts and matrix using a wide variety of terms. Murchison in 1839 introduced the term "drift" for the stiff, bouldery clays and other poorly-sorted sediments associated with glaciers. Murchison's ideas on the origin of these facies, as deposits of drifting ice masses in a glacial sea, were soon proved to be only locally correct but the term stuck as an umbrella term for Pleistocene glacial deposits irregardless of origin. The term "till" is of ancient Scottish derivation (used well before its origin as a glacial deposit was understood) and is literally translated as "stony ground". It was described and defined scientifically by James Geikie in 1863 as "a tough, strong clay resuiting from the grinding action over the country of an ice sheet". Its origin was exhaustively reviewed in the subsequent editions of his textbook "The Great Ice Age" (e.g. Geikie, 1896). The term "tillite" identifies its lithified equivalent and was first used by Penk in 1906 (Harland et al., 1966). Since

N. EYLES

little was then understood of sediment gravity flows and other non-glacial processes that also produce poorly-sorted sediments, the labels till and tillite have since been applied indiscriminantly. Wegener (1929) was one of the first to criticize this approach in his attempts to use true tillites as precise chronostratigraphic markers from one continent to another. Wegener used the term "pseudoglacial conglomerate" for till "look-a-likes", the depositional origin of which had nothing to do with glaciers. The term "tilloid" was subsequently introduced for "till-like deposits of doubtful origin" (Blackwelder, 1931); the work of Dott (1961) and Crowell (1957, 1964) on matrix-rich conglomeratic mass flows deposited on submarine slopes, clearly focussed attention on the problem of facies equifinality. Pettijohn (1957) redefined the term tilloid to refer to a "nonglacial conglomerate mudstone resulting from downslope slumping or mudflow"; the term, while well-meant, was flawed by the continued explicit reference to glacial action (tilloid). Schwarzbach (1975) suggested that the term be dropped from future usage. Schermerhorn (1966) introduced the term mixtite for a coarse, poorly-sorted clastic rock characterised by a "sparse to subordinate coarse fraction" (e.g. pebbly mudstones) without regard for composition or origin. Whilst the term was widely used it did not lend itself to the description of unlithified sediments nor a wide range of poorly-sorted facies dominated by the coarse fraction. Flint et al. (1960) introduced the terms diamicton and diamictite for respectively, unlithified and lithified poorly-sorted, non-calcareous terrigenous sedimentary rocks. The root of the words is the Greek diamignymi meaning to mingle thoroughly. Harland et al. (1966) used the term "diamict" as an umbrella term for both unconsolidated and consolidated facies but this practice did not enjoy wide circulation; N. Eyles et al. (1983) used this term for unconsolidated facies and retained "diamictite" for consolidated facies and introduced a now widely-employed scheme for

EARTH'SGLACIALRECORDAND ITS TECTONICSETTING

7

on rigorous field criteria emphasizing facies approaches. These schemes also suffered from a strongly reductionist approach that emphasized textural, mineralogical or other laboratory criteria reflecting the use of "tills" as mappable stratigraphic units rather than as depositional facies for reconstructing palaeoenvironments (e.g. Legget, 1976 and refs. therein). In the last ten years great strides in the field of sedimentology and dynamic stratigraphy has refocussed emphasis on detailed reconstructions of glacial envi-

facies descriptions of core and outcrop exposures. The original terms of Flint et al. (1960) are still also used by some workers. This paper adhers to the scheme of N. Eyles et al. (1983) where the term till(ite; lithified) is reserved specifically for diamict(ite)s deposited directly by glaciers. Ideas of till genesis have undergone a lengthy evolution but genetic classifications of various till types are still largely based on a priori assumptions of processes together with what such sediments ought to look like rather than ®

~ ) GLACIATED VALLEY

SUBGLACIAL

(~) GLACIOFLUVIAL

Br~ded fluvial, gianlolacustrine

(~) GLACIOLACUSTRINE

Debris flow

~tve sands ;esociated

Sandy channel system

Cross-cutting till sheets

.*, & ~.

l

ic facies

Debris flow

X-stratifieb sands

dites

i~i~::ii~"

Braided fluvial

Boulder pavement

~ & _ . _ & I~.--"3 Lodgemant Till

~.. ,~.~

Resedimented diamict

c

dites with itones is flow

~

~

s

Superimposed gravel bars

g

.

Debris flow

C

out' and s flow s

s

Lodgamant till

"

ictites

c

s

Massive gravels

bE~ rdock/steet°~Sm%n t

o

S

S

g

/ " l [~



%

%

( ~ FIORD

(~) SHELF Turbiditas

Bioturbated

muds Debris flow

(~) SUBMARINE CHANNEL Turbidites

~

~

(~)SLOPE

Bioturbatad diamicts

r'urbidi~es .arge slump blocks olis~ostromes)

Channelized massive, graded sands

Chaotic,

3ebris flow

dumped facies

rurbidites

Scoured bedrock surface s

A' I

Turbidites =a a Boulder pavement

Channelizad massive-graded gravels

g

Thick 'rain. out' and resedimanted diamicts Coquinas

Debris flow

Thid~, rain-out, diamictJtes and debris flows with sediment rafts

Bioturbated muds

c

s

s

g

c

s

s

g

Fig. 3.1. Principal glacio-terrestrial (top) and glacially-influenced marine (bottom) depositional systems with representative facies associations.

8

ronments and facies modelling. A whole new suite of tools is now available to the sedimentologist for describing and interpreting poorly-sorted clastic facies (N. Eyles et al., 1983; Koster and Steel, 1984; Levell et al., 1988; Nemec and Steel, 1988; Franca and Potter, 1991; Smith and Lowe, 1991). The term "till" is now restricted to diamict facies deposited by the direct agency of glacier ice which in practice, identifies diamict(ite)s deposited subglacially in an ice-contact setting. It is worth re-emphasising that the principal role of glaciers in Earth history has been to supply large volumes of glaciclastic sediment to sedimentary basins where it is then dispersed by "normal" non-glacial processes. As Schermerhorn (1983) stressed, glaciclastic (or "glacigenic"; the two terms are essentially synonomous) sediment may constitute the bulk of non-glacial rocks or formations but this does not make them glacial deposits. The term "glacially-influenced" was used by C.H. Eyles et al. (1985) for deposits or basins where glaciers contributed to the sediment flux or created changes in water depths either as a result of glacio-eustatic sea-level fluctuations or by glacio-isostatic loading of the crust. 3. GLACIAL DEPOSITIONAL SYSTEMS The sedimentology of modern and ancient glacial depositional systems has been the subject of several recent symposia and reviews (Syvitski et al., 1987; Van der Meer, 1987; Croot, 1988; Powell and Elverhoi, 1989; Dowdeswell and Scourse, 1990; Anderson et al., 1991; Brodzikowski and Van Loon, 1991). N. Eyles and C.H. Eyles (1992) show that such systems can be divided into glacio-terrestrial and glacially-influenced marine types and briefly reviewed depositional settings and facies (Fig. 3.1). As related above, major problems remain with the discrimination and interpretation of diamictite and till facies and the purpose of this present section is to review the depositional settings and stratigraphic "context" in which these problematic sediments occur.

N. EYLES

3.1 Glacio-terrestrial depositional systems Glacio-terrestrial depositional systems can be sub-divided into those that occur below glaciers (subglacial), those which occur at the margins of glaciers transporting large amounts of surface debris (supraglacial), those along confined glaciated valleys (glaciated ualley), in large ice-dammed lake basins (glaciolacustrine) and those associated with extensive, fluvially-dominated braid plains (glaciofluvial). The deposits of these systems have been reviewed most recently by Brodzikowski and Van Loon (1991) and the reader is referred to this work for detailed facies descriptions. The long term preservation potential of terrestrial glacial facies is low especially in areas of active tectonism and high relief where primary glacial deposits are resedimented downslope or are reworked by braided rivers such that a glacial setting may be very difficult or impossible to identify. The following section emphasises the processes and deposits of the subglacial and glaciolacustrine depositional systems which, because of their large areal extent, may be selectively preserved. Subglacially-deposited sediments of Pleistocene age cover more than 15 x 106 km 2 of North America; glaciolacustrine deposits alone extend over more than 2 x 10 6 km 2 of which about half was deposited in a single, albeit complexly changing, ice-dammed lake body (Glacial Lake Agassiz; Teller and Clayton, 1983).

3.1.1 Subglacial depositional system Subglacial environments are recorded in the rock record either by scoured, deeply dissected or glacitectonised bedrock surfaces or by deposits of tillite and associated strata. There are three fundamental processes that lead to the accumulation of till deposits below glaciers (a) melt-out, (b) lodgement, (c) deformation. In reality, there is a continuum between all three but it is convenient to compartmentalize these processes. Whilst past emphasis has been on melt-out and

EARTH'S

GLACIAL

RECORD

AND

ITS

TECTONIC

9

SETTING

lodgement, recent work points to the greater importance of subglacial deformation as a till-forming process.

3.1.1a Lodgement till The term lodgement till refers to a stiff, heavily overconsolidated diamict formed by the subglacial aggregation of englacial debris released from the base of a moving glacier (Figs. 3.2, 3.3). Debris is released from the ice base by pressure meltout as englacial particles are swept over the underlying bed. Rates of deposition, confining pressures, requisite ice velocities and field characteristics have been documented by Boulton (1975), N. Eyles and Sladen (1981), Dowdeswell et al. (1985), N. Eyles et al. (1982a, 1987b). A necessary requirement for lodgement is a "stiff" subglacial bed on which to accrete englacial debris; the process can be likened to smearing "crunchy" peanut butter across toasted bread. Typical deposiLODGEMENT TILL:

'STIFF BED'

DEFORMATION TILL:

TOTAL SURFACEMOVEMENT P T1

Basal sliding

tion rates are less than 10 cm a year (Boulton, 1975). A most important feature is that a lodgement till deposit is not uniformly massive, but displays a crude stratigraphy composed of shingled, cross-cutting till units. Lodgement till deposition requires fairly specific glaciodynamic conditions and the intraformational disconformity surfaces that separate different till units record intermittent non-deposition and erosion of the subglacial bed. Stratigraphic "complexes" of lodgement till, containing the channelled deposits of subglacial streams (Fig. 3.2), rest regionally on a marked unconformity and are typically of restricted thickness ( < 20 m) because thicker deposits cannot be accomodated under the ice sheet without being reworked and eroded. Lodgement tills characteristically contain glacially-shaped clasts. Clasts characteristically have a fiat-iron or "bullet"-like ap'DEFORMING-BED'

TOTAL SURFACEMOVEMENT ""'"~...~o ~,T1

-I

Internal

Sediment Internal deformation deformation

deformation

ICE Basal slip and lodgement of englacialdebris "-~.~ / subglacialstream ~--~j~_. _ deposits

":?~~,-~. ~ ~ ~ ~

~.~' ~::!i~

.~ N

-,x.. :.:..,>

x~. :

~

.~i~!~i~: ,

x .~ " .,..~. <: ::.::::.:~-Z~:~:...

assivediamict units in erosional contact with underlyingunits. Showsevidenceof successive accretion at ice base.

......

~

Massive diamict ('fully-assimilated facies') emplacoden masse

~

assivediamict with raftsof underlying sediment ('partially-assimilated facies')

Fig. 3.2. Subglacial deposition of lodgement and deformation tills. Lodgement requires a stiff, resistant bed; deformation till results from intense shearing of deformable substrate sediments or weak rocks. Deposition rates of lodgement till are less than 10 cm/yr; in contrast very large volumes of sediment may be transported by deformation and the resulting "till streams" may be highly erosive leading to the cutting of drumlins across unconsolidated sediments (Fig. 3.3H) and overdeepened bedrock valleys.

10

N. EYLES

1l

EARTH'S GLACIAL RECORD AND ITS TECTONIC SETYING

0

-•~__., " .~. ~,

"~

~ r.l~

o

o

~, ~

C:~ ~

~

~1~

IDD

~ o ~

~ ~ _~ "~ ~., ,-- ~ ~ ~.~ ~ o'= ~

~

'~

~

ur~ ,.,~.~ ,,-,

0

~

~

0

~. '.,~ ~

~

,..~

~00"~

0

0

e" ~

e-, ~ - . . , - -

i ~..~,

~.

~

~ "~' ~

"

.~.

t~

•~

~

~

~

~

~)

~

~,-'~-

~

.



~"

--

=

!

0



m ~

~

~

~

~

~

~_~

.

12

pearance (Fig. 3.4) and are the unique trademark of clasts that have experienced transport and lodgement at an ice base (Boulton, 1978). Primary glacial deposits are commonly reworked by mass-flow processes or by fluvial or marine erosion resulting in the dispersal of glaciclastic sediment. In these instances, the occurrence of glacially-shaped, fiat-iron clasts may be the only evidence of such a glaciclastic contribution to the sedimentary environment. The emphasis here is on the presence of glacially-shaped clasts because striated clasts are produced in nonglacial settings (Section 10.4.2).

3.1.1b Melt-out till Melt-out till forms by the aggregation of englacial debris when dirt-rich glacier ice stagnates in situ. Melt-out deposits may form either on the surface or below downwasting ice masses (supraglacial and basal melt-out till). This simple model has a long history of use in the interpretation of both Pleistocene and pre-Pleistocene diamictite successions that are thick and laterally extensive. Paul and Eyles (1990) reviewed the melt-out process at modern glaciers and the large supporting literature based on investigation of ancient deposits

N. E Y L E S

alleged to be melt-out deposits. They concluded that subglacial melt-out cannot produce thick ('0 re's), widespread ('000 km2), diamict(ite) deposits given insufficient debris within a column of downwasting ice and a low preservation potential. Instead, melt-out till is more likely to be preserved as restricted lenses within complexly-structured morainal landforms recording the downwasting of stagnant ice margins (e.g. Shaw, 1979; Kaszycki, 1987; Fig. 3.3). In these settings, the ice margin is under severe compression and locally enhanced volumes of englacial debris are created by compressional folding and repeated refolding of basal debris layers (e.g. Boulton, 1972). The long term preservation potential of melt-out till even if deposited, is very low in view of the fact that most deposits will be reworked by fluvial processes as the ice downwastes. This statement has to be tempered by the observation that very large areas of Arctic Canada and the former Soviet Union are still underlain by thick ice bodies that are currently melting in the interglacial climate (e.g. Grosvald et al., 1986; Astakhov and Isayeva, 1988; French and Harry, 1990 and refs. therein). The origin of the ice is not

2

Fig. 3.4. "Bullet" boulder on upper surface of lodgement till; iceflowfrom left to right.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'ITING

well known (permafrost or even less is known of the posits and diamict facies slow wastage of such ice in

3.1.1c

glacier ice?) and sedimentary dethat result from a cold climate.

Deformation till Boulton (1987) showed that tills could be generated, transported and deposited en masse by shear stresses applied by glacier movement over unconsolidated sediments; the "deformingbed" model (Boulton and Hindmarsh, 1987; Fig. 3.2). Geophysical work at the margins of fast flowing glaciers suggests that a large component of ice flow is achieved by deformation of substrate sediments (Alley et al., 1987). Deformation till can be regarded essentially as a sedimentary melange resulting from the remoulding and mechanical mixing of sediment or bedrock particles below the ice base (Figs. 3.5, 3.6). Work is ongoing to establish field criteria for recognition of deformation till and to test the wider applicability of the model but on the balance of the available evidence it is probably true to say that subglacial deformation has been the most important till-forming mechanism in Pleistocene glaciated terrrains. This is somewhat ironic because the concept accords with some of the original ideas on the formation of till put forward by Geikie as early as 1893. Geikie's description of a " . . . tightly packed but flowing gritty mass ... being dragged forward by the ice . . . . in which stones and boulders are striated" is a good summary of the deformation process. The thickness of the deforming till layer reported from the Antarctic by Alley et al. (1987) is 13 m; some Pleistocene deformation tills are as much as 50 m thick but it is not clear whether this represents the true thickness of the deforming layer or is the result of multiple episodes of deformation and deposition (Boyce and Eyles, 1991). Even this problem was anticipated by Geikie when he wrote " . . . h o w thick a stratum of boulder clay was kept in motion by the ice we cannot tell ... but a stage would be reached when the lower part

13

of the deposit impeded by friction, would come to rest." (Geikie, 1893, p. 73). As noted by Geikie, deposition of deformation till occurs when the deforming mass dewaters and "freezes". The deforming bed model of subglacial deposition is widely applicable to the interpretation of Pleistocene paleoenvironments where, in many cases, lobate margins of the continental ice sheets, surged rapidly over marine or lacustrine sediments. Large-scale, en masse transport of deforming sediment ("till streams") below the ice base may be responsible for cutting glacially-overdeepened valleys (Boulton and Hindmarsh, 1987; N. Eyles et al., 1991) and large streamlined bedforms such as drumlins and delivering large volumes of sediment to ice margins (Elverhoi et al., 1983; Solheim and Pfirman, 1985; Alley et al., 1987; Hughes, 1987; N. Eyles and McCabe, 1989a, b; N. Eyles et al., 1991). In turn, the presence of relatively soft substrates may result in lubrication of ice flow allowing chaotic collapse of ice sheets (MacAyeal, 1992). Recognition of deformation till depends on the presence of glacitectonized substrates, together with rafts of substrate sediment not yet fully "kneaded" into deformation till. Partially and fully-assimilated deformation till facies can be recognized (Figs. 3.5, 3.6) recording the evolution from pre-existing, but mechanically-deformed sediment, to a massive till. It has to be emphasised that there is both a process and facies continuum between deformation and lodgement. In addition, detailed field descriptions of deformation tills are as yet lacking so the two types can easily be confused in the ancient record. Deformation tills will show evidence of relatively rapid en masse emplacement involving mechanical mixing of underlying sediments; lodgement tills will show evidence of much slower bedby-bed deposition (Fig. 3.2). The deposits of subglacial melt rivers can be an important stratigraphic component of lodgement till successions; these occur as "shoe-string" bodies of sand and gravel aligned parallel to

14 N. EYLES

O

15

E A R T H ' S G L A C IA L R E C O R D A N D ITS T E C T O N I C SETF1NG

o

~ ~

.--

0

~ ~ oo

~

16

N. EYLES

17

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

ee

~-~O

'~.~ 0

=

~

~.~ ~ =~:_~ , ~ ~~

'~

~.~.~ •

,-

18

N. EYLES

ice flow direction (see N. Eyles and Sladen, 1981; N. Eyles et al., 1982a). These deposits are unlikely to be present below glaciers flowing over deforming substrates; Boyce and Eyles (1991) noted an absence of subglacial stream deposits (e.g. eskers) in areas of deformation till which they attributed to drainage of the ice base by the mass flux of deforming till. Provisional data also suggest that clast fabrics in deformation tills are akin to that shown by debris flows (J. Boyce, pers. commun., 1991; Boyce, 1992) such as reported by Lindsay (1968), Mills (1984), N. Eyles and Kocsis (1988), N. Eyles et al. (1988a); lodgement till fabrics are described by Dowdeswell et al. (1985) and Rappol (1985). The value of using clast fabric data for distinguishing different diamict facies is potentially great but not yet realised; the procedure is not easily amenable to investigation of lithified strata and the published data base is not yet extensive nor matched by a precise understanding of the origin of the deposits for which data are reported.

Subglacially-deposited till(ite)s will rest on a regional unconformity that shows evidence of local or extensive glacitectonic disruption, have a strongly lensate, "thickening and thinning" regional geometry, be of restricted thickness and be associated with coarsegrained, ice-contact glaciofluvial facies or the deposits of glacial lakes. Because of the ability of deforming till streams to erode underlying substrates, and thereby create their own accomodation space, deformation tills may be locally thicker ( < 50 m?) compared to lodgement tills ( < 15 m). The upper surface of subglacially-deposited diamict facies may show an undulating relief created by drumlins. While much work remains to be completed on the characteristics of subglacial tills, it can also be stated fairly safely that diamict(ite) deposits with a thickness much greater than 50 m are not likely to have been deposited in a subglacial setting. This is because greater thicknesses are very unlikely to be accomodated below the ice base. This

PLUME OF SUSPENDEDSEDIMENT

EFFLUX JET

ICE BERGAND SCOURS o

PUSHRIDGE

-UNDERFLOWS ----..._~.~

DEFORMATIONTILL OR DEFORMEDSUBSTRATE (Fig. 3.2)

LAMINATED, MASSIVE MUDS WITH DROPSTONES PITTED FAN SURFACE FROM MELTOF TRAPPED ICE

v

t

SEDIMENT GRAVITY FLOW

Fig. 3.7. Ice contact depositional system at the margin of a glacier terminating in water. After N. Eyles and McCabe (1989a) and Powell (1990).

19

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'Iq~ING

provides a useful "rule of thumb" for interpretation of pre-Pleistocene diamictite deposits that approach several hundreds and even thousands of meters in thickness (Section 3.2).

3.1.2 Glacial lake basins Glacial lakes come in many different shapes and sizes and consequently show a wide range of depositional settings. Basins range from dynamic, high energy ice-contact

'CLASSICAL VARVES' 1 WINTER

2 v

3

~

DRAPED LAMINATION

Decreasing current velocity Increasing suspended load

CLIMBING RIPPLES WINTER

CLAY - -

--

DRAPED LAMINATION AND CLIMBING RIPPLES

91 Type A

Increasing bed load Type B

Draped

Lamination

J

Fan Delta

Seasonal Ice Cover

TURBIDITES RELEASED BY FAILURE OF DELTA FRONT

Fig. 3.8. Depositional model for formation of seasonally-deposited varves in non-ice contact glacial lake basins. Based on N. Eyles and Miall (1984), Ashley et al. (1985). Top shows sequence of ripple types produced by fluctuating underflow currents.

20

N. EYLES

21

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E ' I T I N G

~

°

<~ . ~ ¢'~ N ,--

•~

~ .~-.~,

>0='~

.

~ .~ ~

~

~-~

22

N. EYLES

lakes characterised by large and rapid influxes of poorly-sorted sediment (e.g. Fig. 3.7) to non ice-contact lakes where the glacier plays a more restricted role, sedimentation rates are much reduced and depositional processes are driven by seasonal meltwater discharges (Fig. 3.8). Depositional processes and predominant facies types have been summarised by Ashley et al. (1985), Brodzikowski and Van Loon (1991), N. Eyles and C.H. Eyles (1992c). This section emphasises the origin and characteristics of fine-grained

laminated glaciolacustrine deposits (varves) and the difficulty of distinguishing these from other laminated facies deposited in marine settings. 3.1.2a Varves The term "varve" was originally defined by De Geer (1912) essentially as a descriptive term for a "periodical iteration of layers". This word literally translated in Swedish, identifies a turn, revolution or a lap and is related to the English word "whorl" (see Antevs, 1925). De Geer was

LAMINATION TYPES 1

graded fall-out layer from tidally-influenced plume of suspended sediment

z

graded sediment gravity flow

massive sediment gravity flow winter fall-out massive, weakly-graded summer underflow layer

4a

winter fall-out repeated summer sediment gravity flows and multiple-graded layers

4b

TIME FRAME

CYCLIC (TIDAL OR

NON-CYCLIC, OR

SEASONAL)

CYCLIC

CYCLIC (SEASONAL "VARVES")

4a __

LAMINATION TYPES

4b

bioturbation c~ top of summer layer only

[= 0

10 DEPOSITIONAL

PROCESSES

FALL-OUT FROM T I D A L L Y INFLUENCEDPLUMES OF SUSPENDED SEDIMENT (e.g. Phillips et al, 1991)

MARINE

DEPOSITION FROM FINE-GRAINED,LOW CONCENTRATION TURBIDITY CURRENTS (Fig. 4.2)

MARINE/ LACUSTRINE

ALTERNATING SUMMER SEDIMENTATION FROM UNDERFLOWS (STEADYFLOW TURBIDITY CURRENTS; SECT. 4.3) AND WINTER SUSPENSION FALL-OUT (e.g. Ashley, 1975; Fig. 3.8)

MARINE/ LACUSTRINE

Fig. 3.10. Laminated mud facies produced in various glacially-influenced environments. Discrimination between "varved" and non-varved facies in ancient successions is not straightforward; terms such as laminites or rhythmites should be employed where origin is unclear. Note differing extent and position of bioturbation.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'ITING

struck by their resemblance to the annual growth rings of trees and argued for an annual rhythm of deposition driven by the summer melt of glaciers. Modern practice has been to ignore the descriptive intent of De Geer and to use the term in a genetic sense for an annually-produced couplet of contrasting grain sizes resulting from vastly different conditions of sedimentation in lakes during winter and summer (e.g. Bates and Jackson, 1987). The recognition of seasonal banding in sediments was a major paradigm in geology. Students of De Geer correlated ("teleconnected") seasonally-banded Pleistocene deposits from wide-ranging geographic localities thereby creating a powerful new tool for chronological investigations (Antevs, 1953; Coleman, 1941). The model also carried major implications for the recognition of glacial deposits in the ancient record; pre-Pleistocene "tillites" were widely identified by their stratigraphic association with supposed "varved slates" (e.g. Sayles, 1919) though the anomalous thickness of such strata compared with Pleistocene examples was noted (Antevs, 1925). Not all teleconnections were valid (e.g.. De Geer, 1926; Coleman, 1929) although De Geer's work on postglacial ( < 12 Ka) varve chronologies in Scandinavia has withstood the test of radiometric dating (e.g. Stromberg, 1983; Boyle, 1993). Notions of a "tillite/varvite" facies association, however, have been supplanted by recognition that many such strata are the product of debris flow and turbidity currents in deep marine settings (Sections 4.0, 10.4.2). The tendency to label thinly-bedded and laminated turbidite sequences as "varved" still continues (e.g. Levell et al., 1988; de Castro, 1989). Unfortunately, the term varve (lithified; varvite) is widely used as a routine term for any laminated or thinly-bedded deposit in glaciated basins (see many entries in Hambrey and Harland, 1981). This practice is particularly dangerous as it implies that the successions are continental, glaciolacustrine in origin and thus may convey an erroneous

23

palaeogeographic message. Unfortunately, while the depositional processes responsible for varves are well known (e.g. Ashley, 1975) it is not easy in ancient successions to discriminate varves from other fine laminated facies deposited from turbidites or contour currents. A varve couplet, as classically recognised by geologists, comprises a relatively coarsegrained summer layer (typically fine grained sand a n d / o r silt) and a winter clay layer (Figs. 3.9, 3.10). It is worth noting, however, that no specific grain-size is implied by the term "varve". Proximal equivalents may include varved gravels but these are difficult to recognize because of bed amalgamation and the difficulty of recognising any winter mud lamina that might be preserved. Use of the term "varve" presupposes the central importance of a strongly seasonal sedimentation regime comprising supraglacial melting of the ice margin in summer, and the release of large volumes of meltwater to lakes, alternating with winter freezing and the development of a closed, quiet water lake basin (Fig. 3.8). In summer, sediment-laden, higher-density meltwaters move down the front of fan delta lobes as density underflows (i.e. steady flow turbidity currents; Section 4.3). Less dense suspension of water and mud move across the lake as interflow or overfow plumes within the water column (Fig. 3.8). In glacial lakes that are dominated by overflows/interflows, thin ( ~ mm's), delicate laminae are deposited by vertical settling of silt and clays from suspended sediment plumes (Ashley et al., 1985). Silt laminae show grading but no traction current structures, with generally sharp contacts between successive laminae. Smith (1978) identified the importance of down-lake thinning of individual laminae and across-lake variation in thickness caused by Coriolis forces acting on plume dispersal. Thin, winter clay laminae recording many months of slow suspension settling under an ice cover may show internal grading. These couplet types are directly comparable to the diurnally-produced tidal

24

laminations deposited by fall-out from suspended sediment plumes in ice-contact marine settings (Fig. 3.10) and may be indistinguishable in outcrop or core. Several supposedly "varved" laminites have been re-interpreted as tidal in origin (Cecioni, 1981; Williams, 1988, 1989) and tidally-influenced muds are an important stratigraphic component of Pleistocene and pre-Pleistocene icecontact and distal glacially-influenced marine deposits (Section 4.5.1). In glacial lakes dominated by density underflows, a distinct succession of sandy lithofacies is deposited each summer on the midand lower-portions of delta foreslopes. Sands are ripple cross-laminated and commonly show an upward progression of ripple types from type A, through B to C recording the onset and subsequent decay of density underflows during the course of the summer melt season. Toward the end of the summer, the rate of ripple migration is slowed and the rate of vertical aggradation increased ultimately producing draped lamination (Fig. 3.8). One or several such sequences can be produced in a melt-season recording uniform or unsteady current flows. Further downslope, and out onto the basin plain, sand bed thickness is reduced and rippled facies are replaced by massive or graded sand or silt facies. These may contain multiple grading combined with parallel or cross-lamination, such as starved ripples, recording deposition by a pulsating or intermittent density underflow or from a series of solitary turbidity currents produced by delta front slumping or floods (Duringer et al., 1991). Summer deposits are capped by a winter clay layer deposited from suspension (Banerjee, 1973; Ashley, 1975). The dark clay unit (winter layer) of a varve is organic-rich and shows normal grading that records the fallout of successively finer-grained suspended sediment under the ice cover of a closed lake. Clay layer thicknesses are generally uniform across the basin but such layers are known to be interrupted by thin beds or laminae of massive or graded sands and silts

N. EYLES

which record turbidity currents produced by delta foreslope slumping during the winter (Shaw, 1977) or major meltwater floods. Burrows and trace fossils are commonly present in summer layers but not in winter clays (Fig. 3.10) reflecting the seasonal life cycle and activity of a wide variety of aquatic insects and worms (Gibbard and Stuart, 1974). Many deposits described in the literature as "varved" comprise successions of single or multiple graded units of silt and clay belonging to divisions C, D and E of the classic "Bouma" turbidite sequence (Figs. 3.9, 3.10). They may have been deposited by either surge-type or steady flow turbidity currents (see Section 4.3) with no seasonal control (Gustavson, 1975; Lambert and Hsu, 1979a, b; Duringer et al., 1991). Turbidites can be released year round as a result of continuous slumping along the side slopes of the lake basin. Discrimination of turbidites from classical varves is not straightforward; some workers emphasise the occurrence of distinct couplets in varves, the constancy of clay layer thicknesses in such deposits and the possibly greater variation in clay layer thickness associated with turbidites (e.g. Ashley, 1975; Hart, 1992). However, extensive work on laminated sediments deposited in deep marine basins, from dilute turbidites and contour currents, and on coastal estuaries by fluvial and tidal processes, shows a very wide range of lamination types that includes "varve-like" lamination types (see descriptions and discussions in Lash, 1987; Schieber, 1990; Piper and Stow, 1991; Porebski et al., 1991; Ghibaudo, 1992). Clearly, varves are not characterised by any unique structures or lamination types but are only distinguished by the temporal framework of deposition. Other contextual data, independent of lamination type, are therefore needed to positively identify that laminated facies are varves or varvites. Identification of a seasonal sedimentation regime ultimately relies on detailed bed counts, palaeontological investigations, radiometric, palaeomagnetic and palaeontological dating control and time-

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETFING

series analysis (see Jackson, 1965; RochaCampos and Sundaram, 1981; Pickrill and Irwin, 1983). Stihler et al. (1992) show how the application of radionuclide and tephrochronological dating methods has the potential to identify annually deposited laminae; these workers re-emphasise the need for caution in applying the term varve. Outcrop analysis of thick glaciolacustrine successions (N. Eyles, 1987; N. Eyles et al., 1987a; N. Eyles and Clague, 1991) and highresolution seismic investigations of the bottom stratigraphies of glaciated lake basins (N. Eyles et al., 1991) shows that "varved" deposits can be a minor stratigraphic component of such systems. A wide range of mass flow and rain-out diamict facies (Section 3.2) can be interbedded with coarse- and finegrained deposits of fan-deltas. A seasonal sedimentation regime can be suppressed or masked by many factors, the most important of which is probably sediment supply. In basins receiving large volumes of sediment, year-round slumping of oversteepened slopes and the release of sediment gravity flows will mask any seasonal delivery of meltwater. This situation is typical of ice-contact basins where the ice front sits directly in the lake (Schmok and Clarke, 1989). In deep lakes, the ice margin may be unstable; rapid ice flow, calving and the release of icebergs results in enhanced sediment supply and the rapid deposition of chaotically-bedded facies. In many glacial lakes, varves are deposited only during the very latest stage of basin filling when ice is withdrawing from the drainage basin. The seasonal deposition of varves, typically no more than a few centimetres thick, clearly heralds the onset of basin starvation reflecting a marked reduction in sedimentation rates compared with earlier phases of lake history. In summary, it can be said that processes of seasonal sedimentation in glacial lakes have probably been overemphasized in the interpretation of laminated facies at the expense of other aperiodic processes. The varve model has certainly been too vigorously ap-

25

plied to the interpretation of laminated lithified facies in pre-Pleistocene glaciated basins where it is increasingly apparent that such deposits more likely record deposition in marine basins affected by hemipelagic and sediment gravity flow processes or by tides (Section 4.5.1; Fig. 3.10). Given the difficulty of positively identifying a deposit to be varved, it is preferable to use purely descriptive terms such as rhythmite or laminite and apply the detailed non-genetic, descriptive schemes that have been developed for such facies (e.g. Pickering et al., 1989; Schieber, 1990; Ghibaudo, 1992).

3.1.3 Cold climate facies and structures Before concluding this brief review of continental glacial environments and facies it is worth emphasising that the record of cold climates in Earth history is not restricted to glacial or glacially-influenced strata but is also recorded by periglacial facies and sedimentary structures. The term periglacial is something of a misnomer in that such deposits may accumulate in cold climates with or without the development of glaciers; it may, for example be too arid for glaciers to form. A rich legacy of cold climate phenomena survive from Pleistocene glaciations of the mid- and high-latitudes testifying to the very widespread development of cold conditions. This is reflected in a very substantial literature (e.g. Washburn, 1979); the modern day distribution of perennially frozen ground (permafrost) is similarly considerable (e.g. Fig. 23.3). In comparison there is a restricted number of published descriptions of comparable pre-Pleistocene cold climate structures and facies (see Spencer, 1971, 1975; Nystuen, 1976; Epshteyn, 1978; Edwards, 1979, 1984; Frakes, 1979; Deynoux, 1982; Eyles and Clark, 1985; Williams et al., 1985; Williams, 1986, Gostin, 1986; Fischbein, 1987; Collinson et al., 1989; Deynoux et al., 1989; Vaslet, 1990 and refs therein). Even here, however, the origin of many pre-Pleistocene structures is far from clear and it is probable that soft-sediment and cold climate deformation

26

N. EYLES

27

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C SE'VFING

~='=,

~

~

.~ ~ "0 ~ - ~

.=~

.~

'~

~

~a

~

~

28

N.EYLES

structures have been confused. Given the pronounced preservational bias of the rock record to glacially-influenced marine strata (Section 5) the apparent rarity of reported cold climate structures is not unexpected but the deposits of seasonally or perennially cold coastlines should be common. Again, however, the pre-Pleistocene literature is sparse (see Section 17.1). Many present day midlatitude shorelines are characterised by the reworking of large volumes of glacial sediment deposited during Pleistocene glaciations (e.g. Forbes and Taylor, 1987; Carter et al., 1990). These authors clearly show that existing sedimentary models for non-glacially influenced shorelines are inappropriate as a result of rapid sea-level changes and abundant coarse sediment. Existing sedimentary models for glacially-influenced coastlines, where glaciers discharge large volumes of glaciofluvial sediment to the shoreline, are limited to a few examples along modern glaciated coasts (e.g. Nummedal et al., 1974; Boothroyd and Nummedal, 1978). These conditions were commonly developed during and after pre-Pleistocene glaciations but these deposits, if they were preserved, currently go unrecognised.

3.2 Glacially-influenced systems

marine depositional

Complex ice-contact depositional systems are constructed where glaciers reach sea-level (Fig. 3.7). These systems are identified by their geomorphological and sedimentological complexity resulting from deposition in a dynamic, high-energy marine environment. Rapid lateral and vertical facies changes, highly irregular sediment geometries and a great range of facies types characterize these deposits. Descriptions of these settings and representative deposits can be found in Powell (1981, 1984, 1990), Solheim and Pfirman (1985), Syvitski et al. (1987), N. Eyles et al. (1988b), N. Eyles and McCabe (1989a,b), Anderson and Ashley (1991), Brodzikowski and Van Loon (1991) and Lonne and Mangerud

(1991). The long-term preservation potential of ice-contact glaciomarine deposits is slight as they are commonly scoured out by advancing ice, resedimented into deeper water as mass flows or are reworked by prograding fan deltas or by shallow marine processes (see Visser, 1983; Visser and Loock, 1987). Ice-contact facies pass offshore into glacially-influenced marine deposits of the shelf and slope (Fig. 3.1) which have a much greater chance of long term preservation. Diamictite facies generated in these settings predominate in the glacial rock record and have been deposited by two major processes. The first involves downslope resedimentation as sediment gravity flows either of glacial debris released from the ice front or of debris produced by downslope mixing of preexisting fine- and coarse-grained sediments (Section 4.2). Another process involves the accumulation of a wide range of "rain-out" diamict facies by the ice-rafting of debris into areas of mud deposition (Fig. 3.11). These diamict forming processes are not mutually exclusive. Very substantial accumulations of massive and weakly stratified diamictite, in some cases up to 3000 m thick, are widely interpreted as the product of combined "rain-out" and resedimentation processes in Late Proterozoic, Late Palaeozoic and late Cenozoic glaciated basins (e.g.C.H. Eyles et al., 1991; Matsch and Ojakangas, 1991; Young and Gostin, 1991; Visser, 1991). The fundamental control on deposition of such facies is that of regional climate. Temperate, wet-based glaciers, in areas of humid climate produce very large volumes of mud and meltwater resulting in "high input" glaciated continental margins characterised by very high sedimentation rates (e.g.C.H. Eyles et al., 1985; Anderson and Ashley, 1991; Henrich, 1991 and refs., therein). In colder arctic and antarctic settings this supply is virtually shut-off and deposition rates are an order of magnitude lower. Given the great importance of sediment gravity flow processes to the generation of diamictite facies in high input, glacially-in-

29

E A R T H ' S GL AC IAL R E C O R D AND ITS T E C T O N I C S E T T I N G

CHARACTERISTICS

PROCESSES

DEPOSITS

\ - ~ . ~ o~ n ~ . . ~

Rock F a l l

Olistostrome Avalanchedeposit

Creep Slide

~

Creep deposit Slide

Slump

, Debris Flow

Slump

" ~ Debrite

FDensity-modified r ~ "-~ .~ i grain flow L ~ . 9 ~ Turbidity Flow~Fluidized flow ~:~-J)"~_ /Liquefied flow ~ LHyperconcentrated flow

Turbidite (coarse,medium, fine grained)

Fig. 4.1. A continuum of sediment gravity flow processes and deposits--see Fig. 4.2 for facies. After Stow (1985). Turbidity flows, with the exception of hyperconcentrated flows are exclusively subaqueous processes. All other processes occur in both subaerial and subaqueous settings.

fluenced marine settings these processes are briefly reviewed in the following section. 4. SEDIMENT GRAVITY FLOW PROCESSES AND DEPOSITS

Glaciers, like volcanoes, can be regarded as large geological dump trucks. In most cases, the role of glaciers in the rock record Slump

has been to deliver very large volumes of poorly-sorted sediment to marine settings where glaciclastic sediment is reworked downslope by debris flows and turbidites. An understanding of sediment gravity flow processes and the predominant facies types is therefore important in unravelling the "glacial" record. Figures 4.1 and 4.2 provide a summary of the most important subaqueous Coarse-grained Turbidite

Debrite

Medium-grained Turbidite (classical)

Fine-grained Turbidite

_ f°-4Lr~

C~"-J

Jog

1

.]~v

uJ ~ O :£ ~~

COHERENT SLUMPING

COHESIVE MATRIX STRENGTH AND FLUIDBUOYANCY

TURBULENCE HIGHCONCENTRATION

LOWCONCENTRATION

Fig. 4.2. Subaqueous sediment gravity flow facies (see Fig. 4.1). After Stow (1985) with additions. Bed thickness and sediment volume generally decrease from left to right.

30

N. EYLES

31

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

o

_

~

o.~

-~_

~

c-

~

~o

~

~

='~_~

~-~

.

~_~,,,~o

-~

~_~

o

o

~

"~.~

~

~.,

o.~'~

.

•-' ,~ ~ G



.

~)

c~

~"~

r~,~

=

.

0

~

~

32

Y. EYLES

EARTH'S

GLACIAL

RECORD

AND ITS TECTONIC

33

SETTING

0~0

•~

~

0"0

~

~o

~

0

~

"~ o

o>.-~=



. - °

•~

~

.~,.

~

•r~ "0

m " 0 ,-~

~'~

"~ "0

~

~

34

N. EYLES

O

35

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T F I N G

~.q

09

O . -

~o~

O

O

°~

.~

@

.~

~a o °~ .~

..~ ~ ..~

o

rl

,~

~"

.1~I ~

°

0

~,

-~ ~ ~ ~-~

~ .~ ~~

o

36

processes and sedimentary facies associated with downslope resedimentation emphasizing the downslope evolution from poorlysorted to better sorted deposits. Full reviews can be found in Postma (1986), Nemec and Steel (1988), Pickering et al. (1989), Ghibaudo (1992) and Walker (1992).

4.1 Creep and sliding Downslope mass flow is often initiated by the slow movement of intact sediment accompanied by internal deformation (creep). This is identified by complex internal deformation of any original lamination, stratification or bedding. Creep processes are often initiated by high sedimentation rates at the top of depositional slopes and resultant "oversteepening" or by seismic activity. Downslope creep of intact strata can be likened to a carpet, of varying thickness, being "rucked up"; folds may be generated in the subsurface and diapirs may also develop in response to the vertical movement and "loading" of one sediment body into another. Continued creep of cohesive muddy sediments will ultimately lead to sediment failure and the formation of fault-bounded, slide blocks and eventually chaotically-bedded olistostrome deposits (Coleman and Prior, 1988; Fig. 4.3). Sediments may remain relatively undeformed within the individual blocks (olistoliths), with downslope movement accomodated along basal slide planes and faulted margins. The dimensions of slide blocks may be such that individual field outcrop or cores may be too small to identify that the strata had been displaced downslope (e.g. Heubeck, 1992). A hummocky surface relief is created on slopes that have experienced slumping; this may be identified in ancient strata by the complex outcrop geometry of facies that were deposited around slump blocks.

4.2 Debris flow Accelerated sliding results in the disaggregation of individual slide blocks and the for-

N. EVLES

mation of poorly-sorted debris flows that resemble and behave as wet concrete. An ability to transport large rafts of intact sediment or boulders is the result of the cohesive strength provided by any fine-grained matrix present in the flow together with buoyancy ("lift") forces provided by the relatively dense matrix and admixed pore water. Together these keep larger grain sizes aloft within the flow thereby producing massive, chaoticallystructured diamict facies (Figs. 4.4, 4.5). Dispersive pressure, created by repeated collisions between tumbling clasts, may provide an additional support mechanism and promotes mixing and homogenization within the flow. Typically, the flow moves as a semi-rigid mass, or plug, riding along on the top of a rapidly-deforming basal zone where shear stresses are greater than the strength of the debris/water mix. Debris flows "run-out" downslope and stop ("freeze") when excess porewater pressures in the basal deforming layer are sufficiently reduced such that its shear strength begins to exceed applied shear stresses. The term debris flow is used to describe both the sedimentary process and the resulting deposit though some workers prefer the term "debrite" for the deposit (Pickering et al., 1989). Debrite beds are poorly-sorted diamictites lacking any pronounced internal stratification. Diamictites may show crude grading of clast sizes reflecting limited internal sorting by dispersive pressures (Fig. 4.5). As a subaqeous debris flow moves downslope, water and mud may be taken up by the flow as a result of shear between the top of the flow and the overlying water column or as the flow moves over and incorporates muddy sediments. Flow acceleration and the onset of internal turbulence is recorded in debris flow sequences by the appearance of facies showing more frequent clasts in their basal or upper parts (Fig. 4.5); this is described as "coarse-tail" normal grading or inverse grading respectively since only the clast component of the flow is graded. Further turbulence will result in the formation

37

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SETTING

of a fully-developed turbidity current (Section 4.3). 4.2.1 Outcrop characteristics Good exposures of late Cenozoic debris flow facies, deposited within broad (500 m wide) channels cutting the outer edge of a glaciated continental shelf, are described by C.H. Eyles (1987). They comprise "chaotic" diamictite facies (Fig. 4.6) having a coarse sand to muddy sand matrix texture with abundant floating rafts of mud, gravel and sand and large boulders. Clasts fail to show any preferred orientation and include large numbers of angular and striated clasts recording the downslope movement of proximal, coarse-grained glaciclastic debris from an ice sheet margin that reached the shelf edge. These upper slope facies have a channelized cross-sectional geometry recording substrate erosion by rapidly-moving, and highly-erosive debris flows issuing from point sources along the ice margin. There is a marked tendency for the coarser-grained debris flows to occupy narrow, steep-sided, Vshaped channels and for less coarse-grained flows to fill broader U-shaped channels. Debris flow facies are intimately associated with coarse-grained turbidite facies (Section 4.5; Fig. 4.6). In contrast to the channeled "pointsource" debris flow facies described above, N. Eyles and C.H. Eyles (1989) and N. Eyles (1990) described laterally extensive, lenticular "line-source" flows deposited on a smooth midslope portion of a Late Proterozoic glaciated continental margin. Debris flows are recorded by massive diamictites up to 25 m thick and many kilometers in lateral extent, having a sandy to pebbly mudstone texture (Figs. 4.7, 4.8). Rare stratification is the result of thin, discontinuous interbeds of laminated mudstones or sandstone that record pauses tgetween successive debris flows. Diamictite bed bases are planar, generally non-erosive and conformable with sandy and muddy turbidite facies. Flows originated from extensive sediment slumps on

the upper slope. Flows are, intimately interbedded with turbidite facies and with slumped horizons (Figs. 4.6, 4.7). An important matrix component of air fall tephra suggests a volcanically-influenced setting. A glacially-influenced setting is recorded by striated clasts and dropstones. Aksu and Hiscott (1989, 1992) describe the geometrical characteristics of shingled Late Pleistocene debris flow units resulting from the repeated downslope failure of glaciclastic sediment along the eastern Canadian continental margin. Submarine debris flow facies can be produced by mixing of coarse and fine sediments, most commonly when gravel-rich turbidity currents run-out onto a muddy substrate (Crowell, 1957; Nemec et al., 1984). Gravitational "loading" of gravel into mud results in partial mixing which may be completed if the sediment mass is unstable and moves downslope as a debris flow. A Late Proterozoic succession of submarine slope facies, showing transitions from conglomeratic turbidites to deformed "disrupted" mixtures of conglomerate and mudstone to pebbly mudstone diamictite is described by N. Eyles (1990) from the Granville Formation of Normandy (Fig. 4.9). Diamictites retain evidence of their "mixed" origin in the form of small mud clasts akin to raisins in a fruit cake; the term "poudingue" is used, appropriately enough, by local geologists. In glacially-influenced marine settings, rapid deposition from suspended sediment plumes may blanket slopes and slope channels with mud; the incorporation of mud within coarse-grained turbidity currents produces debris flow facies (see C.H. Eyles, 1987). 4.3 Turbidity currents Turbidity currents are turbulent fluids with sufficient density to transport sediment particles in suspension (Middleton, 1993). They are produced by various mechanisms such as slope-failure and liquefaction, by sediment suspensions cascading down delta fronts or

38

N. E Y L E S

0

' ~ ~N Q

t.~

o~

0 '~ ~2 ~ -~

.~

~-~

,..< ~ =

~.~

°,4."'~ .~ . ~

~z~.~



E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

39

40

N. EYLES

o o

...~

. ~ a~ ~

!~.~-~_~ .=N

~.~.~ ~ ~ "~

~

~ ~ ~.~""

~d.~ . ~ . ~~a

~.~ '~

~.~Xz

~

41

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

within canyons, or by the dilution of a debris flow (see Walker, 1992 for a full discussion). All of the processes are important in glacial settings where sediment is carried to the continental shelf edge and spills over into deep water repositories. Turbidity currents may be short-lived events lasting minutes or a few hours (surge-type flows) or exhibit uniform flow over days or weeks (steady flows). In glacial settings, surge-type flows typically

develop from the dilution of debris flows released by the slumping of heterogenous glacial sediments. Steady flows are synonomous with glacier-fed lakes and tidewater ice margins where large inputs of meltwater produce strong underflows that may last the entire summer melt season (Section 3.1.2). Sediment is deposited from turbidity currents that begin to decelerate and "deflate" in areas of low slope. The character of the

I

/ /

/

I

I

I

DROOK FM. ¢

M~TAKEN PT.FM. FM. GASKIERS FM. MALLBAYR4. O~OOK

G A S K I E R S

O~.~ICTn'E

• IAAA~' 50m

f

25

0

F STRATIFIED DISRUPTED T~ES NUOSTONE

-~

Fig. 4.8

SUJWOEID$ D FAULT

MALL BAY FM.

Fig. 4.7. A classic Late Proterozoic glaciclastic facies association; subaqueous debrites and turbidites of the Gaskiers Formation, Newfoundland recording sediment gravity flow dominated deposition in basins receiving large volumes of glaciclastic and volcaniclastic sediments (Figs. 4.8, 10.6). After C.H. Eyles and N. Eyles (1989). The glaciclastic component forms the lower deep water part of a thick, coarsening-upward basin till (top left). See also Figs. 8.2, 8.3, 11.2, 11.4 and 12.2.

42

deposit is in part dependent upon sediment concentration in the flow and it is common practice to distinguish high concentration turbidity currents from low-concentration flows. This does not reflect the grain size of the suspended size but the density of the suspension. High concentration currents can be either gravelly or muddy. The density of the flow does, however impact on the ability of the current to transport coarse grain sizes. The classic Bouma (1964) sequence applies to relatively fine-grained, low concentration turbidity currents and shows an upward progression of bedforms deposited from a turbidity current as it loses momentum (Fig. 4.2). In contrast, within glacially-influenced settings where large volumes of mud or gravel are bulldozed to the head of steep slopes, turbidites facies having different characteristics are deposited from high concentration turbidity currents.

4.3.1 High concentration, coarse-grained turbidity currents High concentration, coarse-grained, turbidity currents deposit a variety of facies ranging from massive conglomerates and sandstones to coarsening-upward (inverselygraded) and fining-upward (normally-graded)

N. EYLES

facies (Fig. 4.6). Massive beds have been described in the past as "grain flows" (Lowe, 1976) where the inferred support mechanism was thought to be dispersive pressures created by grain-to-grain collisions. This process is now known to operate only on steep slopes and the key role of small amounts of cohesive matrix in generating a buoyant uplift within the flow has been recognized ("density-modified grain flows"; Lowe, i982). More recently it has been suggested that massive beds are simply the product of rapid en masse settling of dense grain concentrations from a high concentration turbidity current (Pickering et al., 1989) and this reasoning is followed herein. Significant size sorting is probably prevented by rapid transport, short transport distance and rapid loss of flow competence (e.g. Postma et al., 1983). Consequently, these facies are intimately associated with debris flow facies and, in general, are characteristic of "proximal" settings near to sources of poorly sorted sediment. They pass "distally" into inversely and normallygraded conglomerate and sandstone facies. The origin of inverse-grading is not well understood. It may result from the upward dispersal of clasts in response to clast-to-clast collisions at the base of a flow or some form

Fig. 4.8. Interbedded turbidites (near hammer) and thick (6 m), massive debris flow facies (below hammer). Late Proterozoic Gaskiers Formation, Newfoundland, Canada. See Fig. 4.7 for location.

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C SE'I~I'ING

of "kinetic seiving" where smaller clasts are able to slip down between larger clasts. Irregardless, preservation of inverse grading requires rapid "freezing" of a high concentration, coarse-grained flow. C.H. Eyles (1987) showed that these facies occupy the axial portions of glacier-fed submarine channels recording higher clast concentrations and flow velocities in those positions. These facies passed both laterally and downslope into normally-graded facies deposited from low concentration flows.

4.3.2 Low concentration, fine-grained turbidity currents The deposits of low concentration, muddy turbidity currents typically comprise only the upper divisions of the Bouma sequence and a number of separate descriptive schemes are available for such deposits (see Stow and Piper, 1984; Stow, 1985). These facies are likely to be encountered in terrestrial glacial settings (in lakes), locally on continental shelves and commonly in deep water slope and base of slope settings. Because the same facies accumulate in all these settings discrimination of lacustrine from marine, shallow from deep water, can be very difficult and is reliant once more, on contextual information forthcoming from associated facies. Together with diamictites, thin, fine-grained turbidites provide major problems of interpretation for the glacial sedimentologist. Their deposition in glacially-influenced marine and lacustrine settings is briefly reviewed below. Those of glaciolacustrine origin have been discussed in Section 3.1.2. Laminated muds are a common stratigraphic component in glacially-influenced marine settings. In proglacial environments, diurnal (tidal), seasonal and weather-induced changes in meltwater discharge from tidewater glaciers control the volume of mud transported by and deposited from suspended sediment plumes. Seasonally-deposited, laminated glaciomarine muds resembling glaciolacustrine varves, were described by Domack (1984) from Pleistocene strata. Diurnally-

43

produced laminations have been identified by Cowan and Powell (1990) and Phillips et al. (1991) from modern Alaskan fiords; those resulting from sudden storms were identified by Cowan et al. (1988). The identification of a seasonal control on marine sedimentation close to glaciers is not straightforward because the interaction of tidal fluctuations with suspended sediment also produces laminated muds. Buoyant, mud-laden meltwater, released as part of an efflux jet (Fig. 3.7) produces brackish surface water layers both at the surface of the water column (overflows) and at intermediate depths (interflows). At low tide, suspended sediment is released from the bottom of these layers to form thin (< 1 cm) graded laminae of sand and silt laminae (Fig. 3.10); two such laminae can be produced by the diurnal tidal cycle (Phillips et al., 1991). Cowan and Powell (1990) showed that tidal laminae are interbedded with other laminae types produced by both surge-type and steady-flow turbidites. The former record episodic slumping along the ice front; the latter record density underflows generated directly from the incoming meltwater effiux jet. Unfortunately, unambiguous discrimination between laminae produced by these different mechanisms is not possible at present (e.g. Phillips et al., 1991, p. 59); the potential for confusion with glaciolacustrine varves is great. Domack (1990) describes tidally-laminated glaciomarine muds deposited along the Antarctic continental margin and emphasises the need to descriminate laminites composed of terrigenous sediment from those resulting from productivity cycles in diatom populations. Laminated sediments are often a significant component of strata deposited in icedistal settings on continental shelves and slopes. Thick (O's to 00's m) Late Pleistocene deposits of muddy turbidites have been deposited along glaciated continental margins within continental slope prisms or within the distal portions of deep marine submarine fans (e.g. Section 23.3). These deposits have

44 N. E Y L E S

--.m-,

_O

E

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

45

o

.'~

0

E ..~

46

been described in considerable detail from core studies and are summarized by Swift (1985) and Piper et al. (1990). Large volumes of glaciogenic mud were deposited on the continental slope by meltwater plumes that spilled over the shelf break to form dilute turbidity currents ("lutite flows"; Pickering et al., 1989). Currents merge in deep water, with slope-parallel "contour currents". The deposits of dilute muddy turbidity currents and dilute, muddy contour currents ("contourites") are virtually indistinguishable even in the modern marine environment (see discussion in Pickering et al., 1989). However, contourites are a volumetrically-minor component of most continental margin stratigraphies and their stratigraphic importance is likely to be even less in glacially-influenced slope settings where downslope movement of very large volumes of sediments will overwhelm any contribution by currents moving parallel to the slope (e.g. Swift, 1985). This has been clearly demonstrated for the northwest United Kingdom continental margins where Late Pleistocene interglacial sedimentation on the slope is dominated by relatively thin "background" hemipelagite and contourite drifts. Glacially-influenced sedimentation in this region is dominated by debris flow deposits (Stoker et al., 1991).

N. EYLES

With the vastly increased extent of seismic coverage in the ocean basins surrounding the glaciated portions of North America it is now possible to make realistic estimates of the volume of glaciclastic sediment produced by Late Pleistocene ice sheets. These data can be used to constrain the average glacial erosion rate on continental surfaces under different tectonic settings. The North American continent hosted two large ice sheets during successive glaciations; the Cordilleran Ice Sheet in the west (2.8 × 10 6 k m 2 ) , which grew by the amalgamation of valley glaciers, and the Laurentide Ice Sheet (13.4 × 106 km2), which grew by the thickening and amalgamation of snowfields, principally along the uplifted eastern rim of the Canadian Shield (Fig. 5.1). The Cordilleran Ice Sheet occupies the Pacific Margin of North America characterized by active plate margin tectonism and mountainous topographic relief. The Laurentide Ice Sheet is limited to the relatively low relief stable craton and episodically extends onto the tecton-

5. TECTONIC CONTROLS ON SEDIMENT PRODUCTION, D E P O S I T I O N A L FACIES AND PRES E R V A T I O N P O T E N T I A L IN G L A C I A T E D BASINS

The following section attempts to make some quantitative generalisations, based on well studied Pleistocene glaciated basins, regarding the production of glaciclastic sediment under different tectonic settings and the long term fate of such sediment. This analysis then allows assessment of the relative importance of marine and terrestrial settings to accomodate and preserve glaciclastic sediment. In turn, it is possible to predict the relative preservation potential of different glacial facies types and their probability of occurrence in the rock record.

D~IVED MARINE SEDIMENT

Fig. 5.1. Extent of Cordilleran and Laurentide ice sheets in North America with type and distribution of Pleistocene glacial sediments.

47

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

ically "passive" divergent margin of the western Atlantic and Arctic Ocean basins.

5.1 The eastern passive margin of North America Figure 5.2 shows the volumes of Pleistocene glaciclastic sediment preserved along the shelf, slope and abyssal plains of the western Atlantic Ocean. This sediment is derived from the adjacent continental surface where regional glaciation is assumed to have begun at about 2.5 Ma (Section 18.3). The largest volume of sediment is present on the Sohm Abyssal Plain (110 × 103 km 3) with a further substantial volume present within the continental slope; relatively little is stored

60*W

on the shelf which reflects the lack of subsidence along the slowly-uplifting passive margin (Section 23.1). Along most of the continental shelf of eastern Canada, deposits of the latest glacial maximum ( < 30 Ka) rest on a regional unconformity cut across Mesozoic and Tertiary strata. Older glacial deposits have not been preserved because the shelf has not subsided (Section 23.2). The total volume of Pleistocene glaciclastic sediment stored along the western Atlantic margin of North America is about 300 × 103 km 3. The source area from which this is derived has been estimated to be 4000 × 103 km 2 (Piper et al., 1990). The mean total thickness of the craton and shelf eroded by successive Pleistocene ice sheets is therefore about 75 m.

40 °

w

TOTAL PLEISTOCENE GLACICLASTIC SEDIMENT ( x l O ~ k m 3 )

SHELVES

[]

SCOllAN SHELF GULF OF MAINE GULF OF ST. LAWRENCE

1S

GRANDBANKS

3

NE NEWFOUNDLAND AND LABRADOR

9

HUDSON STRAIT AND BAFFIN SHELF

a O

35 SLOPE AND RISE

[]

SOD n 'AN

36

GRANDBANKS

30

LABRADOR

60

BAFFIN

5

131 ABYSSAL PLAIN SOHM NEWFOUNDLAND NOR33"IERNLABRADOR

BASIN (BOW)

BAFFIN BAY (50%)

[] 110

1Z 10 Z.5

134.5 TOTAL SEDIMENT - 301 (xlO 3 km 3 )

Fig. 5.2. Volume of Pleistocene glaciclasticsediment along eastern (passive) Canadian margin of North America (data from Andrews, 1990; Piper et al., 1990).

48 The volume estimate of glaciclastic sediment stored in the Western Atlantic has not been corrected for compaction nor for the large volume of sediment eroded chemically. In addition, a larger volume of sediment may in fact be stored on the continental surface than has been previously assumed ( e . g . N . Eyles et al., 1991). Bell and Laine (1985) suggested as much as 19 m of rock had been chemically removed from the area glaciated by the Laurentide ice sheet during the last 3 Ma. These workers concluded that almost 140 m in total had been removed by physical and chemical processes. This figure is much larger, almost twice as great, as previous estimates made by Flint (1971), Gravenor (1975) and Kaszycki and Shilts (1980). Comparison of volumetric data for the same west Atlantic area by Bell and Laine (1985) and Piper (1991) shows the later estimate of Piper's to be less than 50% of that made on the basis of older data by Bell and Laine (1985). According to Piper (pers. commun., 1991) the older data set included a significant component of underlying pre-glacial

N.EVLES sediment. Despite these uncertainties it is safe to conclude that Pleistocene ice sheets stripped between 60 and 120 m of rock from eastern North America with the real value somewhere closer to the lower figure. Regional glaciation commenced c. 2.5 Ma before present (Section 18.3) suggesting an averaged erosion rate of between 24 and 48 m/Ma.

5.2 The western active margin of North America The data base for the Pacific Ocean basin is much more restricted compared with that available for the Atlantic and considerable uncertainty surrounds estimation of the volume of glaciclastic sediment stored along the active continental margin (Fig. 5.3). Over most of its extent along the British Columbia and Oregon coast, there is a restricted thickness of glaciclastic sediment ( < 100 m) reflecting a compressional and transpressional plate margin regime and associated uplift. The major repositories of Pleistocene glaci-

300

Fig. 5.3. Volume of Pleistocene glaciclastic sediment (X 10 3 knl3) along western (active) margin of North America (data from Plafker, 1987; Stephenson and Embley, 1987). Same key as in Fig. 5.2. A further 11 x 10 3 k m 3 is stored onland (Fig. 5.4) and about 100 × 10 3 k m 3 has been lost to subduction. By way of comparison, the volume of Bengal Fan (Fig. 5.S) is 4 x 10 6 k m 3 (Bouma et al., 1985).

EARTH'S GLACIALRECORD AND ITS TECTONIC SE'Iq'ING

clastic sediment occur in the north Pacific Ocean adjacent to the heavily glacierized Gulf of Alaska, Kodiak Margin and eastern Aleutian Arc areas. Glaciation of this sector is thought to have began in the late Miocene (c. 6 Ma; Section 26.1) and resulted in a flood of glaciclastic sediment to the Gulf of Alaska and Kodiak forearc basin, the Aleutian Trench and the associated deep sea fan systems (e.g. Surveyor Fan; Fig. 5.3). The total volume of sediment stored in the forearc basin is not well known but at least 135 x 103 km 3 is estimated to be present along the northern Gulf of Alaska based on isopach maps that show up to 5 km of glacially-influenced marine strata (Plafker, 1987); broadly similar sediment thicknesses are each recorded for the Kodiak and eastern Aleutian Arc margins. A total of about 300 x 103 k m 3 is estimated to be contained within the shelf margin (Fig. 5.3). The volume of sediment within the slope is estimated as 100 x 103 km 3 (Fig. 5.3). Complexity is created by the observation that the slope comprises glaciclastic sediment derived both from the continental margin together with that component of sediment scraped off the subducting Pacific plate and now part of a thick accretionary wedge. It has been suggested that some 500 km of Pacific plate has been subducted along the Aleutian Trench since the start of Late Miocene glaciation (Stephenson and Embley, 1987). This is equivalent to an area of about 50% of the present Surveyor Fan and indicates that as much as 100 x 103 km 3 of sediment may have been lost. A total of 1218 x 103 k i n 3 of late Cenozoic glaciclastic sediment is estimated to be present in the north Pacific Ocean (Fig. 5.3). The area covered by that sector of the Cordilleran ice sheet which drained to the Pacific Ocean is approximately 1.4x 10 6 km 2. By direct comparison with the area of the Laurentide Ice Sheet and the amount of continental glacial sediment stored within that area (Fig. 5.4), about 11 x 103 km 3 of glaciclastic sediment is estimated to remain

49

on land in this portion of the ice sheet. Thus the total volume of glacigenic sediment produced along the active Pacific margin is about 1229 x 103 km 3. This figure has to be increased by about 100 × 103 km 3 to account for sediment lost to subduction. The glaciated source area is about 1.4 x 106 km 2 and it can be suggested that an averaged thickness approaching 1000 m of bedrock has been eroded by late Cenozoic glaciers along the western continental margin of North America. This figure has not been corrected for compaction or for the volume of rock lost by chemical erosion; the latter may be substantial within temperate glaciated catchments (e.g.N. Eyles et al., 1982b). If correct, the value of about 1000 m is an order of magnitude larger than the averaged depth of glacial erosion apparent along the passive margin of eastern North America (60-120 m; Section 5.1). Given that regional glaciation was initiated at about 6 Ma then a glacial erosion rate of about 166 m / M a is indicated.

/

~c'nco~,. (.t

f

~

170(.)~. ~

~' \ -~. w.A'n~N~C



SOURCES; (*) (1) (Z) (3)

BELLAND ~ (1985): PROBABLEO V E ~ T E PPERetd(lggO); RGS.Z. BELL& ~ (1985) WeS s-nJDY: FiG. 5.3.

C

(SEETEXT).

Fig. 5.4. V o l u m e of Pleistocene glaciclastic s e d i m e n t

remaining on the continent compared with that preserved in the marine setting (see text for details and Fig. 5.5 for schematic representation).

50

N. EYLES /,,'/@~,9.\vx~, /

2271 x 103kin 3

127 x 103km 3

TOTAL VOLUME OF PLEISTOCENE GLACICLASTIC SEDIMENT = 2398 x 103km 3

Fig. 5.5. Volume of Pleistocene glaciclastic sediment preserved on the North American continent (6% of total) compared with that preserved offshore in marine settings (94%). See Fig. 5.4. This gives a clue as to the relative proportions of marine and terrestrial glacial strata preserved in the ancient rock record.

5.3 Preservation potential of glacially-influenced marine vs. terrestrial glacial sediment in earth history; discussion From Figs. 5.4 and 5.5 it can be shown that less than 6% of the total volume of glaciclastic sediment produced by late Cenozoic ice sheets in North America remains on land (Bell and Laine, 1985). The bulk of the sediment flux has been directed to the marine environment and deposited in deep water continental slope and submarine fan settings. This has probably been a long term characteristic of glaciation throughout Earth's history and provides an important quantitative estimate regarding the relative proportions of marine and terrestrial "glacial" strata in the ancient record. A second observation relates to the importance of tectonic setting to sediment yield; contrast the order of magnitude larger volume of sediment eroded from the active tectonic margins of the Pacific with that produced from the passive Atlantic margin (Fig. 5.4). The Pacific margin is dominantly high relief and characterized by easily-eroded sedimentary strata that have been deeply dissected by "high yield" valley glaciers and piedmont ice lobes. The Atlantic margin, in contrast, is dominated by resistant Archean and Proterozoic shield rocks which have been

scoured by an extensive but "low yield" ice sheet. Thus there is no simple relationship between the size of the glaciated area and the volume of glaciclastic sediment produced. The Pacific margin is characterized by temperate oceanic conditions of high precipitation where glaciers have a high "mass transfer". This is to say that these temperate glaciers experience a high snow fall during the winter that is balanced by large mass losses during the melt season. These glaciers therefore release very large volumes of water-the dominant glacial marine sediment released by these glaciers is mud. It appears very likely that the single most important influence on sediment yield in glaciated basins is not the size of the glaciated area but the tectonic setting which in turn constrains substrate relief. Regional climate and the availability of meltwater (i.e glacier thermal regime) are important secondary influences. If the data derived from Pleistocene glaciated basins (Figs. 5.1, 5.2, 5.3, 5.4, 5.5) have a wider significance it can be argued that the bulk of the Earth's glacial record consists of glaciclastic sediment that has been deposited and reworked in marine settings. Furthermore, most of this sediment flux has probably been completely reworked by marine processes thereby eradicating any evidence of a glacial influence on sedimentation. Given that the bulk of sediment is transferred to deep marine submarine fans, such as shown in Fig. 5.6, it can be predicted that the principal stratigraphic record of glaciation in Earth history is contained wihin deep water successions of turbidites and debris flows. Indeed, this appears to be the case for, as will be demonstrated in what follows, the dominant "glacial" facies types in the ancient record are marine debris flow diamictites and turbidites. 6. OVERVIEW OF EARTH'S GLACIAL RECORD

Figure 6.1 depicts the Earth's glacial record where deposits are reasonably con-

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E T T I N G

strained by radiometric age dating. A large n u m b e r of glaciclastic strata whose age is unconstrained by absolute dating or whose origin is unclear, have not b e e n included. Comprehensive reconstruction of Earth's glacial record is h a m p e r e d by a lack of absolute age dates and uncertainty regarding the glacial origin of many deposits. C h u m a k o v (1981b, 1985) argued that glacial deposits are unevenly distributed in the stratigraphic column and this is borne out by Fig. 6.1. O t h e r workers perceive evidence of either a 300 Ma (Keller, 1973) or a 150 M a periodicity in glaciations (John, 1979, p. 223). A 300 Ma periodicity identified by Keller (1973) was linked to the duration of a galactic or cosmic year by Steiner and Grillmair (1973). In contrast, Harland (1981) concluded that the stratigraphic record suggested irregular episodic rather than any regular, cyclic glaciation and this argument is supported by the data depicted in Fig. 6.1. It would appear that glaciation is a "normal",

51

0

~

_~

Ne~ene Palaeo~ene

m



m

Cretaceous

?

?

Jurassic

?

?

Triassic

l'l|-



m

0

200.

300- o ;'~ 400-

Permian Carbonifarous Devonian

~

-

?

l

?



Silurian Ordovician Cambrian Vendian



i

i

-"

.• "IIB]].. m

|

m

,I

Sturtian

a,'

800-

Riphean

1000I:

|

150¢

Animikean 20(X Huronian

,,||

?

25O(

?

? t,

30O0

L

~



<

-

l,i;ii

Fig. 6.1. Earth's glacial record. This figure differs from previous versions in that it depicts only those deposits that contain unambiguous glaciclastic sediment and that are reasonably constrained by radiometric age dating. Many other deposits, reported as glacial in origin, are poorly understood, not well described and very poorly constrained by either relative or absolute age dating. The record of Archean and Mesozoic glaciation(s) is controversial (see text). The Riphean was likely characterized by local ice masses in tectonically-active areas as the Late Proterozoic supercontinent began to assemble but no well-constrained record can be identified (Fig. 9.2). Simplified timescale after Harland et al. (1989). Fig. 5.6. Comparative sizes of Earth's submarine fan systems. Data from Bouma et al. (1985) and Fig. 5.3. Enhanced sediment supply from glaciers, tectonic setting and topography of basin floor has dictated fan size. The bulk of Earth's glacial record has been deposited in such settings.

frequently occurring feature of planet Earth and has occurred at different time intervals. As yet, there is no convincing evidence of extensive Archean glaciation ( > 2500 Ma;

52

N. E Y L E S

Section 7). More convincing, and possibly correlative Early Proterozoic glacial deposits deposited just after 2300 Ma occur in North America and Europe (Section 8) and appear to record accumulation of glacially-influenced mass flows in marine basins along a rifted continental margin; less convincing deposits occur in Australia and South Africa. The role of local, tectonically-driven vs. global controls on the initiation of glaciation and the extent of the ice cover is as yet unresolved. Glaciation at this time appears to coincide with a step-wise jump in oxygenation of the Earth's atmosphere which is increasingly being interpreted as a result of global plate tectonic events (Section 8.2.1). No well-dated glacial deposits are recorded for a lengthy mid-Proterozoic interval from about 2200 to c. 800 Ma and this may reflect non-recognition, non-preservation or relatively high average global temperatures in response to CO 2 induced "greenhouse" con-

1STORDERCYCLES RELATIVECHANGESOFSEALEVEL

ditions (Sections 7.3, 8.2). In contrast, Earth's glacial record during the Late Proterozoic (after c. 800 Ma) and Phanerozoic (after c. 600 ma) is substantial except apparently, for the Mesozoic (Figs. 6.1, 6.2). Having said this, it is very unlikely that the Earth has ever been totally ice free after about 800 Ma. Recent syntheses of Late Proterozoic and Phanerozoic tectonic, oceanographic and biotic evolution all emphasize the importance of a global plate tectonic cycle of fragmentation, dispersal, assembly, stasis and fragmentation with complete "Wilson" cycles lasting about 400 Ma (Worsley et al., 1986). Figure 6.2 combines the global tectonic cycle with the eustatic sea-level curve identified by Vail and his associates (Vail et al., 1977). The Late Proterozoic and Phanerozoic stratigraphic record of glaciation is also shown. It can be clearly seen that there is n o simple relationship between glaciation and the global tectonic cycle; the timing of glaciation

2ND ORDERCYCLES(SUPERCYCLES) PERIODS

RELATIVE CHANGESOFSEALEVEE RISING FALLING

RISIk_~pRESENTSEALEVEL FALLI.~NG

PRESENT

z~

SLOSS CYCLES

~A LEVEL

~--~..~"

c°Li

CRETACEOUS

~

ZUNI

JURASS,C TRIASSIC

"

1

!

1

p

~

.i

DISPERSALOF CONTINENTS

S (~)

ABSAROKA

600

~r~

-200

PERMIAN

300

= , I , , I , ,

-100

-

i

0

0

-QUATERNARY-, ~ ' " " TO T¢ TERTIARY . . . . . I. . . . . . . . . . . . . . . . . . . f .................. I r=~. . . . 0) 1(

NUMBER OF MARINE ANIMAL FAMILIES

GLOBAL TECTONIC CYCLES

~- DEPOSITIONAL

I ~ ~I--$TARTOFRIFTING 4 RNALCOALESLENCE "4'--OF PANGEA 3 O .d

-3oo ~ Z

PENNSYLVANIAN MISSISSlPPIAN D-M

DEVONIAN

O

KASKASKIA

C

o

)4(

-400 ~) SILURIAN TIPPECANOE

ORDOVICIAN

DISPERSAL ~j MAX OF'CONTINENTS

-500

5C CAMBRIAN

co \

SAUK

B REAKU O PF LATEPROTEROZOIC

/

SUPERCONTINENT

Fig. 6.2. Earth's Phanerozoic glacial record with that of sea level (after Vail et al., 1977), depositional sequences (after Sloss, 1963), atmospheric CO 2 (after Berner, 1990), ratio of 87Sr/86Sr in marine waters (after Koepnick et al., 1988), rate of ocean crust accretion (after Gaffin, 1987), global tectonic cycles (after Worsley et al., 1986) and marine animal extinctions (after Raup and Sepkovsky, 1982). See also Fig. 21.1.

53

E A R T H ' S GL AC IAL R E C O R D AND ITS T E C T O N I C S E T I ' I N G

is not in phase with the global tectonic cycle. The concept of such global cycles, of course, assumes a highly simplified view of crustal development and obscures much regional complexity. For example, most Late Proterozoic glaciation(s) occurred in extensional, rift settings during the disintegration phase of the Late Proterozoic supercontinent and so could be related, in very simple fashion, to the breakup phase of a typical Wilson cycle. However, this breakup phase and the related glaciation, was protracted and markedly diachronous, beginning at around 750 Ma along the palaeo-Pacific margin of Laurentia (e.g. Australia, western North America; Section 12) followed by another phase of rifting and glaciation at about 640 Ma which was coincident with the onset of rifting and glaciation along portions of the palaeo-Atlantic margin of Laurentia (e.g. Scandinavia, Britain, Greenland and Spitsbergen; Section 11.1). Thus the long-standing notion of a single "global" Late Proterozoic glaciation is not supported by the rock record and complicates any search for simple correlations with global tectonic cycles; instead, a good case can be made for the repeated, but markedly diachronous, triggering of regional "adiabatic" glaciation on uplifted rift margins from 750 to 640 Ma (Section 9). It should also be noted that the amalgamation and disintegration phases of the Late Proterozoic supercontinent actually overlapped (Hoffman, 1991); during the same time interval local cordilleran ice masses developed along volcanic arcs produced by continued amalgamation of parts of the supercontinent; glaciation persisted until as recently as 550 Ma in these active plate margin settings. The overall picture that emerges from the Late Proterozoic is that of local glaciations triggered by regional tectonic controls. Previous workers have seen evidence in the Late Proterozoic of climatic "crises" recorded by the apparently paradoxical interbedding of "warm" (e.g. dolomites) and "cold" strata; (e.g. tillites). It is the case that these notions have preceded detailed stratigraphic and sedimen-

tological studies and as details emerge it is clear that uniformitarian explanations apply (Sections 13, 14). Early Paleozoic, Late Ordovician glaciclastic strata record a brief phase of glaciation shortly before 440 Ma (Section 15) at a time when continents were widely dispersed. The longest phase of Phanerozoic glaciation is recorded by Late Palaeozoic Permo-Carboniferous glacial strata (350-250 Ma) that were preserved within intracratonic basins across the Gondwana supercontinent (Section 16) at a time of tectonic "stasis". Cenozoic glaciations (50 Ma to present) in contrast, have occurred during a period of advanced continental fragmentation and dispersal (Section 18). It is apparent that no one specific global plate tectonic or palaeogeographic setting is essential for the initiation of glaciation. Glaciation has occurred during phases of supercontinent assembly and during phases of continental dispersal (Fig. 6.2). Regional plate tectonic controls appears to be of much greater importance. Given that glaciation is a "normal", frequently occurring feature of planet Earth, and that global geochemical controls play an important role, the glacial record primarily reflects the regional tectonic conditions under which glaciation is initiated and in which glacial sediments are preserved. It is the principal conclusion of this study that ancient glacial deposits were most commonly deposited and preserved within extensional plate margin or intracratonic rift settings. Complementary uplift of rift shoulders to produce glaciated source areas, combined with rapid subsidence of offshore basins containing glacially-influenced marine strata, is a recurring theme throughout Earth's glacial record. 7. A R C H E A N ( < 2500 MA) G L A C I A T I O N

The timescale used in this section follows that of Harland et al. (1989) where the Archaean/Proterozoic boundary is placed at 2500 Ma. Proterozoic time is generally di-

54

vided into three chronometric subdivisions at 1600 and 900 Ma (or 1000 Ma in some schemes) and the terms Early, Middle and Late, Palaeoproterozoic, Mesoproterozoic and Neoproterozoic, or Pt 1, Pt 2, Pt 3 have been variably employed (Plumb, 1991). What is the oldest known glacial deposit on Earth? This is not an easy question to answer because evidence of Archean glaciation is tentative at best. This is surprising given that the radiative output of the Sun at this time is thought to have been only 75% of that of the present day (Gerard et al., 1992; Fig. 7.1). The beginning of the extensive glacial heritage of planet Earth is traditionally placed at around 2700 Ma (Hambrey and Harland, 1981) at a time when the planet's atmosphere was dominated by a non-reducing mixture of CO 2, N 2 and H 2 0 in which oxygen was present only as a trace gas (Figs. 7.2, 7.3). Oxygen only became an abundant constituent of the atmosphere at about 2300 Ma (Walker et al., 1983; Retallack, 1990). Study of the 3800 Ma banded iron formations exposed at Isua in west Greenland suggests, nonetheless, that sedimentary processes by that time were essentially the same as at the present day (Holland, 1984; Draganic et al., 1991).

N. E Y L E S

1.0

1.0

>-

F0 z

o

~, :~

-~

0.8

E

0.7

N

o.8 o

0.7 sooo

4000

3000

zooo

-<

0.6 o.o

looo

TIME ( M a )

Fig. 7.1. Solar luminosity through geologic time as modelled on left by Endal and Sofia (1981) and on right by Bahcall and Ulrich (1988). Uncertainty arises as to original luminosity.After Crowleyet al. (1991).

Kaapvaal Craton (Fig. 7.4). Sedimentation was initiated between 3120 and 3070 Ma and was coeval with deposition in the Pongola Basin to the south (Armstrong et al., 1991; de Wit et al., 1992; see below). The lower part of the succession (West Rand Group) is dominated by deep water turbidites that shallow up into coarse-grained fan-delta facies of the Central Rand Group. Gold fields of the Central Rand are scattered around

7.1 Southern Africa

Witwatersrand Glaciation ?

1

Diamictites of the Witwatersrand Supergroup in Southern Africa are widely quoted as the oldest known tillites (Hambrey and Harland, 1981; Chumakov, 1985) but their glacial status cannot be clearly demonstrated. Alleged Archean glacial deposits of the Kaapvaal craton, which extends over some 1.2 × 10 6 K1TI2 of southern Africa, are of particular significance because the craton is largely unique in retaining a substantial component of pristine mid-Archean crust. Because of its great mineral wealth this area has been the subject of intense investigation (de Wit et al., 1992 and refs. therein). Strata of the 7 km thick Witwatersrand Supergroup fill an elongate trough on the

co

0.9 0.9

\

~

10'

-

HuronianGlaciation ? N

,,

k

,,

\

&

@

o_

x

\

10 .2

\\ Late Precarnbrian Glaciation 10 "~

i

-XX\

\

\

-

Permo°Carboniferous / Glaciation I0 ~ 450'

I 3500

~

I

i

I

2500

1500

500

/

TM

"4

/ 0

AGE (Ma)

Fig. 7.2. Long term trend of atmospheric C02, after Kasting (1987). Note that model assumes continentalscale glaciation during the Archean (Witwatersrand) and Early Proterozoic (Huronian). This is not confirmed by the stratigraphic record (see text). See Figs. 6.2 and 21.1 for CO2 variation during the Phanerozoic.

EARTH'SGLACIALRECORDAND ITS TECTONICSETTING

55

The precise plate tectonic setting of the initial Witwatersrand basin is unclear but there is agreement that the period from about 3200 and 2700 Ma was characterised by accelerated tectonic activity involving Andeantype arc volcanism, inter-cratonic collision and rifting. Some workers argue that the Witwatersrand Group was deposited in a foreland trough (Burke et al., 1986). In this model the initial basin (West Rand Group) developed as a retroarc foreland basin on the Kaapvaal Craton behind an Andean-type volcanic arc recorded by the Dominion Reef volanics (Fig. 7.4). Other workers have interpreted Witwatersrand strata as the infill of a graben resulting from oblique collision between cratons (Clendenin et al., 1988). Recent data favour a passive continental margin facing an open ocean to the south (de Wit et al., 1992). The turbidites of the West Rand Group record reworking of volcanic

the apices of the former fan deltas along the northwestern margin of the basin (the Limpopo Mountains). Several of the goldbearing conglomerate horizons ("reefs") within the Central Rand Group contain subsidiary diamictites which have been interpreted as "tillites" (Wiebols, 1955; Harland, 1981). They have been recognized in the Kimberley Reefs, Bastard Reef and within the West Rand Group in the West Rand Shale. "Varved" shales, folded, striated and deformed substrates, the constant thickness and lateral persistance of the diamictites have all been cited as supplementary evidence of a glacial origin (Wiebols, 1955). Non-diamictite facies were interpreted, in classic climatostratigraphic style, as recording glacial retreat. Recent identification of the tectonic setting of the Witwatersrand Supergroup suggest that the alleged "tillites" are better interpreted as conglomeratic mass flows.

E/~RLY blID4~DTEIIOZOIC PROTIEROZOIC NOI~I.G!..~aAL GLACIATION INTERVAL

GowgandaFormation

LATE PROTEROZOI¢ GLA~ATION

~-,-~7,,//////////~

-1.4

10-1 .~ 102

~ / / ~

/ Dickinsonia

-1.2

10.3

-1.0

-~ 10.4

e oN I0 s ~ / ' ~~~ / ~// ' ~// / ' 2/ ~ Lj B ~ n d e d t--~Abundant Ir°n F°r t ymat hPhyl P i°r;lankton ~ndant

-0.8

10~3-

~s00

2 s'00

i s'o0

s6o

0.6

Years before present (Ma) Fig. 7.3. Best estimates of atmospheric oxygen over the last 3500 Ma. Shaded zone shows range of partial pressures inferred from geologic data. After Kasting (1991) with additional data from Robinson (1991) and Riding (1992). A l l models are agreed upon a rapid rise of atmospheric oxygen around 2000 Ma. This is coincident, within uncertainity of age dating, with Huronian glaciation. Solid line shows averaged quantity of organic carbon (M) in the crust (after Des Marais et al., 1992). Increased burial results in an increase in 0 2 content in atmosphere. Note step-wise increase in M and association with glaciations in the Early and Late Proterozoic. Intermediated values of M during the Mid-Proterozoic are associated with intermediate 0 2 contents of the atmosphere and a long non-glacial interval. Abrupt increases in M are tectonically-driven, suggesting causal relationship between global tectonics and glaciation (see text).

56

N.EYLES aerial (Central R a n d Group) debris flows, recording proximity to steep basin margins and a ready supply of volcaniclastic sediment (Martin et al., 1989; Stanistreet et al., 1989; Tainton and Meyer, 1990). The work of these authors, together with that of Kingsley (1984) on alluvial fan sedimentation in the Witwatersrand Supergroup, provides an firm basis for further work on sediment gravity flow deposits in the Central R a n d and West R a n d Groups. von Brunn and Gold (1992; pers. commun., 1993) report striated clasts within mass flow

debris and, in the upper part of the successsion, are interbedded with mafic lavas (e.g. Crown Lava, Bird Lava; Fig. 7.4). The overlying Central R a n d Group records terrestrial sedimentation in the foreland trough that was now peripheral not to a volcanic arc but an elevated plateau overlying thickened continental crust similar to that of present day Tibet. By this time the Zimbabwe Craton had collided with the Kapvaal craton at about 2840 Ma. "Tillite" facies of the Witwatersrand Supergroup can be better explained as subaqueous (West R a n d Group) and sub-

TECTONIC

SETTING

PERIPHERAL FORELANDTROUGH

WITWATERSRAND BASIN (CENTRAL RAND GROUP)

KAAPVAALCRATON W T IWATERSRAND BASN I

0 100 I km I

~LIIdPOPO COLLISIONALZONE---=-

WEST RAND GROUP

DO~ N IO IcANNREdF / "~" "

T -.-

U~3OOCr.AN

.,.a'-,_'~," ZIMBABWE CRATON

..,,,,,,/

+y

~sse~ ~

vv

.f

"<, ~ , 5 "_~"".,_:,"~,Z~'

~"~"~lli~ ~

~

P 4

2710

b

~

-

~APV'AAL CRATON

STRATIGRAPHY Ma

/z~ZIMBABW/ C~TON ~ ~

OUP

z84o M,

P

I

?090

/

?~-°''v:~ '1 ~ o ) /

~C'Rw AT~ON " ~

,UP~

............

~

A

",.

3120-3070

~, WITWATERSRAND

~

Ma "

BASIN

~O~R~AN

Fig. 7.4. Tectonic and depositional setting of West Rand (I) and Central Rand (II) strata, Witwatersrand Basin (after Burke et al., 1986). Dates are from de Witt et al. (1992). Adiabatic glaciers may have formed on tectonically thickened and uplifted margins of Witwatersrand Basin (see text).

57

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'ITING

diamictites of the Archean Pongola Sequence of southern Africa. These strata are dated between 3100 and 2940 Ma, and are probably correlative with the lower part of the Witwatersrand Supergroup (Armstrong et al., 1991; Beukes and Cairncross, 1991). Four diamictite units, from 3 to 80 m thick, occur within a thick mudstone interval of the Mozaan Group. Diamictites show a very wide range of clast size and lithology unlike those facies that result from the release of debris flows from tectonically-elevated fault scarps. Moreover, von Brunn and Gold (1992) report the presence of striated and faceted clasts. For these reasons a glacial origin must be seriously entertained. However, the occurrence of striated clasts in ancient mass flow deposits does not unambigously identify a glacial source; as has been stressed by Blackwelder (1930) and Winterer (1963), non-glacial striations are produced either by abrasion during flow or by differential rotation of clasts within a coarse matrix during metamorphism, von Brunn and Gold (1992) infer deposition along the rifted margins of a rapidly uplifting source terrane which could have provided the backdrop for the formation of local "adiabatic" ice masses contributing coarse and fine-grained glaciclastic sediment to surrounding basins. It is worth emphasising in this regard that de Wit et al. (1992) were able to identify rapid uplift of large magnitude (about 20 km) along the Limpopo belt around 2700 Ma. The association of allegedly glacial deposits with active orogenesis gives credence to an "adiabatic" model of glaciation but at that same time further emphasises the probable existence of steep basin margins, an abundance of volcaniclastic debris and the importance of sediment gravity flow processes. There are clear parallels with the depositional settings identified for many "tillites" deposited in Late Proterozoic and Late Palaeozoic active margin basins (e.g. Sections 10.4, 16.4). The problem common to all these basins is to positively identify any glaciclastic contribution to mass flow facies. Detailed comparison

of all these settings would be a very useful exercise given that the Archean tectonic history of the Kaapvaal craton, and related sedimentary basins, is being interpreted in terms of processes similar to those of present day subduction zones, mid-ocean ridges and continental margins (e.g. de Wit et al., 1992).

7.2 Europe and North America Archean diamictites, formerly regarded as tillites, occur in Karelia (the Pebozero "tilloids') and in the Baikal Mountains (the Ol'Khon Group) but are now regarded as sediment gravity flows with no glacial connection (Negrutsa and Negrutsa, 1981). In Montana, highly-metamorphosed diamictites, up to 60 m thick, occur in hornfelsed metasediments just below the base of the Stillwater Complex deposited sometime prior to 2750 Ma (Page, 1981). A glacial origin was suggested by Page on the basis of dispersed clasts in highly-metamorphosed, weaklystratified (foliated?) matrix material; surface features and any original orientation of clasts have been obliterated by metamorphism. The supposed glacial origin of these strata cannot be substantiated on the basis of the available data which in fact, suggests a non-glacial mass flow origin (Page, 1981, p. 822).

7.3 The problem of the missing Archean glaciations. To summarise what has been discussed in the preceding sections, it can be stated that no convincing Archean glacial deposit has been reported to date though data do not rule out the possiblity of local ice covers on uplifted crustal blocks in southern Africa (von Brunn and Gold, 1992). The dearth of Archean glacial deposits is unexpected because Archean crust has an excellent preservation potential as a result of being bonded to a deep, mantle lithosphere keel up to 200 km thick (Kerrich, 1992). Later, post-Archean mantle temperatures were insufficiently hot to generate such a deep keel. It could be

58

argued that so little is known of Archean global tectonics that the apparent absence of glacial sediments may reflect vastly different tectonic mechanisms and a lack of suitable settings in which glacial deposits could be preserved. This argument has some merit. For example, while Archean tectonics appear to have been no different in any major sense from that of the Proterozoic (Burke et al., 1986; Hoffman, 1989; Sivell and McCulloch, 1991; see below) the rate of tectonic activity appears to have been much faster. It is well known that the Archean was characterised by geothermal heat flows very much greater than at present as is indicated by the restriction of high MgO komatiitic lavas to Archean crust (Goodwin, 1991). Campbell and Jarvis (1984) have argued that hotter Archean mantle plumes necessitated much greater rates of oceanic spreading. In turn, Hargraves (1986) suggested that higher heat flows required a considerably greater length of oceanic spreading ridges compared with that at the present day. This would greatly reduce the size of Archean tectonic plates and any potential for climatic cooling in plate interiors thereby decreasing the likelihood of ice sheets. Thus it might be concluded that the absence of widespread Archean glaciation(s), if real, may be a product of the dominant style of plate tectonics. This conclusion is, however, directly contradicted by recent work emphasising "uniformitarian"interpretations of Archean tectonics (de Wit et al., 1992). The paucity of any well-defined Archean glacial record is also surprising because widely-accepted geophysical models indicate that the early Sun, prior to 3000 Ma, was much less luminous than at the present day. Under such conditions, glaciation should have been much more widespread in the Archean than is suggested by the stratigraphic record (the "faint young sun" paradox; Kasting et al., 1984; Kasting and Toon, 1989; Gerard et al., 1992). Estimates of the reduction in initial solar luminosity range from 25% (Endal and Sofia, 1981) to 40% (Bahcall and Ulrich, 1988; Fig. 7.1). If the

N. EYLES

Earth's early atmosphere had been similar in composition to that of today then there would have been permanent glaciation until at least 2000 Ma when the luminosity of the Sun is argued to have to increased. Some models suggest that even with a reduction in solar luminosity of only 5% there would be permanent glaciation (Sellers, 1990) thereby creating conditions similar to those on the planet Mars. If this had been the case, it follows that biological evolution of the planet would have been profoundly different. However, the absence of widespread Archean glacial deposits, together with the "normal" character of the 3800 Ma year-old sedimentary rocks preserved at Isua (e.g. Sagan and Mullen, 1972), show that the Earth was not permafrozen but was partially covered with water. The rock record indicates that Archean glaciation was in fact rare or absent. The apparent absence of Archean glacial deposits is puzzling from another standpoint. The period between 3000 and 2700 Ma is a key phase in Earth's history since it saw the development of large cratons composed of continental crust (Goodwin, 1991). Estimates of the amount of continental crust produced in this time interval are as high as 80% (Taylor and McLennan, 1985; Condie, 1989). Under these conditions, the release of large amounts of easily-weathered crustal materials could have consumed a significant fraction of atmospheric CO 2 resulting in a tendency for global cooling (see Section 21). The solution to the paradox of the "faint young sun" and the missing Archean glaciations could lie with enhanced global warming by atmospheric CO 2 and water vapour (Kasting, 1987). However, this hypothesis does not recognise the possibility that air temperatures could have fallen sufficiently far enough to have allowed the formation of carbon dioxide crystals and ice clouds (Caldeira and Kasting, 1992). These clouds would be very effective in scattering solar radiation thereby accelerating any tendency to permanent glaciation. This clearly did not happen and the Earth remained relatively warm in its

59

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'Iq'ING

early history. The major uncertainty in all these discussions is the real value of solar luminosity. New models, cogniscant of the dearth of Archean glacial deposits, now suggest that the early Sun had a much greater mass and solar flux than had been previously argued (Graedel et al., 1991). The value of solar luminosity may have been closer to that characteristic of the Phanerozoic and present day. Given that glaciation has been a common process over this relatively short period of Earth history, the problem still remains of finding glacial rocks in the much longer Archean. 8. PROTEROZOIC GLACIATIONS (2500-600 Ma)

8.1 Early Proterozoic glaciations (2500-1600 Ma) The first convincing evidence for a glacial-influence on deposition is preserved in the core of the North American and Baltic shield and is dated at c. 2200 Ma (Fig. 6.1). Glaciation took place against a background of a newly-oxygenated atmosphere combined with rifting and strong uplift along the margins of Archean cratons. Glaciclastic deposits appear to be of a subaqeous mass flow origin and are are preserved within thick turbidite successions; deposition along a glacially-influenced rifted margin appears likely. Other supposed Early Proterozoic glacial strata are reported from Australia, South Africa and India but these are highly metamorphosed and have restricted outcrops from which allimportant "contextual" evidence of a glacial origin is missing. Many of these deposits continue to be interpreted as glacial deposits simply on the presence of massive diamictite facies.

8.1.1 Africa In the Grigualand West and Transvaal basins of South Africa, diamictites, shales and sandstones of the Timeball Hill Formation are capped by andesite lavas dated at

2224 + 21 Ma (Visser, 1981). Strata show a crude cyclicity comprising massive diamictites over 50 m thick, overlain by stratified diamictites, pebbly sandstones, shales and iron-rich carbonate rocks. Though facies data are few, sandstones are deformed by downslope slumping which suggests a slope setting; rapid lateral changes in formation thickness (0-500 m) point to a considerable, perhaps fault-controlled, topography consistent with an early stage of rifting and later eruption of andesite lavas. It is difficult to agree with previous interpretations of these strata as a typical platformal association of lodgement tills, fluvioglacial and glaciolacustrine deposits that accumulated under very stable conditions. No striated clasts are reported and the crude cyclicity suggests repeated deposition of fining-upward, mass-flow sequences in a volcanically influenced rift setting. A glacial influence has not so far been substantiated.

8.1.2 Australia In the Hamersley Basin of western Australia, the Meterorite Bore Member has been interpreted as glacigenic based on striated and facetted boulders within massive diamictite. The latter outcrops discontinuously over some 30 km and is of constant (300 m) thickness indicating a remarkably tabular geometry (Trendall, 1981). Trendall inferred an ice-rafted origin but recognized that a glacial origin could not unambiguously be demonstrated. The stratigraphic context of the diamictite does in fact, suggest a mass flow origin since it occurs conformably within the upper part of 3 km thick section of turbidites (Kingarra Formation). An actively subsiding tectonic regime is suggested. The diamictite is only loosely dated as being between 2500 and 2000 Ma and deposition in a submarine slope setting appears likely. A distant glacial influence is suggested by striated clasts. The strata are, however, highly metamorphosed and boulders have been rotated into the plane of cleavage; this is a condition under which "tectonic" striations are known to be

60

N. EYLES

produced on clast surfaces (see Winterer, 1963).

8.1.3 Europe Uncertainty as to depositional setting also characterizes supposed Early Proterozoic (2450-2300 Ma) diamictites on the Baltic Shield. Evidence of a glacially-influenced setting is preserved in the Sariolian Group of eastern Finland which contains diamictites and turbidites with dropstones, having a total thickness of 60 m (Urkkavaara Formation). Marmo and Ojakangas (1984) described gradational contacts between turbidites and diamictites and suggested a glacial origin involving the "rain-out" of suspended fines and ice-rafted debris. These strata may be correlative with diamictite-bearing sequences in north Karelia and on the Kola Peninsula (Marmo and Ojakangas, 1984). It is worth noting that palaeomagnetic work on correlative Early Proterozoic strata in Scandinavia indicate high palaeolatitudes (Pesonen et al., 1989).

~

Paleozoi¢rocks

~

GrenvilleProvincerocks

m

SudburyIrruptivebelt

~ ~

WhitewaterGroup HuronianSupergroup (Fig. 8.2) Archeanrocks

~

8.1.4 North America; Ontario Classic Early Proterozoic glacial strata occur within the 10 km thick Huronian Supergroup in Northern Ontario, Canada (Fig. 8.1). These deposits have been described in considerable detail (Schenk, 1965; Casshyap, 1969; Lindsey, 1969, 1971; Long, 1981; Young, 1981; Miall, 1985; Young and Nesbitt, 1985; Mustard and Donaldson, 1987a, b) since a glacial origin was first identified (Coleman, 1907, 1908). The Huronian Supergroup, deposited sometime between 2500 and 2200 Ma (Krogh et al., 1984), has been subdivided into four groups each characterized by a succession of lowermost coarse clastic strata overlain by fine-grained turbidites that prograde upward into thick continental fluvial deposits (Fig. 8.2). These were deposited on the rifted, south-facing margin of the Superior craton. The Matinenda Formation occurs within the lowermost Elliot Lake Group and hosts uraniferous and pyritic quartz pebble placer conglomerates associated with thick volcanic sequences that mark the onset

I--~-"] Paleocurrenttrend

Fig. 8.1. Distribution of Huronian Supergroup strata in northern Ontario (after Young and Nesbitt, 1985).

EARTH'SGLACIALRECORDAND ITS TECTONICSE'Iq'ING

61

of rifting (Young and Nesbitt, 1985). The overlying stratigraphic groups (Hough Lake, Quirke Lake, Cobalt) commence with conglomerates and diamictites overlain by deep marine turbidites that shoal upwards. Strata thin rapidly to the north, and have been deformed by syndepositional faulting. A glacial (tillite) origin has been previously proposed for diamictites of the Ramsay Lake, Bruce and Gowganda formations though others have favoured deposition as mass flows without any glacial control (Card, 1978). A glacial source of debris can only be positively identified for the Cobalt Group

which contains the Gowganda Formation. This is the most extensive of the Huronian diamictites. Miall (1985) showed this unit to have been deposited as sediment gravity flows on a glacially-influencedsubmarine slope that prograded southwards by mass flow. No icecontact deposits were identified and a deep water setting peripheral to any ice margin was inferred. Mustard and Donaldson (1987a) suggested that some massive diamictite beds at the base of the Gowganda Formation are lodgement or melt-out tiUites but data are sparse and the interpretation is unconvincing given the stratigraphic context of

PALAEOZOIC

FORMATIONS

GROUPS

BAR RIVER

M

F

~o~ooN~t~

COBALT

GOWGANDA

///~F

I

SERPENT

QUIRKELAKE ESPAN~

~RUCE

HOUGH LAKE ~

Y

~

I,~KIM

ELLIOT LAKE

ARCHEAN

c~To.

PARDSOURCEAREAUPLIFT,BASINALSUBSIDENCEANDSEDIMENTPROGPADATION ~

CONGLOMERATE

~

DIAMICTITES

M

TURBIDITES F~

I~

BRAIDEDFLUVIAL ~

VOLCANICS CARBONATES

=F : PROGPADATIONAL,MARINEMASSFLOW(M)TOFLUVIAL(F) CYCLES

Fig. 8.2. Tcctono-stratigraphic "cycles" of the Huronian Supergroup recording

repeated

rifting and tectonic

differention of uplifted source areas, shedding of coarse elastics and basinal subsidence. Stratigraphy after Young and Nesbitt (1985). A record of glaciation (adiabatic?) is preserved in the Gowganda Formation. Compare with Figs. 4.7 and 12.2.

62

N. EYLES

diamictites with graded pebbly sandstones; a typical subaqueous mass flow facies association. This is suggested further by Mustard and Donaldson (1987b) who invoke an iceproximal subaqueous fan setting for deposition of overlying Gowganda strata. Massive and crudely stratified diamicties were interpreted as "undermelt" deposits recording the dropping of basal debris from a floating ice tongue. They remarked, however, on the lack of deformational effects normally associated with ice-contact environments (e.g. Fig. 3.9) and a debrite origin appears much more likely; associated fine-grained facies contain dropstones so a glacial source is probable. Miall (1985) argued that previously recognized glacier advance/retreat cycles inferred from alternation of debrites and turbidites in the Gowganda Formation were simplistic given the wide range of possible controls on sediment delivery and downslope mass flow other than climate. Jackson (1965) showed that previously reported "varvites" had no seasonal control (Fig. 4.7) The three diamictite-bearing groups of the Huronian Supergroup are clearly tectoni-

cally-generated depositional successions (Figs. 8.2, 8.3). Each succession appears to be the sedimentary response to tectonic subsidence and complementary source area uplift. They closely resemble the 1 km thick stratigraphic cycles described from the Late Proterozoic Windermere Supergroup of northwestern Canada by Eisbacher (1985; Section 12.1, Fig. 12.2). Windermere strata consist of thick (00's m) deep water mass-flow diamictite successions that pass upward into shallow marine facies; these successions were explained by reference to repeated glacio-eustatic changes in sea-level by Eisbacher (1985). It is argued below (Section 12.1.3) that a more realistic control is that of repeated large-scale foundering of a rifted craton margin. In this regard, the regional downdip geometry of the Huronian Supergroup suggests repeated "backstepping" of the margin of the Superior craton. A glacial influence can be unambiguously demonstrated for the Gowganda Formation only; underlying diamictites and conglomerates are subaqueous in origin and may have originated as talus debris reworked from degrad-

BACKSTEPPING OFRIFTEDMARGIN

)PLIFTAN{)UNROOFING B

A

......

/

COVERROCKS

~ ....... CYCLEBOUNDARY I ~ ] FLUVIALFACIES PRO~A'noN OF ~ S I - i A U ~ . O W MARINEFACES SUBA(~EOL~FANS ~ T U R B I D I T E ~,1 SOURCEAREAUPLIFT& ~ INFLUXOFDEBRISFLOWS

MASSFLOW _C_OMP_L_EX__ CYCLEBOUNDARY

FACIES COMPONENTS OF TECTONO-STRATIGRAPHIC SEQUENCES A,B,C

Fig. 8.3. Formation of "tectono-stratigraphic" cycles along a rifted, glacially-influenced basin margin. Compare with Figs. 4.7, 8.2 and 12.2. Each cycle starts with basin subsidence and source area uplift followed by an influx of subaqueous mass flows (diamictites) and a prograding turbidite succession. The cycle is completed by shallow marine and subaerial braided river facies.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SET'FING

ing fault scarps (e.g. Nemec et al., 1984; Nemec and Steel, 1988; see below). Regional uplift along boundary faults may have been an important trigger dictating the development of Gowganda glaciation along the southern margin of the Superior craton (see below; Section 8.1.5). The Huronian Supergroup lies on the southern margin of the craton but north of Archean crust exposed in the southern Lake Superior region; Karlstrom et al. (1983) inferred an Atlantic-type passive margin, Goodwin (1991) suggested a rifled intracratonic setting for sedimentation. "Adiabatic" formation of ice caps along rifled margins is shown below to have been a very important mechanism in Late Proterozoic glaciated basins (Section 12.5). Global climate controls cannot be ruled out, however, because the appearance of reddish colouration in the upper Gowganda Formation can be interpreted as evidence of oxygenation of the atmosphere (Roscoe, 1973); the Gowganda Formation and the Roosberg Group in South Africa, dated at between 2500 and 2100 Ma, are known to span this transition (Twist and Cheney, 1986; Fig. 7.3). Increased oxygenation and associated CO 2 drawdown (Fig. 7.2) may have been the key precursors to Huronian glaciation. Yet another influence acting to promote glaciation may have been the opening and propagation of rifled marine basins within a large and hitherto arid continent (Young, 1988). Increased precipitation may have been directed to ice caps growing along the rapidly rising margin of the Superior Craton as has been inferred from Late Proterozoic settings (Yeo, 1981). 8.1.5 Northwest Territories, Quebec and the United States Strata that are possibly correlative with the Huronian Supergroup of Ontario, occur elsewhere in Canada, in the Northwest Territories and Quebec, and in the United States in Wyoming, South Dakota and Michigan. All these strata lie on the margins of large Archean cratons and have been interpreted as recording the "rift to drift" transition

63

during active uplift accompanying the breakup of the cratons (Young, 1988). Many are associated with banded iron formations. The Hurwitz Group of the Northwest Territories contains the Padlei Formation (500 m) composed of lowermost massive diamictites, subordinate conglomerates and sandstones. These facies are overlain by a succession of turbidites, up to 200 m thick. The Padlei Formation occupies narrow depressions in the underlying basement; overlying units overstep older strata suggesting progressive subsidence of fault-bounded basins. No striated clasts are reported and a nonglacial mass flow origin was suggested by Bell (1968); later workers have favoured a glacial origin on the basis of isolated clasts, inferred as ice-rafted (Young, 1970) within turbidites described as "varve-like" (Young and McLennan, 1981). The Hurwitz Group, together with the underlying Montgomery Group strata, contains three crudely-developed cycles each commencing with coarse-grained facies (with volcanics in cycle 3) with successively younger cycles having progressively thicker fine-grained intervals that are overlain by fluvial sandstones. At face value, this probably records punctuated but accelerating subsidence of a continental margin (e.g. Fig. 8.3). This is borne out by the limited distribution of the older strata and the successively greater extent of younger strata of cycles 2 and 3. The Hurwitz Group is variably dated between 2250 and 1770 Ma (Young and McLennan, 1981) or older than 2100 Ma (Patterson and Heaman, 1991). The precise age relationship between the Montgomery and Padlei groups is unresolved (Young, 1975) but there are strong similarities between the stratigraphic cycles of the Padlei and those of the Huronian Supergroup suggesting a common plate tectonic control involving deposition along rifting plate margins (Young, 1988). Correlation of the Gowganda and Padlei formations is further supported by petrographic data (Young and McLennan, 1981). Strong similarity with strata of the

64

Amer Belt and the closely associated Montressor Belt, lying some 300 km north of the Hurwitz Belt, has been emphasised by Young (1988) Long (1974) described interbedded complexes of conglomerates, graded sandstones, diamictites and laminites from the correlative Chibougamau Formation of northern Quebec; these were interpreted as the product of alluvial fans at the margin of an extensive Early Proterozoic ice sheet (see also Long, 1981). Striated clasts are not present, however, and a subaqeous, nonglacial origin as sediment gravity flows appears probable. In Wyoming, within the Cheyenne Belt, the Headquarters Formation of the Snowy Pass Supergroup has been interpreted as glacigenic (Karlstrom et al., 1983). The lower part of the supergroup was deposited in a large intracratonic rift basin. Elsewhere in southern Wyoming, diamictites of the Campbell Lake and Vagner formations occur within thick laminated phyllites. The latter facies are probably metamorphosed turbidites, some of which have "scattered clasts ..... with structural relationships like dropstones" (Houston et al., 1981, p. 797). No soled or striated clasts have been found, however. A glacial interpretation for the diamictites, using the ice-shelf model of Carey and Ahmad (1960), and correlation with the Gowganda Formation was proposed by the above workers. The age of the diamictites is not precisely known, however, and a glacial origin is not immediately obvious from the data published to date. The same qualifications apply to poorly exposed and highly deformed ',metacong!omerates" of the eastern Black Hills of South Dakota that are between 2560 and 1620 Ma in age; no definite glacial indicators are reported by Kurtz (1981). In northern Michigan, laminated slates with dropstones and associated diamicrite facies occur in the Fern Creek, Enchantment Lake and Reany Creek formations. These rocks occur within the Animikie Group toward the base of the Marquette Range Supergroup and are dated to around 2000-

N. EYLES

2100 Ma (Gair, 1981). Lithostratigraphic correlation of these strata with the Huronian of Ontario was proposed by Young (1983). Young (1988) has argued that the Early Proterozoic strata listed above are correlative (see also Young, 1970; Chumakov, 1985; Ojakangas, 1988) and comprise a distinct tectonostratigraphic assemblage recording the breakup of Archean cratons. He showed that the assemblage is characterised by lowermost coarse grained "rift" facies, with glacial indicators, overlain by deep water "passive margin" turbidite successions that in turn, prograde into shallow water facies. Iron-rich sedimentary rocks, particularly turbidites, are present in many of the basins and record the background precipitation and "rain-out" of iron oxides in deep water (e.g. Eriksson, 1983). 8.2 Global cooling in the Early Proterozoic? Young (1991) argued that widespread Early Proterozoic glacial deposits were evidence of a global "anti-greenhouse" effect caused by weathering of labile minerals and a drawdown in atmospheric CO 2 pressures (e.g. Fig. 7.2). In this model, the accretion of large continental nucleii during the Archean and Early Proterozoic kick-started an important Earth surface process involving the weathering of labile minerals (e.g. feldspars) in clastic sediments. Enhanced weathering brought about a reduction in atmosphere CO 2 partial pressures thereby allowing widespread glaciation (Young, 1991). Chemical evidence of deep weathering in the Huronian Supergroup, consistent with the model, was presented by Nesbitt and Young (1982). The advantage of the CO 2 model is that it could also explain the long apparently nonglacial periods in Earth history prior to, and after, 2200 Ma (Fig. 6.1). Widespread magmatic activity in this interval, perhaps accompanying large scale changes in the mantle (Hoffman, 1989), may have contributed to high CO 2 levels and greenhouse warming (Tajika and Matsui, 1992).

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T I ' I N G

8.2.1 Global tectonics, oxygenation of Earth's atmosphere and glaciation Best estimates of atmospheric oxygen over geologic time using many different data sets are shown in Fig. 7.3. Oxygen contents of the early atmosphere can be constrained by reference to sulphur isotopes because isotopic fractionation is generated by sulphate-reducing bacteria (Lambert and Donnelly, 1991). 345 trends in strata older than 600 Ma show narrow ranges for Archean sulphides and sulphates consistent with direct deposition from hydrothermal sources. The first evidence of bacterial sulphate reduction occurs only after 2300 Ma when mean values show large excursions from values typical of the Early Archean consistent with increased free oxygen and the growth of photosynthetic organisms. Other support for this model is derived from the absence, after 2300 Ma, of banded iron formations and detrital uraninite placers together with the increased stratigraphic importance of red beds and copper and uranium ores which require oxidizing conditions (see Lambert and Donnelly, 1991 for discussion; Fig. 7.3). Work on palaeosols by Holland and Beukes (1990) also identifies rapid increases in atmospheric oxygen after 2300 Ma. Further constraints on atmospheric oxygen levels are provided by studies of the biochemistry and physiology of early eukaryotes such as Grypania together with Ediacaran fauna such as Dickinsonia (Runnegar, 1991). Kasting (1991) concluded that the early atmosphere switched from reducing to oxidizing states over a narrow time interval of about 150 Ma sometime around 2000 Ma. It is perhaps significant that this timing is coincident, within the limits of current age dating, with the first unambiguous record of glaciation in Earth history contained within the Gowganda Formation of northern Ontario. Despite uncertainty surrounding the precise timing and extent of Early Proterozoic glaciation, the apparent coincidence of Huronian glaciation and rifting with a stepwise increase in the amount of free oxygen in

65

the Earth's atmosphere invites continued research. This is a potentially fruitful area of investigation because new work increasingly shows the central importance of global tectonic cycles to progressive oxygenation of Proterozoic environments (Des Maris, 1992). It has been a common assumption that expansion and evolution of the planet's biosphere, involving the splitting of water molecules during photosynthesis to produce molecular oxygen (O2), was the principal cause of oxygenation of the atmosphere/ocean system. In turn, the innovation and rapid evolution of life forms provided an important regulatory control on the entire system (e.g Lovelock, 1979). The record of oxygenation of Proterozoic environments between 2500 and 540 Ma has been identified by tracking the growth of the crustal organic carbon reservoir (Des Marais et al., 1992). Well-defined changes in the carbon isotope composition of carbonate and sedimentary organic carbon testify to marked increases in the burial of organic carbon (Fig. 7.3). These data strongly suggest that step-wise increases in oxygenation resulted not from biological innovations but from geological events involving rifting on a global scale, the formation of large anoxic basins and enhanced carbon burial thereby releasing 0 2 for oxygenation of the atmosphere. The first of these tectonically-controlled oxygenation events started at around 2200 Ma and lasted until 1700 Ma (Des Marais et al., 1992; Fig. 7.3). The earliest well-constrained glaciation (Huronian) coincided with the start of this event. Des Marais et al. (1992) suggest that glacio-eustatically lowered sea-level might have in turn, exposed more of the continental surfaces to erosion but once again, the precise configuration and dimensions of such ice masses can only be guessed at. Certainly the sedimentary record is not extensive (Section 8.1) but the geochemical model of Des Marias et al. (1992) allows for widespread glaciation as a result of the drawdown of atmospheric carbon dioxide. The onset of widespread oxidation, enhanced

66

N. EYLES

clastic sedimentation, weathering and soilforming processes could draw down CO 2 partial pressures to a threshold value where glaciers could grow on uplifted rift margins (Young, 1991; Fig. 8.3). Kasting (1987) estimated that CO 2 partial pressures were between 0.03 and 0.3 during the Huronian glaciation (Fig. 7.2) but this is based to a large extent on the assumption of a continental-scale ice cover. This model must be qualified in that the glacial status of strata in

Wyoming, South Dakota and South Africa is not proven and a definite glacial origin can only be demonstrated in Ontario (Gowganda Formation) and Scandinavia (Urkkavaara Formation). These strata appear to have been deposited as mass flows in a glacially-influenced marine environment along rifted margins (e.g. Karlstrom et al., 1983; Zolnai et al., 1984; Fig. 8.2). It is also noted that the palaeomagnetic data base whereby Laurentia and Baltica are placed adjacent to each other

• 8.1,8.2

R OUGAMAU TRESOR

Hurwitz Animikie

Huronian

Cheyenne

~3

_'--D_" _

\

\

\

DLG

////

\ \ \ . \

RIFTED DIAMICTITE

ARCHEAN BASEMENT •

IRONFORMATIONS ~

VOLCANICS

AG; ANIMIKIEGROUP HQF; HEADQUARTERSFORMATION PF; PADLE,FORMATION DLG ; DEEPLAKE GROUP MG ; MONTGOMERYGROUP

Fig. 8.4. Schematic distribution and correlation of Early Proterozoic strata containing alleged glacial indicators. Strata lie in the margins of large Archean cratons and record rifting and the development of passive margin elastic wedges (e.g. Fig. 8.3). Uplift of rift shoulders and the penetration of marine waters into a hitherto arid continent may have set the scene for "adiabatic" glaciation. After Young (1988).

67

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

(e.g. Piper, 1983) is also sparse. An alternative model for Early Proterozoic glaciation may be one of small and widely-scattered glaciers growing on uplifted rift shoulders on the margins of broken cratons (e.g. Figs. 8.3, 8.4), rather than continental ice sheets responding to a global geochemical control. Some combination of local initiation and global feedback may have operated. A model of enhanced orogenic activity and drawdown of atmospheric carbon dioxide is very similar to that proposed by Schermerhorn (1983) for Late Proterozoic glaciations (Section 9). Indeed, the Late Proterozoic was again characterised by a well-defined increase in the quantity of organic carbon preserved in the crust (Fig. 7.3) and renewed glaciation. This followed a long mid-Proterozoic episode of non-glaciation, (mid-Proterozoic non-glacial epoch) from about 2000 to 1000 Ma, which coincides with a decrease in the quantity of organic carbon in the crust, lessened global tectonic activity and relatively stable atmospheric oxygen contents. This long non-glacial interval may be the product of extensive magmatism amd outgassing of CO 2 producing a long-lived green-

house condition. Hoffman (1989) emphasises that this period was characterised by rapid growth of continental crust and elevated mantle temperatures. Possible mid-Proterozoic glaciclastic strata have been described from the former Soviet Union (Chumakov and Krasil'nikov, 1992) but their age is imprecisely determined. 9. LATE P R O T E R O Z O I C GLACIATIONS

"The farther back we go in time the more difficult must it become to detect evidence of ice action. The older systems consist for the most part of deposits which gathered on the floors of ancient oceans" (Geikie, 1896, p. 810). Late Proterozoic glaciclastic strata are reported on all continents including Antarctica (Stump et al., 1988). The widespread distribution of Late Proterozoic diamictites (Fig. 9.1) has stimulated intense debate concerning palaeoclimates and tectonic setting. Paleomagnetic data purporting to show low depositional paleolatitudes (Harland and Bidgood, 1959) led to the notion of a great

Continental Cratons ~---]Orogenic

Belts F--~Glaciclastic

strata

Fig. 9.1. Distribution of Late Proterozoic (c. 800-550 Ma) glaciclastic sediments. After Dott and Batten (1988) and N. Eyles (1990).

68

"infra-Cambrian ice age" (Harland, 1964) involving synchronous growth of continental ice sheets extending into equatorial latitudes (Tarling, 1974; Abouzakhm and Tarling, 1975; Urrutia-Fucugauchi and Tarling, 1983). A corollary of this thesis is that "tillites" can be used for high-resolution stratigraphic correlations from continent to continent (e.g. Chumakov, 1981b, 1985). Reviews of Late Precambrian glaciogenic strata, emphasizing their value for high resolution stratigraphic correlations, were made by Saito (1969) and Chumakov (1981); in contrast, Crawford and Daily (1971) argued that such strata had a limited use as time-stratigraphic markers. Diachronous glaciations in high plaeolatitudes was suggested by McElhinny et al. (1974). In an extensive rebuttal to the notion of a global Late Proerozoic glaciation, Schermerhorn (1974, 1975) argued that many deposits were non-glacial but his arguments were not generally accepted (e.g. Roberts, 1977). Morris (1977) re-emphasised the warning of Crawford and Daily (1971) and rejected the notion of a single, globally synchronous and catastrophic glaciation. Support for a global Late Proterozoic refrigeration is traditionally found in the common stratigraphic association of alleged tillites and dolomites of supposed warm water origin (the "dolomite/tillite paradox"). Many diamictites are indeed either underlain by thick dolomite successions (e.g. Greenland, Norway, Sweden, Scotland and the former USSR), contain dolomitic interbeds (Scotland, Norway) or are capped by dolomite (N. Africa, N. America, Australia). There are many explanations for these relationships and the rapid climatic changes which are suggested. Williams (1975) rationalised the apparent paradox of intimately associated warm and cold indicators by reference to extreme variation in climate resulting from increased obliquity of the earth's tilt. Hambrey (1983) suggested deposition of dolomite in waters "close to freezing" whilst Sheldon (1984a, b) argued for an ice ring circling the equator similar to that possessed by Saturn. Salop

N. EYLES

(1977) identified a supposed relationship between glaciation, outbursts of radiation from supernovae and the shielding of incoming solar energy by gas and dust (see also Hoyle, 1981). Malcuit and Winters (1986) argued that low latitude refrigeration resulted from the lengthening of the seasonal cycle, to thousands of years duration, brought about by an increase in the rate of precession of the Earth's orbit due to Earth-Moon interaction. "Anti-greenhouse" global cooling caused by low CO 2 partial pressures in the atmosphere has also been proposed by several workers (Roberts, 1971, 1977; Fischer 1986; Section 21). The model of global Late Proterozoic glaciation with equatorial ice sheets has not found universal support. It has been suggested that low paleolatitude determinations are the result of tectonic overprinting (Piper, 1981; Stupavsky et al., 1982) or, if real, are the result of rapid postdepositional plate movement into low palaeolatitudes (Crowell, 1983; Meert and Van der Voo, 1992). Crowell (1983) called for further research on rock and mineral magnetics with respect to diagenetic changes. Schermerhorn (1983) suggested that low latitude glaciers could locally develop under conditions of strong regional uplift combined with depletion of atmospheric CO 2 brought about by accelerated weathering of newly uplifted and exposed crust. Some examples of palaeomagneticallyconstrained palaeolatitudes appear to be geophysically sound, however, and allow for the possibility of glacial deposition in mid- to low-paleolatitudes (Chumakov and Elston, 1989; Schmidt et al., 1991). Indeed, whilst continental reconstructions differ from worker to worker (see below) it is widely agreed that most continents occupied a near equatorial to mid-latitude position during the Late Proterozoic such that low to mid-latitudinal glaciation must be accounted for in any geodynamic model of Late Proterozoic glaciation (Fig. 13.1). The results of recent general circulation models emphasise the importance of palaeogeography to global climate and

69

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

show that global temperatures are lowest at times when continents straddle the equator (Section 21.3). The nature of the Late Proterozoic glaciations is currently one of the most enigmatic problems in Earth history. It is possible, however, that the problem can be made much more ordinary but no less interesting, by close consideration of the timing of Late Proterozoic glaciations and their plate tectonic setting.

literature (e.g. Piper, 1985; Hoffman, 1991; Moores, 1991; Van der Voo and Meert, 1991; Dalziel, 1992). The following section makes several assumptions regarding Late Proterozoic paleogeography and is based on the supercontinent model of Hoffman (1991; Fig. 9.2). It is proposed herein that Late Proterozoic glacial deposits, preserved between c. 800 and 550 Ma can be subdivided into two tectonostratigraphic types. These record glaciations under contrasting tectonic and depositional settings coeval with Late Proterozoic supercontinent amalgamation and subsequent disintegration. The long-lived amalgamation phase of the Late Proterozoic supercontinent is recorded by active margin basins in compressive tectonic regimes during the interval 800-550 Ma. Basin fills are now preserved as tectonically deformed stratigraphic slivers in mobile belts that

9.1 Tectono-stratigraphy of Late Proterozoic glaciated basins There is no general agreement as to any one single model of Late Proterozoic continental configuration. Conflicting opinions regarding the existence of a Late Proterozoic supercontinent and the paleolatitudes of its component parts have been voiced in the

Dpre-GrenvillianBeltsCratons

~Grenvillian

~

Glacially-Influenced Basins

~9

MARGIN ( ~ :, 12.4) ~70 Ma ~ -

0

~

O'J

0° GLACIATION 0 DURING COLLISIONA 7_. EVENTS •~, < 8 5 0 M~

It,, tY ...J "ED

~TLANTIC

~,

:lG. 1L1) 0 ) Ma

¢. 7


,p

IF~

~" tO /',-

d' Fig. 9.2. Plate tectonic setting and age of principal Late Proterozoic glacial strata. Late Proterozoic supercontinent is depicted according to model of Hoffman (1991). Note location of ice centres along opposing Palaeo-Atlantic and Palaeo-Pacific rifted margins of Laurentia and along active subducting margin (Avalonian-Cadornian Belt). Local cordilleran ice masses developed during the amalgamation phase(s) of the supercontinent commencing after 850 Ma.

70 record ocean closure between colliding cratons (Fig. 9.2). Deposits record local glaciation of volcanic arcs and cordillera and the downslope reworking of glacial and volcanic debris into deep marine basins. The most characteristic deposit in such settings is the diamictite/turbidite facies association (e.g. Fig. 4.7). The second and most widespread type of tectonostratigraphic succession accumulated in extensional tectonic regimes during a protracted phase of supercontinent disintegration (c. 750-600 Ma; Fig. 9.2). The picture emerges of local ice caps forming on the rapidly uplifted flanks of rift basins. In these settings glacial marine deposits are preserved at the base of and within, thick syntectonic clastic wedges recording active source area uplift (e.g. Fig. 8.3). Ice contact deposits (e.g. tillites) were not widely preserved. Many different configurations have been presented for the Late Proterozoic supercontinent (e.g. Bond et al., 1984; Piper, 1985; Hoffman, 1991; Moores, 1991; Torsvik et al., 1991) but irrespective of the precise palaeogeographic model used, Laurentia occurs as a "keystone" in the supercontinent (Fig. 9.2). The break-out of Laurentia by rifting, starting at c. 750 Ma along its Palaeo-Pacific margin is marked by Sturtian and Marinoan glacial deposits of Australia and equivalent strata in China and western North America. As emphasised by Signor and Moores (1991) in their review of the biogeographic evidence for supercontinent breakup, Australia and Antarctica were the first continents to rift away from Laurentia. This is recorded by the "early" development (c. 750 Ma) of Late Proterozoic glaciation along the uplifted margins of the Adelaide Geosyncline in Australia, the Yangtze Platform in China and along the western margin of North America. Rifting along the opposite (Atlantic) margin of Laurentia only began much later, about 640 Ma, along an early Iapetus seaway (Fig. 9.2). Rifting and complementary elevation of glacial source areas is recorded by widespread glaciomarine and marginal marine

N.EYLES glacial deposits of Greenland, Spitsbergen and Scandinavia.

Scotland,

10. THE GLACIAL RECORD OF LATE PROTEROZOIC BASINS ALONG ACTIVE PLATE MARGINS Collisional, active margin basins developed during the accretionary phase(s) of the Late Proterozoic supercontinent and have been sub-divided on the basis of age and spatial distribution by Murphy and Nance (1991). These workers categorized the orogenic basins into two fundamental types. The first are those within interior orogenies recording continent-continent collisions after 850 Ma that were responsible for cratonization of the supercontinent (e.g. Brazil and southern Africa; Fig. 9.2). The second type, dated at c. 600 Ma and thereafter, is associated with younger peripheral orogenies along the outer margin of the newly assembled supercontinent (e.g. Avalonian-Cadomian Belt; Fig. 9.2). Both these settings are characterized by voluminous volcanic-sedimentary piles many kilometres in thickness. Such deposits are widespread in Africa, Brazil, Saudi Arabia and eastern North America and contain prominent diamictite horizons within thick turbidite successions; a classic Late Proterozoic facies association. The term "tillite" has been widely applied to such strata. Locally, debris flows contain glaciclastic sediment and clasts recording cordilleran glaciation of volcanic arcs but in most cases sediment gravity flows reworked pyroclastic and volcaniclastic debris on the subaqueous flanks of volcanoes with no glacial influence. In those cases where glaciation can be demonstrated it is interesting to speculate about the possible interaction of volcanism, climate and the increased tendency for glaciation during "volcanic winters." Such interactions are suggested by data from Quaternary glaciations (e.g. Rampino and Self, 1992; Section 20.3) but remain untested for Late Proterozoic active margin basins.

71

EARTH'S GLACIAL RECORD AND ITS TECTONIC SETTING

_20 o ATLANOCE~ P

- - I 5° DAI~

'-~

MAURITANIDES OROGENIC BELT

A

B NAURITANIDES

TAOUDENI FORELAND

Fig. 10.1. Tectonic setting of glaciationwithin the West African (Taoudeni) foreland basin. After Robineau (1990). For location of section A - B see Fig. 10.2. Collisional tectonics also created large foreland basins where glacial sediments were preserved in tectonically-downwarped cratons. The extensive glacial deposits of the West African craton were deposited in such a setting (Figs. 9.2; 10.1). 10.1 North Africa Late Proterozoic active margin basins are now preserved as mobile belts that comprise the "welds" between several cratonic blocks. A critical event was the collision and suturing of the West African and Guyana shields at about 675-650 Ma (Murphy and Nance, 1991). This was an essential precondition for the onset of extensive, predominantly conti-

nental glaciation across the West African Craton. At this time West Africa lay close to the South Pole and may have been particularly susceptible to glacierization triggered by uplift in areas of active orogenesis ("adiabatic glaciation"). The same model can be applied to Cenozoic glaciation of Antarctica beginning at 36 Ma (Section 19.1). Caby and Fabre (1981) identified the role of Late Proterozoic collisional orogeny in the growth of Cordilleran glaciers on the uplifted margins of the West African Craton. They emphasized the syntectonic downslope reworking of volcanic and allegedly "glacial" sediments within a wide range of active margin basins now preserved as part of the Trans-Saharan fold belt. These strata are not well known but appear to share similar characteristics with co-eval deposits of the AvalonianCadomian Belt now preserved in eastern North America and Europe (N. Eyles, 1990; see Section 10.4). Recent work within the Anti-Atlas Mountains on the the northern margin of the West African Craton in Morocco has documented the collision and suturing of a volcanic arc along the edge of the craton resulting in the destruction of a forearc basin between 565 and 615 Ma (Hefferan et al., 1992). These workers show that diamictites previously interpreted as tillites within the Tiddiline Formation (Leblanc, 1975, 1981; Caby and Fabre, 1981) comprise part of thick (up to 1600 m) synorogenic conglomerates deposited by sediment gravity flows. Diamictites contain large amounts of volcaniclastic debris; analysis of clast size shows a well-defined decrease away from basin margins that is consistent with a debris flow origin. Associated facies include thick sand and silt turbidites identifying a subaqeous setting. The Tiddiline Basin provides a good example of the evolution of a forearc basin in an oblique-convergent collisional setting. These basins are characterised by the migration of depocentres in front of the moving thrust sheets; thrust faulting results in uplift of successively younger strata and their extensive cannibalisation. Under

72

N. EYLES

these conditions, large fluxes of coarsegrained sediment by sediment gravity flow processes provide ample scope for misinterpretation of debris flows as tillites (N. Eyles, 1990). 10.1.1 Glaciclastic sedimentation in a foreland basin; the Taoudeni Basin In contrast to the non-glacial debris flows of the northern margin of the West African Craton, undisputed glacial deposits of piedmont ice sheet lobes occur across the Taoudeni Basin (Fig. 10.2). The basin covers more than 2 × 106 km 2 of the craton. Extensive, predominantly glacioterrestrial facies accumulated as part of an extensive molassic fill within the West African foreland basin formed by crustal loading below easterly-directed thrust sheets (Fig 10.1). Exposures only occur around the upturned "rim" of the basin. Recent work within the Mauritanide,

Z (.9 r~

<~

Bassaride and Rokelide orogenic belts (Villeneuve and Dallmeyer, 1987; Villeneuve, 1989; Robineau and Ritz, 1990) and new age dating of the Late Proterozoic glacial Jbeliat Group reveals a clear relationship between orogenesis of the western margin of the West African craton and glaciation. The key event is the Pan-African I orogenic event at 660 Ma. Glacial deposits of the Jb61iat Group postdate Rb-Sr dates on clay minerals of 630 and 595 Ma (Clauer and Deynoux, 1987). The restricted thickness of the Jb61iat Group (50-150 m) across the Taoudeni Basin is consistent with a shallowly-downwarped foreland basin on the eastern margin of an active plate margin (Figs. 10.1, 10.2). The Jb61iat Group contains an impressive suite of terrestrial glacial landforms that are unmatched anywhere in the rock record outside of the Late Pleistocene (Deynoux and Trompette, 1976; Deynoux et al., 1978; Dey-

>late Convergence lferred ~acial movement

ce cap centers

L~

LACIAL FACIES ~larine

.-J CL

~'ontinental and Shallow ~arine

~lainly continental

Z

tJJ

(.9 Cz: LU ~Z I ~

O km

500 I

Fig. 10.2. Palaeogeography and distribution of glaciclastic facies across the Late Proterozoic West African Foreland basin. A - B shows location of structural section on Fig. 10.1. After Deynoux (1981).

73

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T T I N G

honoured fashion, are simply interpreted as "tillites" without detailed supporting evidence. However, diamictites appear to be lateral basinal equivalents of lacustrine deltaic strata of the Bouenza Group suggesting a subaqueous origin.

noux, 1985b). Because of their largely passive tectonic setting, these deposits escaped being reworked and are preserved below a thick (1500 m) succession of Cambrian to Ordovician age (Supergroup 2 of Deynoux and Trompette, 1981) recording a later phase of subsidence and sediment accomodation. The Jb61iat Group is capped by baryte-rich calcareous dolomite, and marine shales of the Teniagouri Group that record a basin wide transgression that allowed the development of a restricted and probably clastic-starved, shallow marine shelf. Thicker correlative glacial marine deposits possibly associated with Cordilleran glaciers along the Dahomeyides orogenic belt occur in the Volta Basin and in Sierra Leone (Trompette, 1981; Tucker and Reid, 1981a, b). Poorly dated, but probably coeval Late Proterozoic diamictites are reported from the Comba Basin in the Congo which developed as an intracratonic rift basin within the foreland of the Mayombe mobile belt. Diamictites ("Tillite superieure"; Alvarez and Maurin, 1991) were deposited at a time of rapid subsidence. Unfortunately, there are no detailed facies descriptions of these strata which, in time-

STRATIGRAPHY

SEA LEVEL -

10.1.2 Controls on deposition in glaciated foreland basins The extensive Late Proterozoic glacial deposits of West Africa were preserved in a foreland setting marginal to an active orogenic belt; previous work has tended to perpetuate the notion of glacial deposition across a "stable platform" but new data from Mall shows the importance of basin tectonics to sedimentation and preservation. Thin (~ 50 m) continental glacial deposits of the Jb61iat Group pass southwards toward the northern margin of the Leo Shield into thicker (~ 500 m) deposits of the Bakoye Group. In Mali, the Bakoye Group can be divided into four depositional sequences defined by fluvially-cut unconformities overlain by alternations of continental glacial and aeolian facies capped by transgressive marine deposits (Deynoux et al., 1989). Sequence

FACIES

I +

I

PARASEQUENCE

[-'--]

REGRESSIVEMARINE

. . . .

RAVINEMENTSURFACE

RSENING UPWARDS

200 m

0 0

20 km

Fig. 10.3. Stacked parasequences of the Late Proteroaoic Bakoye Group in Mali (modified fror Deynoux, 1991). Each parasequence is composed of lowermost transgressive marine strata overlain by regressive marine strata capped by continental glacial and aeolian facies. See Fig. 10.2 for location.

74 preservation is clearly the result of episodic tectonic downwarping, subsidence and flooding of the basin. Each depositional sequence can be divided in turn into small-scale parasequences about 50 m thick (Fig. 10.3) argued to record short-term ( ~ 100 ka?) changes in relative sea-level produced by repeated glacio-isostatic depression of the basin (Deynoux et al., 1989; Proust et al., 1990; Deynoux, 1991). Interpretation of a glacio-isostatic control on deposition of the parasequences of the Bakoye Group is not straightforward. It should be noted, by analogy with well-studied Pleistocene examples, that basin subsidence due to glacio-isostatic loading is short-lived and is fully recovered at the end of the glacial cycle when the regional ice load is removed (see Boulton, 1990; N. Eyles and C.H. Eyles, 1992c). Such subsidence by itself cannot create long term accomodation space for sediments because the basin subsequently rebounds and widespread erosion surfaces are created. Moreover the spatial pattern and extent of glacio-isostatic subsidence across glaciated basins is known to be highly complex such that some areas experience uplift associated with forebulge effects (see Scott et al., 1987 and refs. therein). Therefore, unless there is an accompanying background component of tectonic subsidence, repeated glaciation and loading cannot give rise to a simple succession of stacked parasequences because successive strata are not preserved. This is clearly demonstrated by mid-latitudinal continental shelves subject to repeated Pleistocene glaciations. These shelves have failed to accomodate sediments older than the very last glacial cycle because of insufficient subsidence rates (Section 23.1). It can be suggested that the stacked parasequences of the Bakoye Group in Mali are a record of short-term changes in sea-level superimposed on an overall rise in relative sea-level created by subsidence of the foreland basin. These short-term changes may be glacio-isostatic in origin but could also equally reflect glacio-eustatic sea-level fluc-

N.EYLES tuations or changes in the rate of tectonic subsidence in the basin generated by thrust sheet loading or by changes in intraplate stresses (e.g. Cloetingh, 1988; Christie-Blick, 1990; MacDonald, 1991). Unfortunately, the time frame for deposition of each parasequence is not precisely known. Closer resolution of the origin of the parasequences of the Bakoye Group can be achieved by comparison with Late Palaeozoic cyclothemic strata deposited in the Appalachian Basin of eastern North America during the Pennsylvanian (Section 16.13). These strata were deposited co-evally with Gondwanan glaciation and detailed work has successfully discriminated those cyclothems deposited under a glacio-eustatic control from those recording tectonic loading of the basin (Klein and Willard, 1989). It should be emphasised, however, that direct comparison of the Appalachian foreland basin with that of the Tauodeni Basin is not straightforward because eastern North America did not experience direct glaciation. Deposition in a glaciated foreland basin subject to complex changes in water depths and sediment supply created by ice sheet fluctuations, glacio-isostatic loading, glacio-eustasy and tectonism is clearly more difficult to model. The work of Proust et al. (1990) and Deynoux et al. (1991) in applying sequence stratigraphic techniques to deposits of the Bakoye Group represents a major step in this direction.

10.2 Southern Africa The Late Proterozoic mobile belts of southern and central Africa record the destruction of small ocean basins as cratons began to collide to form the Late Proterozoic supercontinent. They provide excellent examples of the interior orogenies of Murphy and Nance (1991). Diamictites occur within the thick volcaniclastic strata of these basins and many have traditonally been regarded as "tillites". Recent work emphasises the importance of sediment gravity flow processes on the margins of volcanic arcs without the

75

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

need to invoke a glacial origin. Local glaciation of the higher volcanic cones is recorded. Martin et al. (1985) describe extensive diamictites within the Pan-African Damara mobile belt of southern Africa which separates the Congo and Kalahari cratons. The older Varianto diamictites ( ~ 40 m thick) of Namibia are not convincingly glacial but are coeval with thicker diamictites (up to 500 m) of the Court and Blaubeker formations that are intimately associated with volcaniclastic sediments and calcareous turbidites which contain striated clasts. Diamictites were interpreted as synorogenic mass flows and were compared with non-glacial olistostromes. Martin et al. (1985) favoured a rift setting for deposition though development of a forearc basin during ocean closure could not be discounted; Miller (1983) had earlier suggested subduction and closure of a small ocean comparable in size to that of the modern day Red Sea (see Stanistreet et al., 1991). Martin et al. (1985) concluded that the diamictite deposits of the Damara Sequence are not directly glacial in origin but allowed that small, alpine-type glaciers may have contributed some debris; the arguments of Schermerhorn (1974) regarding the origin of many Late Proterozoic "tillites" as non-glacial mass flows were upheld. Many other poorly-understood Late Proterozoic diamictites occur within mobile belts in central southern and eastern Africa and are probably of mass-flow origin (e.g. Angola: Schermerhorn and Stanton, 1963; Schermerhorn, 1974; Zaire: Cahen and Lepersonne, 1981a). Binda and Van Eden (1972) described diamictites of the Great Conglomerate (Kundelungu Tillite) in Zambia and similarly identified a subaqeous debris flow origin. These workers emphasised the stratigraphic association of diamictites with turbidites. Kulka and Stanistreet (1991) identify the importance of submarine fan sedimentation and mass flow within the ocean closed by the collision of the Kalahari and Congo cratons between 750 and 505 Ma; this depositional model may be widely appli-

cable to the interpretation of deformed slivers of diamictites preserved within Late Proterozoic fold belts in Arabia and Brazil. 10.3 Arabia and Brazil

A tectono-sedimentary regime similar to that identified by Martin et al. (1985) for southern Africa is recorded on the Arabian shield. Kemp and Young (1981) stressed the importance of downslope resedimentation of volcanic debris (the presence of glaciers was considered only a remote possibility) for the AI Ays and Hadiyah Groups of northwestern Saudi Arabia. Possible tectonic settings vary from intracratonic rifts, back-arc extensional basins to collisional arcs (Murphy and Nance, 1991). In Brazil, co-eval Late Proterozoic diamictites occur with several "interior" orogenic fold belts (Paraguay-Araguaia, Trans-Brasiliano fold belts) in the States of Minas Gerais, Goias, Mato Grosso do Sul, Mato Grosso do Norte and Bahia (Isotta et al., 1969; RochaCampos and Hasui, 1981). Fold belts record closure of interior oceans between colliding cratons during the Brasiliano orogenic cycle (about 900-600 Ma; Pimentel and Fuck, 1992). The geology of the mobile belts is dominated by mantle-derived island arc rocks and accompanying volcaniclastic sediments. Diamictites are not well constrained by age dating, are highly deformed and poorly exposed. Evidence of a glacial setting is weak and traditionally has centred on the presence of, and gross textural characteristics (i.e poor sorting, presence of large clasts), of diamictites; striated bedrock, clasts and ice-rafted horizons can locally be identified. It can be suggested that the bulk of these deposits were emplaced by debris flows within turbidite-dominated active margin basins during Late Proterozoic ocean closure; uplifted cordillera along island arcs may have supported small glaciers. A very similar depositional setting can be recognised for active margin basins that developed along the periphery of the Late Proterozoic superconti-

N. EYLES

76

BRITAIN



fl.~f%~~'~

~

/

/

~

/ [ L NEWFOUNDLAND

,J (/.ova

/

/

RI AN RMMAOSIC

Y

IREL A ~ ~ ~ A

?Y

- SCOTIA

0

.

1000

PIEDI4ONT TERRANE

o km 300

.,

ST.t.AWRENCE TERRACE

- ....

,

,~;;:~:

TAIDONICR , ~

_..,,,:~,~

~'~~"

\

- - / J J C . ~ , , ' J _ _ _ ~ t r ~-,,

" ~ C O N T I N E N T A L S H E L F

,%, v 't~ ,,%~---% ~.~/',,, TALLAHASSEE~ ~"~ SUWANNEE ~ ~ 1 TERRANE / ,D~ EASTMARGIN

BOSTON BAY ---~.-k;.g 6~ouP

~A~Ur~ LA'~HN'rA,N & ~l~/.~ulurn/n&.n m~.lu / i

AV~O.TEmU~e

--~'/,e~oF4_

MOUNT ROGERSFORMATIONS

/

/

/

G~OE, ,E.,,,e

GASKIERS FORMATION t'l~l~ ~. 7~

•. . . . . . .

Fig. ]0.4. Top: Distribution of Avalonian-Cadomian strata on pre-Mesozoic continental reconstruction (after Nance et al., 1991). Numbers identify stratigraphic sections in Fig. ]0.5. Bottom: Occurrence of glaciclastic strata deposited along glaciated volcanic arcs (Fig. ] 0 . 5 ) .

CAMBRO - ORDOVICIAN PLATFORM

MASSACHUSETTS

-

ARC/RIFT

NORMANDY

2

3

"'.

-7:

BOSTON BAY GROUP

DIAMICTITES

--

SIGNAL HILL --- ST, JOHN'S GP

• &&~&&• LYNNMATTAPAN ~"~I++*.~ COMPLEX v vlLv

MAGMATIC ARC

NEWFOUNDLAND

UPPER BRIOVERIAN

HARBOURMAIN GP

w~~ ' ~ ' ~ ....

LOVE COVE GP MARYSTOWNGP

6 0 0 Ma

.'~A,=I

~ v'i=2-! LOWER v ~ ¢ -, BRIOVERIAN rr~ v

v _

u

~e~ v

, --.

CONCEPTION GP

......

vv • ,,v

540

. ~

v

v v

SUBDUCTION

'V ,~, ~- "

I PENTEVR,A~

comD~ s

Fig. 10.5. Correlation of Late Proterozoic diamictites along the Avalonian-Cadomian Belt; see Fig. 10.4 for location. Glacial influence can only be identified in New England and Newfoundland. Based on Murphy and Nance (1991) and Eyles (1990).

77

E A R T H ' S G L A C IA L R E C O R D A N D ITS T E C T O N I C S E T T I N G

DEEPWATER BACKARC BASIN: (]) TD EL~sRIG~ESDVo~i si~EGSE pR O,FV~G SI~T;iLO~wEI g sCH EIV/00%~0 SI GENIjCCTE

I : + ~ ~ I ~

O

G

+ i+ + +/ ~ L E R A N E

N

+ +

,

C

/

,

GLACIERS ONVOLCANIARC C DEEPMARINE BAS,NE,LL:

TURBIDITES []

GLAClOGENI DEBRI C PLOWS S []

Fig. 10.6. Deposition in a glacially-influenced backarc basin; a depositional model for the Late Proterozoic Avalonian-Cadomian Belt. A dominantly transpressive tectonic regime appears likely. See Fig. 9.2 for location of orogenic belt on Late Proterozoic supercontinent and Figs. 4.7 and 4.8 for typical stratigraphic infill. See also Fig. 26.7. nent after 600 Ma along the AvalonianCadomian belt in eastern North America and Europe.

10.4 Eastern North America and Europe; glacially-influenced deposition in volcanic arc basins Late Precambrian diamictites, deposited just before 550 Ma, are a prominent, but volumetrically small, component of the Avalo n i a n - C a d o m i a n orogenic belt preserved in North America, Europe and Africa (Figs. 9.2, 10.4). This belt records the interval from 700 to c. 540 Ma that saw the transition from an extensive Late Proterozoic volcanic arc to a stable platform (Lemos et al., 1990) and is a good example of a peripheral orogeny as defined by Murphy and Nance (1991). Glaciation is associated with the closing phases of Avalonian volcanism along the active margin of the Late Proterozoic supercontinent (Nance et al., 1991) and is well constrained by radiometric age dating. The glaciclastic Boston Bay Group in Massachusetts post-dates 579 + 23 Ma, the Gaskiers Formation in Newfoundland post-dates

586 + 3 Ma and the allegedly "glacial" Granville Formation in Normandy post-dates the Coutances quartz diorite (584 + 4 Ma) but is younger than the 540-550 Ma Mancellian granites (Fig. 10.5). Late stage synorogenic volcanic and glaciclastic sediments were deposited in deep water rift basins, floored by thinned continental or oceanic crust, adjacent to ensialic volcanic arcs (Fig. 10.6). Plate reconstructions vary but there is agreement that sedimentation took place on the margin of the African continent which was then part of the Proto-Gondwana supercontinent (Fig. 9.2). Palaeomagnetic reconstructions suggest a palaeolatitude somewhere between 30 ° and 60 ° (Murphy and Nance, 1991). A dominantly transpressive tectonic regime can be identified comparable to the relationship between the Pacific and North American plates along the west coast of present-day North America (Fig. 10.6). This episode of volcanism records the final assembly of the contracting supercontinent and renewed subduction of its margins as the exterior ocean continued to expand. Sedimentological characteristics of diamictites in North America (Gaskiers Formation

78

N. EYLES

O

EARTH'S

GLACIAL

RECORD

AND

ITS TECTONIC

79

SE'ITING

:::Iv e. t.,

~

e'~



I/3

Or/')

° r~ ~

..*2 ~ h 5

i

.=-

~NE

~~ ~ ~ ~.~_

80

of Newfoundland, Boston Bay Group of the Appalachians), France (Granville Formation) and North Africa (Tiddiline Tilloid) have recently been reviewed by N. Eyles (1990). Other less well understood but possibly correlative basins are represented by the diamictite-bearing Grandfather and Mount Rogers formations of Virginia. Sedimentation occurred within rifted back arc/intra arc basins (Nance et al., 1991; Fig. 10.6) dominated by volcaniclastic sediment (see Socci and Smith, 1987; Keppie and Dostal, 1991). Massive and crudely-stratified diamictites in beds several tens of metres thick, have a simple tabular geometry and are interbedded with both fine-grained and conglomeratic turbidites (Figs. 4.7-4.9). A subaqueous debris flow origin is clear (Blondeau and Lowe, 1972; Lowe, 1976, 1982; Socci and Smith, 1987). Large-scale intraformational deformation is the result of downslope sliding and slumping on seismically-active slopes. Diamictites record remobilization and mixing of large volumes of unstable volcanic and glacial debris (Dott, 1961; N. Eyles, 1990). A distal volcanic setting is indicated by large volumes of reworked air fall tephra redeposited as turbidites (e.g. Gardiner and Hiscott, 1988); a generalized facies model for the glacial/volcanic strata is presented by N. Eyles and C.H. Eyles (1989).

10.4.1 Palaeogeography of the AvalonianCadomian belt The Avalonian-Cadomian tectono-stratigraphic belt extends from New England to Northwest Europe (Figs. 10.4, 10.5) but a glacial influence on sedimentation can only be identified in New England and Newfoundland; glacial indicators have not so far been identified in Nova Scotia, New Brunswick, Britain or France. Correlative units in these areas are dominated by proximal volcanic strata (e.g. the "Eocambrian" succession of New Brunswick, the Main-a-Dieu Group of Nova Scotia) that include large volumes of lahar and pyroclastic flow de-

N. EYLES

posits. A glacial influence may simply have been masked by active volcanism or, more likely, was not present. In Britain, equivalent strata are represented by the UriconianWarren House Volcanics dated at 566 __ 2 Ma (Tucker and Pharoah, 1991) that are closely related to the Longmyndian volcanogenic turbidites. These in turn, are broadly equivalent to the thick volcanicallyinfluenced submarine fan deposits of the Brioverian Supergroup of northern France. No definitive glacial influence can be identified in either. The "poudingues" (diamictites) of the Brioverian contain finely-striated clasts interpreted as glacial (Graindor, 1954; Dore et al., 1985) but Winterer (1963) clearly showed these to be the result of differential movement of clasts and matrix during tectonic deformation. The picture that emerges of AvalonianCadomian glaciation is that of a number of extensional arc basins, flanked by volcanic cordillera on which small ice caps or glaciers were able to grow in response to elevation and latitude (Figs. 9.2, 10.6). If this conclusion is correct, then the the highest parts of the Avalonian-Cadomian orogenic belt lay in New England and Newfoundland.

10.4.2 Modern analogs and problems of discriminating tillites from debris flows Modem analogs for the depositional setting in which mass flow diamictites were deposited in Late Proterozoic active margin basins can be found along any volcanic arc. Pyroclastic and eruptive volcanic debris together with gravelly coastal fan deposits are the dominant sediment types along the steep flanks of many volcanic islands. These heterogeneous sediments are mixed together during downslope collapse of volcanic cones thereby depositing poorly-sorted debris avalanche and debris flow facies in marine and subaerial settings (e.g. Car and Ayres, 1991; Rodolfo and Arguden, 1991; Fig. 10.7). Sigurdsson et al. (1980) describe extensive debris flows, up to 5 m thick, within the

81

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'VFING

Granada Basin which lies on the west side of the Lesser Antilles volcanic arc. Flow deposits are composed of large volumes of volcanogenic and fluvially-reworked pyroclastic sediment and are interbedded with turbites derived from reworked tephra. Superficially, both marine and terrestrial debris flow deposits in volcanically-influenced settings can easily be mistaken for tills (e.g. Crandell and Waldron, 1956). Large subaerial debris avalanches frequently exhibit a hummocky surface topography that closely resembles the "morainic" topography left by glaciers (Porter and Orombelli, 1981). Very useful reviews of the characteristics of terrestrial debris flows in glacial, volcanic and other settings are provided by Schultz (1984), Nemec and Steel (1988) and Smith and Lowe (1991). Volcaniclastic debris flows contain great numbers of fractured clasts produced by the dilation of large rock masses after slope failure such that the final deposit may be monolithologic. These flows may contain large amounts of fresh, glassy volcanic shards though this component is rapidly weathered. Tills will generally contain far travelled extra-basinal clasts. "Hot" debris

flows derived from volcanic eruptions exhibit radially-cracked clasts produced by thermal contraction upon cooling (Fig. 10.7); a vesicular matrix structure records the release of trapped air though this is a feature shown by many terrestrial debris flow facies irregardless of setting. "Hot" debris flows will show the same remanent palaeomagnetic pole orientations from clasts and matrix consistent with a common cooling history; in contrast, both debris flows deposited at ambient temperatures and tills show a random remanent magnetism from clast to clast. Cold debris flows show consistent relationships between bed thickness (BTh) and maximum particle size (MPS; Nemec et al., 1984) which is not a characteristic of tills. Facies that are closely associated with terrestrial debris flows are "hyperconcentrated flow" deposits recording the dilution of flowing debris to a point transitional between cohesive, viscous debris flow and fully turbulent, dilute and non-cohesive streamflow (see Smith and Lowe, 1991, for a full discussion). Hyperconcentrated streamflow deposits can essentially be regarded as the terrestrial equivalent of coarse-grained

S o .,,, P ~. ~

F~.'I

FAN-DELTAGRAVELS

("

"

~

~.~J

LARGEVOLUMESOFUNSTABLERAPIDLYDEBRISMIXEDWrfHGLACIALDEBRIS

,'%

#% A

#k

,.% A

#%

Fig. 10.8. Generalized depositional model for glacially-influencedsedimentation on subaerial margins of large stratovolcanoes. Primary glaciclastic and vnlcaniclastic sediment is reworked downslope by sediment gravity flows across alluvial fans (see Figs. 4.4 and 10.7 for diamict facies). See also Fig. 16.16.

82

turbidites and share the same facies characteristics (Fig. 4.2). Hyperconcentrated streamflows are typically produced during violent floods, triggered by events such as rainstorms, the sudden drainage of caldera or ice dammed lakes and melting of ice caps during volcanic eruptions (e.g. Jackson, 1979; Clague et al., 1985; Smith and Lowe, 1991). An awareness of the need to discriminate till from debris flow diamict facies is particularly important where consideration of basin structure, palaeogeography, substrate slopes or associated facies point to a high relief depositional setting. In such settings, primary glacial sediments are reworked downslope after deglaciation to produce thick valley infill deposits of debris flows that are interbedded with, fan-delta, lacustrine a n d / o r marine facies (Fig. 10.8; Ryder, 1971; C.H. Eyles and N. Eyles, 1988; N. Eyles et al., 1988a; Collinson et al., 1989; Fig. 4.4). A major problem in identifying debris flows derived from the downslope failure of glaciclastic sediment is that striations are rapidly abraded from clast surfaces. The same process, however, also produces scratched clasts in lahars and debris flows which can be mistaken for glacial striations (Fig. 10.7; Blackwelder, 1930; Winterer and von der Borch, 1968; see also Harrington, 1971; Fisher and Schmincke, 1984 and refs. therein). In a well-known paper, Schermerhorn (1974) emphasised the importance of debris flow in generating so-called "tillites" in Late Proterozoic orogenic belts. It is still possible, however, to read statements in the recent literature such as "the glacial origin of the tillite is clearly indicated by the massive, unsorted nature of the deposit." Wehr (1986) provides a cautionary example of seeing Late Proterozoic glaciers where none probably existed. He described the Rockfish Conglomerate and overlying Thorofare Mountain Formation at the base of the Lynchburg Group of central Virginia. Strata were deposited in a steep-sided, deep marine rift basin between 690 and 570 Ma and are probably coeval with the debris flow diamictites and

N. EYLES

associated turbidites of the Late Proterozoic Grandfather Mountain and Mount Rogers formations (Schwab, 1976, 1977). Outcrops show a very well-defined, fining-upward succession nearly 500 m thick, composed of conglomeratic mass flows and graded sandstones. Outsized clasts and thin pebbly beds were readily interpreted as ice-rafted by Wehr (1986) and a glacial setting was inferred. There is, however, an intimate and unambiquous genetic association of these facies with sediment gravity flows indicating an active slope depositional setting. In environments such as these outsized clasts are commonly transported by mass flows; pebbly lag horizons are created by rolling of clasts downslope or by winnowing (see Crowell, 1957; Schermerhorn, 1974; Lowe, 1982; Nemec et al., 1984; C.H. Eyles, 1987; Postma et al., 1988). The above authors provide many references to the occurrence of coarsegrained and poorly-sorted facies in non-glacial depositional settings dominated by sediment gravity flows. The stratigraphic association of diamictites with turbidites provides the interpretative key to the origin of many so-called "tillites" with Late Proterozoic active margin basins (Figs. 4.7, 4.8). The common occurrence of diamictites in these strata reflects an abundant supply of volcaniclastic sediment. Without evidence such as dropstones and striated, glacially-facetted clasts a glacial influence cannot be inferred simply on the basis of diamictites. The use of quartz grain surface textures to identify a glacial contribution to sedimentary deposits (e.g. Whalley and Krinsley, 1974) continues to appeal (D'Orsay and van de Poll, 1985; Mahaney et al., 1988; Mahaney, 1990) but data clearly show that supposedly unique "glacial" textures are found on any grain freshly released from bedrock and subject to repeated grain breakage regardless of sedimentary environment (Solomon, 1986; Gomez et al., 1988; Hodel et al., 1988; Clark, 1989; Gomez, 1990). The same comments regarding the utility of identifying a glacial source based on the surface texture of garnet

83

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E T T I N G

grains (e.g. Gravenor, 1979, 1980; Gravenor and Gostin, 1979) have been made by Bull et al. (1980). Even the use of striated clasts to identify a glaciaclastic source is not straightforward. Given that striations can be cut on clasts in non-glacial settings by clast-to-clast contacts in debris flows (Blackwelder, 1930; Fig. 10.7), by post-depositional, tectonicallyinduced rotations of clasts (e.g. Winterer, 1963) and by aeolian sandblasting it is very important to identify whether glacially-shaped clasts (flat-irons, bullets) are present in a diamict(ite) of questionable origin; this shape cannot be produced in any other terrestrial sedimentary environment (Section 3.1.1) and is the unique "hallmark" of subglacially-handied debris (Boulton, 1978).

the margins of the modern Atlantic Ocean, from "palaeo-Pacific" basins on the west flank of Laurentia that are now found around the modern Pacific Ocean. This classification of Late Proterozoic glaciated basins is employed below.

11.1 Palaeo-Atlantic sector; Late Proterozoic glaciation of eastern Laurentia Late Proterozoic "Varangerian" glacial strata are widespread around the margins of the modern day Atlantic Ocean in Scotland, Greenland and Scandinavia. These deposits accumulated shortly before 600 Ma under extensional crustal regimes around the rifted margins of several cratonic blocks (Figs. 9.2, 11.1). An extensional regime gave rise to

11. THE GLACIAL RECORD OF LATE PROTEROZOIC BASINS IN EXTENSIONAL TECTONIC SETTINGS It is argued below that most Late Proterozoic glacial deposits, found along two belts either side of Laurentia, accumulated as glacially-influenced marine strata within rifted basins (Fig. 9.2). Widespread and diachronous phases of crustal extension accompanied disintegration of the Late Proterozoic supercontinent and set the scene for glaciation of uplifted rift shoulders. Given that strong, localized uplifts of several kilometres can occur along the boundary faults of rifted terrains, the requirement of a glacial source area and sediment repository are thus satisfied simultaneously. Dolomites that occur below many diamictite successions may record clastic-starved sedimentation in restricted warm-water basins prior to onset of uplift along the basin margins. Uplift is recorded by larger volumes of detrital dolomite that occur within diamictite successions. As argued above (Section 9.1) it is possible to subdivide Late Proterozoic extensional basins into "palaeo-Atlantic" basins, developed along the eastern extremities of Laurentia (Fig. 9.1) and now preserved around

. ~ ":J ~ \ ~ ~ (~ ~J(-~'~='./'} ~ ~ . .*r~v. \ ",3-\

\

~ " BARENTSCRATON

+ + X \\ LAURENTIAN i__~ CRATON " . ~ l . . ~ >~ i(

+

+ + .I + + PECHORA CRATON f + -I"

i ~ Fig, 11.2

,.

k~. + + J +

.-.'" RIFT BASINS VENDIANSEDIMENTS VENDIANGLACIALLY• INFLUENCEDSTRATA ~ ICE FLOW DIRECTION + CRATON

~;=~,----"r="=~--N ' : ~ -
/ ~ y - ' _ ; !

i

J t ++/

/ +®

T

"<4

i t +z +

o

I

\

I +

+

bb.

I "3,~

"~)~'~

@l~+i\

If+

+

+)nuss~l

milL,

z____--

/

I I/

/

..~

.

~"

+

~\') A

+



1 + :~--

+

.

; , ' / +"

+ ~ g~' i

~

+ +

+

_

Y--------/7

+

BLACKSEA

0 I

I~a

500 I

+

+

+

Fig. 11.1. Glaciated basins along the rifting palaeaAtlantic margin of Laurentia (see Fig. 9.2). (1) North Greenland; (2) West Spitsbergen; (3) Central Spitsbergen; (4) East Spitsbergen; (5) East Greenland; (6) Finnmark; (7) Urals; (8) Scotland and Ireland; (9) Valdres, Hedmark, Risback, Tossafjallet and Engerdalen basins; (10) Central Sweden Basins; (11) Aulacogens of the Russian Platform. After Nystuen (1985).

84

widespread aulacogenic and rift basin formation accompanied by strong uplift of rift shouders and the development of regional ice centres. 11.1.1 Scotland In Scotland, the Port Askaig Tillite of the Dalradian Supergroup was first described by Thompson (1871). The succession, about 750 m thick, is well exposed in a strike-parallel outcrop from southwest Ireland to northeast Scotland (700 km) and consists of a relatively simple alternation of massive and tabular bedded diamictites conformable with sandstones, minor conglomerates and laminated siltstones. Spencer (1971, 1975, 1985) emphasized repeated episodes of "melt-out till" deposition below a grounded ice sheet. Problems with a melt-out model for the deposition of thick and extensive diamictite bodies have already been discussed (Section 3.1.1b). C.H. Eyles (1988) argued that the geometry and sedimentology of the formation was inconsistent with subglacially-deposited facies and argued a glaciomarine origin involving the "rain-out" of glacially-produced muds and ice-rafted debris (e.g. Fig. 3.11) on a subsiding, shallow marine shelf. She highlighted the structural setting of the Port Askaig Tillite at the base of a deepening-upward succession of basinal turbidites (Argyll Group; Anderton, 1982) deposited in faultbounded blocks and basins on the northern margins of an incipient Iapetus Ocean (Fig. 11.1). Large olistostromes composed of dolomite (e.g. the Great Breccia; Spencer, 1971) record the downslope collapse of shelf carbonates, accompanying the onset of block faulting, earthquake activity and subsidence. Modern analogs are described by Hine et al. (1992). Intraformational dolomitic-rich sediments and locally, large amounts of dolomite matrix in diamictites are detrital (Fairchild, 1983; 1992; see Section 13). Distinctive petrographic characteristics of the diamictites (Anderton, 1982) record progressive exposure of deep crustal rocks (principally granites) on the margins of uplifted ranges. Di-

~. EYLES

amictites record sedimentation from meltwater plumes below wave base and ice rafting; the degree of any downslope resedimentation is difficult to identify and was probably underestimated by C.H. Eyles (1988). Continued downslope sliding and slumping is recorded by deformed sequences of diamictites, siltstones and detrital carbonates ("Disrupted Beds"; Spencer, 1971; C.H. Eyles and N. Eyles, 1983a). Large scale cross-bedded sandstone facies that separate diamictites record shallower, tidally-influenced conditions. Deeper-water conditions appear to have obtained in the Irish sector of the basin where facies are more mud-rich and interbedded siltstones are thicker. Ongoing seismic activity and ground shaking may be recorded by penetrative sandstone deformation structures; a subaerial periglacial origin has also been proposed (Spencer, 1971, 1975; N. Eyles and Clark, 1985). The Port Askaig diamictites are overlain by mixed dolomitic-siliclastic sediments of the shallow marine Bonahaven Formation (Fairchild, 1980, 1989) Deposition of Dalradian sediments in a rifted extensional setting was inferred by Yardley et al. (1982). Soper and Anderton (1984) argued that the Dalradian "slides", a series of thrust-faulted blocks produced during a Late Cambrian Grampian orogeny marking the closure of the Iapetus Ocean, were initiated as extensional faults during the Late Proterozoic opening of Iapetus. The conversion of extensional faults to thrust faults (inversion tectonics) during later regional compression serves to emphasize the importance of Late Proterozoic structures along the margin of an incipient Iapetus Ocean to subsequent tectonic development. 11.1.2 North and East Greenland Vendian "tillites" that are broadly co-eval with the Port Askaig Formation, are represented within the 17 km thick Late Proterozoic of East Greenland. The 1300 m thick Tillite Group is exposed in strike parallel outcrop over some 300 km between 72° and

EARTH'S GLACIAL RECORD AND ITS TECTONIC SEqq'ING

74°N (Fig. 11.1). This sequence consists of two diamictite horizons variably called Lower and Upper Tillite (or Ulves6, Storeelv formations respectively) separated and capped by thin-bedded marine mudstones and crossstratified sandstone facies. The glacial and associated marine strata rest on predominantly dolomitic shallow water sediments of the Eleonore Bay Group. Thick carbonate platform deposits below glaciclastic strata are a distintive feature of both East Greenland and northeast Svalbard (see below). Herrington and Fairchild (1989) suggested that these two areas lay in the same depositional basin and identified an abrupt phase of basin extension, marked by deep water carbonate slope deposits, prior to glaciation. The first input of glaciclastics in East Greenland records the initiation of uplift and subsidence along the eastern margin of Laurentia along the margin of an incipient Iapetus Ocean (Figs. 9.2, 11.1). The Lower Tillite (Ulves6) in East Greenland contains clasts composed almost exclusively of yellow weathering dolomite (see Section 13.2). An upward increase in basement lithologies has been interpreted as a record of orogenic uplift and unroofing (Higgins, 1981). Massive diamictites with sandstone rafts (debris flows?), associated with thin to medium bedded turbidites with dropstones, comprise the dominant glacial facies; as in Scotland (see above), penetrative deformational structures may record ongoing seismicity. Diamictites and asociated facies have been briefly described by Hambrey and Spencer (1987), Moncrieff and Hambrey (1988) and Moncrieff (1989) and interpreted as having been deposited in an environment transitional between glaciomarine and glacioterrestrial below a floating ice shelf. It would appear that both glacially-influenced shallow marine shelf and resedimented slope facies are present; several alleged "tillite" horizons, as with so many other Late Proterozoic strata, occur within thick turbidite successions (e.g. Fig. 4.7); the archtypical record of glacially-influenced deposition in a rapidly-subsiding

85

basin flanked by an uplifted glaciated source area. Similarity of the East Greenland and eastern Svalbard (Wilsonbreen Formation) glacial successions has been emphasized by Spencer (1975, 1985), Hambrey (1983) and Herrington and Fairchild (1989). This suggests a close geographic juxtaposition of the two areas perhaps within the same basin (Fig. 11.1). Strata of west Svalbard are thicker however, indicating either accelerated subsidence or that the succession originated in another basin and was juxtaposed against central and east Svalbard by later strike-slip movement (see Harland and Wright, 1979). In northern Greenland the relationship between rifting, uplift and glaciation is especially clear. In northern Greenland and Ellesmere Island Late Proterozoic rifting of the east-west oriented Franklinian Basin is dated by basic volcanics at 635-640 Ma (Trettin, 1989). This basin lay perpendicular to the main north-south oriented Iapetus margin of east Greenland and so may represent an aulacogen (Surlyk, 1991). Uplift and block faulting accompanied rifting and not surprisingly therefore, syntectonic glaciation is recorded (Sonderholm and Jepsen, 1991). In southern Peary Land, the Moraeneso Formation consists of diamictites, breccias and conglomerates that fill palaeovalleys incised several hundreds of metres into underlying Middle Proterozoic strata. A glacial source is indicated by striated and facetted clasts but Collinson et al. (1989) conclusively showed that the diamictites record postglacial reworking of older glacial sediments in a cold subaerial setting not unlike that of the present day. Fluvial and aeolian facies fill the axial portions of palaeovalleys. Valley infills closely resemble the so-called "paraglacial" strata typical of glaciated mountain valleys that record the early postglacial resedimentation and fluvial reworking of unconsolidated glacial sediments (e.g. Ryder, 1971; N. Eyles et al. 1988a, b; N. Eyles and Kocsis, 1988). The paraglacial valley fills of Peary Land are capped by domal stromatolites per-

86

N. EYLES

haps deposited in cold lake basins similar to those of the Dry Valleys of the present day Antarctic. A shallow marine origin cannot be ruled out however. Reworked diamictites of the Moraeneso Formation are probably correlative with the "lower tillite" horizon (Ulvesco Formation) of the Tillite Group of East Greenland (Sonderholm and Jepsen, 1991). 11.1.3 Scandinavia and Eastern Europe A particularly clear picture of the relationship between pre-Iapetus crustal extension, uplift and glaciation is emerging from Scandinavia. The importance of block faulting during early rifting of the Iapetus Ocean, sometime after 640 Ma, is strongly apparent with regard to the Late Proterozoic glacial deposits on the west-facing margin of Baltica

A

B

o~ I . / ~ M O R T E N S N E S

z

,"r

. . . . .

~

.





=

SMALFJORD •



. . . . . . . . . . . . . .

n" I , . ~ '..9 O BIGGANJARGA TILLITE ~

(Fig. 9.2). These deposits, and the correlative episode of glaciation, are collectively referred to as the "Varangerian", a term derived from the Varangerfjord region in Finnmark where such deposits were first recognised in the late nineteenth century (Nystuen, 1985, p. 211). The term "Laplandian" has also been used (Chumatov, 1992; Knoll and Waiters, 1992). Late Proterozoic glaciated basins in western Norway (Tossafjallet and Engerdalen basins; Fig. 11.1), central Sweden and other subsidiary basins in southern Norway (the Valdres, Hedmark and Risb~ick basins; Nystuen, 1985; Fig. 11.1), together with those of Finnmark (Gaissa Basin; Fig. 11.2) share a common structural framework involving deposition in "aulacogenic" or failed-rift basins (Bjorlykke, 1978; Kumpulainen and Nystuen, 1985). Other less well

C R A T - "

+

+

TURI~DITES

(~BRISFLOW

N

o L

50 I UPLIFTED EDGE OF BALTOSCANDIAN CRATON

+

Fig. 11.2. Speculative palaeogeography and stratigraphy of Gaissa Basin on eastern margin of lapetus Ocean (Fig. 11.1) showing deposition of glaciclastic strata on extensional rifted margin south of the Komagelv Transfer Fault Zone. Based on Gayer and Rice (1989). A - B (inset) shows regional stratigraphy. Compare overall tectonic setting with that on Figs. ll.6B and 19.9. For facies see Fig. 11.3

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E T T I N G

understood deposits are associated with aulacogens on the Eastern European Platform (Chumakov and Semikhatov, 1981). Chumakov (1992) provides a lengthy summary of these deposits emphasing their association with grabens. A useful review of the "classic" literature on the Late Proterozoic Varangerian glacial strata of northern Norway is provided by Spjeldnaes (1964) and Bjorlykke (1967). The first modern work on these strata was that of Reading and Walker (1966) who concluded that a terrestrial depositional setting was very unlikely; deposition in a glaciomarine environment was suggested by associated facies. Reading and Walker (1966) commented on the lack of published data regarding glaciomarine settings and were unable to offer a detailed assessment of depositional processes particularly in regard to the origin of massive diamictites; the then current model (e.g. Carey and Ahmad, 1960) emphasised deposition in quiet water below a floating ice shelf. Edwards (1975) reinterpreted these strata as the product of terrestrial glaciers employing then current models of sedimentation along sub-polar glaciers in Spitsbergen (e.g. Boulton, 1972). Strata have been interpreted in classic climatostratigraphic style by Foyn and Siedlecki (1980) where diamicites are axiomatically interpreted as tillites and other intervening facies are taken to indicate glacier retreat. More recent and ongoing work confirms the glaciomarine setting invoked by Reading and Walker (1966) and offers new insights into the origin of massive diamictite facies (T. Warman, pers. commun., 1993).

The Gaissa Basin of Finnmark In Finnmark, Late Proterozoic glacially-influenced sedimentation occurred on the western rifted flank of Baltica, adjacent to the Timanian Aulacogen which separated Baltica from the Barents Craton to the north (Fig. 11.1). The structural configuration consists of an approximately west-east oriented transfer fault (Trollfjord-Komagelv fault) with westwarddirected block faulting on its southern flank

87 (Fig. 11.2). Thick (450 m) glacial deposits of the Vestertana Group (Smalfjord, Nyborg, Mortensnes formations; Fig. 11.2) began to accumulate in the Gaissa Basin following uplift of shallow water stromatolitic dolomites (Porsanger, Grasdal formations); uplift at least 1500 m is inferred along the eastern boundary of the Gaissa Basin but nearly 8 km of strata were removed from areas north of the T - K fault (Gayer and Rice, 1989). Evidence for localized uplift in areas north and east of the basin accord with striations and palaeoslope data that indicate westwarddirected ice flow from small ice caps on uplifted blocks. A restricted ice cover is indicated by the absence of correlative glacial strata in the co-eval Lokvikfjell Group of the Timanian Aulacogen and the "Baltoscandian miogeocline" to the west (Gayer and Rice, 1989; Fig. 11.2). It is highly significant that time-correlative strata in these basins that were closely adjacent to the Gaissa Basin are represented by alluvial fan, fluvial facies and shallow marine facies; these contain diamicrites formerly regarded as glacial (Siedlecka and Roberts, 1972) but now interpreted as alluvial fan debris flows (Laird, 1972; Gayer and Rice, 1989). These findings are especially pertinent to the "tillites" of the Vestertana Group in the Gaissa Basin. New studies show that the Smalfjord Formation comprises strata deposited within a large fan-delta complex as well as deeper water debris flow and ice-rafted deposits (Fig. 11.3). Diamictites of the Smalfjord Formation are rich in dolomite clasts and record downslope slumping and sediment gravity flow of coarse-grained glaciofluvial gravels and glacial debris. In the southern part of the basin these facies are interbedded with shallow marine, storm-influenced sandstones. Offshore deposits in this area comprise turbidites. The famous Bigganjarga Tillite and associated striated surface (Reusch's Moraine; Reusch, 1891; Holtedahl, 1918; Fig. 11.3) lies at the base of the Smalford Formation and appears to be a small remnant of previously more extensive diamictite facies de-

88

N. EYLES

89

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'VI'ING

_==o

~~;: i!~i~iiiii!i~, Li!~!!!~!iii!i!~ i ii~!~!~ii!~i ~ ~:

"0

.=- ~'~

,...~ ~,R

"'~

~ ~ "~ =--'~

-~ee$

~,,~

.

~

~

~

~

90

N. EYLES

posited as debris flows (Crowell, 1964) having been largely reworked in shallow water during fan delta progradation. Northern exposures of the Smalfjord Formation show "banded tillites" (Edwards, 1975, 1978) that comprise variably coloured debris flow facies, "rain-out" diamictites, laminated and massive mudstones and abundant ice-rafted debris (Figs. 4.5, 11.3). These facies record sedimentation from suspended sediment plumes and floating ice with secondary mass flow. These facies suggest a relatively deep water setting (below storm wave base) some distance from tidewater ice margins supplying large volumes of mud. Laminated siltstone facies interpreted as marine by Reading and Walker (1966) were interpreted as aeolian "loessites" by Edwards (1979).

The Nyborg Formation comprises a thick succession of turbidites, recording enhanced fault-controlled subsidence. A shallowing trend is observed toward the top of this unit just below the base of the overlying Mortensnes Formation. This formation consists of massive and stratified diamictites (rain-out and mass flow facies) interbedded with fine grained, laminated and thin-bedded turbidites. The Mortensnes Formation noticeably "oversteps" underlying units to the west and east indicating progressive foundering of the continental margin and regional transgression. Initial extension and rifting of the continental margin is marked by dykes dated at 640 Ma (Beckinsale et al., 1976). These intrusions are part of a wider network along the margin of Baltoscandia interpreted as

r-

.o ~Cambrian Shale 7 o~ Ringsaker Qtz.ite I_

kre Shale Massiveand bedded debrites

~!i~!iii~~:~i~i:~!'~!Vardal ~i Sandstone/ c~

i~i:i:i:i~:~:ii:iii:i%l ~Ekre

~ Shale 5-200m

Moelv Tillite 0-160m

lassive debrite

.....

RendalFormation

luddy turbidites

::;::

OsclalConglomerate

;andstone turbidites vith dropst0nes

~

Elta Formation

luddy turbidites with ropstones and thin ebrites DIAMICTITE

~ot Exposed

MASSIVE STRATIFIED I ~

FINE-GRAINED TURBIDITES SANDSTONE TURBIDITES

==?i=:=

DROPSTONES

Bedded debrites

Massive debrite. Sandy matrix

Base not:exposed

Fig. 11.4. Stratigraphic and facies summary of Moelv "Tillite" in the central "Sparagmite" basins of Scandinavia (Fig. 11.1). Modified from Nystuen and Saether (1979) and Bjorlykke et al. (1976). Note similarity of stratigraphic succesion of Moelv "Tillite", Ekre Shale and Vardal Sandstone to stratigraphic cycles produced by rifting (e.g. Fig. 8.2). Compare also with Figs. 4.7 and 12.4 showing common association of debrites with turbidites.

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T ' r I N G

syn-Iapetus rifting intrusions; turbidites from the Nyborg Formation have been dated at 654 Ma (Hambrey, 1983). Glacially-influenced sedimentation in the Gaissa Basin is inextricably linked to the regional tectonic setting involving source area uplift, probable adiabatic glaciation and an influx of poorlysorted glaciclastic sediment to a rapidly subsiding basin. No simple "climatostratigraphic" interpretation of ice advance and retreat (Edwards, 1975, 1978) should be sought from the lower and upper glaciclastic strata of the Vestertana Group.

The sparagmite basins of central Scandinavia Late Proterozoic glacially-influenced "fandelta" settings, very similar to that identified for the Vestertana Group in the Gaissa Basin, occurred in other Norwegian basins along the former Baltoscandian continental margin (e.g. Bjorlykke, 1978; Nystuen and Kumpulainen, 1981; Kumpulainen and Nystuen, 1985; Tietzsch-Tyles, 1989; Fig. 11.1). The same tectonic setting of glaciation can be identified. In central Scandinavia, the term "sparagmite" refers to conglomeratic arkoses with subsidiary limestones, quartzites and shales deposited in graben-like basins (Bjorlykke, 1966). In southern Norway five main basins (Risb~ick, Hedmark, Valdres, Engerdalen, Tossasfjallet) have been identified. These were displaced several hundreds of kilometres to the east during later Caledonide foreshortening (Kumpulainen and Nystuen, 1985). Nystuen (1987) recognized a common succession of lowermost diamictites overlain by laminated mudstones with decreasing quantities of ice-rafted debris upward in section. Preglacial deposits are dominantly fandeltaic in origin and include significant thicknesses of debris flows that relate to active faulted basin margins; these provide important clues to the origin of later "glacial" deposits. Massive and stratified diamictite facies, conventionally described as "basal tillites" (e.g. Moelv Tillite and correlative Langmarkberg Formation; Nystuen, 1976,

91

1985; Fig. 11.4) are interbedded with fine and coarse-grained turbidites; a debris flow origin is clearly indicated by this association. Shales that are interbedded with the Moelv diamictite have yielded a Rb/Sr whole rock age of 630 Ma (Nystuen, 1976, p. 8). No striated pavements below the diamictites have been reported and perhaps none should be expected given that the preserved record is a basinal one. In the Hedmark Basin, massive diamictites of the Moelv Tillite rests on the Osdal conglomerate which was deposited by "braided" high energy rivers in an alluvial fan setting (Nystuen and Saether, 1979, p. 241). The so-called "tillite" is intimately and conformably interbedded with conglomerates and is of a restricted thickness (0.5-20 m; Bjorlykke et al., 1976). It appears to have been deposited as part of a coarse clastic wedge deposited on the steep margin of a graben basin. A local ice centre, on upstanding horsts and basement escarpments, probably fed detritus to surrounding fan-deltas as mass flows. Nystuen and Saether (1979) identified a close correspondence in clast content between the Moelv "Tillites" and underlying Osdal Conglomerate suggesting a glaciofluvial origin. Dropstones in laminated shales may record rafting by winter ice or from local tidewater ice margins that were able to reach sea-level. There is no firm evidence to support the concept of a single large ice cap across the Scandinavian basins; the existence of multiple caps on uplifted blocks is more likely given the tectonic setting and the associated basin and range topography consisting of uplifted crustal blocks separated by deep water basins. It is significant in this regard that carbonates and black mudstones of the Birs Formation which underlies the Moelv Tillite record deposition in isolated, block-faulted basins (Saether and Nystuen, 1981) providing further contextual evidence of the structural setting of overlying diamictites. A widespread allegedly "postglacial" marine transgression (e.g. Ekre Shale of the Hedmark Basin;

92 Nystuen, 1982; Dreyer, 1988) more likely records continued fault-controlled subsidence of the western Baltoscandian margin. Nystuen (1987) recognised that the greatest thicknesses of the Moelv "Tillite" lie adjacent to faults indicating the importance of syndepositional tectonics. The palaeolatitudinal position of Baltica during Late Proterozoic glaciation is not well constrained but workers are agreed on a mid- to high-latitude position (30-60°; Pesonen et al., 1989; Pesonen, 1990; Torsvik et al., 1990, 1991). The relationship between the Varangian glacial strata of Scandinavia with those of Eastern Europe is not known. Poorly understood Late Proterozoic deposits of alleged glacial origin are associated with northwest and northeast trending grabens (aulacogens) on the Eastern European Platform (Aksenov et al., 1978; Aleinikov et al., 1980; Fig. 11.1). Correlative deposits occur in Central Asia, in Tien-Shan (Zubtsov, 1972). Chumakov and Semikhatov (1981) report K - A r ages on glauconites of 630-650 Ma for "tillites" of the Vil'chitsky Group of the Orsha Basin west of Moscow, and dates of 665-660 and 610 Ma for those of the Urals (see also Chumakov, 1992).

11.2 Extensional tectonics, uplift and glaciation along the palaeo-Atlantic margin of Laurentia; discussion The wide distribution of "tillites" in the North Atlantic region led Chumakov and Cailleux (1971) and Spencer (1975) to infer an extensive European ice sheet, several thousand kilometres in diameter, stretching from western Britain to the Urals (Fig. 11.5). Spencer (1985) and Hambrey and Spencer (1987) have argued that the area covered by the ice sheet was of very low relief and close to sea-level; the upward vertical evolution of petrographic suites from sedimentary to crystalline in many glacial successions was construed as evidence of large scale glacial erosion or unroofing. The tectonic setting and sedimentology of the glacial deposits, in fact,

N ZYLES

1

I

n

Fig. 11.5. Reconstruction of Late Proterozoic PanEuropean ice sheet along the eastern margin of Laurentia (Fig. 9.2). After Spencer (1975). suggests a different model focussing on the central role of adiabatic uplift during regional extension. The North Atlantic region of the Late Proterozoic supercontinent saw the widespread development of rift basins peripheral to an early Iapetus Ocean (Fig. 11.1). The climatic threshold for the initiation of ice covers was crossed as rift margins underwent rapid uplift ("adiabatic" glaciation). Figure 13.1 depicts the average elevation of the regional snowline with present day latitude; these data serve to emphasise that regional uplifts of several kilometres in mid- to high-latitudes can play a critical role in the initiation of regional glaciation. In contrast to Mesozoic "Atlantic" margins, not much is known of the precise structural geology of the North Atlantic "Iapetus" rift basins and continental margin. Irregardless, large scale extension and thinning of the crust is associated with topographic "arching" on the margins of the extended terrane (e.g. Keen and Beaumont, 1990). This satisfies the need for an uplifted glacial source area and a

93

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C SE'IT1NG

depositional repository. Many areas undergoing extension show broad topographic domes separated from the active rift by a faultbounded topographic escarpment. The modern Red Sea rift and associated updomed Arabian Shield which locally reaches over 3500 m in elevation, is a good example. The Death Valley area west of the Sierra Nevada also shows a marked topographic doming; the lowest ( - 86 m asl) and highest (4419 m) points in the mainland US occur in this area (Wernicke, 1985). Wernicke (1985) characterized the surface topography across extensional terrains, from the active rift zone to uplifted flanks, as a "sine-wave". Lister et al. (1991) identify two fundamental types of extensional plate margin; upper and lower plate (Fig. 11.6). The upper plate margin type is the most significant in terms of geodynamic controls on adiabatic glaciation because of topographic arching and the generation of "passive margin" mountains and plateaux. As cooler, upper crustal strata are pulled out from below an "upper plate" margin so the base of the plate undergoes warming and is exposed to the asthenosphere; magmatic underplating and enhanced crustal bouyancy results in isostatic uplift and the creation of extensive highlands inboard of the faulted coastal margin. The deeper crustal and sub-crustal structure below extensional terrains has been modelled by many workers and is reflected in (A)

Lower-plate

several different schemes (Lister et al., 1986). This work is not reviewed here. I wish instead, to emphasize the magnitude and timing of the uplift processes associated with extension--these after all, are the keys to the development of "adiabatic" glaciation along passive margins. Topographic doming is known to extend several tens of kilometres away from extensional structures and continue many tens of kilometers parallel to strike. The Death Valley Basin in California is characterized by a central graben bounded by an active range-front normal fault; uplift of between 7 and 17 km has occurred over the past 6 Ma accompanied by deposition of over 3 km of sediments along the graben floor (King and Ellis, 1990). Unroofed midcrustal rocks are exposed along the fault. Further examples are widespread across the Basin and Range province of the western U.S. which is characterized by high-angle, planar fault-bounded ranges and basins. Uplifts of up to 20 km within timespans as short as 10 Ma appear typical (King and Ellis, 1990). Naturally. the exhumation of older crustal strata does not necessarily imply absolute uplift with regard to the average long term position of sea level (e.g. England and Molnar, 1990) but the association of topographic doming and passive margin uplifts is well established. The Cenozoic West Antarctic rift system provides an excellent analog for conditions Upper-plate

(B)

margin

Narrow

margin

Continental drainagedivide

Sedimentarysag basin Outer dse cean

(J

Asthenosphere

Asthenosphere .L

Fig. 11.6. Extensional plate margins (after Lister et al., 1991). Note uplift and topographic arching associated with upper-plate margin. This may be an important precursor to glaciationin mid-latitudes(see Figs. 19.6, 19.7, 19.8, 19.9 and 20.1).

94

N. EYLES

along the eastern rifted margin of Laurentia during the Late Proterozoic. The rift system extends for almost 3000 km (Fig. 24.2) with a width of about 750 km and is therefore, comparable to the Late Proterozoic early Iapetus Ocean (Fig. 11.1). The eastern rift

VA o,,0"

boundary is marked by a rift-shoulder scarp up to 7 km high extending from Victoria Land to the Ellsworth Mountains. Fitzgerald (1992) identified as much as 6 km of uplift over the last 60 Ma; a m e a n uplift rate of 100 m / M a . However, recognition of the ver-

,3"0" :11(

! /

\/ •

i

/

IIii II

• Outcrop of thick diamictites

\\ \

V Outcrop of volcanic rock Isopach (contour interval 1500 m). After Stewart, 1972. 1. Corbitt and Woodward, 1973 2. Miller, 1985; Link et al., 1993 3. Ojakangas and Matsch, 1980; Cdttenden et al., 1983 4. Link, 1983 5. Aalto, 1981 6. Eisbacher, 1985 7. Eisbacher, 1981, 1985; Yeo, 1981 8. Young, 1982 9. Mustard, 1991

/ / ///1

40~

0

~1/1~/ / i

30*

\ ",4

J o

3oo

i J /KILOMETRES

Fig. 12.1. Late Proterozoic diamictites of the western Cordillera and thickness of related strata deposited along rifted Palaeo-Pacific margin of Laurentia (Fig. 9.2).

95

EARTH'S GLACIAL RECORD AND ITS TECTONIC SETTING

The key to the wide extent of Late Proterozoic " V a r a n g e r i a n " glacial deposits around the Palaeo-Atlantic sector of Laurentia may be found in the widespread growth of small ice caps on uplifted crustal blocks. R a t h e r than a monolithic ice sheet controlled by one or two ice centres, such as favoured by recent Soviet workers (Chumakov, 1985), a loose collective (commonwealth?) of i n d e p e n d e n t ice caps is probably a better model for Late Proterozoic Varangian glaciation around the margins of the evolving Iapetus Ocean. Strongly localized uplift of rift shoulders on the margins of subsiding extensional basins provides a

tical displacement of Late Pliocene glacial deposits allowed B e h r e n d t and Cooper (1991) to infer episodic uplift rates of as high as ~ 1 k m / M a . Tectonic uplift along the rift shoulder has b e e n directly implicated in the initiation of the Antarctic ice sheet sometime around 40 Ma and the changing thermal regime of the ice sheet during the late Cenozoic (Section 19.1). In addition, the role of accelerated tectonic uplift in the Himalayas, the Cordilleras of North and South America and around the margins of the North Atlantic Ocean can be shown to be a first-order control on global cooling over the last 50 Ma (Section 19).

o

1__,

E

-

O

(Sh~p~a- /

,/

Cycle 2 I / '"" :"' Cycle1 ~ S h " - - ~ - , ~ / ' ~ ~ . ~ u ~ ' f " ~ - P ~ /

/r

-

• ; : > ~ 1 - - " ~ ' ~

(Rapitan)

-

~

.

.

I

"

.... . / ~

~

-Y

"

el

. km

0 i

A

i

200m

lO

'Iv

DIAMICTITE

~

SHALE

I ~

SILTSTONE

SANDSTONE

DOLOSTO~

B / Shezal

;,-~-~

:0.~o

"~ ~ - - " ~

o o,~ . . . . . . . . . . . . ~

o Co ° . ~ _ . ~ - ~ o o Z ' . X ,

• •

. o



. •

l

km

;

o/ /

o * °°~ ~e "~ o 0 " ~% o

. e -o

"~ %°

Ao 0g~e ' o~*o * o oe • • • oQoo~

I

Fig. 12.2. A. Restored rifted margin of Late Proterozoic Windermere Supergroup basin showing the three stratigraphic "cycles" of Eisbacher (1985) with glacially-influenced strata of Cycle 1 (Sayunei, Sa; Shezal, Sh) and Cycle 2 (Ice Brook Formation, Ib). Tw, Twitya Formation; Ke, Keele Formation. After Eisbacher (1985) and Aitken (1991). B. Mass flows and associated turbidites of the Rapitan cycle with C. down dip section through infill of diamictites and turbidites. Strata record intermittent uplift of source area glaciation and progradation of debris flows into turbidite basin. Glaciation may be correlative with Sturtian and Marinoan glaciations in Australia (Figs. 9.2 and 12.5). Compare with Fig. 8.2.

96

mechanism for initiating glaciation and preserving glaciogenic sediments. The uplift "model" predicts the occurrence of paleokarst features at the unconformity surface between uplifted carbonate successions and overlying Late Proterozoic glacigenic strata. Data are sparse however, probably because of a lack of observation or misinterpretation. Mustard and Donaldson (1990) describe karsted surfaces and associated breccias at the base of the late Proterozoic Windermere Group in northwestern Canada that contains glaciclastic strata (Section 11.2.1). Their data are of considerable significance because the same surface has been previously interpreted as the product of glacial erosion (Eisbacher, 1985).

12. P A L A E O - P A C I F I C SECTOR; L A T E PROT E R O Z O I C G L A C I A T I O N O F WESTERN LAURENTIA

The argument has been made above that the widespread Late Proterozoic diamictites of the circum North Atlantic region record pre-Iapetus rifting, uplift and glaciation of basin margins starting at around 640 Ma. A genetic relationship between rifting, initiation of glaciation and the preservation of a glacial record is also apparent along the palaeo-Pacific margin of Laurentia. These deposits are now represented by Late Proterozoic diamictites and associated strata that are now dispersed around the margins of the Pacific Ocean in Australia, China and western North America (Fig. 9.2). These strata record repeated phases of rifting and glaciation shortly after c. 770 Ma and again after about 650 Ma though the dating control is weak. Stratigraphic similarities between the glacial record in northwestern Canada and Australia have been emphasised by Eisbacher (1985) and Young (1992); Figure 9.2 shows the close juxtaposition of these areas astride the palaeo-Pacific margin of Laurentia. No simple correlation with the tripartite glaciclastic strata of China can be made as

N. E Y L E S

yet however (Fig. 6.1), despite close proximity during the Late Proterozoic.

12.1 Western Canada Eisbacher (1985) described Late Proterozoic glacial strata of the Windermere Supergroup in western Canada that were deposited along the rifted, palaeo-Pacific margin of Laurentia (Figs. 9.2, 12.1). The Windermere Supergroup consists of three shallowing-upward stratigraphic "cycles" (Rapitan, Hay Creek, Sheepbed-Backbone Ranges; Fig. 12.2) each about 1 km thick. Until fairly recently glaciclastic sediments were thought to be resticted to the lowermost (Rapitan) cycle deposited sometime after 770 Ma. Recently, Aitken (1991) has reported glaciomarine diamictites from the upper part of the Hay Creek "cycle" more than 1 km above the Rapitan glacials. Hofmann et al. (1990) describe diverse and well preserved Ediacaran fauna below the uppermost glacial strata of the Hay Creek cycle; these are older than any previously reported Ediacaran remains which elsewhere postdate Late Proterozoic glacial strata (Kaufman et al., 1992; Section 14). Kaufman et al. (1992) suggest that the age of these unique fossils can be no younger than 620 Ma. Correlation of the two Windermere glacial horizons with Sturtian and Marinoan glacial strata in Australia is likely (Eisbacher, 1985; Aitken, 1991; Young, 1992).

12.1.1 Rapitan stratigraphic cycle The Rapitan stratigraphic cycle is composed of a complex of massive and stratified diamictites up to 500 m thick that are interbedded with and that pass down dip into fine-grained and thin-bedded turbidites. Diamictites contain striated and glaciallyfacetted clasts and were emplaced by mass flow processes on the steep flanks of fault controlled troughs. Diamictites are thickest immediately adjacent to the back walls of faults and thin downslope into thick successions of basinal turbidites (Fig. 12.2). These

97

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETFING

strata rest unconformably on underlying Middle Proterozoic carbonates; diamictites contain large detrital carbonate clasts, some many metres in diameter, and an abundant carbonate matrix typical of many Late Proterozoic diamictites (Section 13.2). Fold noses and other soft-sediment deformations typical of mass flow deposits are commonly displayed. Occasional beds of volcanogenic banded-iron formation suggest rifting and renewed subsidence. The Rapitan is abruptly overlain by a thick (100 m) black shale (Twitya Formation) which record renewed submergence associated with relative starvation of coarse grained glaciclastic debris and the start of the overlying Hay Creek stratigraphic cycle. Diamictite facies were attributed by Eisbacher (1985) to mass flow processes reworking glacial debris delivered by an advancing ice shelf. Some massive diamictites were attributed to subglacial processes near the grounding line of the ice shelf. An absence of conglomerate and other glaciofluvial facies was argued to be evidence that the ice shelf was cold-based. Large outsized boulders and clusters were attributed to icebergs; clast-rich diamictites to unspecified "meltout" processes near the ice shelf grounding line. Subglacially-deposited lodgement tillites were also identified. Implicit in the model of Eisbacher (1985) is the assumption that massive diamictites are neccessarily of a sub- or proglacial origin. In contrast, diamictite facies of the Rapitan cycle occur within thick ('00 m) downslope-thickening wedges that terminate upslope against faults and downslope within basinal turbidites (Fig. 12.2). A glacial source of sediment is clear but any evidence of an ice-contact setting has been destroyed by downslope mass flow. The depositional setting appears to have been characterised by west facing half grabens, steep substrate slopes and an abundant supply of coarse grained glaciclastic debris that was delivered to the basin by sediment gravity flows (e.g. Fig. 8.3). A marine basin is suggested by the great thickness of the succes-

sion and its considerable areal extent. Syndepositional rifting and a minor volcanic influence can be identified. Rifting along the Windermere rift system was probably accompanied by active source-area uplift and the formation of a regional "adiabatic" ice cover. Correlation with the Sturtian glacial strata of southeastern Australia (Section 12.4.2) is suggested by Eisbacher (1985) and Young (1992).

12.1.2 Hay Creek stratigraphic cycle The non-glacial Hay Creek cycle comprises a prograding, predominantly clastic shelf-slope assemblage up to 1000 m thick. The cycle starts with a 100 m black shale and is overlain by a shallow water carbonate succession capped by a laterally persistent detrital dolostone ("Tepee dolostone"; Aitken, 1991). Carbonates show evidence of downslope mass flow and rapid thickness variations recording an irregular fault-controlled topography and the downslope collapse of rifted, platformal carbonates. Mass flow facies are characterised by oligomictic carbonate rudites, diamictites and olistostromes embedded in a fine grained carbonate matrix; these facies are very similar to those of the underlying Rapitan cycle but until recently were thought not to contain any component of glaciclastic sediment. Aitken (1991) demonstrated a glacial influence on mass flow sedimentation in the form of very rare striated clasts, dropstones and clusters of debris probably rafted by icebergs (till pellets; Ovenshine, 1970). Glacially-influenced strata are in erosional contact with underlying strata but lack any extrabasinal clasts. Aitken (1991) recognised that evidence of a glacial influence is not substantial, given that striated clasts are produced in debris flows and that ice-rafted debris is also a characteristic of seasonally cold but non glacial coastlines (e.g. Section 17.1). This raises the significant problem of distinguishing diamictites released from degrading fault scarps in areas of active rifting from those produced by the resedimentation of glaciclastic sediment.

98 Both settings are characterised by large downslope fluxes of coarse grained debris and deposition of sediment gravity flow facies such that identification of a specific glaciclastic sediment component may not be straightforward. This problem pertains to many Late Proterozoic "glacial" successions given their preferred depositional setting in extensional rift basins but is particularly appropriate to consideration of diamictites of the Windermere Supergroup. Mustard (1991) has recently described thick (1100 m) mass flow diamictites and conglomerates from Late Proterozoic strata exposed in the Ogilvie Mountains of the Yukon Territory. These strata are argued to be equivalent to lowermost Windermere Supergroup rocks. Mass flows were generated as talus along active fault scarps and deposited on the margins of a shallow water body; no glacial influence can be detected and diamictites were interpreted as the direct result of synsedimentary normal faulting (e.g. Fig. 8.3).

12.1.3 Origin of stratigraphic cycles The Rapitan, Hay Creek and SheepbedBackbone Ranges stratigraphic cycles record repeated flooding and shoaling along the palaeo-Pacific margin. Eisbacher (1985) interpreted these cycles as a record of fluctuating sea level and inferred a glacio-eustatic control (see also Young, 1992) but such a mechanism is very unlikely (Schermerhorn, 1975). The growth of Pleistocene continental ice sheets for example, resulted in glacio-eustatic sea level fluctuations no greater than about 175 m (Fig. 16.21). Glacioeustatic fluctuations moreover, consist of repeated sea level draw down from a mean interglacial sea level and so cannot promote the accomodation of thick sediment bodies on continental margins; because there is no long-term transgression and flooding of continental margins there can be no increase in accomodation space. It is difficult to see how stratigraphic cycles in the Windermere Supergroup, each having a compacted thickness of over 1 km

N.EYLES and representing substantial time slices, could be accommodated without an overriding tectonic control on basin subsidence. The rapid lateral thickness changes and shoalingupward character of the stratigraphic cycles in the Windermere Supergroup are better interpreted as the product of relative sealevel change caused by large-scale faulting and successive phases of "backstepping" along the continental margin. The shoaling upward cycles identified within the Windermere Supergroup probably record initial subsidence of the continental margin and the progradation of mass-flow dominated slope prisms (see Figs. 8.2, 8.3). A further indication of an active rift setting for glaciclastic sedimentation is the very common occurrence of banded iron formations and cherts within the diamictite successions (e.g. Young, 1976). Yeo (1981) and Young (1988) have reviewed the origin of glacigene-related Late Proterozoic iron formations in western Canada and concluded that deposition from metal-rich hydrothermal brines was the most likely mechanism. Associated glaciation was seen as a critical factor in the generation of both massive hematite-jasper bands associated with diamictites and thinly laminated ironstones interbedded with turbidites. Brines were subject to upwelling along glaciated coastal margins in response to displacement by cold bottom waters derived from nearby glaciers, leading to the direct precipitation of iron and silica. The distinct association of ferruginousrich sediment and diamictite is a further expression of glacially-influenced marine sedimentation in a rift setting.

12.2 United States and Mexico Late Proterozoic glacially-influenced strata, possibly co-eval with those of the Windermere Supergroup in western Canada, are widespread in the United States and Mexico (Aalto, 1971; Stewart, 1972; Corbitt and Woodward, 1973; Okajangas and Marsch, 1980, 1990; Christie-Blick, 1983; Crittenden

EARTH'S GLACIAL RECORD AND

ITST E C T O N I C

99

SE'VI'ING

of c. 725 Ma (Knoll, 1991; Knoll and Walter, 1992). P. Link (pers. commun., 1993) reports ages between 780 and 730 Ma for Late Proterozoic glacial rocks of the western U.S. Cordillera. These strata are correlative with the Sturtian glacials of the Adelaide "Geosyncline" in Australia (e.g. Young, 1992; Section 12.4.2). The widespead extent of thick, tillite-like deposits, dated at c. 750 Ma, along the western margin of North America suggests ice covers of continental dimensions but it is worth emphasising that a direct glacial influence can be demonstrated in few cases. The overall tectonic setting is one in which large volumes of coarse, poorly sorted sediments are produced along fault scarps (e.g. Nemec et al., 1984; Nemec and Steel, 1988). The rifted palaeo-Pacific margin of Laurentia probably consisted of complexly-linked basins with local ice caps on uplifted crustal blocks and rift flanks rather than extensive ice sheets.

et al., 1983; Miller, 1985; Ross, 1991; Fig. 12.1). All these deposits show sedimentary facies that indicate a relatively deep water rift setting dominated by mass flow processes. These strata are thick (1 km + ) and comprise monotonous associations of thick mass flow diamictites (e.g. the 450 m thick diamictites of the Surprise Member of the Kingston Peak Formation; Miller, 1985) and turbidites. Evidence of syntectonic deposition occurs in the form of rapid lateral thickness variations, a wedge-like, down-dip geometry, a marked fault-controlled topographic relief on underlying formations and complex paleogeographies in addition to the presence of slumped horizons and displaced megablocks. Incipient rifting is also recorded by associated volcanic strata and iron formations interpreted as forming in rift basins with a restricted circulation (Young, 1976; Yeo, 1981). The oldest Late Proterozoic glacial strata of western North America have a mean age

70*

100"

/

130 °

I

\

,.,

,,

(/

/.,._1- ~\

\\

1~"

I

~\

,"'7 I

40:, /

N

.40"

30

\~

~ o

,,~,,~ .30"

~LU~UAN

"-/

(620-600 Ma)

I N A N T U O (720-680 Ma)

~CHANGAN ( 8 0 0 - 7 6 0 Ma)

)

)

/ _/

'~- , ~,"

e~il~ °

K •

..~"

/

/

B K

BEUING KUNMING

H

~ N ~

III~'~'~'~'"

,,---~:=S" t~

"~, c--'"

~.,~ f -Q'~

J

C CHANGSHA Z ZHENGZHOU

Fig. 12.3. Distribution and age of tripartite Late Proterozoic glaciclastic strata in North (N) and South (S) China. (Ater Yuelun et al., 1981 and Sognian et al., 1985.)

100

12.3 Asia

Late Proterozoic diamictites are widespread throughout Asia and the Russian platform (Section 11.1). Though stratigraphic and sedimentological data are few, there is a particularly clear relationship between diamictites and episodes of rifting and the infilling of aulacogens. Late Proterozoic diamictites comprise part of the stratigraphic fill of narrow, elongate troughs (e.g. Fig. 11.1; Section 11.1.3) that are dominated by arkosic rocks and bimodal volcanics dated between 750 to 600 Ma (Ilyin, 1990). These are commonly overlain by thick and laterally extensive carbonate sediments recording the rift-to-drift transition. An Asian continental ice sheet was suggested by Ilyin (1990) but the data are sparse and regional ice masses responding to local palaeogeographic controis cannot be excluded. Late Proterozoic glaciogenic strata of China fall into three principal successions: Changan (c. 800-760 Ma), Nantuo (c. 720680 Ma) and Luoquan (c. 620-600 Ma). The outcrop belts relative to the principal tectonic zonations of China are depicted in Fig. 12.3. Locally-named glaciogenic strata are widespread and are used for inter-regional correlations. Sedimentological details are few and strata are still rigidly interpreted using a climatostratigraphic approach (Sognian et al., 1985). The North and South China blocks were separated during the Late Proterozoic by the Ouinling marine realm (Wang and Oiao, 1984; Zhao et al., 1992); the two areas occupied different paleolatitudes (Fig. 9.2). The oldest glaciogenic horizons (Changan, Nantuo) occur along the southern margin of the south China block (the Yangtze Platform). This margin was the site of middle and late Proterozoic island arcs recorded by thick turbidite and ophiolite succession, e.g. the late Proterozoic volcaniclastic Banxi Group (Wang and Qiao, 1984). This phase was succeeded at about 840 Ma by the development of parallel, linear aulacogens. The precise

N. EYLES

depositional setting of Changan glacial strata that rest both conformably and unconformably on Banxi strata, and elsewhere on carbonates, has yet to be determined, but they are regarded as mostly glaciomarine (Wang et al., 1981). Yianping et al. (1991) report low palaeolatitude magnetic remanence from strata that include Changan glacial sediments; deposition within a belt 8°+ 6° north or south of the equator was suggested. High paleolatitudes were reported however by Zhang and Zhang (1985). The main outcrop belt of Changan strata lies southwest of Changsha (Fig. 12.3). The total thickness approaches 4 km and diamictites are interbedded with "varve-like" slates and "glacial conglomerates" that pass laterally into marine sandstones and siltstones. Large volumes of intermediate to basic submarine lava and tufts are present and record a rifted volcanic setting. These volcanogenic strata strongly suggest that the associated "glacial" diamictites are mass flow in origin though they contain large numbers of glacially-shaped and striated clasts. Changan and Nantuo strata are separated by 700 m of sandstones and shales (Fulu Formation) for which facies information are not available but a marine origin is probable. The Nantuo Formation contains great quantities of volcanic debris, tufts and andesites; in places, reworked pyroclastic debris and airfall tephra make up an appreciable part of diamictite matrix. This formation is the most extensive Sinian "glacial" unit extending some 1500 km from Kunming to Hangzhou (Fig. 12.3) and contains diamictite beds that approach 40 m in thickness. Again, details of depositional settings await further study but deposition in elongate rifted marine basins with strong volcanic and glacial influence is clearly indicated. The youngest Sinian glacial deposit (Luoquan Formation) outcrops over a large area of the North China block (Fig. 12.3). Zhaochang et al. (1993) describe the depositional record of the Luoquan glaciation preserved along the west margin of the North China

EARTH'S

GLACIAL

RECORD

AND ITS TECTONIC

101

SETTING

= ,~1

Zhengmuguan

= ~1 IQQI(~LJan South Ten~

Tuerkeng ~ =Ol

/i

-

ti

,!

i

-

-

-

L

c c P=G •

~

=

~a °

i,

A 01

~J

L2

I~DOLOMffE

Fig. 12.4. Eitbofacies logs through the [.,ate Proterozoic Zhengmuguan Formation, NingxJa Province, China (Fig. 12.3). Modified from Zhaochang eta]. (1993). Note simple association of diamictites containing rafts of dolomite and turbidites suggesting a subaqueous mass flow origin for diamictites. Compare with Figs. 4.7 and 8.3.

1 2 3 4 5 6

KIMBERLEY OFFICER ABADEUS NGALIA GEORGINA \ ADELAIDEGEOSYNCLINE 0

km ~51500

X'

Fig. 12.5. Continent-crossing Late Proterozoic intracratonic rift zone and associated glaciated basins in Australia; basin outlines are generalized. X - X ' line of section shown on Fig. 12.5. See Fig. 9.2 for geodynamic setting within Late Proterozoic supercontinent.

102

N.EYLES

Block along the Helan mountains near Ningxia (Fig. 12.3). According to these workers, the Zhengmuguan Formation comprises diamictites, up to 75 m thick and containing dolomite clasts, that rest unconformably on thick-bedded dolomites. Diamictites are thick and massive in their lower part passing up

X

into stratified diamictite facies with dropstones. In turn these pass up into a thick sequence of thin-bedded and laminated siltstones (turbidites) with common trace fossils. (Fig. 12.4). This succession is overlain by Cambrian shales and is representative of much of the Luoquan glacial record of the

X

TORRENSHINGEZONE

!

t

GAWLER¢RATON

~

ADELAIDE

GEOSYNCLINE

0-

~ i [ ~ P E N A G R O U P ~ - - . . . . . . ~ ~ 6 7 6

w,,PENA

+/- 204

GROUP

2-

+ \4\'%,\\

~

~

<

46-

POSSIBLE /

~ 10-

1%~13,X:1~,

,%-----'~.= . . . .

UNCONFORMY I' jr Ii ~v.L~; ~ i ~=.__...~,i ~ I~

T"t;.'F/'J

x

I ~

~

~',i,Jl~ \ / l~l~

B U R R G R o R ~,

_o

~7

12-

14-

~ -I-

~

~ | I ~

.

'

7

/ fl//~x/~

A ."I[-~.~ ~ i~ ' ~'~--

~A GROUP

16-

18-

+

+

+

.... ::::'~:':: +

2O

+

+

+ +

I'~

MARINOANGLAC1ALSI - ~ l

STURTA INGLACA ILSI ~ SL ITSTONE~ SANDSTONE~~-----]TURBD IT IES

p~

LIMESTONE

DOLOMITE

O z-

+

46E o 10-

EVAPORITES~

BASICVOLCANICS

POST-RIFT ~ , . ~

+

+

14-

16-

~

jPALE OSHELFEDGE ~-~Jg-"~ T

~ • ~ ~f "~--.--~-~ + ~,7,~ ~"-/_.~LI_~_~.~, ~ i /

CONTINENTAL BASEMENT

~-

[--~1 IRONSTONES~

'

+

+

+ ~ +

+

+PRE."~ ~'~k\+

~

--

--U K - -

--

" ~ - SED!MENTS~ " LJ

+ " " ~ " ~2' ~" ;"- ~"~

OCEANICBASEMENT

+ I

100

200

I

300 KILOMETRES

I

400

I

S00

Fig. 12.6. Top: Northwest/southeast cross-section through Adelaide Geosyncline (Fig. 12.5) with bounding age-dates. Length of section is about 500 km. After Gostin (1986). Bottom: Generalized Mesozoic-Cenozoic cross-section through United Utates Atlantic continental margin. After Klitgard et al. (1988).

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

North China Block which Zhaochang et al. (1993) interpret as a record of initial subglacial conditions followed a period of ice sheet oscillation and postglacial transgression. Guan et al. (1986) reported striated pavements and infer a considerable, probably faulted, topographic relief; highly variable striation directions were argued to be consistent with a marine margin dissected by fiords (Guan et al., 1986). This might suggest that the Luoquan diamictites depicted in Fig. 12.4 provide a record of glacially-influenced mass flow rather than subglacial deposition. The intimate association with turbiditic facies, uplifted dolomites and evidence for a high relief setting provide important contextual clues as to possible depositional settings. Zhao et al. (1992) report Early Palaeozoic palaeomagnetic poles from the North China block consistent with a low ( < 35) palaeolatitude for glacial deposition. The data base is very limited however, and the possibility of widespread Early Palaeozoic remagnetisation is real; the palaeogeographic positioning of the North and South China blocks at this time remains controversial. These blocks may have been part of Laurentia, Gondwana, Pacifica or were widely separated (Hsu et al., 1990); some workers note strong palaeomagnetic and stratigraphic affinities with Australia during the Late Proterozoic (Barrett et al., 1989). 12. 4 Australia

Late Proterozoic intracratonic basins of southern, central and northwestern Australia, contain two prominent diamictite successions interpreted as glaciogenic (Figs. 12.5, 12.6). "Cap" dolomites provide important stratigraphic markers (Coats and Preiss, 1980). In the Adelaide Geosyncline the oldest glaciclastic horizon (Sturtian) is older than 750 + 150 Ma and younger than 802 _+ 10 Ma (Young and Gostin, 1989). The age of the youngest horizons (Marinoan) is constrained by a R b / S r date of 676 _+ 204 Ma from overlying shales (Coats, 1981). Similar ages are

103

recorded in the Kimberley region of Western Australia (Coats and Preiss, 1980). The Late Proterozoic glaciclastic strata of Australia may be directly correlative with the two phases of glaciation recorded in northwestern Canada (Young, 1992a, b). The overall tectonic setting in the Australian sector consisted of subsiding graben basins on the margins of glaciated cratonic blocks (Figs. 9.2, 12.5); very substantial thicknesses of glaciclastic strata have been accomodated (Fig. 12.6).

12.4.1 Sturtian glacial strata of Central and South Australia The most well known Sturtian diamictites occur within the Yadnamutana Subgroup of the "Adelaide Geosyncline" of South Australia; the group is thickest and most well-exposed in the Mount Painter area of the northern Flinders Ranges (Coats, 1981) and was deposited in a west-east orientated rift system. Correlative units were deposited in subsidiary rifts to the south in the central Flinders Ranges, to the northeast along the margins of the Broken Hill and Euriowie blocks of New South Wales (Tuckwell, 1981; Young 1992b), in the Ngalia, Georgina and Amadeus basins adjacent to the Arunta Block of central Australia (Walter, 1981; Wells, 1981; Lindsay, 1989) and around the margins of the Kimberley and Sturt blocks in Western Australia and the Northern Territory (Plumb, 1981). A picture emerges of syntectonic deposition in rapidly subsiding, fault-bounded, deep water basins on the margins of uplifted, glaciated crustal blocks similar to that identified for many Late Proterozoic glaciated basins (Figs. 8.3, 12.5). Young and Gostin (1989) describe the Bolla Bollana "Tillite", exposed around Mount Painter in South Australia, which locally approaches a thickness of 2500 m (Coats, 1981); even greater thicknesses (3300 m) are reported for the Pualco "Tillite" in the central Flinders Ranges. The "Tillite Formation" is distin-

104 guished from the underlying Fitton and overlying Lyndhurst Formations by being "diamictite dominated" (Young and Gostin, 1989, p. 836). Massive diamictites units (up to 30 m thick) occur conformably with graded diamictites, conglomerates, sandstones (some up to 10 m thick) and thin-bedded and laminated silty-mudstones with occasional dropstones; thickening-upward and thinning-upward sequences are common. It is clear that a great range of facies has previously been included under the term "tillite" (Coats, 1981; Walter, 1981; Wells, 1981) though striated clasts are reported. Facies types, associations, sequences and thicknesses of the Sturtian tillites record deposition in deep troughs supplied by large volumes of coarse and fine debris. Diamictites are predominantly of mass flow origin and occur intimately within thick turbidite successions; a slope/base of slope setting appears likely but a specific paleogeographic setting (submarine fan, fan delta, point source line source, etc.) awaits detailed study. Young and Gostin (1989) recognized the importance of mass flow into deep basins and suggested that diamictites included a large component of "rain-out" sediment derived from suspended sediment plumes and floating ice as described by C.H. Eyles et al. (1991) from Alaskan late Cenozoic shelf deposits. This sediment was then reworked downslope by mass flows to accumulate as very thick diamictite successions in the deeper parts of the basin. A direct glacial contribution to the sediment flux is recorded by striated clasts but this source could have been of secondary importance to poorly-sorted sediment supplied from fault-bounded uplifted blocks (Fig. 8.3). Sturtian "tillites" record mass-flow on the unstable margins of a complex, continentcrossing rift system (Figs. 9.2, 12.5); the rift was broken into subsidiary basins and depocenters by a considerable fault-controlled relief. "Tillites" mark the initiation of intense subsidence and corresponding marginal uplift sufficient to accommodate impressive

N.EYLES stratigraphic thicknesses and to generate "adiabatic" glaciation along rift flanks. The Ross Sea/Weddell Sea Rift System of Antarctica provides a good modern analog (Fig. 24.2). Sedimentological details of younger Sturtian "glacial" strata are few and prohibits detailed interpretation but speculation is in order. In South Australia, the Bibliando, Merinja, Hansborough, Appila and Sturt "tillites" are generally mud-rich and occur at the base of siltstone, mudstones and sandstone successions which are locally rich in dolomite and have "cap dolostones". "Tillites" grade laterally into thick shallow and deep marine facies (e.g. Wilyerpa Formation; Coats, 1981). Ice-contact facies are absent (see Wells, 1981, p. 523); "tillites" most likely record renewed episodes of subsidence and tectonically-triggered mass flows possibly on fan-deltas prograding into enclosed basins. Erosion of dolomites in the hinterland as recorded by abundant clasts in the diamictites, suggests dolomites are detrital. Capping dolomites may record clastic-starved sedimentation in basins of restricted circulation receiving reworked dolimitic-rich sediment. Lindsay (1989) discussed the nature of depositional controls on different glacial facies associations in the Areyonga Formation contained in the 800 km long Amadeus Basin in central Australia (Fig. 12.5). These strata are probably correlative with the Sturtian glacials discussed above but are poorly constrained by radiometric age dating to between 800 and 700 Ma. The purpose of the study was to show how sequence stratigraphy (e.g. Vail et al., 1984) could be used to define a generalised model for basinal sedimentation that would clarify understanding of global climatic events in the Late Proterozoic. Sedimentological details were not presented but Lindsay (1989) was able to recognise a distinct "Areyonga depositional sequence" composed of genetically-related facies bounded above and below by unconformities (i.e sequence boundaries). The lower se-

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E ' I T I N G

quence boundary in the Amadeus Basin is a glacial erosion surface, complete with subglacially-cut channels, overlain by tillites, ice marginal marine and distal glacially-influenced marine rocks. The upper part of the sequence is composed of carbonate-rich anoxic shales capped by tidal deposits. The sequence was interpreted as the result of a single cycle of ice advance and retreat recording initial glacio-isostatic sea level fall, ice advance onto a shelf platform, glacio-isostatic crustal loading and increased water depths followed by ice retreat, crustal rebound of the basin and erosion of deep water deposits in shallow water. Comparable depositional sequences are well known from Pleistocene glaciated terrains (e.g. Andrews, 1978; Boulton and Deynoux, 1981; Domack, 1983; Scott et al., 1987). Lindsay (1989) assumed that the initial sea level fall recorded the formation of "globally distributed Proterozoic ice sheets" and further, that the growth and decay of local ice masses on the Arunta Block was sychronous. Both these assumptions are clearly difficult to substantiate but Lindsay (1989) concluded that "the identification of such a depositional sequence in the stratigraphic record would indicate ... a global ice age event regardless of whether correlations had been established with glacial sediments elsewhere on the globe". Certainly, there is no question that intrabasinal analysis of very poorly dated Proterozoic strata is best approached through a sequence stratigraphic framework but it is difficult to agree with the assumption that such sequences, where defined, are globally correlative. Lindsay (1989) stressed the importance of basin subsidence in allowing accomodation of glacial deposits and ascertained that the Areyonga deposits are associated with a well-defined phase of crustal extension. Distinction of depositional sequences resulting from regional or intrabasinal tectonics from those of global scope resulting from world-wide changes in global sea level is clearly the major problem (e.g. Miall, 1992)

105

12.4.2 Marinoan glacial strata of South Australia Marinoan "tillites", dated at c. 670 Ma, were deposited under a similar tectonic regime as identified above for Sturtian strata. In the eastern part of the Adelaide Geosyncline, the Pepuarta Tillite of the Yerelina Subgroup is equivalent westwards to the Elatina Formation which was deposited on the Stuart Shelf. "Tillites" contain very few clasts, most of which are unoriented, a dolomitic or calcareous matrix and are interbedded with laminated siltstones. Associated facies include conglomeratic sandstones, massive and laminated siltstones and "cap" dolomites (e.g. Nuccaleena Formation). An eastern source area is indicated for diamictites, a predominantly western source for associated facies on the Stuart Shelf. "Tillites" may be mass flows triggered from an eastern uplifted source area or "rain-out" facies where clasts have been dropped into shelf muds by floating ice; detailed facies studies await completion. Periglacial structures, such as ice-wedge casts, are reported by Williams (1986) from the landward margins of the Stuart Shelf suggesting that clasts may have been rafted by winter pack ice.

12.5 Extensional tectonics, uplift and glaciation along the palaeo-Pacific margin of Laurentia Discussion so far has emphasized the thick and well-exposed Sturtian and Marinoan successions of the southern part of the continent-crossing rift system (Fig. 12.4). The same facies and successions are seen, with local differences, to the north and west across Australia. Strata may not be strictly co-eval given that deposition occurred in widely separated basins. The presence of cap dolomites, stromatolites and shallow water depths inferred from many successions (e.g. Plumb, 1981) suggests the importance of restricted brackish or lacustrine basins. A detrital origin for dolomites can be inferred given that

106 some strata contain dropstones (Plumb, 1981, p. 510). In some localities, volcanic horizons are intimately associated with Sturtian glacials (e.g. Wantapella Volcanics in the Sturtian succession of the Officer Basin; Shaw et al., 1991). Volcanic activity may explain local enrichment in base metals within laminated siltstones (e.g. Walter, 1981, p. 527) and banded iron formations (e.g. that associated with the Chambers Bluff Tillite and the Braemar Ironstone facies of the Pualco Tillite; Coats, 1981). These facies, together with dolomites, suggest the existence of enclosed fault bounded basins with a very restricted circulation and is further indicative of a rift setting. Lindsay et al. (1987) presented tectonicsubsidence curves for the Late Proterozoic to Early Cambrian basins of central Australia (Bonaparte, Georgina, Amadeus, Officer and the Adelaide Geosyncline). These workers identified two periods of accelerated subsidence, at 900 Ma and 600 Ma, which identify failed rifting events within the Australian sector of the Late Proterozoic supercontinent. Lindsay et al. (1987) suggested that these basins are not simply intracratonic in character and were affected by other tectonic events that are at present only poorly understood. This important qualification is supported by Myers (1990) who argued that the Proterozoic record of Western Australia does not in fact, support the notion of an intact Late Proterozoic Australia with orogenesis limited to rifting of intracratonic basins. Instead, Myers (1990) argued that crustal aggregation continued until 600 Ma. A major collisional event (Paterson Orogen; 750-550 Ma) on the eastern margin of the West Australian craton resulted in deep synorogenic foreland basins; the relationship of this event with the Adelaide geosyncline and associated basins in central Australia is not known but clearly complicates any simple "rift" model for these basins (Myers, 1993). Irrespective of the precise tectonic configuration in which the thick Sturtian and Marinoan glacially-influencd strata were deposited, the impor-

N.EYLES tance of syndepositional tectonism can be emphasized. In summary, the Late Proterozoic "tillites" of Australia may provide a detailed record of pulsatory uplift and subsidence along a complex glaciated rift system from the northwestern part of Western Australia to South Australia. Accomodation of very thick successions dominated by mass flows and turbidites requires very active source area uplift and subsidence. The application of sequence stratigraphic techniques to these glacial strata would be extremely valuable and would aid correlation of the Australian, Chinese and North American successions. Several authors have pointed out stratigraphic similarities along the western Late Proterozoic margin of Laurentia and invoked ice sheets of continental dimensions and associated cycles of glacio-eustatic sea level change. Eisbacher (1985) suggested a broad correlation of the glacial record of northwestern Canada (Section 12.1), Australia (Sections 12.4.1, 12.4.2) and China (Section 12.3) based on recognition of a lower event dated at about 750 Ma and younger glacial strata deposited sometime around 650 Ma (Fig. 6.1). Young (1992) and Hutchinson (1992) reported new stratigraphic studies from southeastern Australia and northwestern Canada that suggest close juxtaposition of these two areas (Fig. 9.2). The very great thicknesses of glacially-influenced strata in these areas, dominated by mass flow diamictites, turbidites and olistostromes, indicates a common record of source area uplift and deposition in rapidly subsiding rift basins. The widespread distribution of such deposits on the pacific margin of Laurentia suggests a common tectonic control. The formation of ice covers on uplifted crustal blocks and the preservation of glacially-influenced marine strata within thick shoaling-upward stratigraphic "cycles", could both be the response to repeated phases of lithospheric extension accompanying the initial rifting and eventual break-up of the Late Proterozoic supercontinent.

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SETTING

13. THE TILLITE/DOLOMITE ASSOCIATION AND PALAEOMAGNETIC CONSTRAINTS; A LATE PROTEROZOIC PARADOX? "As a rule, the more bizarre a thing is the less mysterious it proves to be." Sherlock Holmes, A. Conan-Doyle; The Red-Headed League (1891) Descriptions of Late Proterozoic mixed carbonate/glacial successions, where carbonate strata either occur underneath, within or on top of glacial sediments, have generated a wide range of hypotheses to explain rapid alternations between warm and cold climates. Several workers envisage enigmatic "palaeoclimate crises" by assuming conformable interbedding of "warm" limestones and primary dolomites with "cold" glacial strata (Spencer, 1971, 1975). Roberts (1971, 1976) argued for a global Late Proterozoic glaciation produced by the drawdown of atmospheric CO 2 by widespread deposition of dolomite, producing an "antigreenhouse" effect (see Section 21). Other workers have argued for rapid plate tectonic movements (Piper, 1985), low latitude glaciation (Harland, 1964; Williams, 1975; Section 13.4), very warm interglacials (Spencer, 1971), cold water deposition of carbonate as seen in modern Antarctic lakes (Hambrey, 1982; Walter and Bauld, 1983), deposition from supersaturated marine brines under ice shelves (Carey and Ahmad, 1960), subglacial carbonate deposition (Page, 1971; Deynoux, 1985; Souchez and Lemmens, 1985), a non-glacial origin for associated clastic sediments (Schermerhorn, 1974) and a detrital origin for associated carbonates (Fairchild, 1983, 1989; C.H. Eyles and N. Eyles, 1983a; Fairchild et al., 1989). Discussion of mixed carbonate and glacial successions has been retarded in the absence of detailed sedimentological descriptions of both "glacial" and "carbonate' components. Fairchild (1992) reviewed the stratigraphic association of such strata and showed that carbonates occur (1) below many diamictites, (2) within diamictite successions, either as

107 clastic interbeds or as glaciolacustrine precipitates, and (3) above as "cap dolostones". 13.1 Carbonates below diamictite successions

Fairchild (1992) recognized that carbonates below glacial strata in Greenland, Scotland and Spitsbergen show an upward change from deep water limestones to shallow water dolostone facies. In general, the shallowest late Proterozoic carbonates are dolomitized whereas deeper water, offshore facies are dominated by limestones. Inferred paleoclimatic requirements include warm temperatures and active evaporation. The stratigraphic relationship between diamictites and the underlying carbonates is of particular significance. Where diamictites rest on shallow water dolomites the contact is unconformable. Conversly, in those areas where diamictites rest on offshore, deeper water limestones the contact is commonly conformable. Fairchild (1992) suggested that the shallowing-upward sequences of limestone to dolomite found below many diamictites was the result of glacio-eustatic sea level lowering but he was unable to find any independent evidence of climatic change and glaciation. He stated that "the paucity of direct evidence of climatic deterioration is remarkable" (Fairchild, 1992). Ice sheet growth was attributed to a reduction in global temperatures or, less likely, rapid continental drift to higher palaeolatitudes. An adiabatic glaciation model involving tectonic uplift, not glacio-eustatic sea-level lowering, provides a ready explanation for the stratigraphic relationship between dolomites, limestones and overlying diamictites. Upward-shallowing carbonate successions lying unconformably below glacial deposits, can be interpreted as the record of tectonic uplift accompanying the onset of an extensional tectonic regime. Continued uplift is recorded by progressive unroofing of carbonates, exposure of basement lithologies and by abundant clastic carbonate within overlying diamictites. In offshore areas dominated by deeper water

108 limestones, the effects of uplift were either much less or replaced by renewed subsidence such that glacial strata, most commonly glaciomarine sediments, are superposed conformably on top of the limestones. In contrast, the shallower water dolomites were more susceptible to uplift and erosion and are therefore, unconformably overlain by glaciclastic strata.

13.2 Carbonates within diamictite successions It follows from the arguments given above, regarding the enhanced susceptibility to uplift and erosion of shallow water dolomites, that dolomites within diamictite successions are overwhelmingly of a detrital origin. Most dolomitic interbeds can be classified lithologically as dolarenites and record the reworking by ice and gravity of uplifted faultbounded carbonate massifs. Some strata, however, contain primary dolomites and modern day lakes of high salinity in cold, arid settings in Canada, Siberia and Antarctica provide good depositional analogs. Fairchild and Spiro (1990) and Fairchild et al. (1989) report primary lacustrine precipitates from Late Proterozoic strata in Spitsbergen that are directly comparable to lacustrine carbonates deposited in the "arid" saline lakes in present day Antarctica (e.g. Walter and Bauld, 1983). Again, however, the availability of large amounts of reactive detrital carbonate derived from the glacial erosion of underlying carbonate strata can be emphasised (e.g. Fairchild and Hambrey, 1984).

13.3 Capping carbonates "Cap" dolostones lying above Late Proterozoic diamictites are common in Australia, West Africa and North America (Williams, 1979; Deynoux, 1985). Fairchild (1992) stressed the lack of detailed data and could not discriminate between the effects of diagenesis or primary deposition based on currently available descriptions. This situation is

N.EYLES not surprising given that such successions have not been investigated in detail and that the dolomitization process is still not well understood by carbonate petrologists and geochemists. Brathwaite (1991) has emphasized the consensus view that even where direct precipitation has been shown to occur dolomite is probably a synsedimentary alteration product of other carbonates. Furthermore, there is as yet no well-identified relationship between petrographic facies and any particular geological model. As stressed in the section above, probably the most important factor in regard to Late Proterozoic "cap" dolostones was the production of large amounts of fine-grained and highly reactive carbonate rock flour by glacial abrasion of carbonate substrates. Early post-glacial sedimentation in restricted rift basins may have been characterised by "starved" conditions and the reworking of relatively fine-grained carbonate-rich glaciclastic sediment. Fairchild et al. (1989) demonstrate the great utility of oxygen isotope studies of Late Proterozoic carbonates for placing constraints on depositional palaeolatitudes. Initial data suggest that some carbonates are highly enriched in 180 relative to typical values from Late Pleistocene glaciomarine carbonate typically deposited in seawater depleted in 180 (Sections 18.1.1, 20.1). Relative enrichment in the heavier oxygen isotope may suggest deposition in low palaeolatitudes but unqualified comparison with Pleistocene conditions is not warranted given the lack of understanding of diagenetic effects in Late Proterozoic strata. These and other problems associated with carbon, strontium, sulphur and oxygen isotope compositions in Late Proterozoic strata are discussed by Knoll and Waiters (1992). In summary, recent work on mixed carbonate and glacial strata in Late proterozoic successions shows that the availability of large amounts of reactive silt-sized detrital carbonate can explain many of the paradoxical relationships between carbonate and glacial strata. It can be suggested that the wide-

109

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E ' r F I N G

margins of Laurentia. However, there is no consensus as to the precise arrangement of continents at this time and many widely divergent tectonic models have been proposed. Discussion centres on the quality of palaeomagnetic data used to constrain Proterozoic tectonics (e.g. Pesonen et al., 1989; Torsvik et al, 1990, 1991; Van der Voo and Meert, 1991). Concern with the quality and quantity of existing data is especially pertinent given the widely-held view that Late Proterozoic glaciation occurred in low, equatorial, palaeolatitudes (Fig. 13.1). The apparent paradox of glaciations at low latitudes was first identified by Harland and Bidgood (1959) and Bidgood and Harland

spread occurrence of mixed glaciclastic and carbonate successions in Late Proterozoic glaciated basins is essentially a tectono-stratigraphic "motif" reflecting the widespread development of elastic-starved, carbonate platforms in restricted basins to be followed by the sudden input of detrital elastic carbonate from uplifted and glaciated massifs.

13.4 Palaeomagnetic constraints on Late Proterozoic glaciations The preservation of Late Proterozoic glaciogenic strata within extensional rift settings has been demonstrated for the western (palaeo-Pacific) and eastern (palaeo-Atlantic)

RIPHEAN

S

V

C O S D C P T J I

CPN

I~111

60 ° N,

uJ 30", 0".

5

30*.

60* S,

14b0

1zb0

10b0

l

800

660

,

• • • • •

460

North America West Africa North China SouthChina Fennosc~dia Z00

[ I I I I 0

Ma

7 _

,~ 5E 4-

21 N g 0 o I~0o 7=0° 6=0° 5=0° 4=0° 3=0° 2bo i=0° ~o i=0° 2=0° 3=0° 4=0° 5=0° 6=0° 7~)o 8=0° os Latitude Fig. 13.1. Top: P a l a e o l a t i t i u d e of continents and subcontinents and Eartb's glacial record over the past 1500 Ma. M o d i f i e d f r o m C h u m a k o v and Elston (1989). Palaeomagnetic data for Late Proterozoic are assumed to be p r im a r y (see text). Timescale after Harland et al (1989). Bottom: Present day variation of snowline elevation as a function of

latitude. Note rapid lowering of snowline above 30° latitude. After Wegener (1929) and Skinner and Porter (1987). These data emphasize the importance of "adiabatic" glaciers initiated along rifted basin margins undergoing rapid uplifts of several kilometres (e.g. Figs. 8.3, 9.2, 19.1, 19.7, 19.9).

110 (1961) and has been subsequently supported by many workers (Chumakov and Elston, 1989). Possible explanations include a nonglacial origin for the supposed "tillites" (Schermerhorn, 1974), "anti-greenhouse" cooling of the planet following widespread deposition of dolomites (Roberts, 1971, 1976), rapidly-moving tectonic plates and secondary remagnetization in low palaeolatitudes or during burial metamorphism (Crowell, 1983; Morris, 1977; Stupavsky et al., 1982), widespread global glaciation (Coats and Preiss, 1980; Hambrey and Harland, 1985), reduction in the radiative heating of the planet either by an Earth-encircling ring of ice (Sheldon, 1984b) or outbursts of gas and dust from supernovae (Salop, 1977), an increase in the inclination of the Earth's rotational axis (Williams, 1975) and a lengthening of the seasonal cycle to several thousand years brought about by an increasing rate of precession of the Earth's orbit (Malcuit et al., 1986). The most compelling evidence for a low depositional palaeolatitudes of Late Proterozoic glaciogenic strata is forthcoming from the Marinoan succession within the Adelaide Geosyncline in South Australia (Section 12.4.2). These strata comprise the uppermost part of the glacially-influenced Umberatana Group and are dated to about 650 Ma. Paleomagnetic study of fine-grained tidal laminites (Elatina rhythmite member of the Elatina Formation; Williams, 1989) suggest that an original remnant magnetization is present and records a low inclination (< 10°) consistent with deposition in the belt between 20°N and 12°S of the palaeoequator (Schmidt et al., 1991). Co-eval strata, deposited close to sea level on the Stuart Shelf, show permafrost structures such as ice-wedge casts, that under modern conditions indicate a severe climate (mean annual temperatures below -4°C; Williams, 1986) characterised by strong seasonal changes in temperature. The former existence of a severe climate at low, equatorial palaeolatitudes when taken in conjunction with a seasonal temperature

N.EYLES cycle at sea level, clearly cannot be explained by reference to any "normal" configuration of global climate belts. In a fine piece of understatement, Idnurm and Giddings (1988, p. 81) commented that " . . . t h e Australian Precambrian palaeolatitudes do not fit uniformitarian concepts of palaeoclimate'. Other evidence for glacial deposition in low palaeolatitudes ( < 30°) is much less convincing given debate over exactly when magnetization was acquired. Stupavsky et al. (1982) showed, for example, that supposed "low latitude" primary magnetization of the Scottish Port Askaig diamictites reported by Abouzakhm and Tarling (1975) is the result of secondary acquisition during burial metamorphism. Piper (1981) has also, though in more general terms, queried the alleged prmary origin of other Late Precambrian sedimentary strata; such doubts are reasonable given the poor state of current understanding of the primary and post-depositional magnetization histories of sediments that have undergone complex burial histories. Embleton and Williams (1986), recognized that reasonable doubts remain regarding the validity of palaeomagnetic data for certain Late Proterozoic glacial horizons. They argued that if such data are real, then three hypotheses could be suggested to explain low depositional palaeolatitudes. (1) The first, involving glaciation from pole to pole (e.g. Harland, 1964), was rejected since a "frozen-over Earth" would presumably, require an unrealistic increase in solar luminosity to thaw out. As can be appreciated from Fig. 7.1 such an increase is not allowed by any current model of solar luminosity (Newman and Rood, 1977; Endal, 1981; Bahcall and Ulrich, 1988). (2) A second hypothesis is that the present day axial geocentric dipole model of the Earth's magnetic field is invalid for the interval in Late Proterozoic time when glacial deposits were accumulating. Better geophysical data and models are needed to test this. (3) The third hypothesis is that there was a

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'FFING

considerably increased obliquity of the ecliptic in Late Proterozoic time. In this argument, low latitudes will receive less solar radiation than high latitudes, seasonality will be intensified in the equatorial belt and Milankovitch-type changes in insolation might be intensified leading to abrupt changes in climate as evidence by glacial deposits at sea-level in equatorial palaeolatitudes and their association with dolomites (e.g. Williams, 1979). There are some interesting parallels between this hypothesis and the recent suggestion of chaotic behaviour of Earth's planetary motions resulting in dramatic changes in global climate (Laskar et al., 1993). The argument of low-latitude glaciation is weakened by new work on the dolomite/glacial association (see above) and by the demonstrated non-synchroneity of Late Proterozoic glaciation (Fig. 9.2). Supposedly enigmatic associations of glacial and carbonate strata, together with widespread coeval glaciation, have hitherto been major planks in arguments for the existence of equatorial ice sheets. These concepts do not bear rigorous resting against the rock record. Further testing of the hypothesis of low depositional palaeolatitudes for glacial strata is ultimately dependent on more and better paleomagnetic data and geophysical modelling. This is particularly true given intense discussion of the configuration and geographic location of Late Proterozoic continents (Hoffman, 1991; Moores, 1991). These workers offer very different views of Late Proterozoic palaeogeography but are agreed that there were rapid plate motions following the breakup of the Late Proterozoic supercontinent. The establishment of a magnetic remanence in old sediments subject to long and complex burial histories is not well known; secondary remagnetization in low paleolatitudes following rapid plate migrations was emphasized by Crowell (1983). New work bears out the correctness of Crowell's caution; Meert and Van der Voo (1992) show that North America occupied high latitudes during the latest Proterozoic and that previous data on which low

"111

latitude positioning had been inferred is the result of a later Cambrian overprint when North America occupied low latitudes. A high latitude position was followed by rapid drift (9-15 cm/yr) into low latitudes. It can be emphasised that most Late Proterozoic glacial deposits accumulated in extensional, rifted margin basins during the break-up phase(s) of the Late Proterozoic supercontinent (e.g. Fig. 9.2); if arguments for "low palaeolatitude glaciations" are to be taken seriously it needs to be very clearly established that the deposits in question were not affected by compressional deformations and remagnetization at low palaeolatitudes during ocean closure in the Early Palaeozoic (Fig. 13.1). It is also worth noting that the Late Pleistocene ice covers over North America extended as far south as 35 ° latitude; Figure 13.1 shows, for the present day situation at least, the rapid fall in the average elevation of regional snowlines above 30° latitude. These data may not be strictly applicable to the radically different palaeogeographies of the past but they do highlight the importance of tectonic uplift and the generation of "adiabatic" glaciations at "low to middle" palaeolatitudes. These arguments, together with the recent findings of Meert and Van der Voo (1992), throw into sharp relief the need for geophysicists to improve the quality of the Late Proterozoic palaeomagnetic database before exploring non-uniformitarian models of equatorial glaciation. 14. LATE PROTEROZOIC GLACIATIONS: A SYNTHESIS It has been argued above that Late Proterozoic glaciations were neither global in scope nor synchronous but a diachronous response to regional tectonic events (Fig. 9.2). Late Proterozoic glaciations are associated with two principal tectonic settings: that of extending continental margins characterized by uplifted glaciated rift shoulders and compressional continental margins characterized

112 by glaciated volcanic cordillera. The oldest (Sturtian) record is preserved along the palaeo-Pacific margin of the Late Proterozoic supercontinent (i.e. Australia, western North America, China; Fig. 9.2) which was the first sector to undergo rifting and attendant uplift sometime around 750 Ma with a later (Marinoan) phase at about 650 Ma; the Varangian glacial record of the palaeoAtlantic glaciated basins on the far plate margin of Laurentia commenced with initial opening of the Iapetus Ocean at c. 640 Ma recording the final breakout of Laurentia from the supercontinent (Hoffman, 1991). Along the active compressional margin of the Late Proterozoic supercontinent, a late phase of cordilleran glaciation occurred along the Avalonian-Cadomian belt after c. 550 Ma; an earlier glaciation of the West African foreland basin, dated between 630 and 595 Ma, was the regional response to orogenic uplift along the western margin of the West African craton (Figs. 6.1, 9.2, 10.1). Because the timing of Late Proterozoic glaciation appears to be dictated by regional tectonic controls it follows that glaciclastic strata have limited potential for intercontinental and global lithostratigraphic correlations (cf. Chumakov, 1985, 1992). Most Late Proterozoic "tillites" appear to have been deposited either as rain-out facies or glaciogenic mass flows, or some combination of the two, within marine basins adjacent to glaciated volcanic cordillera or uplifted glaciated blocks. Schermerhorn's (1974, 1975) arguments that tectonic differentiation, involving complementary uplift and subsidence, set the scene for the deposition of many Late Proterozoic "glacial" successions, are fully supported by the data presented above. The term "tectogenic" has been used for coarse clastics shed from areas of active tectonics (Steidtman et al., 1986) and the term could be extended to include many Late Proterozoic glacial successions. A local or regional tectonic control on glaciation clearly begs clarification of the precise role of global feedback mechanisms such as at-

N.EYLES mospheric CO: (Young, 1991) and is ultimately reliant on better age dating and correlation. A fundamental problem in many basins is to determine the proportion of the "glacial" component within marine diamictite successions and to identify the regional extent and palaeogeography of glacial activity on the landward margins of the basin where no depositional record survived. The model of a monolithic "global" ice sheet with ice covers on the tropics, acting in unison with large glacio-eustatic sea-level changes, is not supported by the rock record. Nonetheless, the notion that Late Proterozoic diamictites are "tillites" and are the product of rapid and globally-synchronous climatic changes extending to low depositional palaeolatitudes will be a hard one to dispel. This model has been dominant for several decades and is implicit even in recent discussion of the geologic timescale. The Precambrian Subcommission of the International Commission of Stratigraphy (ICS) for example, have proposed that the period between 850 and 650 Ma be called the "Cryogenian" based on the Greek derivation Cryos for ice and Genesis for birth in reference to supposed "global glaciation" (Harland et al., 1989, p. 17). These authors maintain that tillites have the potential for precise correlation from continent to continent. These ideas conflict with the new picture of Late Proterozoic glaciation, that of widely-dispersed ice centres responding diachronously to regional tectonic controls. Recognition of the importance of regional tectonic controls on Late Proterozoic glaciation (Fig. 9.2) is in keeping with what is currently understood of the fossil record. The Late Proterozoic biostratigraphic record is dominated by unicellular procaryotes, eucaryotes and multicellular algae. Thereafter, soft-bodied metazoans (Ediacara faunas) are succeeded in the lower Cambrian by invertebrate groups (Knoll and Waiters, 1992). Glaciation has been implicated in the appearance and disappearance of these organisms involving abrupt anoxic transgressions

E A R T H ' S G L A C IA L R E C O R D A N D ITS T E C T O N I C S E T T I N G

and cooling regressions as a result of global climate deterioration and major glacioeustatic sea level fluctuations resulting from the growth of "super" ice sheets (Brasier, 1982; Solokov and Fedonkin, 1986). It is widely believed that the appearance of Ediacaran faunas was a response to early Cambrian climatic amelioration following "global" Vendian glaciation after 640 Ma (Brasier,

113

1982; Solokov and Fedonkin, 1986). It has recently been shown however, that the first appearance of such faunas predates Vendian glaciation (Hofmann et al., 1990) and that the precise timing and magnitude of any climate or glacio-eustatic changes in sea-level is very uncertain. An important influence on the evolution of ecological niches was regional plate tec-

PA

Fig. 15.1. Three possible apparent palaeomagnetic polar wander paths during the Lower Palaeozoic shown by solid, dashed and dotted lines. (after Van der Voo, 1988; Scotese and Barrett, 1990; Butler, 1991). See text for details. Bottom: Migration of Palaeozoic glacial centres across Gondwana (after Caputo and Crowell, 1985).

114 tonic changes accompanying the breakup of the Late Proterozoic supercontinent (Brasier, 1989). Rising sea-levels in the latest Proterozoic and into the Cambrian (e.g. Fig. 6.1) are not the product of large scale deglaciation of a global ice cover but are a direct record of continental breakup and the resulting thermal subsidence of newly-generated passive margins (Bond et al., 1988). Superimposed on rising relatiue sea-levels created by subsidence were changes in absolute (eustatic) sea-level generated by renewed growth of ocean spreading centres and the consequent decrease in the volume of the world's ocean basins (Bond et al., 1984). The contribution to sea level change made by Late Proterozoic glaciers was probably not great, was shortlived and is difficult to identify. 15. EARLY PALAEOZOIC GLACIATIONS It is common practice to distinguish Early from Late Palaeozoic glaciations and this division is followed herein. It is stressed however, that apart from a short episode between the mid-Silurian to the Late Devonian, there is a near continuous record of glaciation, on some area of the planet, extending from the Late Ordovician to the Late Permian (Fig. 6.1). Further scrutiny of the rock record may well fill this gap and also extend the record back into the Cambrian (Crowell, 1983, p. 247). The Ordovician ice age is of especial interest because this is currently the earliest time period for which accurate plate tectonic reconstructions can be made (Scotese and McKerrow, 1991). Early Palaeozoic (Ordovician, Silurian) glacial deposits are widespread in northern Africa and correlative units of supposed glacial origin also occur in several basins in South America, North America and Europe. Early Palaeozoic glaciation is directly implicated in several well-defined extinctions of marine organisms (Fig. 6.2). Late Ordovician and younger strata record migration of ice centres from central northern Africa westward into Brazil and southward

N.EYLES into Bolivia, South Africa and northern Argentina by the early Silurian (Crowell et al., 1980). Grahn and Caputo (1992) identify diachronous early Silurian glaciations in the Parana Basin (latest Ashgillian), the Amazonas Basin (early middle Llandovery) and Parnaiba Basin (latest Llandovery to earliest Wenlock). No glaciation is recorded from the mid-Silurian to the Late Devonian (Famennian). Caputo (1985) documents well-constrained evidence of Late Devonian glaciation in the Amazonas, Solimoes and Paranaiba basins of Brazil; Crowell (1983) mentions poorly known Late Devonian strata in Africa (Niger). Correlative glaciomarine strata, thought to be of Late F a m m e n i a n / Early Tournasian age, are represented in the Titicaca Basin of Bolivia by the Cumana Formation (Martinez, 1991; Vavrdova et al., 1991). Early Palaeozoic glaciation of Gondwana can be attributed to a high latitude position, the rapid drift of Gondwana across the south pole at about 6 to 8 cm a year (Caputo and Crowell, 1985; Scotese and McKerrow, 1991; Fig. 15.1) and tectonic uplift along the western margin of the Afro-Arabian platform (e.g. Deynoux et al., 1985). Palaeomagnetic and biogeographic data show that the Gondwanan supercontinent lay in high southerly palaeolatitudes and was separated from Laurentia by a wide ocean (Van der Voo, 1988; Scotese and McKerrow, 1991). Thereafter the two continents commenced a collision course until their final docking in the Early Devonian (the Acadian Orogeny) when the Late Proterozoic glacial strata of the AvalonianCadomian Orogenic Belt (e.g. Gaskiers Formation; Section 10.4.1) was transferred, as part of Avalonia, to North America. According to this model, shown by the solid line in Fig. 15.1, the south magnetic pole was located over North Africa in the Late Ordovician, over Chile in the Silurian, somewhere in the vicinity of present day equatorial Africa in the Devonian and in southeast Africa in the Carboniferous. This suggests a drift rate of about 1.5 degrees of latitude every million

EARTHS'GLACIAR LECORDANDITSTECTONIC SETTING

115

years and is in reasonable agreement with what is known regarding the distribution and age of Early and Late Palaeozoic glacial deposits (see Crowell, 1981, 1983; Caputo and Crowell, 1985; Fig. 15.1). However, as emphasised by Butler (1991), the drift history of Gondwana from the Ordovician to the Car-

boniferous is not completely constrained by palaeomagnetic data. In particular, the existence of a mid-Palaeozoic loop, from Brazil to southern Chile and back up to a Late Devonian pole in central Africa, has been questioned. The beginning (Ordovician) and end (Carboniferous) of the drift history is

C NOUAKCHOTT

~

m

OUTCROPOF ORDOVICIANGLAQAL STRATA SUBSURFACEEXTENT OF GLACIALLY-/ INFLUENCED MARINE 1 I - r ~ SUBSURFACEEXTENT OF CONTINENTALGLACIAL

~NIAMEY

[]

"~-~':~:~iii

r~

+

/

PALAEOVALLEYS(FIGS. 1S.4, / 15.5)

[ ] .oB,LEBELTS

/ /

RELATWESEALEVEL "el"-"--/+ I~ 1 --7 (

WIL FM ~

5

UBAHFM WBAN MBR..

rAMADJERTFM

RAHFM I ~ ~FM SIMFM ~

/

:~FM

A (~)

B

C

Lower Devonian fluvial sandstones

(~)

Lower Silurian transgressive shales

(~

Late Ordovician continental to marine glacials

1~)

Cambrian to Late Ordovician continental to marine sandstones and shales

Fig. 15.2. Late Ordovician (c. 440 Ma) glaciclastic strata of western and northern Africa and Arabia. Compiled from Deynoux (1981), Biju-Duval (1981) and Vaslet (1990). Glaciation follows uplift of the African and Arabian platform and cutting of widespread subaerial erosion surface. Uplift is associated with tectonism across W. Africa (see text).

116 reasonably certain and the location of the south magnetic poles at these times is confirmed by the location of the major Ordovician and Carboniferous ice centres in north and south Africa, respectively. The outstanding problem lies with the correct identification of the Silurian and Late Devonian pole positions which are arguable (e.g. Schmidt and Morris, 1977; Morel and Irving, 1978; Van der Voo, 1988). It is possible that the apparent polar wander path connects north Africa and central Africa directly without any mid-Palaeozoic loop through South America (Butler, 1991). This path is depicted by the dashed line in Fig. 15.1. Another argument, advanced by Scotese and Barrett (1990), accepts a Silurian pole in South America but rejects the idea of the Late Devonian pole in central Africa. In this model the apparent polar wander path goes directly from South America to southern Africa (dotted line; Fig. 15.1). New work reported by Chen et al. (1992) confirms the central African position of the south pole during the latest Devonian but argues that it remained there until the Early Carboniferous. More palaeomagnetic data are clearly needed before the Early Palaeozoic drift history of Gondwana can be said to be well established.

N.EYLES

15.1 North Africa and Saudi Arabia Early Palaeozoic (Late Ordovician) glacial strata, spanning some 35 Ma, are a prominent component within the cover rocks of the West African platform (Taoudeni Basin), the Hoggar and Tibesti massifs of the central Sahara and in west central Saudi Arabia on the central Arabian Shield (Fig. 15.2). Possible glacial deposits are also reported from Libya and Egypt (E1-Nakhal, 1990; Vaslet, 1990) and Sierra Leone (Tucker and Reid, 1981b). The first descriptions of Late Ordovician glacial strata, from southern Algeria, were reported by Beuf et al. (1966, 1971). Polar glaciation of Gondwanaland, with the South Pole located just south of the Sahara in the Guinean Gulf, is suggested (Deynoux, 1985a, b; Fig. 15.3). Early Palaeozoic glaciclastic strata are of considerable economic interest; gas is produced from Ordovician glaciofluvial outwash deposits in eastern Jordan (Risha field) where organic-rich "postglacial" shales act as both a seal and source (Beydoun, 1991). Elsewhere, in central Saudi Arabia, the same organic-rich shales are an important source rock for several Lower Palaeozoic siliclastic sequences that include Late Ordovician glaciclastics (e.g. Mahmoud et al., SUBDUCTION

Fig. 15.3. Top: Palaeotectonics and continental positioning of the Late Ordovician (modified from Scotese and McKerrow, 1991).

117

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'I~FING

1992; McGillivray and Husseini, 1992). There is considerable interest in the subsurface extent and character of these Late Ordovician glacial deposits in that they may not only contain hydrocarbons but also by virtue of poorly-sorted, muddy horizons they also may prevent the migration of hydrocarbons from

~ ~ ~

Oceans

I

lice-cap

overlying Early Silurian shales to underlying formations. The consensus view of Early Palaeozoic glaciation is one of repeated radial fluctuations of a large ice sheet, comparable in size to the modern Antarctic ice sheet, across the Afro-Arabian platform fringed by extensive

Emerged land Epicontinental seas

Direction of glacial movement (striae, paleovalleys) o

Presumed glacial pole (PG)



Presumed magnetic pole (PM)

Fig. 15.3. Right: Palaeogeography of Late Ordovician glaciation from a polar perspective (modified from Vaslet, 1990).

118

N. EYLES

mediately prior to glaciation at about 440 Ma is recorded by a widespread metamorphic event along the western margin of the platform in Mauritania and Senegal (Deynoux et al., 1985), in Morocco (Destombes et al., 1985), Algeria and Tunisia (Legrand, 1985). Deynoux et al. (1985) suggest that West Africa resembled a giant, northward-dipping inclined plane with the ice sheet centred over high ground in southwestern Morocco and the present day Canary Islands. In the Taoudeni Basin of western Africa, Late Ordovician glacial deposits of the Tichit Group are capped either by Silurian shales

glaciofluvial outwash fans (Biju-Duval et al., 1981; Deynoux and Trompette, 1981, 1985a, b; Vaslet, 1990). The tectonic setting appears to have consisted of broad-scale tectonic uplift across the platform in response to Taconic orogenesis along the western African plate margin. Uplift, acting in combination with a high palaeolatitude may have been the trigger for glaciation across the platform. In northern Africa and Saudi Arabia glaciallyinfluenced strata rest on a regional disconformity surface that shows an angular discordance with underlying Early Palaeozoic strata. Uplift and tilting of the platform im-

~

42° 28* I"1

60*

40*

I

$halaniyah

I

300

[]

Jauf Formation

[]

Tawil Formation

I ~ ] Tayyarat Formation Jal as Saqiyah

+ +

+

zT*-~

+

+

[]

Sarah Formation

[]

Zarqa Formation

[]

O.asimFormation

[]

Saq Sandstone

[]

Proterozoic Basement

+ + +

4-

-t-

+

~

+

0 ,

10 i

20 ,

30 ,

40 =

"

50 km ,

Salasil

\

"~

C~ Rawd

I

+

~ \ ~

+

+

~

"

"ii~' ~

~

~i '. Unayzah :



26%

Fig. 15.4. Late Ordovician "palaeovalleys" in central Saudi Arabia (see text). After Vaslet (1990). Arrow is location of Fig. 15.5.

119

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

"~.>.

O

~E

O

120 or fluvial and shallow marine sandstones of Devonian age (Fig. 15.2). The lower boundary of the glaciclastic succession is less well defined and varies from Caradocian to Upper Ashgillian. Outcrops occur on the margin of the Reguibat and Leo shields. The basal glacial unconformity is planar on a basin-wide scale but locally shows major palaeovalleys. These can be traced over as much as 50 km, are between 2 and 10 km wide and several hundred metres deep. Valleys are floored by thin diamictites described as tillites and plugged by thick sandstones capped by lacust r i n e / m a r i n e shales. Tucker and Reid (1981a, b) describe correlative glaciomarine deposits from the southern ice sheet sector in Sierra Leone. Valley fill sandstones are massive and present a "ruin-like" outcrop appearance locally showing characteristics consistent with a fluvioglacial origin. "Cordons" comprise elongate outcrops of sandstone that have been interpreted as eskers or valley fills. New data from central Saudi Arabia suggests that these may be controlled by faults and tectonic lineaments in the underlying basement (see below). Deynoux (1981) drew the comparison with the extensive outwash plains along the margin of the present day Vatnaj6kull Ice Cap in Iceland. Glacitectonic structures, thin tillites and striations occur locally at the base of the major palaeochannels. Apparent palaeovalleys of the same scale are a marked feature of the glacial Tamadjert Formation exposed within the sandstone plateaux ("tassilis") around the margins of the Touareg and Toubov Shields in the central Sahara, and within the co-eval Zarga and Sarah Formations on the margins of the Arabian shield in Saudi Arabia (Fig. 15.2). They are commonly aligned along persistent structural lineaments in the underlying mobile belts and are filled with considerable thicknesses of sandstones. "Valleys" are of especial interest to petroleum geologists given their thickness and extent and probable occurrence in the subsurface below organic-rich Silurian shales. Palaeovalleys in the Sahara

N.EYLES and Taoudeni Basins contain thin diamictites, periglacial structures and cross-bedded fluvioglacial sandstone facies. Vaslet (1990) mapped numerous palaeovalleys along the belt of Early Palaeozoic strata that onlap against the margin of the central Arabian arch that affects the Arabian Shield in western Saudi Arabia (Figs. 15.4, 15.5). Valleys are reported as being filled with continental tillites, glaciofluvial sandstones and lacustrine facies and are subdivided into the Zarqa and Sarah formations on the basis of a separating unconformity (Fig. 15.5). These valleys were regarded as the product of erosion and deposition by glaciofluvial outwash rivers draining the eastern margin of the Late Ordovician ice sheet. "Pingos" (Fig. 15.5), rare granite boulders and striated pavements were cited as further evidence of subaerial periglacial conditions along the margins of the ice sheet (McClure, 1978; Vaslet et al., 1987; Williams et al., 1986). The Late Ordovician of central Saudi Arabia is the subject of ongoing re-investigation by geologists from Saudi Aramco in terms of hydrocarbon potential and exploration models and is based on outcrop and subsurface seismic investigations. This work is not yet completed but initial data suggests that there a restricted record of glaciation in central Saudi Arabia. "Palaeovalleys" in several cases are fault-bounded mesas composed of shallow marine shelf sandstones that rest on a widespread erosion surface cut across marine sediments of the Saq and Quasim formations. This surface has a maximum erosional relief amplitude of about 500 m and may record uplift of the central Arabian Shield, subaerial fluvial erosion and subsequent deposition of shelf sandstones during marine transgression. The presence of "pingos" (Fig. 15.5) suggests a rigorous periglacial climate (e.g. Mackay, 1979) but isolated granite boulders previously considered as glacial erratics may have a local non-glacial source (AI-Laboun, 1986). It appears that the influence of the Late Ordovi-

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SE'ITING

cian ice sheet in central Saudi Arabia was limited to the delivery of large volumes of sediment to large coastal and shelf sand bodies. This finding is critical to the development of depositional models and hydrocarbon exploration strategies (McGillivray and Husseini, 1992). There is some information that suggests more ice-proximal conditions in southern Saudi Arabia consistent with an overall coarsening of facies from north to south around the margins of the exposed shield. The adjacent shield attains elevations of over 3500 m adjacent to the Red Sea Rift and there is limited data that suggests this area has been a persistent topographic high that may have supported a local ice cover during Late Ordovician glaciation. In Oman which lies well east of the glaciated area, the "preglacial" uplift event is recorded by subaerial exposure and peneplanation of marine strata; Late Ordovician sandstones of the Ghudun and Safiq formations are of Llanvirnian to Ashgillian age and are overlain by the Sahmah shales of Llandoverian age. The Early Silurian Llandoverian shale can be traced over virtually the entire Afro-Arabian platform (Beydoun, 1991) recording diachronous marine transgression (Vaslet, 1990; Mahmoud et al., 1992). The picture emerges of a Late Ordovician ice sheet that was largely restricted to the African platform with an outlying ice centre in southern Saudi Arabia and adjacent Sudan. The formation of the ice sheet may have been in response to high palaeolatitude (Figs. 15.1, 15.3) combined with "Taconic" uplift and crustal thickening along the western rim of the Afro-Arabian platform. Much further work clearly needs to be done in regard to the origin and precise extent of glacial deposits across the platform but it appears that the overall plate tectonic setting of glaciation is that of large foreland basins. Tectonism around the margin of the platform may not only be instrumental in promoting "adiabatic" glaciation across the interior of the

121

platform but also creates broad forelands in which areally extensive spreads of glaciallyinfluenced coastal and shallow marine sands can accumulate. The great extent of such facies has implications for the thermal regime of the ice sheet. The very considerable thickness of sandstones within the Late Ordovician deposits of Africa and Saudi Arabia is at odds with the notion of a polar ice sheet (Fig. 15.3). The characteristics of these strata are more in keeping with a wet-based ice sheet releasing large volumes of meltwater. The modern polar setting of Antarctica for example, is characterized by severe refrigeration and ice sheet margins that release little or no meltwater to the surrounding oceans. Sedimentation rates of less than 1 c m / k a are typical for much of the Antarctic continental margin (Section 19.1.1). Admittedly, as much as 90% of the Antarctic coastline consists of ice shelves where the nature of sub-ice shelf deposition is hidden from view. Some areas are known to produce large fluxes of sediments to the marine environment (e.g. Alley et al., 1987) but this is not reflected in sedimentation rates on the Antarctic shelf as a whole (Anderson et al., 1983; Elverhoi, 1984; Domack, 1988). Such "sediment-starved" characteristics are not apparent from the sedimentary record of Late Ordovician glaciation which suggests the release of large volumes of meltwater and sediment. It is very possible that the large Early Palaeozoic ice sheet was much more complex in form than that depicted in Fig. 15.3. The ice sheet may have consisted of multiple ice centres indented by shallow interior seaways. Interestingly, the early, pre-Pleistocene Antarctic Ice Sheet probably exhibited a similar palaeomorphology (e.g. Fig. 19.4) and was also temperate in character; recent refrigeration after about 2.5 Ma, appears to be the result of uplift along the Transantarctic Mountains (Section 19.1). It is probable therefore, that the extensive Late Ordovician glacial record of the Afro-Arabian platform

122 records the final decay phases of the ice sheet when climate was much more temperate.

15.2 South Africa Late Ordovician (Late Caradocian to Early Ashgillian) glaciogenic strata occur in the Cape Supergroup of S. Africa. Glaciated pavements, cut across soft sediments are associated with larger scale folding of underlying sandstones attributed to subglacial "dragging" below a large (200 × 103 km 2) ice sheet at about 35°S palaeolatitude (Bell, 1981; Rust, 1981; Visser, 1989b). Published sedimentological and tectonic data are too few to attempt identification of a tectonic or depositional setting.

15.3 Bolivia and Brazil The Bolivian segment of Gondwana may have wandered across the south polar region during the Early Silurian (Fig. 15.1). Crowell et al. (1980) describe thick mass flow deposits such as olistostromes and debris flows of the Cancaniri Formation which locally reaches thicknesses of up to 700 m. This deposit outcrops over an extensive area of southeastern Peru, much of central Bolivia and extends into northernmost Argentina. The Cancaniri Formation is dominated by mud-rich diamictites that record downslope resedimentation of fine-grained glacially-influenced marine strata in an active tectonic setting. Significant ice covers and sediment sources may have been present on the Guiana and Brazilian (Guapore) shields. Grahn and Caputo (1992) have argued for strongly diachronous Early Silurian glaciation in Brazil as a result of rapid migration of a South Polar ice cap. They identify predominantly continental glacial deposits of four glaciations successively preserved in the Parana, Amazonas and Paranaiba basins. The southern margin of the Parana Basin of Brazil (Fig. 15.1) exposes a thick Silurian interval of continental, braided river facies (Furnas Formation) resting on a thinner (16 m) diamic-

N.EYLES tite interval (Iapo Formation) that in turn, rests unconformably on Late Ordovician rhyolites and conglomerates of the Castro Group (Rocha-Campos, 1981a; Caputo and Crowell, 1985). Correlative diamictites of the Vila Maria Formation outcrop around the northern reaches of the basin. The age of these units is uncertain and may be latest Ashgillian a n d / o r earliest Llandovery (A.B. Franca, pers. commun., 1991; Grahn and Caputo, 1992) and the precise depositional origin of the diamictites is not clear. Nonetheless, a moderately high palaeolatitude is suggested at this time (Fig. 15.1) and a cold, periglacial setting appears likely. Similar uncertainty as to a precise glacial origin and tectonic setting surrounds correlative units preserved in the Amazonas Basin (Fig. 15.1). Diamictites inferred to be of subglacial origin (Caputo and Crowell, 1985; Grahn and Caputo, 1993) occur in the Upper Ordovician-Lower Silurian Trombetas Formation of the middle and lower Amazonas Basin. Recent work suggests an early middle Llandovery age (Gregarius graptolite zone; Grahn and Caputo, 1992). The Nhamunda Member consists of thin (15 m max) diamictites interbedded with cross-stratified sandstones and richly-fossiliferous, bioturbated horizons indicating a near-shore marine setting. The Nhamunda contains three stratified diamictite horizons which have a markedly lobate plan form across the basin with a lensate cross-sectional form. These have been interpreted as subglacially-deposited tills filling submarine channels but no indication of a glacial setting has been identified (RochaCampos, 1981b). Diamictites are conformable with shales and sandstones showing slump structures and a debris flow origin appears likely. The youngest record of Early Silurian glaciation is present in the Parnaiba Basin (Fig. 15.1), where the Late Ordovician to Early Silurian (early late Llandovery to earliest Wenlockian; Grahn and Caputo, 1992) Serra Grande Group has been argued to record a glacial influence on sedimentation

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

in the form of coarse-grained, bouldery sandstones (Ipu Formation). Other glacial deposits of this age are recognised in the Paranaiba, Jatoba and Cariri Valley basins. Diamictites within the more well studied Parnaiba Basin are a minor component of the stratigraphy indicating either a marginal glacial influence, extensive reworking of tillites or an alternative non-glacial depositional setting involving deposition of coarse grained sediment on a fan delta. Faceted and striated clasts are reported by Caputo and Crowell (1985). The tectonic setting of the basin is not well understood (Gabaglia and Milani, 1990). The Serra Grande Group (up to 900 m thick) is dominated by marine shales and sandstones and strong subsidence is suggested by the thickness and short time frame of deposition. This may have been accompanied by source area uplift along steep, faultbounded basin margins. Reactivation of Late Proterozoic mobile belt structures that underlie the basin was likely an important control as was the case with subsequent Late Palaeozoic glaciations (see Section 16.2.1). Assuming that the Early Silurian strata of the Brazilian basins contain true glacial deposits then rapid migration of ice covers from central North Africa (late Ashgillian) to the central part of Brazil in the middle Llandovery can be suggested (Grahn and Caputo, 1992). 15. 4 Europe and North America Several basins are alleged to contain Early Palaeozoic glacial sediments both in eastern North America (Schenk, 1972; McCann and Kennedy, 1974; Pickerell et al., 1979; Schenk and Lane, 1981) and northwest Europe (Steiner and Falk, 1981; Fortuin, 1984; Dore et al., 1985) but the data are controversial. Schenk (1972) described metamorphosed diamictites in the Late Ordovician White Rock Formation of Nova Scotia and attributed finely polished and striated clasts to glacial activity on the margins of a North African ice

123

sheet at a time when eastern Canada and North Africa were contiguous. Other independent evidence for glaciation is not forthcoming and the association of diamictites with tufts and other volcanogenic strata and marine shales strongly suggests a non-glacial origin; the character of the "glacial striations" accords with those described by Winterer (1963) as being produced by clast rotation during metamorphism. Long (1991) identified a non-glacial mass flow origin for diamictites of the Cosquer Formation in Brittany and suggested a similar origin for alleged tillites reported from northern Newfoundland by McCann and Kennedy (1974) and Pickerell et al. (1979). Robardet and Dore (1988) and Brenchley et al. (1991) argue for extension of the Late Ordovician ice sheet into peripheral areas such as Portugal and the former Czechoslovakia on the basis of Hirnantian diamictites interpreted as the product of direct depostion from glaciers. The published data are not compelling given the description of such facies as being laminated and commonly graded (see fig. 2 in Brenchley et al., 1991) and associated with marine mudstones and volcanics. These characteristics suggest a sediment gravity flow origin; the presence of dropstones at least indicates a seasonally cold coastline. So-called "glaciomarine" pebbly mudstone facies, close to the Ordovician/ Silurian boundary, occur within the Orea Shales of northeast Spain (Fortuin, 1984) and in the "Lederschiefer" of Thuringia (Steiner and Falk, 1981). Both are characterised by floating clasts and sediment rafts attributed to rafting by icebergs but a seasonally cold, non-glacial origin cannot be ruled out (e.g. Reimnitz et al., 1992). Fortuin (1984) refers to the association of such facies with storm deposits in the Orea Shales and suggested a shelf setting subject to glacioeustatic-driven changes in sea level. Further detailed sedimentological study, along the lines of Long (1991) is clearly warranted before any realistic statement regarding the extent of Early Palaeozoic ice covers in Eu-

124

N. EYLES

rope and North America can be attempted. 15.5 Late Ordovician glaciation and mass extinctions

Late Ordovician glaciation has been implicated in the mass extinction of many marine families toward the end of the Ashgillian, close to the Ordovician/Silurian boundary (Spjeldnaes, 1976; Barnes, 1986; Stanley, 1988; Albritton, 1989; Brenchley, 1989; Barnes and Williams, 1991). The Early Ordovician is characterized by a major biotic radiation when skeletonized suspension-feeders such as brachiopods, bryozoans and corals became dominant. These faunas became especially diverse by the mid-Ordovician as they took advantage of the full geographic range of shelf environments on widely separated continents. Ocean closure after the midOrdovician, particularly in the Caradocian,

System Series

Age (Ma) Z <

Stages p ~

Graptolite zones

Glacio-Eustatic sea-level

the coalescing of previously distinct biogeographic realms and lowered diversity set the scene for a series of extinction events in the Late Ashgillian when many families, genera and species of trilobites, brachiopods, corals, graptolites, conodonts, echinoderms and acritarchs disappeared. About 12% of all families became extinct (Fig. 6.2). The Late Ordovician extinctions have been recognized as one of the best examples of ecological collapse (Brenchley, 1989). Their timing is tightly constrained by reference to standard graptolite zonations; evidence of major sea-level excursions occurs in the Hirnantian, just below the base of the Glyptograptus persculptus Biozone at about 440 Ma (Fig. 15.6). This zone can be widely correlated across the extensive carbonate platforms of North America, Baltica and Gondwana by karst surfaces and shallowing-upward sequences (Stridsberg, 1980; Leggett et al., 1981; Fortey, 1984; Kobluk, 1984). The

Environmental change

Biotic changes

sea-level fall

E P. acuminatus

439-

G. persculptus "Major extinction of FINALDEGLACIATION, p4ankton( conodont$, acritarchs) FLOOOINGOFSHELVES and of shallow shelf biota (brachiopods, corals and ostracods)

c .1C. extraordinarius

FULLYEMERGENT SHELVES, OCEANICOVERTURN?

o P. pacificus ~

~ -

o

E

D. anceps

GROWrHOF GONDWANANICE SHEET

Major extinction of deepshelf benthic faunas (trilobites, cystoids)

Major extinction of graptolites Extinction of conodon~sa~J acritarchs in temperate regions

441

Fig. 15.6. Late Ordovician glacio-eustasy and environmental changes; the role of glaciation in ecological collapse. (After Brenchley, 1989.) Estimates of sea-level variation range from 20 to 75 m (see text). Age is approximate (based on Harland et al., 1989).

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

first phase of extinction is seen at the base of the Hirnantian when nearly 75% of all trilobite genera and 25% of brachiopod genera were wiped out during sea-level regression. A subsequent sea-level rise toward the base of the Glyptograptus persculptus zone in the mid-Hirnantian (Fig. 15.6) saw dramatic reductions in conodonts, brachiopods and achritarchs; for a short period in the early Silurian there were few or no coral reefs (Brenchley, 1989). The first extinction event at the base of the Hirnantian can be directly related to a dramatic glacio-eustatic sea level fall marking expansion of the Late Ordovician ice sheet. Recent work on the oxygen isotope composition of carbonates in central Sweden establishes a substantial shift in 180 consistent with glaciation and a reduction on ocean water temperatures (Middleton et al., 1991). Estimates of the magnitude of sea level fluctuation vary from 20 m (Bjorlykke, 1985) to 75 m (Brasier, 1989) to 350 m (Chen, 1991). Comparison with Pleistocene analogs suggests that the lower figure is more correct. The total extent of the Late Ordovician ice sheet appears to have been comparable with that of the Late Pleistocene Antarctic Ice Sheet (Fig. 15.3). Growth of the latter accounts for about 14% (c. 25 m) of a maximum Late Pleistocene glacio-eustatic sea level lowering of about 175 m (see Fig. 16.21). Abrupt glacio-eustatic exposure of continental shelves, acting in combination with climatic changes in the sub-tropics and influxes of cold bottom waters to the world's oceans, combined to promote mass mortality among plankton and benthos (Sheehan, 1973; Wilde and Berry, 1984). These diverse fauna had hitherto, enjoyed a long period of climatic stability and many occupied very narrow environmental niches. This situation, of abrupt environmental change, contrasts with that for late Cenozoic glaciations where a 50 m.y. long period of climate cooling and sea-level fluctuation preceded the major Late Pleistocene glacio-eustatic regressions (Jablonski, 1985; Brenchley, 1989). The result of the

125

Lower Hirnantian extinction event was the appearance of an impoverished, cool water Hirnantian fauna on extensive tropical carbonate platforms. This fauna was, in turn, wiped out by the subsequent marine transgression during the mid-Hirnantian, which resulted in deposition of black, graptolitic shales recording the spread of warm anoxic waters (Thickpenny and Leggett, 1987). This set the scene for early Silurian radiations. It is worth emphasizing that the abrupt and short-lived glacio-eustatic sea level changes of the Late Ordovician are clearly identified and correlated by faunal extinctions on stable carbonate platforms. Considerably less success has characterized attempts to document coeval depositional sequences on siliclastic shelves essentially because they were limited to areas of tectonic instability. For example, correlations between supposed ice volume fluctuations and facies sequences have been argued with regard to the Appalachian Basin of eastern North America. Dennison (1976) concluded that basinward progradation of the Queenston Delta clastic wedge in the Richmondian has been promoted by glacio-eustatic sealevel fall (see also Fortey, 1984) but more recent work stresses the role of basin tectonics in controlling depositional sequences (e.g. Quinlan and Beaumont, 1984; Bradley and Kusky, 1986; Keith, 1989; Leighton and Kolata, 1990). These and other workers recognize the probable influence of glacio-eustasy but, unlike the case with Late Paleozoic glaciations (Section 16.13), have not been able to isolate this effect because of insufficient high-resolution dating. Global correlations in the Late Ordovician are still poorly established largely because of faunal provincialism; correlation of K-bentonite ash beds holds great promise (Huff et al., 1992).

15.6 A role for glaciation in Late Devonian extinctions? One of the best documented ecological extiction events is that of the Late Devonian.

126

This event lasted between about 7 and 10 Ma and saw a series of slow, gradual extinctions during the Givetian and Early Frasnian followed by an abrupt event at the end of the Frasnian when 60% of prexisting shallow marine benthic fauna disappeared (Goodfellow et al., 1988). Glaciation has been implicated in the development of "cold oceans" (see Copper, 1986; Raymond, 1987) but seems improbable given uncertainty over the extent of Late Devonian glaciers (see above) and the probability that such ice masses were small and unlikely to have exerted any global effects on sea level (e.g. Caputo, 1985; Martinez, 1991). Thompson and Newton (1988) downplay any role for ice in the extinction events and instead, present evidence that increasing ocean temperatures and expansion of oxygen minimum zones played a major role. Other possible mechanisms, including bolide impacts, are reviewed by Goodfellow et al. (1988) and Albritton (1989). 16. LATE P A L A E O Z O I C GLACIATIONS

"The ancient boulder clays of the southern continents bring us face to face with... the fiercest and longest winter of the ages". Coleman, 1916, p. 186. Late Palaeozoic glaciation of Gondwana lasted for about 100 m.y. (350-250 Ma) and thus represents the longest period of continuous glaciation in the Phanerozoic (Figs. 6.1, 6.2, 16.1). Glaciclastic strata occur in a large number of active margin and intracratonic basins (Fig. 16.2). Glaciation of the southern continents coincided with the "greatest episode of coal deposition in Earth history" (Robinson, 1991) and many Late Palaeozoic basins are of considerable economic importance with regard to hydrocarbon resources (e.g. Runnegar, 1984; Martini and Glooschenko, 1985; Williams et al., 1985; Martini and Johnson, 1987; Levell et al., 1988; Goldstein, 1989; Franca and Potter, 1991). There is active exploration of many Late Palaeozoic glaciated basins designed to further under-

N. EYLES 1 --250-

2

3

4

5

6

7

8

9

tu TATARIAN

253. ~ KAZANIAN 258-260- Z KUNGURIAN 263-
ASSELIAN

286

-290

STEPHANIANi - - . - -

296. -300

>, ~

WEST-

PHALIAN

-81o

315. iT" W

-32o- _ _ . -324 -330

~ on n"

"340

NAMURIAN

~

352.

VISEAN

-

-

TOURNAISIAN --360

, $

FAMMFNIAN

I

I~

DEVONIAN

Fig. 16.1. Timing of Late Palaeozoic glaciation across Gondwana. After Veevers and Powell (1987) and unpublished data from Petrobras and Yaciementos Petroliferos Fiscales Bolivianos and A.B. Franca, Pers. Commun. (1991, 1992). Numbers as on Fig. 16.3. Note a Carboniferous start in the west and a largely Permian record in the east. A, Titicaca Basin; B, Amazonas, Paranaiba, Solimoes Basins. Data from Paran~ Basin relate to age of preserved glacial record; a lengthy period of glacial erosion may precede preservation. Shading does not imply continuous glaciation (e.g. Gonzalez, 1990).

standing of complex and little-understood glacial reservoirs. Recently-available subsurface data identify the important role of regional tectonics in controlling basin subsidence and the age and sedimentary character of the glacial infill. No one single continental ice sheet developed during the Late Palaeozoic; instead, complexes of ice masses waxed and waned diachronously across the supercontinent (Crowell, 1983). In general, the glacial record "youngs" eastward across Gondwana and is generally of Carboniferous age in the west and Permian in the east (Fig 16.1). Glaciation appears to have been triggered by tectonism and uplift first along the Argentine

127

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

sector of the supercontinent (Fig. 16.3). The earliest record of extensive Late Palaeozoic glaciation appears about in the Early Visean

(c. 352 Ma) of Argentina and a fading glacial influence, or at least seasonal, cooling, is recorded by dropstones in Late Kungurian

,//

' ) ...o...-.... ,..

v ':':'.".::". .~.

I~4.~ ~ ~

" 21 M~ROCON ~ T S

OROGENESIS

-

~

O

~

[]ICECOVER BERGS.4=. ICE

Fig. 16.2. Top: Late Palaeozoic glaciated basins. (See opposite.) Bottom: Maximum extent of glacially-influenced marine conditions across Gondwana during Late Palaeozoic glaciations. Ice centres migrated from west to east across the supercontinent; no one single continental ice sheet developed. 1. Cauvery-Palar, 2. Godavari, 3. Mahanadi, 4. Damodar, 5. Satpura-Son, 6. Himalayan, 7. Thai-Malay block, 8. Perth-Carnarvon, 9. Canning-Officer, 10. Denman, 11. Arckaringa, 12. Pedirka, 13. Galilee, 14. Cooper, 15. Troubridge, 16. Renmark, 17. Sydney, 18. Tasmania, 19. Transantarctic, 20. Ellsworth, 21. Falkland, 22. Karoo foreland belt, 23. Kalahari, 24. Parana, 25. Congo, 26. Gabon, 27. Zambezi, 28. Chaco (Bolivia)-Tarija (Argentina), 29. Titicaca, 30. Chaco (Paraguay)-Parana, 31. Sierras Australes (Sierra de la Ventana), Colorado, 32. Paganzo, 33. Calingasta-Uspallata/Rio Blanco, 34. San Rafael, 35. Central Patagonian (Tepuel-Genoa)

128

N. EYLES

ation between tectonics, basin formation and the preservation of a glacial record. These basins developed on top of complexly-structured Late Proterozoic basement consisting of cratons (shields) and mobile belts recording the amalgamation phases of the Late Proterozoic supercontinent. Basins formed by reactivation of structures within the mobile belts and are often "cradled" around their periphery by relatively stable shields. Structural lineaments commonly "compartmentalize" the basins in piano-key fashion and have dictated the successive margins of the basins; steep, faulted basin margins were typical. In this way thick successions of glacial strata, containing significant thicknesses of resedimented deposits, were accomodated on so-called "stable platforms". "Far-field" stresses, propagated several thousand kilometres into the interior of Gondwana from its active compressional margins, dictated basin subsidence. Glacial deposits occur on

strata (c. 255 Ma) along the opposite margin of Gondwana in Australia (Fig. 16.1; Frakes, 1979; Veevers and Powell, 1987). Still earlier glaciation of Late Devonian age may also reflect uplift along the western rim of Gondwana (Caputo, 1985; Fig. 16.3). Whilst plate tectonic controls provide a satisfactory explanation for the initiation and subsequent development of Late Palaeozoic glaciations it is premature to ignore a global geochemical control (e.g. Crowley and Baum, 1992). The period from 360 to 250 Ma is conspicuous in Earth history by extremely low values of atmospheric CO 2 (Figs. 6.2, 21.1). A consistent theme in the Late Palaeozoic glacial history of Gondwana is the preservation of glaciclastic strata within embryonic intracratonic basins defined by underlying basement structures (e.g. Fig. 16.4). Late Palaeozoic glaciclastic strata comprise the lowermost stratigraphic component of many Gondwanan basins suggesting a close associ-

7ORO6EN~I8 A

[ ] u~,~r~o uocx

COROILERANGLAC~TION

LATESTDEVONIAN c. 360 Ma

{~

~

C ICESHEETS ANDEXTB~fT OFGLAC P.?4.AS~CSTRATA ~

~

/ ~

3

/

LATEST VISEAN c. 330 Ma

00-- " - - ~ ~~ ~ ~ ~ i

0 LATESTWESTPHALIAN c. 300

B

~.

Ma

\

Fig. 16.3. Orogenic initiation of Late Palaeozoic glaciation around Palaeo-Pacific margins of Gondwana and

subsequent spread of ice covers over the interior. Modified from Veevers and Powell (1987). Numbers refer to age-dated s e c t i o n s o n Fig. 16.1. See text (Section 16.10) for discussion of distinction between age of preservation and timing of glaciation.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

the floors of many basins; typically, younger glacial strata overlap older deposits recording continued subsidence thereby producing a "steer's head" basin infill. In many basins, glacial overdeepening has probably been overemphasized at the expense of syndepositional faulting.

16.1 Initiation of Late Palaeozoic glaciation; polar, adiabatic and global geochemical controis The lack of synchroneity between the infill of the widely-scattered Gondwanan glaciated basins has been known for some time. Du Toit (1927) introduced the hypothesis of "eccentric ice capping" involving migration of ice centres in South America and southern Africa. Crowell (1981) and Caputo and Crowell (1985) enlarged upon the theme of migrating ice centres and argued that the pattern and timing of Early and Palaeozoic glaciation of Gondwana could be explained by the migration of the supercontinent across the South Pole (the "polar" model; Fig. 15.1). During the Ordovician-Silurian-Devonian, ice centres moved from northern Africa to southwestern South America with a notable absence of any glacial record during the Middle Silurian to early Late Devonian. Devonian marine sediments are widespread in South America and contain a low diversity, cold water Malvinokaffric Realm fauna typical of high latitudes (Copper, 1977). There is no evidence of glaciation during this time period perhaps reflecting extensive marine conditions and quiescent tectonic conditions (Barrett and Isaacson, 1987). Late Devonian tectonism along the Andean margin of South America, associated with the collision of "Chilinia", is recorded by a widespread unconformity and, perhaps not surprisingly, with renewed glacial conditions in the Amazonas Basin of Brazil and in the Titicaca Basin of Bolivia (Caputo, 1985; Martinez, 1991). Caputo and Crowell (1985) argue that Late Paleozoic glaciation subsequently spread to southern Africa, Antarctica and

129

Australia as Gondwana rotated and travelled across the south polar region (Figs. 15.1, 16.1, 16.2, 16.3). In contrast to a "polar" model, Powell and Veevers (1987) and Veevers and Powell (1987) argued that widespread Late Paleozoic glaciation of Gondwana did not start until the Namurian and was largely restricted to the mid-palaeolatitudes. They suggested further, that earlier Devonian glaciations in Brazil and those of the Visean in southern South America were also mid-latitude events indicating that a polar position was not the most important control on Late Palaeozoic glaciation. Powell and Veevers (1987) concluded that the first order control on Late Paleozoic glaciation was the closing of the Palaeo-Tethys Ocean, as Gondwana and Laurentia collided, and subsequent orogenesis along the southwestern (South American) and eastern (Australian) margins (Fig. 16.3). Strong Namurian uplift in southern South America and in eastern and central Australia was alleged to be the trigger for extensive "adiabatic" Permo-Carboniferous glaciation during the Late Westphalian (c. 300 Ma) when ice covers extended onto the other Gondwana continents. New data presented by Pique and Skehan (1992) demonstrate similar compressional tectonics along the northwestern margin of Gondwana associated with closure of the Theic Ocean. Elevation of continental areas, either by tectonothermal thickening of the continental crust or by direct continental collisions has also be identified as a precursor to Early Palaeozoic and late Cenozoic glaciations (Sections 15.1 and 19). The application of theoretical palaeoclimatic modelling to Late Palaeozoic glaciations suggests that global geochemical controls may also have played a fundamental role. Crowley and Baum (1992) assessed the relative importance of several factors such as estimated atmospheric CO 2 levels, changes in palaeogeography resulting from plate movement, and the then estimated value of solar luminosity. In addition, fluctuations in

130

N. EYLES

orbital insolation values typical of Late Pleistocene glacial and interglacial conditions were also considered. The model proposed by these workers successfully approximates the overall eastward shift in ice from South America to India and Australia and provides insights into the extension of ice covers across the centre of Gondwana following the initiation of glaciation around its uplifted palaeoPacific rim (e.g. Veevers and Powell, 1987; Fig. 16.3).

INDIA Raniganj Basin ,

Jharia (central) Basin 1

km

2 Jharia (eastern) Basin

Giridih Basin ° 0 0.3 km

l_km.. 0.3_ l

16.2 South America

AUSTRALIA Cooper Basin

0

u2

0 I

~

GondwananSediments

20 I

40 I

km

--q~Basement

60 r

Late Paleozoic glaciated basins of southern South America are of considerable economic importance and there has been renewed study of their stratigraphy and age relationships. In Bolivia, Carboniferous glacial deposits of the Bermejo, Palmer, Montecristo, Madrejones, Sanadita, San Alberto, La Pena, Tita-Techi and Santa Cruz fields are important oil producers (Oca, 1989); correlative strata in Argentina host the Duran, Leva and Madrejones oil fields. In Brazil, the Parana Basin has significant but as yet subcommercial gas shows in the Permian glacigenic Itarar6 Group; significant coal deposits occur in early postglacial strata (Zalan et al., 1990). Frakes and Crowell (1969) provide a comprehensive review of what was then known about the tectonics, characteristics and origin of the Late Palaeozoic glacial record of South America. These workers stressed the impor-

BOMI~

Fig. 16.4. Top: Representative, simplified cross-sections through Gondwanan intracratonic glaciated basins showing structural control by basement faults. Glaciclastic strata occur at the base of many basins recording an intimate relationship between glaciation, subsidence and preservation of a glacial record. Bottom: Principal Late Palaeozoic glaciated basins of the Indian sub-continent (Fig. 16.2). Basins are controlled by faults and structural trends in basement. After Acharyya (1975), Chowdhury et al. (1975), Ghosh (1975) and Battersby (1976).

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

tance of glacially-influenced sedimentation in marine basins and the importance of downslope resedimentation. Veevers and Powell (1987) argued that glaciation was initiated during the Namurian (see above) but new data suggests a much earlier start, perhaps as early as the Late Devonian. Figure 16.1 is a compilation of new age data for Gondwana glaciated basins and shows that Carboniferous glacial deposits began to accumulate much earlier in the southwestern sector of South America during the Visean, at about 350 Ma. Glaciation is also recorded however, in the Late Devonian of the Amazonas, Paranaiba and Solimoes basins of Brazil (e.g. Crowell, 1983; Caputo, 1985). It has been thought that this Devonian glacial record was restricted to Brazil in response to local Hercynian orogenesis along the northwestern margin of Gondwana (Caputo and Crowell, 1985) but new work further south, in the Titicaca Basin of the Bolivian Altipiano has also identified glaciomarine deposits of Tournasian age (Vavrdova et al., 1991). There, the Ambo Group rests with a marked unconformity on underlying Devonian (Fammenian) shales and commences with lowermost deep water mass flows and submarine channel deposits deposited under glacial influence (Martinez, 1991). The succession contains abundant plant microfossils indicating a Tournasian age and possible correlation with diamictites and dropstone facies of Brazil. Thereafter, there is a hiatus in the South American record of glaciation which recommences with glacial diamictites of Visean age in the Tarija Basin of Argentina and the coterminous Chaco Basin of Bolivia (Helwig, 1972; Cuerda and Azcuy, 1986). Carboniferous glacial sediments of the Paganzo, Callingasta-Uspallata, C h a c a Parana and Central Patagonian basins commonly exceed 2000 m in thickness (Gonzalez, 1990; Gonzalez-Bonorino, 1992; F. Wiens, pers. commun., 1992). These deposits cannot be adequately reviewed herein because they are subject to ongoing study, discussion and revision; recent investigations stress the im-

131

portance of structural controls on sedimentation by Late Proterozoic terrane boundaries and lineaments. Previous work stressed the development of alpine glaciers associated with forearc basins inboard of coastal subduction zones resulting from terrane accretion along the Gondwana margin (Frakes et al., 1969; Helwig, 1972). Increasing emphasis is now being placed on relatively oblique collisional processes, the rotation of large foreland blocks and the generation of intracratonic basins that began to receive glaciclastic sediment from uplifted areas of crustal thickening (e.g. Paganzo Basin; Mpodozis and Kay, 1992). Gonzalez (1990) suggested a basin and range landscape for much of western Argentina at this time. The docking of large terranes against Gondwana not only resulted in regional uplift and the opening of intracratonic basins in the interior, but also greatly increased the area of land in high southerly latitudes. It is argued here that one of the principal causative factors in the initiation of Late Palaeozoic ice covers over South America was the southward migration of large continental blocks, such as the Patagonian Terrane, the Malvinas Plateau and Antarctic Peninsula, and their emplacement against southern South America (Ramos, 1989). The accretion of these large continental blocks, essentially completed by the latest Devonian, had the effect of bringing extensive land areas into high southerly palaeolatitudes and thus set the scene for regional glaciation. The growth of ice covers was subsequently "triggered" by rapid uplift along the western edge of these terranes. The onset of glaciation in the Visean of Argentina and Paraguay has been directly related to Hercynian orogenesis along the proto-Andean margin of southern South America (Gonzalez, 1982; Cuerda and Azcuy, 1986; Herve et al., 1987). Figure 16.2 is a very simple representation of the principal paleogeographic zonation at this time with a coastal subduction complex and interior ranges separating independent and partially-linked marine basins opening into a

132

N.EYLES

palaeo-Pacific Ocean. A good example is the Chaco-Tarija Basin of Bolivia and Argentina that was initiated by Late Devonian "Chanic" orogenesis when "Chilinia" collided with the Gondwanic craton. Glaciomarine sediments began to accumulate in this basin and others (e.g. Paganzo Basin) during the Late Tournasian (c. 352 Ma) and were sourced either from Cordilleran-type, coastal ice masses or from small plateau ice caps on interior massifs. To the east, away from the active Chilean and Argentine plate margin, but still in close geographic proximity, glacial sediments only began to accumulate on the Brazilian margin of South America during the latest Westphalian (c. 300 Ma; Fig. 16.1). The Parana Basin of southern Brazil is the largest (1.6 x 106 km 2) intracratonic basin in southern

Fig. 16.6. Isopach map of Late Palaeozoic Itarar6 Group Paran~i Basin. Contours in metres N-S shows line of section Fig. 16.8.

Fig. 16.5. Paran~ Basin, Brazil and outcrop belt of Late Palaeozoic glaciclastic strata (Itarar6 Group). After Franca and Potter (1991).

South America and contains a thick (1400 m) Late Paleozoic glacial succession (Itarar6 Group; Franca and Potter, 1991; Figs. 16.5, 16.6). This succession is poorly constrained by radiometric age dating, and there is as yet no firm ground for establishing the Carboniferous-Permian boundary within the basin (Rocha-Campos and Rosier, 1978). However, a new pollen biostratigraphic zonation indicates glacial deposition between the latest Westphalian (300 Ma) to Early Kungurian (260 Ma; Fig. 16.1). Three formations (Lagoa Azul, Campo Mourao and Taciba) can be recognized within the Itarare Group (Fig. 16.7) each recording a renewed phase of basin subsidence as a result of rifting along steeply-dipping struc-

EARTH'S

GLACIAL

RECORD

AND

ITS TECTONIC

tL ~ zx ,c, z~

I

TACIBA Fm.

A

R

E

CAMPO

N

MOURAO

Fm.

SEGREDO

Mb.

MASSIVE + GRADED SANDSTONES; TURBIDITE DEPOSITS, RARE SHALLOW MARINE INDICATORS.

, ', ', .c.,

R

CHAPEU DO SOL Mb. BLANKET-LIKE MASSIVE DIAMICTITE; RAIN-OUT AND MINOR RESEDIMENTATION. RIO

@

T

A

133

SE'VrlNG

...:.:.,..;~.....: ,.,..,.......,:.~,.......;

G

CAMPO MOURAO Fm. INTERBEDDED STRATIFIED DIAMICTITES AND GRADED/MASSIVE/SLUMPED SANDSTONES AND CONGLOMERATES; TURBIDITE AND DEBRIS FLOW DEPOSITS. RAPID SEDIMENTATION. FIRST MARINE FOSSILS.

'h' "i~'" A

0 U P

LAGOA AZUL Fm.

~

TARABAI Mb. STRATIFIED DIAMICTITES; DEBRIS FLOWS AND MINOR RAIN-OUT.

~

A

~

"

CUIABA PAULISTA Mb. MASSIVE AND POORLY GRADED SANDSTONES, MINOR CONGLOMERATE; TURBIDITE DEPOSITS

PARANA GROUP

Fig. 16.7. Stratigraphy, facies and depositional setting of Itarare Group strata. From C.H. Eyles et al. (1993). The Itarar6 Group has a maximum thickness of over 1300 m (Fig. 16.6).

where (e.g. Gibbs, 1984; Klein and Hsui, 1987; Daly et al., 1989). Each formation constitutes a "deepening" succession composed of lowermost sandy turbidites, recording the reworking of deltaic outwash sediments downslope, overlain by muddy debris flow and rain-out diamictites and shales (Fig. 16.7). The lowermost Lagoa Azul Formation contains continental gymnosperms; overlying

tural lineaments (Fig. 16.8). These structures were produced during the major collisional orogenies responsible for cratonization of the Late Proterozoic Afro-Brazilian supercontinent (Murphy and Nance, 1991). Progressive subsidence of the Parana Basin and accomodation of the Itarare Group was controlled by faulting and reactivation of basement structures similar to that documented else-

+ ~---

120 km

I

+ 75

]

+ 75

I

+ 62"--"4--62

+

+ I

112

I•

+ ~ ----I~

+ 5o-~I~

+ 7s

SOHTH

I

NORTH

FdUlL

Fig. 16.8. Stratigraphic cross-section showing relationship between Late Proterozoic lineaments and successive overstep of younger formations within Itarar6 Group. From C.H. Eyles et al., 1993.

134

N. EYLES

I

20*E

CONGO 'BASIN' and glaciclastic infill of rift basins

-F0*

KAOKOVELD ICE MARGIN

PARANA BASIN

~:~

.__ICE SHEET

WINDHOEK

j

HIGHLANDS /

"x

-...

40* E

+20os

/

PL,,O.E g O EOALONO.ROCTORAL ~"~STRUCTORAL LINEAMENTS(Figs. 16.8, 16.10) gr)-)22~I EASTERNOUTCROPBELTOF ITARARE'GROUP ~/JJ/~;~EXPOSINGSTRIATEDBASEMENTAND TERRESTRIAL GLACIALFACIES [~'~GLACIALLY-INFLUENCEDMARINEFILLOF PARANA I IBASIN(Figs. 16.7, 16.8) C CURITIBA S P SAOPAULO H B HAUBBASIN

Fig. 16.9. Palaeogeography of Late Palaeozoic (c. 280 Ma) glaciation from N. Eyles et al. (in prep.) and Horsthemke et al. (1990).

formations are progressively more marine in character as the Parana Basin opened to an epicontinental seaway between Brazil and southern Africa (Figs. 16.2, 16.9). During deposition of the Itarar6 Group the principal ice centres feeding the basin were not primarily from the west, where orogenesis had been largely completed and glaciated basins filled, but from the southern African (Kaokoveld) margin of Gondwana (Fig. 16.9). 16. 2.1 Tectonic controls on preservation

The apparently delayed onset of glaciation in Brazil, compared to adjacent sub-Andean basins to the west (Fig. 16.1), is the result of the later subsidence and preservation of Permian glacial sediments in the Paranfi Basin

rather than any eastward migration of ice centres (see also Gonzalez, 1990). It is probable that a lengthy episode of Carboniferous glaciation is unrecorded in the basin (see also Section 16.3). Subsidence of the Paran~ Basin appears to have begun when tectonic stresses from the active western margin of Gondwana were communicated into the plate interior (e.g. Fig. 16.10). "Far field" reactivation of intracontinental basement structures, involving the development of extensional rifting by collisional processes at a distant continental margin, is a common process affecting intracratonic basins (e.g. Jorgensen and Bosworth, 1989; Uliana et al., 1989; Howell and van der Pluijm, 1990; Daly et al., 1991; see below).

135

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SEFI'ING

i

'¸¸¸¸'i,

Late Proterozoic Amalgamation And Cratonization c. 800-600 Ma

Rifted Intracratonic Basin / _._____7

Accorr~ation Zone

+ ------

I

+ ÷

+

+

ice Cover

÷ + + + +

Glaciclastic Strata (e,g. Fig. 16.8)

I <. . . .

.'

Late Palaeozoic Extension c. 300 Ma

Fig. 16.10. Development of Late Palaeozoic rifted intracratonic basins by "far-field" reactivation and extension along Late Proterozoic structures. This was a common process across Gondwana and allowed accomodation of substantial thicknesses of glaciclastic strata in the interior of the continent. After N. Eyles and C.H. Eyles (1993).

SILURIAN - DEVONIAN 0-

PERMO-CARBONIFEROUS GLACIATION 1200-

k"~/"

JURASSIC CRETACEOUS

f

I

A

E

3400 -

BACKSTRIPPED

ILl O

3500 -

Itarare G r o u p ~ i i i i ~ i ~ i !

liiiiiili!i!iiiiiiiiiiii:iiii

4800 -

I

6000

i I i

, 425

I

350

'

I

TOTAL SUBSIDENCE

I I

275

I

~

I

200

'

I 125

'

I 50

T I M E (Ma)

Fig. 16.11. Accelerated subsidence of the Late Palaeozoic Paran~ Basin and accomodation of thick (1300 m) glaciclastic strata of the Itarare Group. After Oiiveira (1989).

136 Figure 16.11 shows an example of a total subsidence and backstripped curve typical of wells that pass through the complete Parana Basin infill. This figure clearly shows that preservation of a thick glacial record in the basin is the result of accelerated tectonic subsidence. Maximum rates of total subsidence during deposition of the Itarar6 Group were about 70 m / M a (Oliveira, 1989; Zalan et al., 1990) though this is an average value that ignores episodes of accelerated subsidence associated with faulting of the basin margins. Basin subsidence most likely approximates to simple "Airy-type" isostatic compensation where the local lithosphere below the basin is faulted and has no lateral strength. This is a safe assumption given that the Paran~ Basin is fault-bounded and that glacial sediments in the basin reflect the importance of sediment gravity flow along fault-controlled slopes (Figs. 16.7, 16.8). However, the effects of any sea-level rise and associated water loading have not been included in the analysis. It can be assumed that glacio-eustatic sea-level fluctuations did occur during deposition of the Itarar~ Group but these were of insufficient duration to greatly affect basin subsidence. Data from the Paran~ Basin have implications for modelling glaciations in adjacent areas of South America. For example, Gonzalez-Bonorino (1992) has argued for a midCarboniferous ice sheet, comparable in size to the present day Antarctic Ice Sheet. The centre of the inferred ice mass lay over present day Buenos Aires, Argentina which is immediately south of the Parana Basin. This interpretation is based on identification of thick (60 m) massive and crudely-bedded diamictites in central Patagonia (Tepuel Group) as the deposits of grounded ice; it should be noted that previous work interpreted these strata as sediment gravity flows (Frakes and Crowell, 1969; Lopez Gamundi, 1989). It has to be said, however, that the glacial stratigraphy of both Brazil and southern South Africa is largely of Late Carboniferous and Permian age (Fig. 16.1) with no

rq.EVLES direct sedimentary record of older events. The work of Gonzalez-Bonorino (1992) is of considerable significance because it highlights the problem of reconstructing ice sheet histories, and therefore Late Palaeozoic climates, from glacial stratigraphic successions whose age is controlled by tectonic subsidence. Clearly a substantial part of the history of the ice sheet may go unrecorded. The key to a complete understanding of Late Palaeozoic ice sheet history probably lies with the study of depositional sequences of nonglaciated coastal margins that experienced glacio-eustatic changes of sea-level (Sections 16.13, 16.14). This can be well demonstrated for late Cenozoic glaciations (Section 18).

16.3 Southern Africa Miller (1984) argued for near continuous orogeny along the southern margin of the Brazilian shield from the Late Proterozoic to the Triassic but recognized a well-developed Early to Late Carboniferous orogenic cycle broadly equivalent to the Hercynian of Europe (Dalmayrac et al., 1980; Zalan, 1991). As related above, Hercynian orogenic activity along the western Palaeo-Pacific margin of Gondwana not only initiated Cordilleranstyle glaciation in that area but also promoted "far-field" intracratonic rifting to the east in Brazil (Fig. 16.10). The same tectonic constraints on preservation of a glacial record can be identified in Southern Africa. Late Paleozoic glaciation of the Southern African landmass was initiated at about 275 Ma on uplifted highlands flanking a southern foreland basin (Karoo) and a smaller intracratonic basin (Kalahari) to the north (Figs. 16.2, 16.9, 16.12). A palaeolatitude of at least 50°S is indicated. Visser (1991) suggested a long (20 Ma) phase of glacial erosion on the fiord-like margins of the highlands prior to the start of deposition at about 300 Ma. This erosional hiatus is most likely the expression of the. delayed subsidence of the Karoo Basin and the inability to accomodate and preserve a glacial record. Coeval

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

$ECTK~4 ,~I.¢3NG60"S LINE OF LATITUDE

KAROOFORELANDBASIN SOUTHERNHIGEILANOS

KALAHARI

~

INTRACRATONIC BASIN H~HLANDS I~IIVE~qOEK HIGHLANDS

Fig. 16.12. Top: Palaeogeographicand palaeotectonic (bottom) setting of Late Palaeozoic Dwyka Formation. Modified from Visser (1989).

subsidence is apparent in the Paran~ Basin of southern South America (Fig. 16.1) emphasising a common structural and tectonic control (see below). The principal Dwyka depositional basin consisted of a broad foreland basin downwarped on its southern margin, under the load of the Southern Highland tectonic arc, and faulted on its northern margin against the Cargonian Highlands (Fig. 16.12). The basin may have been closed eastwards on the Antarctic margin of Gondwana but opened westwards where it extended north as a broad epicontinental seaway between South America and southern Africa (Figs. 16.2, 16.9). It is likely that this seaway continued northwards as the Paran~ Basin of Brazil. A large, downfaulted intracratonic basin (aulacogen?) extended eastwards as the Kalahari Basin.

137

The broader tectonic setting of southern South Africa and coterminous South America is less clear. The southern Andes and West Antarctic peninsula formed an active plate margin at this time (Uliana et al., 1989; Fig. 16.2). As already suggested for the Paran~ Basin, the presence of a broad epicontinental seaway between Brazil and southern Africa probably reflects crustal extension inboard of the active Hercynian margin, involving rifting of pre-existing mobile belt structures (Fig. 16.10). These same structures were to be subsequently exploited during the Early Cretaceous opening of the Atlantic (Uliana and Biddle, 1987). The Southern Highland tectonic arc (Cape Fold Belt) formed as a fold and thrust belt inboard of the active margin and defines the southern margin of the Karoo foreland basin (Storey et al., 1988; Fig. 16.12). Development of the Cape Fold Belt and its extension as the Serra de la Ventana in Argentina and the Ellsworth Mountains in Antarctica, postdates about 280 Ma, which is close enough to the onset of subsidence both in southern Africa and Brazil (Fig. 16.1) to warrant further study.

16.3.1 Glaciogenic facies of the Dwyka Formation Sedimentary facies within the Dwyka Formation can be related very closely to the overall tectonic and palaeotopographic setting (Visser, 1987). A fiord-indented coastline on the margins of the highlands, passed offshore into the shelf and slope setting of the foreland basin. Palaeo-fiords contain considerable thicknesses of ice-proximal glaciomarine sediments, often resting on striated basement (e.g. Crowell and Frakes, 1972; Visser and Loock, 1988; von Brunn and Marshall, 1989). Fiord lengths of up to 250 km with widths as narrow as 5 km, are directly comparable to modern examples. Martin (1981) describes exhumed Late Paleozoic fiords, up to 1500 m deep, cut into the margins of the Windhoek Highlands along the Kaokoveld coastal margin of Namibia

138

N. EYLES

IZ°E I

.

, .

with what is understood of sedimentation in late Cenozoic fiords where pre-existing sediments tend to be cleaned out by glaciers and deposition only occurs during final ice retreat (e.g. Hughes, 1987; Andrews, 1990; N. Eyles et al., 1991). Older, thicker (800 m) and laterally more extensive glaciogenic deposits occur southwards away from the fiord-indented margin of the Dwyka Basin (Fig. 16.14).' The southern basin is characterized by homogenous successions of massive diamictites up to 100 m thick, interpreted as dominantly of "rainout" origin locally modified by downslope resedimentation (Visser and Loock, 1987; Visser, 1991). Some units contain large sandstone rafts suggesting the importance of downslope slumping. An absence of shallow water indicators suggests that the depositional setting may have consisted of a gently sloping deep water shelf subject to large-scale mass flow; an analogous setting can be identified in the Gulf of Alaska (Fig. 26.6). Mudrock units up to 50 m thick, containing occasional dropstones, form distinct stratigraphic marker beds that can be traced as much as 900 km across the basin and were interpreted as interglacial by Visser (1987). These strata are directly comparable to the extensive mud blankets that accumulate adjacent to glaciated coastlines when ice has receded inland. At these times, glacier-fed deltas pump large volumes of glaciogenic mud into the marine setting (e.g. Galloway, 1976; Section 26.1). Visser (1989a) formally recognized two interglacials, the Kenhardt

140E 1

.

.

.

NAM, ,A

..

i 180 S

i0o 0 I

Im

I II

_

\

Jj

\H 0B--

F20os

,

I

GLACIALLY-OVERDEEPENED FIORDBASINS I

Fig. 16.13. Late Paleozoic fiord basins along the coast of southwestern Africa. A f t e r M a r t i n (1981). See Fig. 16.9 for location.

(Figs 16.9, 16.13; see also Section 16.4). The dominant facies association in these valleys is that of interbedded debris flow diamictites and turbidites recording downslope reworking of proximal glacial sediments into deep water (Visser, 1983). Occasional brackish water conditions are recorded by low-diversity trace fossil assemblages, mostly orthropod trackways, similar to those found in Pleistocene glaciolacustrine settings (Gibbard and Stuart, 1974; Anderson, 1975, 1981). Stratigraphic relationships between the fiord fillings and deposits of the main basin to the south indicate that the former were deposited during the last phase of Dwyka glaciation (Visser, 1991). This is in keeping SOUTH

[]

NORTH GLACIALLY-INFLUENCED ~ ICE PROXIMAL SLOPE FACIES ~ F I O R D FACIES

POSTGLACIAL (WHITEHILL FORMATION)SHALE

[]

SHALES

CARGONIANHIGHLANDS

, ..... ::: PRINCEALBERT SHALES r

-,.~-

..................

P

A ~ A ~ p ( o] ° °' '''°' '°' °' "' ' ' ~-<'". . . . . . . . . . . . iiiiiiiiiiiii

.~v.~. : : : : : : : : : : : : : :

DWYKA interglacialJL ~ rnudstones

"

~

~

° ° •~ .............

Fig. 16.14. Generalized stratigraphic relationship between basinal slope facies and ice-proximal fiord facies within Karoo Basin. Note overstepping of basinal facies to the north (e.g. Fig. 16.8). See Fig. 16.12 for location. Modified from Visser (1989).

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T T I N G

and Hardap, that can also be identified in the Kalahari Basin (Martin, 1981). The passage from Dwyka deposits filling the high relief coastal margin to the thicker offshore deposits coincides with a palaeoescarpment in the underlying basement. This probably delineates a major basin margin fault demarcating the uplifted but deeplydissected Cargonian highlands to the north from the deep, subsiding southern basin. A north-south cross-section shows progressively younger stratigraphic units overstepping northwards onto the faulted basin margin (Figs. 16.12, 16.14). This is typical of a subsiding and expanding basin and is directly comparable to the infill of the Parana Basin (Fig. 16.8).

16.3.2 Comparison of the Brazilian and South African glacial records The Permo-Carboniferous stratigraphies of the Parana and Karoo basins are broadly similar though there are significant differences which relate to the specific tectonic setting of each basin. Visser (1989a, b) argued for a long (20 Ma) period of glacial erosion across the Karoo Basin prior to the onset of deposition at about 290 Ma. This phase of extensive glaciation is recorded by glacially-sculpted bedrock; the same conditions appear to have obtained in Brazil when the Kaokoveld ice margin may have overrun the entire basin. Glacially-scoured Precambrian basement is widely exposed along the eastern outcrop belt of the Parana Basin and indicates westward-directed ice flow from southern Africa (Gravenor et al., 1984; Fig. 16.9). The change from glacial erosion to deposition of thick basin infills is most easily explained as the result of co-eval subsidence within the Parana and Karoo basins. The main factor appears to have been renewed thrust loading along the active Palaeo-Pacific margin of Gondwana (see above; Section 16.3). This triggered extensive downwarping of the Karoo foreland basin at the same time that the Parana Basin underwent widespread

139

extension and subsidence (Figs. 16.10, 16.11). The two basins formed extensions of a shared e p i c o n t i n e n t a l seaway (Martin and Wilczewski, 1970, 1975; Figs. 16.2, 16.9). Episodic, step-wise downfaulting and expansion of the Parana Basin is recorded by three well-defined successions of lowermost sandstones and overlying shales within the Itarar6 Group (Fig. 16.7). No such simple scheme can be recognized in the Dwyka Formation with the exception of two widespread interglacial mudstones (Fig. 16.14). Both basins show thick postglacial shales and deltaic sequences containing cool-climate coals (Rio Bonito, Prince Albert Formations; e.g. LeBlanc-Smith and Eriksson, 1979). Tankard et al. (1982) present a picture from the northeast Karoo Basin of small prograding outwash fans lapping onto till or bedrock highs typical of a poorly-drained tundra setting. Coal deposits, up to 50 m thick, are characterised by an absence of seat earths that records a cold climate and very low rates of chemical weathering. Alternating vitrinite and inertinite-rich layers suggest fluctuations in the watertable and the degree of oxidation. In Brazil and South Africa coal deposits are capped by extensive organic-rich marine shales of the Mesosaurid-bearing Irati Shale (Brazil) and the Whitehill Shale (South Africa). This unit was deposited co-evally at 258 Ma along the epicontinental seaway (Oelofsen, 1987) and records maximum marine flooding and continued thermal subsidence. Elsewhere in southern and central Africa, Late Paleozoic glacial deposits are locally preserved in elongate, glacially-overdeepened and probably fault-controlled valleys. Good examples are widespread in the Congo Basin of Zaire (Lukuga Tillite; Cahen and Lepersonne, 1981b), in Angola (Lutoe Series; Rocha-Campos and dos Santos, 1981), in Zimbabwe and Zambia (Dwyka Series; Bond, 1981) and in Namibia and Botswana (Dwyka Series; Martin, 1981). Frakes and Crowell (1970) provide a good regional overview of the stratigraphy of these areas.

140

The extent of these strata typically define narrow, linear basins with sharply-defined faulted margins. Outcrop and subsurface data, gathered from exploration of coolclimate coals that overlie glacial strata, show a predominance of laminated (varved?) shales and pebbly-mud diamictites and deltaic sandstones typical of subaqueous glacial deposits in flooded, graben-like, fiord valleys. That many of these basins developed as continental rifts is suggested by the work of Wopfner and Kreuser (1986) who demonstrate a clear relationship between Early Permian rifting and the formation of ice covers on uplifted fault shoulders in southern Tanzania.

16.4 Australia; glacially-influenced deposition in active margin basins of the palaeo-Pacific margin The pattern and timing of Late Palaeozoic glaciation in Australia has been documented by Frakes and Crowell (1971) and Dickins (1985a, b; 1993). The Australian margin of Gondwana during the Late Carboniferous comprised, from west to east, an active magmatic arc comparable to that of the present day Andes, a forearc basin (Sydney-Bowen basin) and a subduction complex (Roberts and Engel, 1980; McPhie, 1987). This sector of Gondwana experienced rapid southward movement; an Andean-type volcanic arc system was established by at least 340 Ma (Little et al., 1992). The presence of local cordilleran glaciers along this margin is recorded by reworked glacial facies interbedded within thick volcaniclastic successions. Glacial deposition first began in the Early Namurian within the Tamworth forearc basin of eastern Australia in response to strong orogenic uplift along the magmatic arc (Veevers and Powell, 1987). The heterolithic and predominantly fluvial Spion Kop Conglomerate was deposited on the subaerial flanks of a volcanic arc and contains striated clasts and dropstone units (Veevers and Powell, 1987). This unit can be considered

N. EYLES

equivalent to the thick (1500 m) "glacial" succession of the marine Kullatine Formation which is predominantly of mass flow origin containing abundant acid and intermediate volcanic debris (Lindsay, 1966). Massive diamictites within thick volcaniclastic strata have been interpreted as tillites (e.g. McKelvey, 1981) but the published evidence for direct glacial deposition is not convincing (see Benson, 1981; Dickins, 1985b, 1993). Crowell and Frakes (1971a, b) report the presence of striated clasts indicating a glaciclastic source for some of the debris but were unable to identify a direct influence on sedimentation. The presence of thick ash fall tufts and conglomeratic fluvial sediments is consistent with a depositional setting close to a volcanic arc dominated by mass flow processes. Controversy still surrounds the origin of supposed "varvites" possibly deposited as waterlain volcanic tephra (Coombs, 1958; Dickins, 1985b). There are strong similarities between the earliest Australian "glacial" sediments of the Early Namurian and the deep marine and terrestrial mass flow strata that record the initiation of late Cenozoic glaciation in the Gulf of Alaska (C.H. Eyles and N. Eyles, 1989). In this high relief setting the growth of local mountain glaciers on rapidly growing volcanic peaks is recorded by coarse-grained and poorly-sorted mass flow deposits interbedded with volcanogenic strata (Figs. 4.4, 10.8). The Namurian palaeogeography of New South Wales consisted of a north-south arc of andesite stratovolcanoes flanked to the east by an extensive braid plain (Fig. 16.15). Seaward, a shallow marine shelf passed offshore into deep water. Thick ignimbrites, debris flow facies and airfall tufts of the main volcanic arc pass eastward into sandy braidplain deposits. The latter contain interbedded ignimbrites several tens of metres thick, that travelled many tens of kilometers from their source (Fig. 16.16). Extensive debris flows and flood deposits may indicate melting of cordilleran ice masses on volcanic peaks. McPhie (1987) compared the

141

E A R T H ' S GLACIAL R E C O R D AND ITS TECTONIC SETTING

Late Carboniferous glaciated arc to the oceanward flank of the Andean arc in northern Chile (Fig. 16.15). In this setting, glaciclastic sediment is reworked downslope as debris flows that are subsequently preserved within lavas and ignimbrite flow sheets. Distal glacial deposits are recorded by extensive braidplain gravels along the arc flanks. In these settings, glaciclastic sediment is a subordinate component within volcaniclastic strata. The same glaciogenic/volcanigenic association is also preserved within the 1800 km long foreland basin referred to as the Sydney-Bowen Basin (Fig. 16.2) where volcanism was widespread during the latest Carboniferous and Early Permian. Glaciated highlands, centred at latitude 50°S, supported valley glaciers which cut eastward LATE CARBONIFEROUSNEW ENGLAND FOLD BELT (310-300

A

Ma)

--B 31° Z4° S

300S

MODERN ANDEAN ARC

3

< OLDER STRATA >

[],ON,,BR,TESTE.RA VAF.OWS [i --?IOLAC'AL'A 'ES

I IFAC,ES

VOLCA",C ENTR

Fig. 16.16. Regional Late Palaeozoic facies distributions on the glaciated margin of the New England orogeny (see Fig. 16.15 for location). After McPhie (1987). See Fig. 10.8 for depositional model and Figs. 4.4 and 10.7 for diamict facies.

draining "palaeovalleys" up to 500 m deep, 3 km wide and 100 km in length (Herbert, 1981). These are filled with glaciofluvial conglomerates up to 250 m thick (Tallong and Yadboro conglomorates; Herbert, 1981) together with minor diamictites. Laminated lacustrine sediments with dropstones are locally preserved. Channels may have been overdeepened by ice streams similar to the elongate "fiord-lake" basins of western Canada and Switzerland (e.g.N. Eyles et al., 1991) and are similar in scale to the Late Paleozoic "palaeofiords" of northwestern Namibia and Angola (Fig. 16.13).

16.5 Australia; glacially-influenced deposition in intracratonic basins [~]

OlderStrata

I~

/ ~)/

VolcanicCentres Palaeocurrents

[]

l~

Ignimbrites & Lavas

[]

Braid Rain Facies

Volcanidaatic & reworked glacial

[]

ShallowMarine Facies

facies

[]

Turbidites

Fig. 16.15. Comparison of Late Palaeozoic palaeogeography of New England Orogeny with that of the present day Andes. After McPhie (1987). A - B defines stratigraphic cross-section shown on Fig. 16.16.

Away from the active plate margin of eastern Australia Late Paleozoic glacial deposits are preserved within several interior intracratonic basins (Fig. 16.2). These basins are floored by Late Carboniferous glacial sediments that rest on a high relief unconformity (e.g. Fig. 16.4) widely interpreted as glacially-overdeepened. Campana and Wil-

142 son (1955) showed that the overdeepened floors of several Permian basins in South Australia (e.g. Wilga, Collie coal basins) are comparable in their dimensions to Lake Geneva in Switzerland. Many basins, however, are structurally-controlled and their high relief floors are also likely to be the result of differential subsidence during faulting (e.g. Wopfner, 1972, 1981; Youngs, 1975; Veevers, 1984; see Storey, 1991 and Section 24.2). Veevers and Powell (1987) suggested that ice began to spread westwards and invade these basins during the mid-Namurian though the onset of widespread glacial deposition only commenced in the Stephanian. The apparent delay was attributed by Powell and Veevers (1987) to an early phase of non-deposition by a cold-based ice sheet; they argue that deposition began when the ice sheet changed to a temperate, wet-based condition. An alternative explanation is that the onset of a glacial depositional record was tectonically-controlled as in South America and southern Africa (e.g. Sections 16.2.1, 16.3). The onset of renewed subsidence and rifting within the interior basins of Australia allowed the accomodation of substantial thicknesses of glacial strata. The age difference between the mid-Carboniferous glacial successions of the active eastern Australian margin and the Permian successions of the intracratonic basins to the west is probably a function of the diachronous migration of ice centres across Australia acting in combination with delayed transmission of far-field stresses to the interior of the Australian plate. Glaciation was initiated along the uplifted palaeo-Pacific margin of Australia. The relative timing of glacial deposition along the active plate margin compared with that of the intracratonic basins of the interior has already been identified with regard to the South American margin of Gondwana (Section 16.2.1). Marine waters were able to penetrate some of the interior intracratonic basins. Glacially-influenced marine strata are present in the Denman, Ackaringa, Troubridge and Den-

N.EYLES mark basins, are absent in the Pedirka Basin and only rarely developed in the Cooper Basin (Wopfner, 1981). Fully marine conditions obtained at the northern, seaward reaches of the Canning and Carnarvon basins with brackish water in its interior southern portions (Fig. 16.2). The infills of the Arckaringa, Officer, Troubridge, Pedirka, Denman basins include mass flow deposits (graded conglomerate, sandstone and siltstone facies, slumped horizons, debris flow diamictites and conglomerates) with evidence of direct glacial activity restricted to the basin margins (e.g. Thornton, 1974; Battersby, 1976; Jackson and Van de Graaf, 1981). Terrestrial, subglacial facies and striated surfaces were probably widely developed on surrounding highlands but are not extensively preserved (Wopfner, 1972; Thornton, 1974; Youngs, 1975; Redfern, 1991; O'Brien and Christie-Blick, 1992). The well-known striated bedrock surfaces exposed along the Inman Valley and at Hallett Cove in South Australia were first reported by Selwyn (1859) and Tate (1878) respectively (see Milnes and Bourman, 1972). Cross-sections through the relatively wellstudied Cooper Basin show that successively younger glacial sediments overstep underlying units typical of basins undergoing progressive subsidence and expansion by faulting of the basin margins. Wopfner (1972) identified continuous syndepositional tectonic activity during the Permian. The similarity of the central and southern Australian basin fills to the Parana Basin of southern America and Karoo Basin of southern Africa invites a detailed comparison. The central and western Australian basins reveal a strong tectonic control on facies types and their distribution; mass flow facies, that record downslope reworking of glacial sediment, predominate. Very thick sandstone units are a common stratigraphic component (as in most Gondwana basins) and record extensive glacial scouring of upstanding Proterozoic crustal blocks and the focussing of glaciclastic sediment along faulted basin

143

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

margins. Continued downslope resedimentation and repeated reworking of glaciofluvial sediment probably accounts for the textural maturity of many "glacial" sandstones in Gondwanan basins and their importance as hydrocarbon reservoirs (e.g. Goldstein, 1989; Franca and Potter, 1991). Recent work in the Cooper Basin has identified thick poorly-cemented, coarse and fine-grained sandstone dominated by parallel lamination within the Merrimelia Formation; these have been interpreted as paraglacial aeolianites (Williams et al., 1985) but an intimate association with subaqueously-deposited diamictites of mass flow origin and graded silts and conglomerates suggests a turbidite origin cannot be ruled out. Comparable sandstone turbidite facies and successions are present in the Parana Basin of Brazil and comprise the most important target for hydrocarbons (Franca and Potter, 1991; C.H. Eyles et al., 1993). Along the western margin of Australia the Carnarvon Basin and its southern extension (the Perth Basin) together with the coterminous Officer and Canning basins contain the youngest Late Paleozoic glacial deposits in Australia. The Lyons Group in the Carnarvon Basin is the thickest glacially-influenced succession (2500 m) across Gondwana with marine faunas present at all stratigraphic levels (Van de Graaf, 1981). The presence of freshwater palynomorph assemblages, however, may indicate an enclosed basin with a restricted, meltwater-influenced circulation. Permian faulting in the Carnarvon Basin was the precursor to Jurassic rifting and the subsequent separation of India from Australia parallel to the basin axis. This emphasizes the long-lived control on basin evolution by Late Proterozoic basement structures. This structural dependency is also evident in the Canning Basin where ice lay over the Pilbara Block and extended north and east across the basin and into the graben-like Fitzroy Trough. The dominant sedimentary facies within the basin is sandstone, not diamictite (O'Brien and Christie-Blick, 1992). Strong

~/EN BASIN

BRISBANE

CDNEY IASIN

)ACIFIC OCEAN

-- - -i a.

=E ~,

KAZANIAN KUNGURIAN ARTINSKIAN SAKMARIAN

BR OR UY GSHILTTO ON RE MAn' ON B ER SN T 'FO NOWRASANDSTONE

SHOALHAVENGROUP

WANDRAWANIAN SW.TSTONE SNAPPIERPOINT F'M PEBBLEY BEACH F'M WASP HEAD F'M

SUBGROUP

CONJ(]LA

Fig. 16.17. Palaeogeography and stratigraphy of the Permian Sydney Basin. After Jones et al. (1986).

similarities can be identified with the mostly turbidite infill of the Paran~i Basin (Section 16.2; C.H. Eyles et al., 1993). As with the Carnarvon Basin, marine faunas are common (e.g. Foster and Waterhouse, 1988). Evidence of Early Permian glacial influence along the eastern margin of Australia is recorded by large ice-rafted clasts and possible iceberg scours within shallow marine deposits of the Sydney Basin (Fig. 16.17). The Permian-Triassic Sydney Basin formed an elongate, north-south orientated backarc basin on the western, continental side of the volcanic arc of the New England Orogen. The basin fill shows a lowermost predominantly marine, Lower and Middle Permian section overlain by thick, coarsening upward Late Permian deltaic and fluvial deposits that include important coal horizons (e.g. Diessel, 1980; Martini and Johnson, 1987). Ice-rafted horizons occur in shallow marine deposits of the Sakmarian to Artinskian Conjola Sub-

144

N. EYLES

16.6 Antarctica and Tasmania

group and in the Kungurian Wandrawandian Siltstone of the Shoalhaven Group (Fig. 16.17). Sediments contain a rich marine fauna that includes corals, gastropods, bivalves, bryozoans and brachiopods (Dickins et al., 1968; Gostin and Herbert, 1973) together with ice-rafted boulders that are directly comparable to the coquinas reported by Eyles and Lagoe (1989) from the late Cenozoic Yakataga Formation of Alaska. Glendonites indicate low water temperatures (Ramli and Crook, 1978). Alternations of storm-deposited hummocky cross stratification and mudstone record cold climate sedimentation on a shallow shelf subject to repeated changes in water depths. It is not known whether these changes are controlled by regional basin tectonics or by glacio-eustatic fluctuations. Icebergs drifted northwards from a southern, polar ice cap over Antarctica and Tasmania but there is no evidence of direct glaciation in the Sydney Basin. Dickins (1985b) has questioned the source and origin of dropstones preferring a debris flow mechanism. Scotia Sea

\

~o •........

~*

°o

In Antarctica, Early Permian glacial sediment occurs at the base of a thick Permian to Jurassic succession (the Victoria Group) and rests unconformably on a widespread erosion surface (Frakes et al., 1971). Outcrops are widely scattered along the Transantarctic Mountains from Victoria Land to Dronning Maud Land (Fig. 16.18). The principal outcrop areas are in Victoria Land (Darwin Tillite, Metschel Tillite, Pagoda Formation), in the Ellsworth Mountains (Whiteout Tillite), in the Wisconsin and Ohio Ranges (Buckeye Tillite) and along the Pensacola Mountains (Gale Mudstone). Considerable thickness variations along the outcrop belt are evident with massive diamictites up to several hundreds of metres thick reported in the Ellsworth Mountains (Whiteout Tillite; Matsch and Ojakangas, 1991; see below). These thicknesses, the presence of interbedded mudstones and transitional upper formational contacts with mudstones containing 31

4

2

I

Antarctic

Peninsula Weddell Sea

--900W

~

OhioRangei~ West

Antarctica

[ ] OUTCROP

I

EIIsworthMtns~~

"~

/

Ice Shelf

Ross Sea

Darwin Tillite 3 Whiteout Tillite 4 Buckeye Tillite

18o°

/

VICTiRIALAND 90°E--

;,~_.

4 ~ '>'=~ ~

1 Gale Mudstone 2 Pagoda Formation Metschel Tillite

ELLSWORTH MTNS. East

2 .

Antarctica

~~ n d ~/ Jictoda

~-

0,

500,km

I

Fig. 16.18. Outcrops of Late Palaeozoic glaciclastic strata of Antarctica. Inset shows thickness of diamictites (after Miller and Waugh, 1991).

145

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'I'I'ING

ice-rafted debris, provide unambiguous evidence of a marine setting in a rapidly-subsiding basin or basins. Kemp (1975) reported acanthomorph acritarchs from diamictites of the Ohio Range suggesting marine conditions; Frakes and Crowell (1975) also suggested a marine origin on the basis of diamictite geochemistry (see below). Dalziel et al. (1987) suggested the presence of several discrete sedimentary basins on continental crust undergoing crustal extension and rifting. Collinson (1991) recognized several stratigraphic basins and suggested that these were components of a larger foreland basin behind an active volcanic arc marking the palaeo-Pacific margin of the East Antarctic craton. The regional stratigraphy of these basins has been reviewed by Barrett (1982). The most detailed descriptions of the Antarctic diamictites are provided by Matsch and Ojakangas (1991) who report thicknesses of more than 600 m for individual diamictite units in the Sentinel Range of the Ellsworth Mountains. Massive diamictites are separated by thin (m's) of laminated shales and mudstones with dropstones and, in the Meyer Hills area, by extensive boulder pavements. The latter are one clast thick and show a consistent striation orientation from outcrop to outcrop. Matsch and Ojakangas (1991) recognized no unequivocal evidence for subglacial deposition of diamictites and argued that they were largely the result of "rain-out" of mud and debris below floating ice; boulder pavements record fluctuation of ice margin without major erosion of the sea-floor. The overall succession is similar to the late Cenozoic Yakataga Formation of Alaska; the upper part of which consists of thick (00's m) monotonous "rain-out" diamictites; boulder pavements separate and occur within diamictite units and were produced as shallow water lag deposits at times of lowered sea-level. These were subsequently overrun and striated by ice lobes that crossed the shelf (C.H. Eyles, 1988a, b). The considerable thickness of massive diamictites in Antarctica together with geo-

chemical and achritarch evidence suggests rapid subsidence and sediment accommodation on the seaward margin of uplifted glaciated highlands. The same argument can be made for Tasmania where the Parmaneer Supergroup (Late Carboniferous to Late Permian) commences with a prominent glacial diamictite horizon (Truro, Wynyard, Stockers "tillites") up to 170 m thick. The underlying surface is of high relief (800 m) and has been interpreted as the result of glacial-erosion but rifting is more likely since the thickest successions lie along pronounced NW-SE trending structural lineaments (e.g. Tamar Fracture System). The overall succession shows a lowermost glaciclastic component overlain by marine muds and cool water limestones deposited in a polar setting (Rao, 1981; Rao and Green, 1982; Domack, 1988) affected by repeated glacio-eustatic changes of sea-level (Banks and Clarke, 1987). 16. 7 Southeast Asia

The precise palaeogeography and tectonic setting of the northeastern "passive" margin of Gondwana (Fig. 16.2) is still poorly understood (Sun Dong-li, 1993). Late Paleozoic glaciomarine diamictites can be traced along an outcrop belt 200 km wide over more than 1000 km from Malaya (Singa Formation), Thailand (Phuket Series) to Burma (Mergus, Martaban series; Stauffer and Mantijit, 1981; Stauffer and Peng, 1986; Metcalfe, 1988). The age of the diamictites is not well-constrained (Waterhouse, 1982). Glaciogenic strata are dominated by laminated (turbiditic) muds and shallow marine siltstones and sandstones. Matrix-rich diamictites contain well-developed Cruziana ichnofacies assemblages (Stauffer and Peng, 1986) typical of glaciated continental shelf areas dominated by muddy substrates (N. Eyles et al., 1992b). Hummocky cross-stratified sandstones record storm-influenced shallow waters; widespread slump structures suggest repeated episodes of downslope resedimentation. There are very strong similarities be-

146 tween the southeast Asian glaciomarine diamicitite successions and those of the late Cenozoic Yakataga Formation of Alaska. The latter record large influxes of glacially-derived m u d and ice-rafted debris to a shallow ( < 150 m) shelf; extensive, blanket-like "rain-out" and resedimented, diamictite facies are interbedded with shallow-marine

N.EYLES sandstones and show a diverse range of ichnofacies (N. Eyles et al., 1992b). The diamictite belt of northwestern Gondwana belongs to the Western Southeast Asia Block which subsequently rifted off from the northern margin of the supercontinent somewhere b e t w e e n northwest Australia and the Arabian Peninsula during the late Permian.

Fig. 16.19. Top: Generalized west-east cross-section through the greater Arabian basin (see below) showing pronounced Hercynian unconformity. Glacial deposits rest on the unconformity surface in Oman (AI Khlata Formation) and in Saudi Arabia and Yemen (Wajid Sandstone). After AI-Laboun (1988). Bottom: Generalized isopach map of Permo Carboniferous strata in the greater Arabian Basin showing influence of "Hercynian" topography and outcrops of glacial deposits. After Al-Laboun (1988).

147

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

Traditional reconstructions of Gondwana depict a large area of oceanic crust (Palaeo Tethys) northward of Australo-India and Eurasia. In contrast, Lin and Watts (1988) and Smith (1988) argue from paleomagnetic and paleobiographic evidence that a "Cathaysian" landmass (south China block, ThaiMalay terrane, West Burma terrane, Indochina terrane) occupied this large triangular area and did not become separated from Gondwana proper until the middle Permian. Their reconstructed Cathaysian landmass resembles the "Cimmerian" continent of Seng/Sr (1984) and Seng/Sr et al. (1988). During the Asselian and Sakmarian the southern margin of Cathaysia lay in mid-paleolatitudes and was characterized by relatively impoverished brachiopod and coral forms associated with the glaciomarine deposits described above. The location of the glacial ice centres feeding mud and ice-rafted debris to the "Cathaysian" shelf is not well-constrained; a Himalayan source area seems most likely but further fieldwork is needed. 16.8 Arabian Peninsula Late Paleozoic glaciclastic strata in Yemen, Saudia Arabia and Oman infill an unconformity surface recording Hercynian orogenesis and uplift (Fig. 16.19). In Oman the relationship between tectonics and glaciation is very clear and there are parallels with the glaciclastic record preserved in the intracratonic Paran~ Basin of Brazil. In Oman, the glacigenic AI Khlata Formation (locally up to 750 m thick) was deposited at about 50% paleolatitude on the northern margin of Gondwana some time between the late Westphalian to Sakmarian (Braakman et al., 1982; Grandville, 1982; Levell et al., 1988). A rifted epicratonic basin (800 km × 200 km), bounded northwards by a highland region in the area of the present day Oman Mountains (Fig. 16.20), extended into Saudi Arabia and onto the Indian Plate; Arabia and India later separated along the axis of the basin. Glaciclastic strata are underlain by Lower Paleozoic sandstones of the Haima

54*

58°

22°.

2 0 °-

I

l J

HUQF

HAUSHI

I

UPPERDA I WC I TT I ES / HALE COUPLEX

~

LOWERCONGLOMERATE MEMBER

AL KHLATA

Fig. 16.20. Top: Outcrop area of AI Khlata formation, Oman with generalized depth contours in metres on top of formation. Oil fields are shown in black. Inset shows outcrops of Late Palaeozoic glaciclastic strata around the western and eastern rim of the Arabian Peninsula. After Helal (1963), McClure (1980), Levell et al. (1988). Bottom: Stratigraphy and structural crosssection across eastern flank area of Oman. After Levell et al. (1988).

Group which are poorly understood but clearly record Hercynian uplift along the basin margin and a massive influx of clastic sediments onto Late Proterozoic a n d Cambrian platformal limestones and evaporite deposits of the Huqf Group. An erosional contact separates the two successions with the depth of erosion increasing eastward toward the uplifted basin margin. Figure 16.20 shows a cross-section through the Marmul oil field and shows westerly-dipping strata on the flank of the Huqf Anticline which approximates the present day coastline of Oman. Outcrops of glaciclastic strata and glaciated pavements occur at AI Khlata close

148

to the axis of the anticline; the thickest glaciclastic strata occur in the vicinity of Marmul in the Eastern Flank Area (Fig. 16.20). Stratigraphy and facies of the A1 Khlata Formation are directly comparable to that shown by the Itarare Group in the Paran~ Basin in Brazil, and further investigation may well reveal a similar tectonic control on sedimentation and preservation. The Itarar6 Group contains three formations (Fig. 16.8) which show lowermost coarse-grained sediment gravity flow facies overlain by a blanket-like complex of shales and silty diamictites; these overstep the underlying conglomeratic strata recording basin subsidence. The same fining-upward succession of facies is shown by the A1 Khlata Formation. Lowermost conglomerates, pebbly sandstones and sandstones show graded facies with massive sand beds as much as 70 m thick. These strata show abundant liquefaction structures indicating rapid deposition and were interpreted as "glaciofluvial" by Levell et al. (1988); a glaciofluvial sediment s o u r c e appears likely but a subaqeous, sediment gravity flow origin is strongly indicated. Brackish marine conditions reflecting the input of glacial meltwaters into a restricted rifted basin, are suggested by an absence of marine fossils apart from algae (Tasmanitids) and achritarchs. Coarse facies of the lowermost AI Khlata Formation are overlain, as in Brazil, by finegrained facies comprising complexes of silty diamictites, massive and laminated siltstones described as "varve-like" by Levell et al. (1988). These facies form a regionally extensive, sheet-like unit that "oversteps" underlying conglomeratic facies. Massive and stratified diamictite facies are unambiguously subaqeous in origin and record the "rain-out" of suspended sediment and ice-rafted debris, with local downslope mass flow and turbidite deposition. The diamictite/siltstone complex passes upward into laminated shales with dropstones and a postglacial shale unit (Rahab Shale) that forms an extensive cap below Early Permian platformal carbonates

N. EYLES

(Haushi Limestone). The overstepping relationship of the upper fine-grained facies of the A1 Khlata Formation over underlying coarse facies has previously been explained as the consequence of erosion by an extensive ice sheet and a reduction in relief of the basin; subsidence of underlying evaporites was also cited (Levell et al., 1988). It can be suggested, in contrast, that a more fundamental tectonic control is apparent where basin subsidence allows preservation of progressively deeper water facies that overstep underlying strata. The glaciclastic deposits of Oman are of especial interest in that they contain significant oil deposits (more than 3.5 billion bbl) discovered in 1956. Heavy, viscous oils (1134 ° API) migrated from algal source rocks of the underlying Huqf Group. The Rahab Shale and Cretaceous Nahr Umr Shale form regional seal units. The main prospective belt occurs in coarse-grained facies preserved around the former basin margin; intraformational diamictites and siltstones form internal seals. The discovery of significant hydrocarbon resources in Late Palaeozic glaciclastic strata in Oman has stimulated interest in the extent of correlative strata elsewhere in the Arabian Peninsula. The extent (and even age) of Late Paleozoic glaciation elsewhere in southern Saudi Arabia and coterminous Yemen is controversial. In Saudi Arabia, bouldery conglomerates of the Haushi Formation (Wajid Sandstone) in the A1 Quasim area have been interpreted as either fluvial (Hadley and Schmidt, 1975) or glacial (Helal, 1964; McClure, 1980; Ad-Dabbagh and Rogers, 1983; McClure et al., 1988); the presence of striated and shaped clasts clearly indicates a glacial source. Sedimentological details of the Wajid Sandstone are few and work is ongoing by Saudi Aramco geologists to decipher the depositional record. Especial controversy centres around the origin of "floating" granite boulders; these appear to have been emplaced within sediment gravity flows but were previously interpreted as ice-rafted drop-

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T F I N G

stones. AI-Laboun (1987) has described Late Carboniferous/Early Permian clastics of the Unayzah Formation in central Saudi Arabia; no direct glacial influence could be identified and the formation appears to be dominated by braided river fluvial facies deposited in a periglacial environment (McGillivray and Husseini, 1992). This unit clearly shows the relationship of sedimentation to the underlying regional Hercynian unconformity. The Unayzah Formation infills a considerable palaeotopographic relief created by Hercynian orogenesis (Fig. 16.19). Elsewhere, in the Yemen Arab Republic the Akbra Shales (130 m) contain silty diamictites interbedded with rhythmically-laminated siltstones containing abundant trace fossils indicative of a glaciolacustrine setting. Correlation with the Endaga Arbi Tillites of Ethiopia can be suggested (Kruck and Thiele, 1983; EI-Nakhal, 1990). In summary, whilst the Arabian Shield is traditionally regarded as tectonically "stable", the preservation of glacially-influenced facies within a high-relief Hercynian unconformity suggests a more complex tectonic setting. The overall tectono-stratigraphic picture of Late Palaeozoic glaciation in Arabia (McClure et al., 1988), Yemen (Kruck and Thiele, 1983) and Oman (Levell et al., 1988) is one of intracratonic subsidence accompanying, and immediately postdating, Hercynian orogenesis and uplift. Tectonic setting, stratigraphy and facies are directly comparable to those Late Palaeozoic strata of the Parana and Karoo basins. A major priority of further research is to determine the precise relationship between the structure of Late Proterozoic basement (e.g. Stoeser and Camp, 1985; Husseini, 1988, 1989) and the configuration of overlying Late Palaeozoic basins; although an intimate relationship cannot be demonstrated as yet, such a dependency can be predicted. 16.9 India and Pakistan

An active, collisional tectonic setting is indicated for Late Palaeozoic glacial deposits

149

in the extreme northwest of India, in Pakistan and Nepal (Fig. 16.2). Palaeogeographic reconstruction of northern Gondwana (de Wit et al., 1988) places these areas in close proximity. The glacigenic record reflects active collisional processes where the Indian sub-continent abutted against Laurasia resulting in upthrusting of the Himalayas and closure of the Paleo-Tethys Ocean. The "Blaini Tillites" of the Lesser Himalaya, India, the Rangit Pebble-Slate and Lachi Formation of Nepal and the eastern Himalaya in India are correlative with diamictites of the Salt Range in Pakistan (Tobra Formation). Diamictite facies occur at the base of the Gondwana sequence and are correlative with those of the intracratonic "Talchir" glacial basins of Peninsula India to the south (Fig. 16.2). Most of the Himalayan diamictites contain substantial volumes of pyroclastic matrix material and are interbedded with pyroclastic deposits and basaltic flows. Acharyya (1975) argued that the diamictites are laterally-persistent, tectonically-generated submarine mudflows deposited close to the rising Himalayan Mountain front (see also Alavi, 1991). Kemp (1975) reports Cymatiosphaera acritarchs indicative of marine conditions. Widely-separated Gondwana basins in Peninsula India contain significant coal deposits (98% of India's reserves) and their subsurface geology is better understood (Chowdhury et al., 1975; Casshyap, 1977; Tewari and Casshyap, 1982, 1983). These basins occur along several linear trends developed in the underlying Proterozoic basement (Fig. 16.2). Systematic drilling shows that the glacigenic Talchir Formation forms the basal unit within many basins (Fig. 16.4). The basins are elongate intracratonic halfgrabens that commonly show on one side a steeply-dipping boundary fault parallel to the basin axis resulting from the reactivation of east-west and northwest-southeast trending Proterozoic structures (Ghosh, 1975). The other margins of these basins typically show "overstepping" strata generated by progres-

150

sive subsidence. Internally, the basin fills shows networks of intrabasinal faults that have an arcuate planform and decreasing dip at depth. Basu and Shrivastava (1981) showed that the basins are "cradled" by Late Proterozoic and Archean fold belts between cratonic blocks and ascribed the initition of rifting to Hercynian orogenesis. Again, "far field" reactivation of Late Proterozoic basement structures by distant collisional forces may have been a critical factor in the onset of intracratonic rifting, local uplift, glaciation and the preservation of glaciclastic strata in embryonic basins (e.g. Fig. 16.10). The most detailed study of Talchir sedimentation to date is that of Banerjee (1992) who reexamined strata in the Giridih, Sahajuri, Jayanti and Daltonganj sub-basins within the Damodar Basin. The importance of faulted basin margins and steep palaeoslopes adjacent to uplifted highlands, the localised preservation of subglacial tillite facies and the predominance of sediment gravity flow facies in basin centres is clearly apparent. Fossil bearing and bioturbated strata record marine conditions at the margins of regional piedmont ice covers that experenced multiple phases of expansion of advance and decay. Interbasinal correlations within the glacial interval are not clearly evident in response to diachronous ice margin fluctuations and differential tectonic subsidence in rifted sub-basins. 16.10 Gondwana glaciation; a tectono-stratigraphic model The widespread preservation of Late Paleozoic glacial sediments across Gondwana from 350 to 250 Ma may reflect the diachronous shifting of ice centres as Gondwana moved across high palaeolatitudes (Caputo and Crowell, 1985). The growth of these ice masses (termed herein the "Gondwana Ice Sheet Complex") was likely in response to locally varying combinations of both "polar" and "adiabatic" controls influenced by global geochemical fluctuations (see below).

N. EVLES

Along the palaeo-Pacific sectors of Gondwana (e.g. northern India, eastern Australia, Argentina), glacial sediments are intimately associated with volcanogenic sequences in a variety of active margin basins. In most other areas (e.g. Brazil, southern Africa central Australia, Antarctica and India), glacial sediments occur in intracratonic basins and in many cases, are associated with clear evidence of syndepositional "rift" tectonics accompanying the renewed subsidence of such basins. As emphasised above, high relief floors of several basins (e.g. Figs. 16.4, 16.8) can be better attributed to the effects of faulting than to glacial erosion, though the impressive palaeo-fiords of southern Africa and eastern Australia testify to enhanced glacial erosion along basin margins. The generation of steep-sided intracratonic basins, in turn, dictated the style of glacial deposition in the basin and therefore the nature of the depositional record (Fig. 16.7). In most cases, glaciclastic sediment was delivered to the basin by ice margins, sited around the basin periphery, and was reworked downslope from steep, faulted basin margins across large fan deltas or turbidite fans into predominantly marine settings (Fig. 16.2; see below). The onset of subsidence and the generation of deeper water basins may have restricted ice covers to upstanding basin margins. The demonstration of intracratonic rifting immediately prior to and during deposition of a glacially-influenced infill in many Late Palaeozoic basins (e.g. Fig. 16.11) carries very significant implications in regard to identifying the timing of glaciation across the supercontinent. As related above, the age and distribution of glaciated basins suggests strong west-east diachroneity (Fig. 16.1) and migration of glacial centres as Gondwana rotated under the southern pole of rotation (Caputo and Crowell, 1985). The case can also be made on the other hand, that this diachroneity is tectonically-driven and also reflects the successive timing of intracratonic subsidence as "far-field" stresses were communicated across Gondwana. It is necessary

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

therefore, to carefully distinguish between the duration of glacial climatic events across Gondwana and the age of the resulting stratigraphic record. For example, many millions of years of glacial erosion and non-deposition occurred prior to subsidence and accomodation of a glacial record in southern Africa and Brazil (Section 16.3.2) and in Australia (Section 16.5). Much closer examination of the relationship between the origin of Gondwana intracratonic basins, their structural framework and the precise age of their glacial depositional record is needed along the lines of that identified principally for North American intracratonic basins by Klein and Hsui (1987) and Leighton et al. (1990). These workers identify synchronous subsidence of widely separated basins and implying a common tectonic control. The fundamental underpinning of these ideas has been the identification and intercontinental correlation of "megascale" depositional sequences and their bounding unconformities (Sloss, 1963, 1988, 1990 and refs. therein). Space does not permit a detailed review of the origins of intracratonic basins and only a few brief comments can be made here. Leighton and Kolata (1990) summarise a large literature ("a profusion of plausible models") listing the most commonly cited mechanisms as (1) thermal doming and later sagging, (2) stretching of continental crust, (3) loading by thrust sheets inboard of active collisional margins, (4) subcrustal phase changes, (5) partial melting of the asthenosphere accompanied by subsidence of overlying crust and (6) subsidence controlled by in-plate, horizontal stresses created along collisional margins. According to Leighton and Kolata (1990), the last named offers the key to cratonic basin evolution and centres on the propagation of far-field stresses generated along active plate margins deep into plate interiors. These stresses activate previously existing structural inhomogeneities in the craton. This model accounts for both the episodic nature of intracratonic subsidence and co-sympathetic responses of widely-sep-

151

arated basins. More importantly, in the context of the present study, it also allows for the generation of glaciers around the uplifted compressional margins of Gondwana (Fig. 16.3) and their subsequent growth into regional ice centres (Fig. 16.2) in response to latitudinal and global geochemical controls. Subsurface data are now increasingly available from Late Palaeozoic glaciated basins; better geochronologic and core control, combined with tectonic subsidence curves and seismic profiles (e.g. Fig. 16.11), will henceforth constrain the relationship between glaciation and tectonic subsidence across Gondwana. Tectonic control on the preserved Late Paleozoic glacial record of Gondwana has its origins during assembly of the Late Proterozoic supercontinent. The inherited Late Proterozoic structure of Gondwana is that of cratons (i.e. crustal blocks) separated by mobile belts (i.e. welds) that record complex collisional histories during supercontinent amalgamation. The structure of blocks and welds dictated the subsequent structural history of Gondwana during Late Paleozoic intracratonic extension (Fig. 16.10). Extension appears to have been in response to the propagation of "far-field" stresses into the plate interior directed from the actively subducting southwestern (South American) and southeastern (Australian) and northern (Laurussian) margins (Fig. 16.2). In North America, the resulting Late CarboniferousEarly Permian Ouachita-Alleghenian orogeny controlled the accumulation of depositional sequences in North American intracratonic basins (the Absaroka I and II depositional sequence; Sloss, 1988; Leighton and Kolata, 1990; Fig. 6.2). Clearly, a high priority of future work should be to identify correlative tectono-stratigraphic glacial depositional sequences across the interior of Gondwana. 16.11 Gondwana glaciation; a depositional model

Figure 16.2 is a speculative map showing the paleogeography of the Late Carbonifer-

152 ous/Early Permian glacial culmination across Gondwana. Specifically, it is intended to depict conditions at the margins of the South American, South African, Australia and Indian sectors of the Gondwana Ice Sheet complex. The depiction is correct in terms of depositional setting for each of these zones but is incorrect in showing synchroneity of glaciation across Gondwana; as disscussed above, the preservation of glacial deposits was markedly diachronous from west to east (Fig. 16.1). Many Gondwana glaciated basins are elongate (Fig. 16.2) reflecting an underlying Late Proterozoic structural control; this would promote the development of a restricted hydrodynamic circulation dominated by meltwater inputs from large glacier-fed deltas. A lack of faunal evidence for a marine setting in many basins is therefore not unexpected. Along the South American sector, Fig. 16.2 emphasizes the connection between the paleo-Pacific rim, characterised by active margin basins and a large epicontinental seaway to the east shared by the Parana Basin and the Karoo Basin. Geochemical data provide strong support for the paleogeographic synthesis shown in Fig. 16.2. Frakes and Crowell (1975) showed that diamictites collected from Bolivia, Brazil, South Africa, Antarctica and the Falkland Islands are deficient in total Fe compared with modern tills; similar low values are typical of modern day and late Cenozoic glaciomarine facies where there is a low potential for oxidation of iron. The geochemical abundance values reported by Frakes and Crowell (1975) are compatible with sedimentological, biostratigraphic and structural evidence for widespread glaciomarine conditions. 16.12 G o n d w a n a glaciation; an ice sheet m o d e l

The large dimensions of Gondwana may have produced a continental climate characterized by extreme seasonality (Crowley et al., 1987). Crowley et al. (1991) argued that

N. ZVLZS

seasonal warming over the continent was so extreme that an ice cover could not survive and they suggested that additional factors, such as changes in solar luminosity and orbital (Milankovitch) driven changes in summer insolation, were required to allow ice sheet growth. Significantly, atmospheric carbon dioxide values were near levels typical of Pleistocene glaciations (Berner, 1990; Figs. 6.2, 21.1). Uncertainties in climatic modelling of the Gondwanan ice cover lead Crowley et al. (1991) to invoke "topographic considerations" such as an increase in average elevation across the continent. This had earlier been suggested by Hay et al. (1981), who presented data consistent with an average continental surface elevation of 1.2 km, and is implicit in the tectonic model of "adiabatic" glaciation proposed by Veevers and Powell (1987). It is also likely that presence of large intra-continental seaways (Fig. 16.2) would moderate the tendency for strong seasonal effects by funnelling moisture to interior ice caps on uplifted terranes. Important global climatic information can be gained by comparing the Gondwanan ice cover with that developed during Pleistocene glaciations (Crowley and Baum, 1991a; Crowley et al., 1991). These workers have assumed that the geographic extent of glacial deposits across the supercontinent can be employed to define the dimensions of the ice sheet. Other data regarding the former dimensions of the Late Plalaeozoic ice cover can be gained from consideration of the magnitude of glacio-eustatic sea level fluctuations recorded on continental shelves outside the glaciated area. Much information is contained in the "cyclothemic" strata of eastern North America and Europe that record cyclic changes in sea level that accompanied the waxing and waning of Gondwanan glaciers. Crowley and Baum (1991a, b) modelled three different ice-sheet conditions for the late Carboniferous; ICE I, II and III. The minimum reconstruction (ICE I) depicts an ice cover across the entire Paran~ and Karoo basins, over much of India and across central

153

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

and eastern Australia. This ice cover has a total volume of about 40 x 10 6 knl3; the maximum model (ICE III) has a volume of nearly 110 × 10 6 k m 3 which is almost double the maximum global ice volume in the Pleistocene (somewhere b e t w e e n 52 and 69 x 10 6 km3; Fig. 16.21). Given that the maximum glacio-eustatic sea-level fall during the Pleistocene was anywhere b e t w e e n 90 and 175 m (Fig. 16.21), then the much larger volume of I C E III would have resulted in glacio-eustatic sea-level fluctuations of almost 300 m (Fig. 16.21). Several lines of evidence allow closer constraints to be placed on ice volume estimates during Late Palaeozoic glaciation(s). The first is that the ice cover across G o n d w a n a never constituted a single ice sheet; the ice cover most likely consisted of several different ice centres that waxed and w a n e d diachronously across the supercontinent (Crowell, 1978). Secondly, the bulk of G o n d w a n a n glacial strata that are used to define the extent of the ice cover by Crowley and Baum (1991a,

b), are glaciomarine in origin and do not define the former area of terrestrial ice cover. This m e t h o d of reconstructing palaeo-ice sheets is akin to drawing the dimensions of the Late Pleistocene Antarctic or Laurentide Ice Sheets on the basis of the distribution of ice-rafted glaciomarine deposits in the surrounding oceans. Figure 16.21 suggests that ice volumes during any one phase of Gondwanan glaciation approximated the minimum reconstruction (ICE I) envisaged by Crowley and Baum (1991a, b). This ice cover has a total volume of about 40 × 106 km 3 which by comparison with Late Pleistocene relationships b e t w e e n ice volume and sea level suggests a maximum Late Palaeozoic glacioeustatic fall of about 70 m (Fig. 16.21). This value, and the reasoning on which it is based, suggests that changes in sea-level, inferred from Carboniferous strata in North America and E u r o p e of up to 200 m (Ross and Ross, 1985, 1988), are not glacio-eustatic in origin but, if correctly identified, are the result of tectonic subsidence. An important constraint

AGE LPRESENT DAYJ

ICE VOLUME (1 0 6 km ~ ) 32 ~

INTERGLACIAL

(MAX)

IL.

PLEISTOCENE I GLACIATION
LAURENTIDE ICE SHEET ANTARCTICICE SHEET EXTIMATES OF GLACIOEUSTATIC SEA-LEVEL FALL

(MIN)

69'

52 ~

34.2 ~ 9.8 ~

25 z

1753 150" 1305 t25 '~ 90 z

]L. PALAEOZOIC I GLACIATION c.300 Ma

1108

408''

ESTIMATED GLACIO-EUSTATICSEA-LEVEL FALL

2787o

697o

IE. PALEOZOIC

I

GLACIATION c. 440 Ma

ESTIMATED GLAClO-EUSTATIC SEA-LEVEL FALL

10" 25.4 ~o

Fig. 16.21. Ice volumes and glacio-eustatic sea level fall (in metres) for Phanerozoic glaciations. 1 Denton and Hughes (1981). If present day ice masses were to melt, global sealevel would rise by about 70 m; about 60m of this sea-level rise would be due to the melt of Antarctica's ice sheets (Andrews, 1992); 2 Paterson (1972); 3 Veeh and Veevers (1970); 4 Chappeli (1983); 5 Chappell and Shackleton (1986); 6 Curray (1965); 7 Dillon and Oldale (1978); s Crowley et al. (1991); 9 This paper, Fig. 16.3; 10 Note: Because ice volumes and sea-level fall for the Late Pleistocene are not well constrained (e.g. Andrews, 1992; Colhoun et al., 1992) no unique solution exists for relationship between ice volume and magnitude of sea-level change in pre-Pleistocene basins. This complicates estimates of sea-level based on volume of pre-Pleistocene ice covers. Sea-level estimates here are based on a Late Pleistocene global ice volume of 69 × 10 6 km3 giving 175 m of sea-level fall; 11 Vaslet (1990), Fig. 15.6.

154

N. EYLES

on ice volume determinations for Late Palaeozoic ice covers is provided by the detailed record of sea level change that has been identified in Pennsylvanian strata of North America.

16.13 Late Palaeozoic glacioeustasy Well-defined cyclic sequences of alternating transgressive and regressive strata have long been recognized in the Pennsylvania of North America (c. 330-280 Ma) and were described as "cyclothems" by Weller (1930). Their origin has been controversial and varyingly linked either to basin tectonics (Tankard, 1986) or global glacio-eustatic sea-level

,~X~,~..,d~

0

,,_ .... 5oo ,

( ILLINOIS

~""



_,~ R" /~

o"""<./o / ~ , / . ~ "

<> - , - >

\ A: KANSAS-TYPE

B: ILLINOIS-TYPE

C: APPALACHIAN-TYPE

l

Fig. 16.22. Tectono-stratigraphic model for Pennsylvanian (c. 330-280 Ma) cyclothems in eastern North America. Based on Klein and Willard (1989).

changes (Crowell, 1978; Ross and Ross, 1985; Powell and Veevers, 1987; Veevers and Powell, 1987). Three types of cyclothems can be identified (Fig. 16.22). These range from those composed of marine limestones and black shales (Kansas type), those of the Michigan and Illinois basins that are composed of mixed marine and non-marine facies (Illinois type) and those of the Appalachian and Arkoma basins which show alternating marine and non-marine clastics containing thick coal deposits (Appalachian type). It is now widely agreed that different processes have been responsible for producing the three cyclothem types. The non-marine, coal-bearing clastic cyclothems of eastern North America (Appalachian type) have been interpreted as the result of episodic loading by thrust sheets of the Ouachita and Appalachian orogenies on the margins of a foreland basin. The basic tectonic model was presented by Quinlan and Beaumont (1984). Thrusting results in renewed downwarping and marine transgression followed by regression and subaerial exposure as deltaic sediment wedges prograde from areas of renewed uplift. Each cyclothem records a change in relative sealevel rise, of up to 30 m, resulting from renewed crustal downwarping in the foreland basin (Tankard, 1986; Klein and Willard, 1989; Klein and Kupperman, 1992). Klein and Willard (1989) show that Illinoian-type cyclothems record deposition in a setting intermediate between the largely stable continental interior and that of the tectonically active foreland basins along the active continental margin to the east (Fig. 16.21). Marine carbonate cyclothems (Kansas type) appear to be the sole product of glacio-eustatic transgressive/regressive cycles driven by Southern Hemisphere Gondwana glaciations of the Late Palaeozoic (Fig. 16.22). Tectonic influences within the continental interior are considered of minimal importance and even small-scale cycles are

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E ' I T I N G

regarded as correlative across large areas of the North America northwards from Texas to Kansas and Iowa using first, last, sole, or acme occurrences of fusilinids, conodonts and ammonoids (Boardman and Heckel, 1989). Milankovitch forcing was argued by Heckel (1986). Klein and Kupperman (1992) highlighted problems in applying specific Milankovitch frequencies, derived from investigations of the Pleistocene record, to analysis of Pennsylvanian cyclothems but nonetheless recognized well-defined glacio-eustatic cycles driven by Milankovitch orbital forcing. The magnitude of glacio-eustatic sea level fall is not well known and recent estimates vary from 20 to 100 m with recent basin modelling suggesting a value closer to the first value (Klein, 1992; Klein and Kupperman, 1992 and refs. therein); glacial geologic data reviewed above (Section 11.11) indicate that a value of about 70 m may be realistic (Fig. 16.21). Weimer (1992) identified glacio-eustatic sea-level changes of up to 45 m from the Lower Pennsylvanian Morrow Formation of the northwestern Anadarko Basin in Colorado and Kansas. Fluvial valley fill deposits that accumulated during low stand exposure of a shallow shelf comprise excellent exploration targets for hydrocarbons. In contrast, neither transgressive or regressive shoreline sandstones were preserved because of the rapidity of water depth changes. In conclusion, it appears from evidence that is presently available that the magnitude of Late Palaeozoic sea-level change attributable to the growth of Gondwanan ice sheets is less than 100 m and probably closer to 70 m. Ice volumes during Late Paleozoic "glaciation" (about 40 x 10 6 kin3; Fig. 16.21) may have been about the same as those of the present day "interglacial" (32 x 106 kin3). Melt of the present day Antarctic ice sheet would result in a sea-level rise of about 60 m (Andrews, 1992). Of course, the absolute magnitude of sea-level change resulting from glaciation is only half the story; the rate of glacioeustatic sea-level change may be of greater sedimentological importance. Rates

155

of sea-level change during Pleistocene glaciations are well constrained. The growth history of Pleistocene ice sheets was markedly asymmetrical characterised by relatively slow growth, punctuated by short-lived periods of more rapid expansion, and catastrophic decay (Sections 18.3, 20.1; Fig. 20.2). During periods of slow growth, sustained over tens of thousands of years, sea-level fell at the rate of about 1 m / K a increasing to between 5 and 6 m / K a during shorter periods of accelerated growth. In contrast, during decay, sea-level rose on average by 10 m every thousand years, with rises of as much as 4 m in as little as 100 years and 20 m in less than one thousand years (Chappell and Shackleton, 1986; Fairbanks, 1989; Bard et al., 1990; Plint et al., 1992). Over any one glacial/interglacial cycle this resulted in relatively slow exposure of continental shelves and rapid flooding which has clear implications for the relative thickness and preservation potential of regressive and transgressive facies along continental margins. It should be emphasised perhaps that rates of Pleistocene glacio-eustatic sea-level change cannot be applied directly to interpretation of the ancient record because controls on glaciation were probably very different.

16.14 Evaporite cyclothems Carbonate and evaporite depositional systems are particularly sensitive to changes in sea-level (e.g Walker and James, 1992) and so may contain high-resolution records of glacio-eustatic changes (Bard et al., 1990). The Amazonas/Solimoes basins of Northern Brazil contains thick (1200 m) PermoCarboniferous strata that show cyclic sequences of evaporates and carbonates of the 1200 m thick Carauari Formation. These rocks range in age from the Early Carboniferous to the mid-Permian (Caputo et al., 1990) and therefore span the timeframe for the main Gondwanan glaciation (Fig. 16.1). Cyclicity results from sea-level fluctuations during an overall rise in relative sea-level;

156

probably the result of glacio-eustatic sea-level change superimposed on basin subsidence (P. Szatmari, pers. commun., 1991) similar to that identified above for Kansan- and Illinoian-type cyclothems of North America. Study of the Caravari Formation may provide a detailed record of ice-volume fluctuations that could be directly compared to that recorded by the North American cyclothems. Evaporite basins are particularly sensitive to changes in the balance between precipitation and evaporation. Detailed study of the Carauari Formation along the lines of Anderson (1982, 1984) on the Permian Castile Formation could provide a wealth of climatic data. 17. M E S O Z O I C G L A C I A T I O N S

"It would hardly be an exagggeration to assert that if geologists had not hitherto discovered the Ice Age ... astronomers would have demonstrated by calculation that Ice Ages must have happened and would ... be urging ... geologists to go and look" (Ball, 1891, p. 44). Mesozoic glacial strata have not so far been found but short-lived, fourth-order cycles of sea-level change recorded in the marine record have been interpreted as astronomically-forced, glacio-eustatic fluctuations (Brandt, 1986; Masetti et al., 1991; Plint, 1991). The possibility of hairy, endothermic dinosaurs roaming a frigid Arctic has generated much controversy as to the possible physiology of these animals (Benton, 1991). In the light of these and other discussions there is intense examination of the geological record for glacial deposits.

17.1 Geological evidence Direct geological evidence of glaciers during the Triassic, Jurassic and Cretaceous has not, so far, been reported though ice-rafted horizons in Jurassic and Cretaceous strata of Siberia and in Cretaceous strata of Australia, indicate seasonally-cold conditions at sea-

N. EVLES

level (Epshteyn, 1978; Frakes et al., 1988). Frakes et al. (1992) designated the late middle Jurassic to early Cretaceous (183-105 Ma, Bajocan to mid-Albian) as a "cool mode" in recognition of evidence of seasonal ice in high latitudes and marked seasonality (Fig. 21.1). This raises the possibility whether glacier ice may have developed in continental interiors positioned at high latitudes (e.g. Antarctica, Siberia) during the middle Mesozoic. E p s h t e y n (1978) d e s c r i b e d pebbly "glaciomarine" clays of Jurassic age from northern Siberia. These deposits contain glendonites that typically form in polar waters with temperatures of ~ O°C (Jansen et al., 1987). On the basis of this report, Brandt (1986) attributed cycles of eustatic sea-level variation, lasting between several hundred thousand years to two million years, recorded in early Jurassic strata (Sinemurian-Pliensbachian) of Southern Germany, to the waxing and waning of an ice sheet in northern Siberia. However, no direct glacial geologic evidence exists for such an ice sheet and it must be emphasized that Epshteyn (1978, p. 54) used a very loose definition of "glaciomarine" which included "material of the beach and shallow-water area, brought in by shore ice floes" and nowhere identified debris rafted by glacier ice. Brandt (1986, p. 267) errs in assuming that Epshteyn (1978) referred to ice berg transported debris released from glaciers. Chumakov (1981a) refers to other studies that argue for rafting of clasts by sediment gravity flow or by driftwood. Analogous pebbly mudstones of Palaeocene and Eocene age have been described from the Tertiary Central Basin of Spitsbergen (Dalland, 1976). Mudstones are extensively bioturbated and contain dropstones from sand size up to 150 kg. Clasts are either uniformly dispersed or concentrated on shallow marine erosion surfaces. A palaeolatitude of about 70°N is consistent with rafting of beach clasts by winter sea-ice. Dalland (1976) remarked on the similarity of these

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

facies with ice-rafted horizons of the Cretaceous Inn Kjegla Member in Spitsbergen. The important role of sea-ice in transporting sediment on shallow, seasonally cold seas is well known (MacCarthy, 1988; Drake and McCann, 1982; Dionne, 1985; Martini and Johnson, 1987; Pfirman et al., 1989; Bischof et al., 1990; Reimnitz et al., 1992). Frakes and Francis (1988) report early Cretaceous ice-rafted horizons in central Australia, and similarly, were unable to demonstrate that such deposits had been emplaced by ice bergs; marine shore ice or river ice was probably responsible. Embry (1984) reports icerafted dropstones from early Cretaceous strata of the Sverdrup Basin in northern Canada; fossil tree ring and isotopic data provide evidence of distinct seasonality at this time (Frakes et al., 1993). An absence of widespread glacially icerafted debris during the Mesozoic is not in itself a conclusive argument for the absence of continental ice sheets because the latter are unlikely to have extended to sea-level. The development of seasonally-cold conditions at sea-level clearly allows for the existence of much more severe conditions in continental interiors. The presence of significant ice covers over the Siberian and Antarctica land masses cannot be ruled out since both these areas lay at high paleolatitudes.

17.2 Invisible ice: Middle Mesozoic glaciers with no stratigraphic record It is worth emphasising that the geologic record of warm Cretaceous paleoclimates is contained in shallow marine strata deposited on the margins of continents. These deposits do not rule out the presence of ice in continental interiors. Oxygen isotope studies, using concepts of late Cenozoic relationships between 160/180, have questioned the existence of any significant Cretaceous ice cover but some of the assumptions behind such work are debateable. Polar ice caps may have little influence on global isotope ratios if they were small or were fed by precipitation

157

derived from warm oceans (Barron, 1989). Palaeobotanic and palaeontological investigations have identified a markedly seasonal cool temperate climate in high palaeolatitudes during the Cretaceous (Spicer and Parrish, 1986; Rich et al., 1988). These data are important because continental land masses at high latitudes, surrounded by warm oceans, satisfy the one fundamental requirement for ice sheet growth, namely relatively mild, wet winters and cool summers. Recent work on the Cenozoic history of the Antarctic ice sheet over the last 36 Ma, is pertinent to consideration of Mesozoic ice sheets. The present day dry-based "polar" characteristics of the Antarctica ice sheet were only established in the Late Pliocene. Before this time the ice sheet was wet-based and temperate in character. Harwood (1991) and McKelvey et al. (1991) concluded that pre-Pliocene Antarctic glaciations were warmer and wetter; the presence of temperate southern Beech (Nothofagus) forests, which persisted through many glacial cycles records a humid, oceanic climate with a limited temperature range (Barrett, 1989). Oerlemans (1982) also showed that with a sealevel temperature of 0°C, i.e. 20 ° above the present temperature, an ice sheet could still develop in the interior; with a sea-level temperature of as much as 5°C mountain glaciation would still occur. Oglesby (1989) concluded that perennial snow fields would develop with sea-surface temperatures as high as 14°C. Thus the present day frigidity of the Antarctic environment is a poor analog on which to base interpretation of a possible Cretaceous ice sheets. Late Cenozoic, but pre-Late Pliocene, Antarctic ice sheets may provide a better model for Cretaceous ice covers on high latitude landmasses in Antarctica and Siberia. Tectonic studies identify significant uplift of the Antarctic landmass during the Mesozoic and this may have some considerable bearing on the existence of former ice covers. The use of fission-track dating to identify the uplift and denudation history of the

158 Transantarctic Mountains shows two periods of strong uplift, in the Early and Late Cretaceous, concident with plate tectonic changes in the region (Stump and Fitzgerald, 1992). This might suggest a direct relationship between tectonic uplift and any Middle Mesozoic ice covers on Antarctica during the early Cretaceous, Bajocian to mid-Albian "cool climate" mode of Frakes et al. (1993) (see Section 19.1 and Fig. 21.1). Middle Mesozoic ice sheets are most likely to have developed under conditions of steep climatic gradients from the coast to the interior of continental surfaces in high latitudes (Barron et al., 1981). Severe climatic gradients are a common present day characteristic of temperate oceanic glaciated coastlines. The Gulf of Alaska in the northern Pacific supports extensive inland ice fields including the largest glacier complex in North America (Bering Glacier; 5800 km2). The presence of warm water offshore and considerable local relief results in very high snowfalls; cool summers ensure the survival of glaciers. Mean annual temperatures along the coast are about 4oc; those just a few kilometres inland are much more severe. Another good example is that of Iceland in the North Atlantic Ocean which despite mean annual temperatures along its southern coastline of 4°C, supports the largest ice cap in Europe (Vatnajokull; 8300 km2). Both the Gulf of Alaska and Iceland are favoured with high precipitation and cool summers and might be good models for Mesozoic ice covers. Axelrod (1984) has identified the presence of temperate conditions at sea-level that are comparable with modern North Atlantic and Pacific cool temperate glacial settings. The one negative feature about these modern settings is that because of their temperate character there is no development of coastal sea ice and thus no flux of sea ice-rafted debris to the marine environment. Consequently, these modern day ice sheets will be "invisible" in the geologic record. Given the demonstrated occurrence of sea ice formation in the Mesozoic and Tertiary (Epshteyn,

N.EYLES 1978; Spicer, 1987) a more continental cool temperate setting is suggested; the likelihood of glacier ice hiding in the interiors of Siberia and Antarctica is thereby strengthened. 17.3 The Mesozoic sea glacioeustatic control?

leuel record;

a

The pattern of past sea-level change may provide proxy evidence of the existence of Mesozoic ice sheets. The growth of large ice sheets results in changes of absolute sea-level that generate globally-synchronous unconformities in the form of ravinement surfaces and flooding surfaces that define individual depositional sequences on continental margins (e.g Posamentier et al., 1988). The growth of Pleistocene ice sheets results in a sea-level drawdown of as much as 175 m and are equivalent to 4th-order sea-level cycles typically lasting from a few tens to a hundred thousand years (Miall, 1990, 1991). Good examples are the depositional cycles of transgression and regression evident in Pennsylvanian coal-bearing strata of the eastern United States ("cyclothems"; Weller, 1930) that record changes in the volume of ice sheets across Gondwanaland (Crowell, 1978; Klein and Willard, 1989; Figs. 16.21, 16.22). Well-defined 4th-order sea-level cycles have been identified in Cretaceous strata deposited in the Interior Seaway of North America (Walker and C.H. Eyles, 1988; Plint, 1991) and in the Jurassic of southern Germany (Brandt, 1986). Circumstantial evidence for the possible existence of a sizeable ice cover during the Triassic also occurs in Middle and Late Triassic strata in the Dolomite region of northern Italy. In this area several workers identified fourth and fifth-order eustatic sea-level cycles (100 Ka and 20 Ka respectively) from platformal limestones and invoked a glacial control (Goldhammer et al., 1987; Masetti et al., 1991). Neither tectono-eustasy, nor tectonic movements of the sea floor operate on a sufficiently rapid timescale to account for the high frequency sea level changes recognised by the above named workers (e.g. Cloetingh,

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'Iq'ING

1988; Harrison, 1990; Cathles and Hallam, 1991) and by default a glacio-eustatic control becomes very appealing. Such a control is not precluded by the oxygen isotope record (Mathews, 1986; Prentice and Matthews, 1988). It is very tempting to ascribe such sea-level fluctuations to orbitally-forced systematic "Milankovitch" changes in Mesozoic ice volume as is well documented for Pleistocene glaciations (e.g Chappell and Shackleton, 1986; Martinson et al., 1987; Section 20). Indeed, it has been argued that Milankovitch-type orbitally-forced insolation cycles can be recognized during most of the Phanerozoic (Fischer, 1986; Berger et al., 1989). However, it is worth emphasizing three points. The first is that small-scale (i.e. < 1 Ma) subdivision of the pre-late Cenozoic timescale is not precise and so the apparent "cyclicity" of 4th-order sea-level changes may not be real. Secondly, even if real, the presence of 4th-order sea-level cycles may in fact, record orbitally-forced changes in precipitation or groundwater storage rather than cyclic fluctuations in ice volumes (Hay and Leslie, 1990). Thirdly, the long-term stability of the Milankovitch orbital rhythms is open to question (Berger and Loutre, 1989; Berger et al., 1989; Laskar, 1989). The problem of explaining 4th- and 5th-order cycles of sea level without necessarily attributing them to Milankovitch-forced fluctuations of glaciers also surrounds interpretation of 5th-order cycles in the Cambrian (e.g. Osleger and Read, 1992; Section 14). Perhaps the stongest objection to invoking a glacial control for the high frequency sea level changes recorded in Cretaceous and Tertiary strata is whether any ice covers were large enough to have exerted a significant glacio-eustatic control. Whereas it may be possible to tuck away some small ice masses in Antarctica or Siberia it is extremely unlikely that they could have had any noticeable effect on global sea levels. Other mechanisms, involving complex feedbacks and threshold effects, will undoubtedly be discovered that will link

159

"Milankovitch" orbital forcing to sea-level change without the need to invoke ice sheets in the Cretaceous (e.g. Park and Oglesby, 1991). Changes in global precipitation, evapotranspiration, greenhouse gases and temperature are obvious contenders (see Fischer et al., 1990; Street-Perrott, 1992). Nonetheless, it still needs to be firmly established that inferred sea-level fluctuations are indeed both cyclic and orbitally-forced (Ginsburg and Beaudoin, 1990) and not the product of tectono-eustatic controls, such as intraplate stress changes and geoidal eustasy (Morner, 1987; Cloetingh, 1988; Leighton and Kolata, 1990; Sloss, 1990), regional tectonics including thrust sheet loading and unloading or even local, autocyclic controls within the depositional basin. The global extent of such sea level changes needs to be clearly demonstrated. A further complication, inherent to all investigations of sea-level history is that of determining the magnitude of sea-level fluctuations from stratigraphic sequences (see Burton et al., 1987). As Christie-Blick (1990) has cautioned, the magnitude of eustatic sea-level changes invoked by Vail et al. (1977) and Haq et al. (1988) may be severe overestimates. Further research on carbonate and evaporative platforms that are more sensitive to sea-level changes than siliclastic shelves (see Schlanger and Philip, 1990), will probably answer for good the debate on glaciers in the Mesozoic. This is borne out by the high-resolution record of glacioeustatic changes recorded on Late Ordovician carbonate platforms (Section 15.5) and on late Cenozoic reefs (e.g. Chappell and Shackleton, 1986; Roop et al., 1992). In conclusion, the question of Mesozoic ice caps at high palaeolatitudes remains open; there are no geological or geophysical data that flatly contradict the presence of ice caps but neither are there hard data that positively identify ice covers during this period. On balance, the available evidence indicates that the Mesozoic Earth was very unlikely to have been ice-free.

160

N.EYLES

18. LATE CENOZOIC GLACIATIONS (36-2.5 Ma) It is now well established that extensive Northern Hemisphere ice sheets were initiated in Plio-Pleistocene time (c. 2.5 Ma) and was the culmination of a long global climatic deterioration that began sometime after 60 Ma. The precise mechanisms acting to cool the planet are not known but late Tertiary tectonic uplift and plate tectonic reorganizations can be identified as first order controls. Initiation of the East Antarctic ice sheet, at or before 36 Ma, is directly related to the progressive thermal isolation of the continent as Australia moved northwards (Kennett and Shackleton, 1976; Kennett, 1982) together with uplift of the Transantarctic Mountains (Fitzgerald, 1992). Well-defined glacio-eustatic sea-level fluctuations, inferred from unconformities present in continental margin stratigraphies and fluctuations in stable isotope ratios record the repeated growth and decay of Antarctic ice masses (Miller et al., 1987). In the Northern Hemisphere, tectonic uplift around the margins of the newly-rifted North Atlantic and the development of broad upwarped plateau there and elsewhere (e.g. Tibet) set the scene for Pleistocene ice sheet growth and decay driven by Milankovitch orbital forcing (Ruddiman et al., 1989). It is arguable whether extensive continental glaciations of the Northern Hemisphere, and concomitant evolution of hominids (see Molnar, 1990), would have occurred without the necessary precondition of tectonic uplift.

18.1 Late Tertiary global cooling

MAASTR PALEO(INIE

-15 ~ -10 ~

.

.

.

.

.

.

.

5

"~-0 60

70

S'O 40 30 20 MILLION YEARS

I0

~ e

0

Fig. 18.1. Deep ocean temperatures as a record of Late Tertiary global cooling. Based on Shackleton(1987). of climatic deterioration on plants and the enhanced erosion of continental surfaces. Shackleton (1987) showed a decrease of nearly 10°C in the temperature of deep ocean waters over the last 60 Ma (Fig. 18.1). Synchronous fluctuations in benthic and planktonic (5]80 values from the Oligocene to the Miocene (36-5 Ma) Atlantic and Pacific oceans were identified by Miller et al. (1987) and linked to Antarctic ice sheet growth and decay (Fig. 18.2). These workers were able to use a new late Tertiary magne8"O 3.0 i

i

i

2.0 i

0

1.0 ,

0.0 I

i

,

k--- Fig.1 8 . 3

10

.:ili

_

•-

?...i;.

ICESHEETS 30

,':'2

-1.0 I

O

ICEFREE OlJgocene INITIATION OF ANTARCTIC ICE SHEET ICl

~1--

~

Z

Eocene

ATLANTIC OC 50

.......

-2.0 i

Plio-Pleistocene

"::

"-iJ40

1

Miocene

..

~[

18.1.1 The Oligocene to Miocene deep marine oxygen isotope record Late Tertiary global climate cooling is expressed in the deep ocean basins by greatly enhanced rates of sedimentation and a decrease in deep water temperatures. Donnelly (1982) identified a six-fold increase in the rate of accumulation of aluminum over the last 15 Ma and attributed this to the effects

PLEISTOCENE PUOCENIE

PACIFIC OCEAN SITES

60

~

Paleocene Cretaceous

70

I

I

I

I

I

I

[

I

I

I

I

Fig. 18.2. 40 million year long record of Antarctic ice volumes identified by synchronous changes in oxygen isotope values from benthic and planktonicmicrofauna from the Pacific and Atlantic oceans. Modified from Miller et al. (1987). The more recent ( < 6 Ma) marine record is shown in Fig. 18.3.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

tobiostratigraphic timescale. The observed record of changing 8180 values was calibrated against global ice volume and a glacio-eustatic sea-level curve was estimated. The commonly quoted calibration of 8180 to sea-level is 0.11%o 6180 per 10 m of sea-level change but this relationship is based on 6180 values typical of large Pleistocene ice sheets where high ice sheet elevations result in lower mean values of 8180. Employing a revised calibration of 0.055%0 per 10 m, Miller et al. (1987) were able to identify glacio-eustatic low stands at 35, 31, 25, 14 and 10 Ma (Fig. 18.2). These fluctuations occurred over short timespans ( < 2 Ma) and were of the order of 30-90 m in magnitude. A relatively ice-free period in the Early and Middle Miocene can be identified (Fig. 18.2); this period was perhaps characterised by the rapid growth and decay of small ice sheets that produced only small changes in 8180 of sea water. The earliest well-dated evidence of late Tertiary glaciation is that from ODP site 748 on the Kerguelen Plateau in the southern Indian Ocean. This site records a short-lived expansion of the Antarctic ice sheet, during the earliest Oligocene at about 36 Ma, which lasted for between 10 and 150 Ka (Zachos et al., 1992). Age control is by reference to calcareous nannofossil and planktonic foraminifera. Pelagic marine sediments at many other sites record a positive shift in 180 values in benthic foraminifera at this time consistent with increased ice volume over Antarctica. Zachos et al. (1992) suggested that the source of ice-rafted debris at site 748 was Prydz Bay where ODP Leg 119 recovered Lower Oligocene diamictites close to the outer margin of the present day continental shelf (Barron et al., 1989). These deposits were interpreted as lodgement tills, implying an Antarctic ice sheet much larger than that of the present day grounded to the shelf edge. This model is of course reliant on correct identification of lodgement till; Zachos et al. (1992) acknowledge that a more restricted ice cover consisting of alpine or valley glaciers cannot be ruled out given the

161

very short timespan for growing such a large ice mass. The development of alpine glaciers, perhaps along the uplifting rim of the Transantarctic Mountains or along the Antarctic Peninsula is in keeping with the record of subsequent Tertiary glaciations (e.g. Birkenmajer, 1991, 1992; Section 26.2). The extent of Tertiary ice cover(s) on Antarctica is important to identifying the magnitude and origin of sea level excursions at this time. Miller et al. (1987) suggested that short-lived, late Tertiary glacio-eustatic sea-level changes, resulting from the growth of substantial ice covers across Antarctica, are recorded by erosional events in the stratigraphy of passive continental margins. Initial results comparing t~180 variation with the "Vail" curve are promising with prominent correlations of significant "offlap" events with glaciation at 10 Ma, 25 Ma and 31 Ma.

18.1.2 The deep marine oxygen isotope record after 5 Ma The Late Miocene represents a critical phase in late Cenozoic climatic change with widespread evidence for glacial activity beginning in and around 6 Ma. Many sites indicate a mid-Pliocene warm interval followed by a resumption of intensified glaciation beginning at about 3.5 Ma. This latter phase marks the start of the "classical" glacial and interglacial cycles typical of the Pleistocene. High frequency fluctuations in 8180 contained in benthic foraminifera can be identified in Late Miocene to earliest Pliocene (6.37 to 3.5 Ma) sediments in the Atlantic and Pacific Oceans (Hodell et al., 1986). Several DSDP sites show a ~ 0.5%o fluctuation in t~180 with a periodicity of between 200 Ka and 400 Ka years as is clearly shown by data from site 588 in the southwestern Pacific Ocean (Fig. 18.3). This change is about one third of the amplitude shown by oxygen isotope fluctuations associated with the "classical" Pleistocene glacials and interglacials after about 2.5 Ma (e.g. Fig. 18.4) and can be attributed to the expansion of ice

162

N. EYLES

over West Antarctica. This is consistent with the marine geologic record of ice-rafting in the southern oceans which shows an abrupt expansion in the northward extent of icerafted debris during the latest Miocene. This event is coeval with the onset of glaciation in southern South America at about 7 Ma recorded by tills and associated basalt flows of the Meseta del Lago Buenos Aires in Argentina (Mercer and Sutler, 1982) and with direct geologic evidence of repeated expansion of the west Antarctic Ice Sheet from offshore seismic records (see below). Short-lived climatic excursions, consistent with extensive Late Miocene glaciation between 5.8 and 4.8 Ma, have also been identified from oxygen isotope studies at ODP site 552 in the northeast Atlantic and ODP site 558 in the southwestern Pacific (Keigwin, 1987) and DSDP site 289 in the western equatorial Pacific (Woodruff et al., 1981). Ciesielski et al. (1991) report a similar scenario from ODP site 704 in the south Atlantic; an episode of strong glacial activity is

c-

,$"o 2.4 Age (Ma)

---]

2.2 Fig. 1 8 . 4 - ~

2.0

1.8 " '

l

"1~

3.88 -

4.77 --

5.35

- -

5.53 -5.68 -5.89 -6.37 --

Fig. 18.3. Late Miocene and Early Pliocene record of Antarctic ice volumes from the southwest Pacific Ocean (after Hodell et al., 1986). Note resemblance to glacial and interglacial cycles shown on subsequent high resolution record after 2.8 Ma (Fig. 18.4).

"o(~) 4.5 3.5

1.6~\

4.5 35

25

I~

5.5

5"0

4.5 3,5 25

,

0

,

o

0.8

I~-R:'71 \ ,.sJ ~ " 1

25

\\

/ 2

891 /

79 2.0

q

60

¢121~'"1 :~ 2.2

9t

2.4

i 99

\

J ~r13118 80

,<

100

t

I ~

103

'°'/

120

io,

I~°7

2,t 'il,

.o

109

115

28

x\ 1% I I.,__/ 57 I

S

NORTH ATLANTIC SITE 607 (Raymo el al., 1989) (Ruddiman et al., t986)

"~

NORTH ATLANTIC SITE 607 (Chappell and Shacklelon,1986)

Fig. 18.4. High-resolution, deep sea oxygen isotope record of Plio-Pleistocene global ice-volumes after 2.8 Ma. See also Fig. 20.2. Compiled from listed sources.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETI'ING

apparent in the Late Miocene (about 6.5 to 5.35 Ma) followed by a weakened glacial influence about 3.4 Ma when vigorous glacial activity was renewed. Tidewater glaciation is recorded in the Norwegian Sea at about 5.5 Ma (Jansen and Sjoholm, 1990). Lagoe et al. (1993) summarize the marine and onshore record of the northern Pacific Ocean and demonstrate initial tidewater glaciation between 6.7 and 5.0 Ma, followed by a midPliocene warm interval from about 4.2 to 3.0 Ma. A major increase in the intensity and extent of glacial influence commences between 3.5 and 2.4 Ma; this is coeval with the initial development of continental ice sheets in North America and Europe (Sections 18.2, 18.3). In summary, a large body of data from both the Northern and Southern Hemispheres indicate that the initiation of late Cenozoic glaciation took place in the Late Miocene and that this event was largely coeval in both hemispheres. A mid-Pliocene warm interval was followed by the onset of widespread glaciation after 3.5 Ma. The marine geological evidence of ice growth is reviewed below.

18.1.3 Glacigenic depositional sequences of the Antarctic continental margin Repeated expansion of the West Antarctic ice sheet is recorded on multichannel seismic profiles across the Pacific margin of the Antarctic Peninsula (Larter and Barker, 1989). At least eight well-defined seismic sequences have been preserved during the last 6 Ma. Each sequence is produced by expansion and grounding of the ice sheet margin out to the continental shelf edge and essentially records progradation ("outbuilding") of the continental slope combined with vertical aggradation ("upbuilding") of the shelf. Larter and Cunningham (1992) show that individual sequences are dominated by diamictites resulting from cross-shelf transport as a subglacial "deforming bed" (Section 3.1.1c) below a grounded ice sheet composed

163

of coalesced ice streams. Alonso et al. (1990) and Anderson and Bartek (1992) identify a similar stratigraphic record, characterised by deeply-scoured erosion surfaces and thick diamictite blankets on the Ross Sea continental margin. Successive sequences are preserved because of rapid subsidence; about 1 km of subsidence has taken place in the last 6 Ma (Larter and Barker, 1991). Excellent preservation along the glaciated margin of the Antarctic Peninsula contrasts with the paucity of the record along glaciated margins that have experienced uplift over the same time interval. Under these conditions, terrigenous sediment bypasses the shelf and a long term record of glacial sedimentation is only preserved on continental slopes (e.g. eastern Canadian continental margin; Section 23.1 and Barents Sea continental margin; Vorren et al., 1989). The recent development of sequence stratigraphy provides a systematic framework for analysing the relationship between sediment progradation and aggradation on continental margins (Van Wagoner et al., 1988; Wilgus et al., 1988). Depositional sequences and their bounding unconformities form in response to changes in relative sea level which is widely assumed to reflect global eustatic sea level fluctuations (Posamentier et al., 1988). The rationale of sequence stratigraphic analysis, together with its limitations, are well-known and are not related here (see Miall, 1986, 1991; Cloetingh, 1988; ChristieBlick et al., 1990). Larter and Barker (1989) extended the concept of sequence stratigraphy to interpretation of the seismic sequences of the Antarctic Peninsula and argued that bounding unconformities on the continental shelf coincide with glacial maxima and their succeeding deglaciations. Expansion of the West Antarctic ice margin to the shelf edge is only possible at times of major volume increases of the Northern Hemisphere ice sheets (Anderson and Bartek, 1992). Therefore, the glacigenic depositional sequences recognised by Larter and Barker should be correlative with late

164 Cenozoic glacioeustatically-controlled depositional sequences identified in lower latitudes (e.g. Ito and Katsura, 1992) and with the glacial record preserved on Northern Hemisphere continental margins. Larter and Barker (1989) were unable to test whether late Cenozoic, high-latitude (e.g. Antarctic Peninsula) and lower latitude depositional sequences were deposited in phase because of uncertainty in the precise age dating of the Antarctic record. Indirect evidence of well-developed glacial and interglacial cycles of Antarctic ice volumes after 6 Ma is borne out by the close correspondence between deep ocean isotopic data and cycles of marine flooding and evaporative drawdown preserved in thick halite deposits of the Mediterranean Sea. During the Messinian (Late Miocene) the Mediterranean was largely isolated from the Atlantic as a result of tectonic closure at the western end of the basin. Global glacio-eustatic sea-level fall and rise are widely argued to be the driving mechanism for repeated basin filling and desiccation during the Late Miocene "Messinian salinity crisis" after 5.5 Ma (Hodell et al., 1986, 1989).

18.2 Marine seismic record of continental glaciation in North America Arguably the most intensely studied nonglacial late Cenozoic continental margin is the northern Gulf of Mexico where more than 3500 m of sediment has accumulated across the Mississippi Fan over the last 6.5 Ma. Beard et al. (1982) attempted to relate the stratigraphy of the northern Gulf of Mexico to late Cenozoic global change and glacioeustatic sea level fluctuations but were hampered by poor dating control and limited stratigraphic data. Feeley et al. (1990) recognised two distinct periods of fan growth; the first from the Late Miocene to the Middle Pliocene (6.6 to 2.4 Ma) and a second from the Late Pliocene to the present ( < 2.4 Ma). The latter interval is coeval with the initiation of continental glaciation across North

N.EYLES America (Section 18.3) and is characterised by a dramatic increase in sedimentation rates. More than 3000 m of sediment, contained in thirteen different depositional sequences, was deposited after 2.4 Ma. The highest sedimentation rate occurs after 0.8 Ma when about 70% of the total fan volume was deposited. Feeley et al. (1990) developed a proxy sea level curve from the deep sea oxygen isotope record for the last 3.5 Ma, using data published to 1984. Glacioeustatic lowstands are recorded by the cutting of erosional channels on the upper fan and the growth of lowstand slope fan bodies composed of slump and debris flow deposits. Glacioeustatic sea level highs (i.e. interglacials) are recorded by hemipelagic mud drapes. The sequence stratigraphy identified by Feeley et al. (1990) should ultimately be correlative with glacigenic depositional sequences preserved on glaciated continental margins as better seismic and age data accrue from these areas. The analysis of Feeley et al. (1990) is based on 8000 km of seismic data with biostratigraphic age dating control provided by two drill holes. Despite this wealth of data, Weimer (1991) criticized several key aspects of the sequence and chrono-stratigraphy developed by these workers arguing among other things that the density and quality of the seismic coverage used by Feeley and his co-workers is too coarse. These discussions serve a particularly useful purpose in demonstrating the difficulties of relating even well studied continental margin stratigraphies to geologically recent glacioeustatic sea level fluctuations. This should be borne in mind with regard to interpretation of pre-Pleistocene strata where the extent of preservation and degree of exposure is much more fragmentary.

18.3 Continental record of Northern Hemisphere glaciation after 5 Ma Direct geological evidence of Northern Hemispheric climate cooling and of glacier expansion after 5 Ma is forthcoming from

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

many areas. Marine strata of the Nuwok Member of the Sagavanirktok Formation exposed on the southern margin of the Beaufort Sea in Alaska contain cold water ostracodes, foraminifera and molluscs that straddle the Miocene/Pliocene boundary. Correlative cold water arctic faunas are also represented in the upper part of the deltaic Beaufort Formation which extends over the Western Arctic coastal plain of Canada (Marincovich, 1990). Arctic permafrost was probably initiated at c. 2.5 Ma and was closely related to the development of perennial sea ice cover though the presence of a sea-ice cover a early as 5 Ma is likely (Clark, 1982 and refs. therein). Glaciation around the uplifted margin of the Gulf of Alaska had started by at least 5 Ma and is recorded by the Yakataga Formation which provides a thick (5 km) and high resolution record of glacially-influenced marine sedimentation (Section 26.1). Glaciation of the interior of Alaska started as early as 10 Ma (Denton and Armstrong, 1969; C.H. Eyles and N. Eyles, 1989). The initiation of Northern Hemisphere continental ice sheets is marked by an influx of ice-rafted debris into the North Atlantic at about 2.4 Ma (Raymo et al., 1989). Volume changes of these early ice sheets was driven by a dominant 41 Ka Milankovitch cycle of orbital obliquity; the 100 Ka cycle became dominant only after 1.6 Ma (Fig. 18.4). After this time the decay and growth phases of the Northern Hemisphere ice sheets (dominantly the Laurentide Ice Sheet in North America) became markedly asymmetric with deglaciation being much more rapid than ice sheet growth. The cause of slow growth but rapid disintegration ("terminations") is not precisely known (Raymo et al., 1989; Eglinton et al., 1992). Northern Hemispheric ice covers may have formed before 2.4 Ma but were not extensive and only locally reached tidewater (e.g. Jansen and Sjoholm, 1990). Anstey (1992) shows that ice rafting had become a significant process in Baffin Bay (ODP site 645) as early as 9 Ma and was the dominant deposi-

165

tional process after the latest Miocene (c. 6 Ma). The oldest ice rafted deposits from the Arctic Ocean are also Late Miocene in age (Thiede et al., 1990). Mudie and Helgason (1983) report the development of Miocene boreal climates in Iceland by about 10 Ma; glaciation commenced about 3 Ma but ice fronts did not reach sea-level until 2 Ma (Shackleton et al., 1984). This timing is of particular significance given that uplift of the Reykjanes Ridge occurred at around 2.5 Ma coincident with a change in spreading direction and rate along most of the North Atlantic spreading system (Vogt et al., 1989). Other regional tectonic controls on the initiation of Pleistocene glaciations of the Northern Hemisphere can be identified and are discussed at length below. 19. TECTONIC INFLUENCES ON LATE CENOZOIC G L A C I A T I O N

Was tectonic uplift a contributory factor in late Tertiary climatic cooling and the initiation of late Cenozoic glaciations? The answer according to some is a definite yes (Kutzbach et al., 1989; Ruddiman et al., 1989; Behrendt et al., 1991; Raymo and Ruddiman, 1992) and to others, only maybe (Barron, 1985; England and Molnar, 1990, 1991; Molnar, 1990; Burbank, 1992; Volk, 1993). The notion of topographic uplift as a precursor to Pleistocene and older glaciations has long been proposed. As early as 1840 Rendu had argued that Pleistocene glaciations were preceded by widespread continental elevation; Brooks (1926) reviewed a then already large literature on the subject and identified an uplifted mid-latitude belt in the Northern Hemisphere. The theme of Tertiary orogenesis and uplift of broad continental surfaces as a precondition to late Cenozoic glaciations was explored by Wegener (1929), Wundt (1944) and Flint (1957). Figure 13.1 shows the averaged global elevation of the present day snowline as a function of latitude; it can be appreciated from these data that rapid and maintained uplift in the

166 mid- and high-latitudes can result in intersection of the snowline and the generation of "adiabatic" glaciation. Emiliani and Geiss (1959) stressed that the major Pleistocene ice sheets, with the exception of Antarctica, grew around the periphery of the northern North Atlantic Ocean as a result of the presence of upstanding landmasses at high latitudes adjacent to a warm ocean current and moist air masses. The first quantitative work on the relationship between late Cenozoic tectonics and climate was that of Barron (1985) who reported results of general circulation modelling designed to examine the role of palaeogeography in Tertiary climate cooling. Positions of continents, averaged continental elevations and the relative area of land and sea with latitude were prepared for three time slices; at 60, 40 and 20 Ma (Fig. 19.1). Barron (1985) concluded that changing palaeogeography did not result in a systematic decrease in globally averaged surface temperatures; palaeogeography in fact accounted for only 0.25 of the recorded global p a l a e o t e m p e r a t u r e decrease during that time. Sensitivity tests, designed to identify the relative importance of individual variables, found that even elevation of the Tibetan Plateau, perhaps the most marked topographic change of the late Tertiary (Figs. 19.1, 19.2) had relatively little effect. This finding is in direct contrast to results of more recent modelling. R u d d i m a n et al. (1989) and Molnar (1990) summarized evidence showing an increase in late Cenozoic uplift rates and absolute elevation of the Tibetan Plateau, Himalayas and the western Cordillera of North and South America (Fig. 19.3). Areas of high topography anchor the long term position of upper tropospheric planetary standing waves which in turn affects the climate experienced on the ground in mid- and high-latitudes. Computer modelling supports the supposed relationship between climate and emerging mountain and plateau areas after 40 Ma. A "no-mountain" model successfully modelled

N.EYLES 9

MIOCENE:

20

0

~

°

EOCENE: 40

900

Ma

Ma

KEYLAT&TUDE FOR MILANKOVITCH FORCINGOf CLIMATE

ANTARCTC I C I ESHEET

0o 30 c

90

90 c

PAl_EOCENE: 60 Ma

30 ~

oo 30 o 60 o

AVERAGEELEVATION

I~-]< 700 rn

[]

700-1500 rn

[]>

1500m

Fig. 19.1. Changes in Tertiary palaeogeography and increased elevation of continental surfaces over the past 60 Ma. After Barron (1985). Note increase in average elevation on margins of North Atlantic at key latitude for Milankovitchforcing of global climate and ice sheet growth (Fig. 19.2).

the pre-40 Ma situation, a "half-mountain" model the situation obtaining during the Late Miocene or early Pleistocene (10-3 Ma) and a "mountain" experiment which most resembled that of the present day. Ice sheets were inserted into the last two models. The models show that lowered atmospheric heating rates, and in turn global temperatures, result in longer winters and cooler, moister summers. A greater snow cover reduces albedos thereby creating favourable conditions for ice sheet growth. A further corollary is that increased weathering of mineral suites released from uplifted areas would take up more carbon-dioxide and provide an antigreenhouse effect (see Sections 8.2, 21). Thus,

EARTH'S

GLACIAL

RECORD

AND ITS TECTONIC

167

SE'VFING

A

Kuhle (1987) argued that a large ice sheet (about 2 × 106 knl2), comparable in size to the present Greenland Ice Sheet, developed over the Tibetan Plateau. Whereas geological evidence for a large Tibetan ice sheet is arguable (H.M. French, pers. commun., 1992; Gupta and Sharma, 1992) the extent of ice across the Tibetan Plateau is a key variable in estimations of global palaeoclimate. Because of its large extent, high elevation (mean: 5 km) and the extreme transparency of the atmosphere the plateau plays a major global role in heating the Earth's atmosphere by re-radiating solar energy. Kuhle (1987) employed the term "heating panel" to describe the role of the plateau. A reduced albedo caused by any significant ice cover on the plateau could result in substantial global cooling of the atmosphere thereby amplifying any tendency toward glaciation in more northerly latitudes. Ruddiman and Kutzbach (1989) placed great emphasis on the global climatic effects of uplifted plateaux such as Tibet, rather than high mountains. While other factors are

II,.CREASE z

2

[ ] OECRE,SE

.<

~

B

~

30 °

60 °

90°S

30 °

60 ~

S

~oo 80 a Z

5

60

40 20

~q

~ 60 °

30 °



/ KEYLATITUDEFORMILANKOVITCHFORCINGOF CLIMATE

Fig. 19.2. Changes in global palaeogeography over the past 60 Ma. Note increased elevation and land area at 60°N. After Barron (1985).

accelerated late Cenozoic uplift results in lowered temperatures and an increased predisposition to Milankovitch-forced glacial cycles (Section 20).

5 Nanga Parbat

4

Late PleiStocene MiddlePleistocene

Eany Plemocem

E 2 ~>

8

v

3000 .~ 2000

00

6 4 2 Age (Ma)

4000 ~"

Late Plk~cene

1

10

5000

Present

Tibet - Himalaya

Eoc*r=

60

50

1000

40

30 20 Age (Ma)

10

1.0

E

0.8 Cordillera Orientale (Andes)

J

0.6

100

I 80

I 60

I

40 Age (Ma)

1000 .~

Sierra Nevada

>

E 0.4 ~"

2000 ~

0.2

o

I

v c

3000 U

I

20

0

50

40

30

20

10

0

Age (Ma)

Fig. 19.3. The record of accelerated Late Tertiary and Late Cenozoic uplift from the Himalayas, Andes and Sierra Nevada (after Molnar, 1990)

168

implicated (e.g. changes in atmospheric CO 2) results of sensitivity experiments on the climatic effects of Tibetan uplift closely match late Cenozoic cooling trends. The important exception is that of the high northern latitudes where an additional mechanism for late Tertiary climate cooling is identified. In that Pleistocene ice sheets were essentially a mid-latitude phenomenon, this is not a major weakness of the tectonoclimatic argument. Raymo and Ruddiman (1992) propose that uplift of the Tibetan Plateau since 40 Ma resulted in deflection of the atmospheric jet stream, a greatly intensified monsoonal circulaytion system, more intense chemical weathering and ultimately lowered atmospheric CO 2 concentrations. They note for example, that Himalayan catchments provide nearly 25% of the total dissolved load reaching the oceans but cover only 5% of the Earth's surface. Volk (1993) has pointed out that an additional source of atmospheric CO 2 is required to counteract a catastrophic drop in atmosphere CO 2 and runaway "ice house" conditions; this could be provided by varying the fraction of carbon buried either as organic material or carbonate (Raymo and Ruddiman, 1993) or by a reduction in weathering resulting from decreased global temperatures. In contrast to the findings of Ruddiman and Kutzbach (1989), Molnar and England (1990) argue that the evidence of accelerated late Cenozoic uplift as an important precursor to glaciation is exaggerated. Their thesis is that a large part of the evidence used to argue for late Cenozoic uplift, mostly derived from paleobotanical and geomorphological studies, can equally result from global climate changes independent of tectonic uplift. These workers argue that such climatic change results in increased mechanical weathering of mountain ranges, thereby giving the appearance of recent uplift. Paleobotanical studies that argue uplift of warm fossil floras into cooler climate zones may be more simply interpreted as the result of climate change rather than uplift. These con-

N. EYLES

cerns are valid but there is indisputable geological evidence of accelerated late Cenozoic uplift from the Himalayas, North America and Antarctica (e.g. Plafker, 1987; Molnar, 1990; Behrendt and Cooper 1991; Fitzgerald, 1992; Stump and Fitzgerald, 1992; Fig. 19.3). However, this uplift may well be due to global climate cooling, increased precipitation and enhanced late Cenozoic erosion resulting in isostatic recovery and uplift of summit areas coincident with an overall reduction in regional elevations (Raymo and Ruddiman, 1992). Burbank (1992) showed differences in the response of Miocene and modern day drainage systems in the Indo-Gangetic foreland basin along the margin of the Himalayas. Pleistocene and present day conditions are characterised by steep mountain fronts, by transverse rivers flowing directly away from the mountains and relatively thin foreland basin fills in response to overall basement uplift. This uplift is argued to be the result of erosion following late Cenozoic climate change. In the case where uplift is generated tectonically by thrusting and crustal thickening, longitudinal rivers drain parallel to the tectonic "grain" of the mountain front and are associated with thick foreland basin fills recording rapid subsidence and loading of an elastic crust along the edge of thrust sheets. These contrasts may not be universal and it still needs to be demonstrated that recent uplift is driven by late Tertiary climate cooling and not thermal uplift or some other, as yet porly understood, means of crustal thickening and tectonic uplift. Nonetheless, the analysis of Burbank (1992), when combined with investigation of fluvial facies, geometries and palaeodrainage patterns in sedimentary basins (e.g. Miall, 1983a, b, 1990) offers exciting possibilities for reconstruction of the relationship between palaeoclimate and tectonics in ancient glacially-influenced foreland basins. Raymo and Ruddiman (1992) conclude that uplift of the plateau was the first-order control on Cenozoic climate cooling. Additional evidence of the close interrelationship

169

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E q T I N G

between tectonics and late Tertiary glaciation is expanded in the following sections first, with regard to Antarctica and then with regard to the glaciated margins of the North Atlantic Ocean.

figuration of Antarctica and Australia. Australia was moving rapidly away from Antarctica allowing the establishment of a strong Circum-Antarctic ocean current. This circulation system led to the progressive thermal isolation of the continent because it decoupled warmer subtropical gyres from cold Subantartic and Antarctic gyres, thereby decreasing meridional heat flows from the equator to the poles (Kennett and Shackleton, 1976; Kennett, 1982). The first significant sea-ice formed around the Antarctic coastline at 40 Ma (the so-called terminal Eocene event) and records the opening of the Drake Passage and the Tasman Seaway and the establishment of a permanent circum-Antarctic flow of cool water. The earliest input of glacial sediment to the Ross Sea margin of Antarctica occurred at about 36 Ma when temperate wet-based glaciers reached the Victoria Land rift basin adjacent to the Cenozoic west Antarctic rift

19.1 Tectonics and late Tertiary glaciation in Antarctica A tectonic control on the initiation and subsequent growth history of the Antarctic Ice Sheet after 36 Ma can be suggested. Behrendt and Cooper (1991) identified the critical importance of rifting along the Transantarctic Rift and associated uplift of the rift shoulder (Transantarctic Mountains). The rift shoulder has been uplifted nearly 6 km over the past 60 Ma with episodic uplift rates as high as ~ 1 k m / M a (Behrendt and Cooper, 1991; Fitzgerald, 1992; Section 11.2). The climatic effects of Cenozoic uplift along the Transantarctic Mountains were amplified by the changing plate tectonic con-

\

I

/

o 0 0

SANTARCTIC /

I , ¢ ~ "~

~-.h~.~_ %%,-,.

)

90ow

\

/

~

-"~-)

o

c>o_

/

,o,,t,~,,

\ o/

'
,."3~ ~

"%..~ a-"r'~

\o

/

__)\ .

A Location of Pliocene Sirius Formation ~ r---]Area higher than l O00m ab. . . . . . .

(~.-, \'--~

t

[~)Area lower than lO00m below current t---Jsea level /

:~

~' /

(/ ~ 180°

/

"d ~

\

(

/j/

~

~,

L=~ ~

,

,o~

Fig. 19.4. Pliocene palaeogeography of Antarctica prior to Late Pliocene "refrigeration" and ice sheet expansion accompanying accelerated uplift along the Transantarctic Mountains. A record of temperate glaciation coeval with the existence of major Transantarctic sea ways is provided by the Sirius Formation (see text). After Webb et al. (1984).

170

N. EYLES

shoulder (Transantarctic Mountains; Barrett et al., 1989; Zachos et al., 1992). A continental ice sheet was probably in existence at this time (though see Section 18.1.1) but a large continental ice mass did not become a permanent fixture of Antarctica until about 14 Ma (Kennett, 1972). New work suggests too, that the continent underwent one or more deglaciations since that time. Recent work has described the occurrence of marine diatoms of mid-Pliocene age from the Wilkes and Pensacola basins of the interior of East Antarctica implying large scale deglaciation (McKelvey et al., 1991; Scherer, 1991). Diatoms occur in tillites preserved on uplifted erosional terraces and are interbedded with fluvioglacial and glaciolacustrine deposits that are associated with in situ Nothofagus (south beech) fragments indicating temperatures some 15°-20°C warmer than at present (McKelvey et al., 1991). The age of this deglaciation has been established by K - A r dating of volcanic ash to be about 3 Ma (Barrett et al., 1992). McKelvey et al. (1991) estimate about 1300 m of uplift at the site since the Late Pliocene. At this time Antarctic ice sheet volumes were drastically reduced with deep continent-crossing seaSEDIMENTATION RATE ('000 yrs):

2-7 cm

ways connecting the Weddell and Ross Seas (Webb et al., 1984, 1986; Harwood, 1985; Fig. 19.4). Conditions at this time may have resembled those characteristic of southernmost Chile at the present day (Mercer, 1986). Current discussion centres on whether the East Antarctic ice sheet collapsed catastrophically or experienced slow decay (see Sugden, 1992). The presence of in situ southern beech fossils now preserved 'at relatively high altitudes is additional evidence for rapid uplift along the Transantarctic Mountain front which is thought to be the principal reason for the present day "polar" thermal regime of the Antarctic ice sheet (Behrendt and Cooper, 1991). The present day "polar" characteristics of the Antarctic Ice Sheet (i.e. dry-based, little meltwater production) were only established during the Late Pliocene at about 2.5 Ma (Webb et al., 1984). The transition from temperate to polar thermal conditions may have occurred as a result of rapid localised uplift and is coincident with the initiation of major Northern Hemispheric ice sheets. Under present day polar conditions glaciclastic sediment is only delivered to the Antarctic continental margin at times of ice
SEI FL(

Fig. 19.5. Sedimentation along a high-relief, clastic-starved glaciated continental margin e.g. present day Antarctica. Starvation of the Antarctic margin, despite a heavily glacierized continent, is the result of tectonic uplift along the Transantarctic Mountains and the development of cold "polar" climatic regime at about 2.5 Ma.

171

E A R T H ' S GL AC IAL R E C O R D AND ITS T E C T O N I C S E T T I N G

sheet expansion when the ice sheet moves out onto the shelf; interglacial marine conditions are characterised by sediment starvation and extremely low rates of terrigenous sedimentation (see Elverhoi, 1984; Anderson and Molnia, 1989; Anderson et al., 1991; Fig. 19.5). The considerable topography and depth of the Antarctic continental margin reflects a long history of glacial erosion, a lack of sediment supply and the failure to deposit the thick sediment bodies typical of more temperate glaciated margins. In turn, the very considerable topography of the inner Antarctic margin and the great depth of inner shelf basins promotes starvation of the outer margin by trapping any glaciclastic sediment that is released from ice shelves and their component ice streams (Anderson et al., 1983; Fig. 19.5). Modern shelf sedimentation is dominated by deposition of siliceous oozes and bioclastic debris (Domack, 1988). Behrendt and Cooper (1991) speculated that the relationship, between Cenozoic tectonism and Antarctic glaciation could have been initiated as early as the Early Oligocene (60 Ma). It should be noted in this regard however, that the West Antarctic rift system was initiated much earlier, during Jurassic rifting of Africa from Antarctica, suggesting that possible Cretaceous ice sheet(s) could also have been generated as a result of adiabatic uplift (Section 17.1)

19.2 Tectonics and late Cenozoic glaciation around the margins of the North Atlantic Ocean Pleistocene glaciations in the Northern Hemisphere are dominated by the growth of the Laurentide Ice Sheet over Canada and that of the European Ice Sheet over Eurasia. The critical position of these two ice masses either side of the northern North Atlantic has already been commented on (Section 19). The Laurentide ice sheet grew by the thickening and amalgamation of perennial snow fields on the Labrador Plateau and over Keewatin ("instant glacierization"). The

0 kms00 LAURENTIDE

~7~

1 ~ SHEET CENTRES AND FLOW LINES

--

POSITIVE, UPLIFTED - - - - BASEMENT TREND

Fig. 19.6. Extent of Laurentide Ice Sheet, location of

ice centres (after Shilts, 1986) and structural arches of the Canadian Shield (after Sanford, 1987). Note coincidence of ice centres and uplifted basement.

ice sheet has traditionally been regarded as single-domed but recent reconstructions identify two main ice centres (Andrews, 1987). It is interesting to note that these two centres exactly coincide with areas of uplifted Archean crust and structural arches across the Canadian Shield (Fig. 19.6). Hoffman (1989) suggested that the restriction of the raised shield area to Archean crust implies that such crust is more buoyant than surrounding Proterozoic crust. It must be admitted however, that any link between doming of the shield and late Cenozoic glaciation is tenuous at best because the origin and timing of uplift is not well constrained. Lithospheric warping as a result of large-scale lithospheric compression may be implicated (Bally, 1989) but data are few. The raised form of the Canadian Shield appears to be the result of long-term uplift and peneplanation during the Phanerozoic (e.g. Ambrose, 1964; Sanford et al., 1985; Sanford, 1987) and is not restricted to the late Cenozoic.

172

N. EYLES

19.2.1 The importance of passive margin uplift

19.7, 19.8). Magmatic underplating around the margins of the North Atlantic can be inferred from the widespread presence of gravity highs along the modern coasts (e.g. Planke et al., 1991). The central importance of passive margin uplift on either side of the North Atlantic is that the locus of Tertiary uplift was centered between latitudes 40 and 60°N (Figs. 19.1, 19.2). This is precisely the latitudinal belt in which the seasonal changes in solar insolation caused by systematic "Milankovitch" variations of the Earth's orbit are greatest (Section 20). Latitude 60° to 65°N is a key latitudinal belt in "Milankovitch" astronomical forcing of glaciations for it is in this belt that variation in summer insolation over a glacial/interglacial cycle is as much as ~ 60 W m -2. The tectonic stability of this zone, and the tendency for uplift or subsidence (e.g. Fig. 13.1), is particulary important for climate change because this belt contains the greatest ratio of land area to sea for any latitudinal belt on Earth. The growth of extensive perennial snowfields on broad, uplifted passive margin plateaux, together with the associated albedo effects, may have been an important "amplifier" of Milankovitchdriven global insolation changes. It is in the latitudinal belt between 65°-80°N that the

A much more significant late Cenozoic tectonic influence has been that of passive margin uplift on the landward flanks of the North Atlantic Ocean. The principal Pleistocene ice centres in Canada, northwest Britain and Scandinavia lie either side of the ocean on extensional plate margins. These margins are fault-bounded and locally, have experienced strong late Cenozoic uplift. Uplift patterns can be constrained by reference to the work of Lister et al. (1986) who recognize two fundamental styles of faulting along passive margins (Fig. 11.6). These two structural styles are controlled by the geometry of the master detachment fault along the continental margin. The upper plate type is of especial interest because as a result of the magmatic inflation and "underplating" the plate margin undergoes uplift resulting in the formation of passive margin mountains and plateaux. The base of the continental margin is warmed by relatively less dense asthenosphere and undergoes uplift inland of a sharply-faulted coastal margin. This model describes the late Cenozoic evolution of the typical North Atlantic coastal margin with interior plateaux and highlands bordered by an abrupt, fault-bounded coastline (Figs.

ii RT"ATLANT'°CEAN C / ~

_

PENEPLAINEDSHELF

~

(

S

T

FIG.23.1 R

A

UPLIFTEDCRATONRIMAND GLACIALLY-DISSECTED AND TILTEDTERTIARYPLANATIONSURFACES

/ N

D

F

L

A

+ ..'..'.:.;-.'-7.:.:..'.:.:.:.:.;.:." OCEANIC CRUST

+ +

++

T

) +

+ +t

+



_p -P

+

.p "P

CONTINENTAL CRUST

Fig. 19.7. Late Tertiary uplift of passive margin plateaux, as a precondition for "Milankovitch-forced" Pleistocene glaciations around the periphery of the North Atlantic Ocean (see also Fig. 20.1).

173

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

first Pleistocene Northern Hemisphere ice sheets grew and it is there that glaciation is re-initiated at the beginning of successive Pleistocene glacial cycles (see Miller and Vernal, 1992). The first critical event is the opening of the North Atlantic Ocean between Greenland and Europe. Rifting was preceded by regional updoming attributed to initiation of the Iceland Mantle plume and upwelling of anomalously hot asthenospheric mantle (White, 1989). The plume created a regional topographic swell, up to 2 km high and more than 2000 km in diameter, that extended from west Greenland to north Norway. UpI

~s7

[]

doming provided gravitational assistance in rifting which was accompanied by a massive outburst of igneous activity at around 60 Ma. Seaward dipping reflectors along the rifted margin from Greenland to Norway, testify to regional uplift and "subaerial sea-floor spreading" (Mutter et al., 1982). The subsequent evolution of the Northwest European continental margin as a large-scale "detachment system" has been reviewed by Gibbs (1989), Tankard and Balkwill (1989; Fig. 19.9). The climatic importance of "upper plate" uplift and the formation of passive margin mountains around the North Atlantic Ocean has already been

SPOT ELEVATION IN METRES ABOVE SEA-LEVEL

/: BATH*METR* ELEVAT,ONSOVER500m

0 ,~,00 300

FAULT ................... ~:'~:~ OCEANIC CRUST

To=r~'r ='T~ OR SHELF

CLIFFED, HIGH RELIEF COAST

80'

70'

60'

50'

90"

80"

70"

60 °

Fig. 19.8. The formation of passive margin plateaux and narrow, deep shelv.esalong the uplifted, rifted plate margins of the Western Atlantic Ocean. Coastline data from Amos (1990).

174

N. EYLES

stressed (e.g. Fig. 19.7). The rifting process created upwarped margins of rifted cratons which subsequently formed the sites for ice sheet growth in Baffin Island, Keewatin, Labrador, Greenland, Northwest Britain and Scandinavia. Late Tertiary uplift is widely recorded around the North Atlantic by marine-cut planation surfaces and broad plateau-like mountains. Geological and geomorphological evidence of such uplift is reviewed below both for the western and eastern North Atlantic regions. 19.3 Western North Atlantic region

Balkwill (1987) identified a late Tertiary (Mid-Miocene) phase of "robust" uplift around the coastal margins of the Labrador Basin between 52° and 66°N. This basin forms the largest topographic element in the western Atlantic comprising a large horseshoe about 1300 km long and 900 km wide. The basin is surrounded by a structurally high cratonic rim expressed as the coastal mountains of Labrador, southern Baffin Island and West Greenland (Fig. 19.8). This heavily glaciated rim is tilted landward and is broken, most noticeably in Cumberland Sound and Frobisher Bay, by transverse MesozoicCenozoic rift basins. Mid-Miocene uplift of

1500 m along this extensive latitudinal belt can be related to renewed extension and subsidence of the Labrador Basin in response to changes along the Iceland-Jan Mayan sector of the mid-Atlantic ridge. On southern Baffin Island, syn-rift marine sediments of Cretaceous to Palaeogene age have been uplifted more than 600 m (Haworth, 1982); total Tertiary uplift may be as much as 2000 m. It has been suggested that post-Eocene uplift around Baffin Bay may be an isostatic response to crustal thickening during the Eurekan Orogen (latest Cretaceous-Palaeogene). Trettin (1989) suggested several cycles of uplift, of the order of several kilometres, during the middle and late Tertiary; rapid uplift of rift shoulders along faulted coastlines is coincident with renewed offshore subsidence. Trettin (1989) highlighted the presence of a Tertiary drainage system in the Arctic that was subsequently incised into fiords. Elsewhere in the western North Altantic region, accelerated uplift of the Torngat Mountains of northern Labrador occurred during the Late Oligocene to the Pliocene. This event is recorded by an extensive clastic wedge (Saglek Formation) that forms the present day outer shelf and slope of Labrador. The Saglek is up to 2000 m thick and consists of poorly-sorted, corn-

LOW[ MAR( (FIG.

CENTRAL GRABEN OF NORTH SEA UPUFTED, GLACIALLY-DISSECTED PASSIVEMARGINPLATEAUX ~ ' ~ ANDPRINaPALCENTRESOF PLEISTOCENEICESHEETS (FIG. 19.7)

~

THICK,PLEISTOCENEGLACIAL SEDIMENTS(FIG. 24.1)

Fig. 19.9. Northwest European continental margin detachment system (after Gibbs, 1989). Note Pleistocene ice centres on uplifted upper plate margins in Scotland, Wales and Norway.

EARTH'S GLACIAL RECORD AND ITS TECTONIC SETI'ING

positionally immature sandstones deposited on prograding fan-deltas (Balkwill et al., 1990). 19.4 Eastern North Atlantic region Vagnes and Amundsen (1993) identified widespread Cenozoic epeirogenic uplift around the continental margins of the eastern North Atlantic and related this to a hot asthenospheric lens associated with the Iceland mantle plume. Evidence for significant late Tertiary uplift around the eastern margins of the North Atlantic is forthcoming from the widespread existence of uplifted and tilted peneplaned erosion surfaces such as the "Paleic" plateau surface of Scandinavia (Gjessing, 1967) and the high level erosion surfaces of western and northwestern Britain (Brown, 1960). These surfaces have been heavily dissected by Pleistocene glacial erosion. Peach and Horne (1930) were the first to recognize that the Scottish Highlands showed concordant summit elevations. These define Tertiary marine planation surfaces subsequently uplifted and dissected by late Cenozoic erosion. One major surface, dipping from 900 m in the west to 650 m in northeast Scotland, post-dates early Tertiary volcanics (George, 1966). No sediment cover survives on the marine cut surface which forms the "bed" of successive British Ice Sheet centres (e.g. Fig. 19.9). The other centre of the British Ice Sheet, in Wales, is also restricted to an uplifted, marine planation surface between 5 to 600 m in elevation. The age of this surface is known to post-date Eocene dyke intrusions dated at c. 50 Ma (Brown, 1960). Successively younger, marine-cut platforms extend to sea-level recording continued intermittent uplift. Upland surfaces also occur in Ireland but are poorly understood (Embleton, 1984). Intermittent uplift of the Fennoscandian Shield is recorded by mountains and broad massifs (fjells in Norway, fjalls in Sweden) with accordant summits (Gjessing, 1967); geologic and geomorphic evidence of rapid up-

175

lift of at least 600 m was reviewed by Morner (1980) and can be related to opening of the Norwegian Sea and passive margin uplift. Late Cenozoic uplift of the Scandinavian land mass is cearly seen on offshore seismic profiles (Vorren et al., 1986). Uplift is related to opening of the Norwegian-Greenland Sea and the progressive construction of a marine sedimentary wedge onto oceanic crust (e.g. Wood et al., 1989) very similar to that of the Labrador margin. Late Tertiary uplift has had a major impact on offshore oil and gas exploration in the Barents Sea area of Norway because reservoir seals have been thinned or removed allowing hydrocarbons to be expelled; Saettern (1992) reports erosion of as much as 2 km of strata since the early Tertiary. Many wells show that favourable subsurface structures were previously filled with hydrocarbons which have been lost in response to erosion of capping strata. In summary, whereas none of the above data can be said to be definitive there is a strong suggestion that Cenozoic uplift and the formation of high-standing continental margins was a critical precondition to the initiation of Pleistocene ice sheets. In combination with the strong uplift of the Tibetan Plateau (e.g. Ruddiman and Kutzbach, 1989; Raymo and Ruddiman, 1992), passive margin uplift may have set the scene for instant glacierization by producing large areas of uplifted plateaux on which perennial snow fields could grow (Fig. 13.1). Subsequently, when mid-latitude temperatures began to fall below a threshold value "Milankovich" climate forcing (Section 20.1) dictated the growth and decay of large ice sheets (see Section 20.3). The most prominent geomorphological feature of the North Atlantic region is the presence of uplifted plateau surfaces of late Tertiary age, dissected by glacially-overdeepened troughs cut along structural lineaments, with their seaward margins characterised by cliffed, high relief coastlines (Fig. 19.7). This very distinctive topography

176

records the interplay between the divergent plate margin setting, late Tertiary uplift, climatic thresholds and the initiation of Milankovitch-controlled cycles of glaciation on upwarped passive margin plateaux. 20. PLEISTOCENE GLACIATIONS ( < 2.5 MA) AND THE ROLE OF MILANKOVITCH FORCING

"Not marble nor the gilded monuments Of princes shall outlive this powerful rhyme." Shakespeare, Sonnets. The terrestrial record of Pleistocene glaciations after 2.5 Ma is complex, incomplete and scattered piecemeal over a large area. Age dating and correlation is uncertain (Sibrava et al., 1986). In contrast, a high-resolution climate signal is retained in deep marine sediments where sedimentation has largely been continuous, where micro-fossils allow reconstruction of water mass characteristics and continental ice volumes can be inferred from study of stable isotopes contained in their skeletons. Reconstructions of the past 1.5 Ma reveal a clear "astronomical" control on the timing and frequency of glacials and interglacials. Alternating glaciations and interglacials are most easily explained by cyclic perturbations of the Earth's planetary tilt and orbit. These "Milankovitch" variables, so named after the Serbian mathematician, are responsible for fluctuations in the seasonal distribution of incoming solar radiation and have been identified as the principal control on the growth and decay of Pleistocene ice sheets (Imbrie and Imbrie, 1979; Ruddiman et al., 1986; Martinson et al., 1987; Raymo et al., 1989).

20.1 Milankovitch orbital rhythms Three principal orbital rhythms (periods given in brackets) can be identified (Berger and Loutre, 1992). For the past 3 Ma these are (1) change in the eccentricity of the Earth's orbit around the Sun (404 Ka, 124

N. EYLES

Ka and 95 Ka; mean; 91 Ka), (2) change in orbital obliquity (the tilt of the Earth's axis with respect to the plane in which it orbits the Sun; 41 Ka) and (3) a wobble (precession) due to the tilt axis sweeping out a cone (23 Ka and 19 Ka; mean 21 Ka). These orbital rhythms act in combination and produce cyclical variations in the intensity and seasonal distribution of incoming solar radiation. In combination, the Milankovitch variables control the length of the summer melt period, such that at times, the winter snow does not melt completely. As snow fields grow in size more incoming solar radiation is reflected and ice sheets of continental dimensions can develop. The prolongation of cool summer seasons is most marked in the Northern Hemisphere at latitudes of between 60 ° and 65°N (Fig. 20.1). This zone coincides with the greatest percentage of land area relative to that for the ocean, for any given latitude (Figs. 19.1, 19.2). An astronomical control on the timing of Pleistocene glaciations has been demonstrated by reference to changes in 180//160 ratios from benthic and planktic foraminifera preserved in deep ocean sediments (Chappell and Shackleton, 1986; Mathews, 1986). During the growth of continental ice-sheets, seawater becomes progressively enriched in the heavier 180 and the ice sheet in the lighter 160. This is because the heavier isotope is less mobile such that when water evaporates, 160 is concentrated in the vapour released when water evaporates. When this water vapour subsequently condenses into rain, the vapour is still further depleted in 180. In this fashion the snow which eventually falls on the elevated surface of a large ice sheet has been relatively enriched in 160, leaving ocean seawater (and any organisms therein) relatively enriched in 180. In contrast, the melting of an ice sheet at the end of any one glaciation is recorded in deep sea sediment cores by abrupt 160 spikes, produced when large volumes of isotopicallylight glacial meltwater return to the oceans. Relative depletion in 180 is expressed as

177

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETFING

ECCENTRICITY ('stretch': 95 K)

PRECESSION ('wobble': 21 Ka)

m

OBLIQUITY ('tilt': 40 Ka)

CRATON RIMS EXPERIENCINGLATE TERTIARY UPLIFT AS PASSIVE MARGIN PLATEAUX (FIG. 19,7)

0 TM

Fig. 20.1. Tectonics and Milankovitch orbital rhythms. Periodicities of rhythms shown in brackets. Note occurrence of Late Tertiary uplift around margins of North Atlantic Ocean north of latitude 60°N. Milankovitch forcing is most effective at this latitude. See also Figs. 19.1 and 19.2.

deviation (d) in parts per thousand (%o) from a standard interglacial mean (SMOW = Standard Mean Ocean Water). The most depleted (i.e. isotopically lightest) ice in the Antarctic Ice Sheet is about -60%o (Chappell and Shackleton, 1986). During the last glaciation, the maximum depletion in the North American Laurentide Ice Sheet was about -30%o (Hillaire-Marcel and Causse, 1989; Schwarcz and Eyles, 1991). Numerous studies confirm that the average amplitude of the d180 signal from glacial to interglacial conditions, as recorded in the deep ocean is about 1.6%o over the last 700 Ka. Over the last 700 Ka, the length of a

glacial cycle as recorded by changes in oxygen isotope ratios, has been about 100 Ka (Fig. 18.4). Between the inception of large continental ice sheets at about 2.4 Ma and as recently as 700 Ka, ice ages were driven by higher-frequency cycles (about 40 Ka). The timing of the cycles fits very well with the Milankovitch periodicities (Ruddiman et al., 1986; Raymo et al., 1989). However, despite a good correlation of 100 Ka long glacial/interglacial cycles with variation in the eccentricity of the Earth's orbit, the changes in incoming radiation that can be attributed to the eccentricity cycle are too small to account for the observed climatic effects. Any climatic effects resulting from the 100 Ka eccentricity signal ought to be minor but instead this cycle dominates the record. Using ideas of the instantaneous frequency of planetary motions, Liu (1992) discovered that the 41 Ka obliquity cycle has a 100 Ka frequency modulation superimposed which could give rise rise to the observed 100 Ka forcing of climate. Far from solving the problem of trying to explain late Cenozoic Ice Ages, the astronomical theory of climate change raises many fundamental questions. One problem relates to the stability of the rhythms and whether the same cycles operated at the same frequency prior to the late Cenozoic (Berger and Loutre, 1989). Another, and perhaps much more difficult problem, stems from the recent realisation that orbitally-driven changes in insolation are insufficiently strong to effect major global climate changes that are synchronous from one hemisphere to another. The fundamental problem here is that relatively weak insolation changes at 60-65°N latitude must be then converted into global climate change. The extremely rapid shift from glacial to interglacial conditions ("terminations"; Broecker and Van Donk, 1970; Fig. 20.2) also cannot be adequately explained by reference to the astronomical theory of climate change. These problems, and possible solutions, are briefly reviewed below.

178

N EYLES

20.2 Stability of Milankovitch rhythms during the Phanerozoic

relief the stability of Milankovitch rhythms during Archean and Proterozoic glaciations. Laskar et al. (1993) show that without the torque exerted on the Earth by the Moon, the Earth's obliquity would be subject to violent and chaotic change as is the case for other planets (Murray, 1993). They conclude that the Moon has acted as a climatic regulator for planet Earth and has prevented dramatic changes in global climate. Finally, it is worth briefly commenting on another problem regarding the effectiveness of Milankovitch rhythms in Earth history. The effects of Milankovitch forcing during the Pleistocene have been most clearly felt at 60°N latitude; it is not known whether this is a long term feature or is simply a result of the greatest extent of continental surfaces present in this latitudinal belt (e.g. Fig. 19.2). Given different palaeogeographic distributions of continents and oceans in the past (e.g. Fig. 21.2) the latitudinal effectiveness of Milankovich variation in solar irradiation may have been radically different and its effects correspondingly amplified or reduced.

The long term geological stability of the Milankovitch rhythms is open to question. Mathematical analysis of the motions of the inner planets of the Solar System (Laskar, 1989) suggests that the Milankovitch rhythms can be accurately calculated only for the last 100 Ma. Over longer periods of time, it is postulated that non-linear amplification of even very small perturbations in planetary orbits, as well as changes in the diameter of the Earth's core and of the planet as a whole, leads to chaotic, unpredictable planetary motions. This "chaotic" view of the Solar System is countered by calculations by Berger et al. (1989), who suggest that Milankovitch rhythms are relatively stable over geologic time. For example, in the Late Cretaceous it has been calculated that the tilt cycle would have been shortened by only about 4% and the precession cycle shortened by about 2%. The eccentricity cycle would be unlikely to change. Berger and Loutre (1989) suggest nonetheless, that by the Early Cambrian it may be impossible to distinguish either precession or tilt cycles. The work of Laskar (1989) and continuing lack of understanding of the Archean and Proterozoic structure of the Earth (Schubert, 1991) throws into sharp ppm 0

200

250

ppb 300

300

500

20.3 The atmosphere and oceans as amplifiers of Milankovitch rhythms Milankovitch cycles in solar irradiation are insufficient to result in the growth and decay

pg per g 700

150

100

8"0(7.)

*C

50

0

-6

-4

-2

0

2

-0.8-0.4

0

0.4 0.8

¢--.

so ..a

E <

100

150



CARBON DIOXIDE

METHANE

ATMOSPHERICD U S T

TEMPERATURE GLOBAL ICE VOLUME

PENULTIMATE GLACIATION

CLIMATE

Fig. 20.2. Global change over the past 150,000 years recorded by the Vostok ice-core from the Antarctic Ice Sheet (after De Angelis et al., 1987; Jouzel et al., 1987; Lorius et al., 1990; Rampino and Self, 1992). See also Fig. 18.4. Note the slow build up of ice volumes during the last glaciation and the rapid loss at the end of the glaciation ("terminations"). Change in ocean circulation and Milankovitch-driven changes in solar irradiation are the driving forces behind rapid global changes (see Broecker and Denton, 1989). See als Fig. 20.3.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

of continental ice sheets. It has been suggested that changes in atmospheric greenhouse gas concentrations may amplify the relatively weak Milankovitch signal (e.g. Broccoli and Manabe, 1987). A very close association between Pleistocene glacial/interglacial cycles and changing levels of atmospheric carbon dioxide and methane is indicated by the Vostok ice-core recovered from the Antarctic Ice Sheet (Lorius et al., 1987; Fig. 20.2). The 2 km long core spans the last 150,000 years and shows a clear coincidence of greenhouse gas levels and atmospheric temperature. The average CO 2 content during glacial episodes is about 200 p.p.m, compared with 300 p.p.m, during interglacials. There is a very good correlation (r 2= 0.79) between atmospheric CO 2 and temperature change, that for methane is 0.78 (Lorius et al., 1990). Even small perturbations in temperature, identifying warmer interstadials within the glacial period, are similarly correlated. The data from the Vostok ice core suggests an averaged global cooling of about 5°C during the last glaciation ( < 125 Ka) of which about 3 ° may be due to changing concentrations of greenhouse gases. The remaining temperature difference could be due to a much greater content of dust in the atmosphere resulting from strengthened circulation systems, continental shelves exposed by sea-level fall and retreat of vegetation covers from mid-latitudes (e.g. Harvey, 1988). Other factors are at work but the role of CO 2 and CH 4 is clearly indicated as an amplifier of otherwise relatively-weak Milankovich-driven changes in irradiation. The fundamental question to be resolved, however, is what ultimately controls the production of CO2 and methane over a glacial/interglacial cycle (e.g. Street-Perrot, 1992). Swamps and wetlands are a common product of glaciation and drainage disruption and enhanced methane production from these areas could accelerate the return to warm interglacial conditions (Adams et al., 1992). Caldeira and Rampino (1992) suggest that

179

glacial-interglacial fluctuations in CO2 are dictated by volcanic emissions that in turn, follow changes in glacioeustatic sea level. The latter, it is argued, affects the loading on the Earth's crust and magma chambers, thereby influencing the timing of volcanic eruptions. Rampino and Self (1992) point to the near-coincidence of the massive eruption of Toba in the Indian Ocean some 74 Ka years ago, the release of about 800 km 3 of ash and accelerated global cooling accompanying the last interglacial/glacial transition. Whether this suggests a general model applicable to other glaciations is open to debate but it could be that climatic changes, resulting ice sheet growth and sea level change, may modulate volcanic eruptions which in turn, modifies climate by way of the resulting "volcanic winter" (Ramaswamy, 1992; Rampino and Self, 1992; see also Section 10.4.3). The existence of a "biological pump" that controls glacial-interglacial fluctuations in atmospheric CO 2 levels has been suggested by Broecker and Peng (1989) and related to fluctuations in biological productivity caused by changing nutrient loading and alkalinity of near-surface ocean waters. In turn, the nutrient chemistry of the oceans is greatly influenced by the distribution of cold bottom waters. The role of such waters in amplifying glacial-interglacial changes in the global heat budget is reviewed by Broecker et al. (1990), Charles and Fairbanks (1992), Eglinton et al. (1992) and Wright (1993). What emerges from this work is that glacial/interglacial fluctuations in atmospheric carbon dioxide are probably driven by changes in ocean circulation and further, that such changes are extremely rapid. The role of ocean circulation in glacial/interglacial cycles is currently the subject of intense scrutiny and it is here that the most effective amplifier of Milankovitch signal is likely to be found. Broecker and Denton (1989) argued that the "missing link" between orbitally-driven insolation and ice sheet fluctuations is orbitally-induced

180

N. EYLES

changes in the amount of freshwater reaching the oceans and resulting changes in the salt structure of the oceans. These workers clearly show that the amount of snowline depression during the last glaciation was similar on all mountain ranges (about 1 km) and that deglaciation began abruptly and was coeval in both Northern and Southern Hemispheres. Such abrupt and synchronous changes in global climate cannot be the product of orbitally-driven changes in solar

irradiation which are effective only within a narrow latitudinal belt (Fig. 20.1). Broecker et al. (1990) introduced the "conveyor-belt/salt oscillator" hypothesis which stresses the climate forcing role of ocean circulation in particular that of North Atlantic deep water (NADW; Fig. 20.3). Warm, equatorial surface waters moving northward into the far North Atlantic Ocean give up heat to the atmosphere, and become part of deep water return flows (NADW) SUMMIT 81, 0(¢.)

,,.s]

.4s

-4o

.s,s

15I1"7 I ~ Ip~-°~E_'i.,L . .

•2.3

VOSTOK "0(~) 0 -0.8 0

J

0.8

50

ANTARCTIC ICE CORES 1 Vostok Z Dome

I00

150

Fig. 20.3. Glaciation and ocean circulation in an opening ocean; the North Atlantic Ocean. Ice sheets at the last glacial maximum (c. 20 Ka) are Antarctic (A), British (B), Barents (BA), Cordilleran (C), Greenland (G), Icelandic (I), Innutian (IN), Kara (/64) and Scandinavian (SC). Present day ocean circulation is shown for the North Atlantic Ocean; oxygen isotope data from ice cores at four sites through Greenland Ice Sheet identify very rapid changes in climate during the last glaciation caused by changes in ocean circulation (inset top right). Vostok record is shown in Fig. 20.2. After Broecker and Denton (1989), Lehman and Keigwin (1992) and Johnsen et al. (1992). NADW: North Atlantic deep water.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

emanating from the Labrador Sea and Norwegian Sea (Lehman and Keigwin, 1992; Fig. 20.3). These move southward down the length of the Atlantic Ocean, traverse across the southern Indian Ocean and thence northwards to the far northern Pacific Ocean. The conveyor belt is powered by the high density of cold, saline waters produced in the far north Atlantic as a result of sea-ice formation and the loss of water vapor from northward-moving equatorial waters. The deep-sea record shows that the conveyor belt system stopped during glacial times (Duplessy et al., 1988) resulting in severe cooling of land areas adjacent to the North Atlantic Ocean. As related above (Section 19.2) these areas consist of tectonically-upwarped passive margin plateaux and so are favoured locations for the formation of extensive snowfields and, ultimately, ice sheets. The rapidity of circulation changes in the North Atlantic Ocean is indicated by data obtained from ice cores through the Greenland Ice Sheet (Johnsen et al., 1992); these show coeval changes in climate around the North Atlantic region involving temperature fluctuations of 12-13°C beginning abruptly over several decades (Fig. 20.3). Identification of the precise sequence of events that may stop the North Atlantic conveyor belt is the subject of ongoing debate (e.g. Bond et al., 1992; Duplessy et al., 1992; Lehman and Keigwin, 1992; Slowey and Curry, 1992) but rapid "mode switches" of the ocean-atmosphere system, brought about by changes in thermohaline circulation, appear to provide the missing amplifier between Milankovitch astronomical climate forcing and ice sheets. The above discussions serve to emphasise the complexity and dynamic nature of causal factors acting to generate Quaternary glaciations (see also Corey, 1991; Jones and Mitchell, 1991; Chahine, 1992). Pre-Quaternary glaciations were no less simple (Socci, 1991). A good question is whether such circulation changes accompanied late Proterozoic glaciations at times of supercontinent breakup and the formation of expanding

181

ocean basins that closely resembled the present day North Atlantic (e.g. Figs. 9.2, 11.1). Unfortunately, unravelling of complex changes in ocean circulation are dependent on excellent palaeontological control. Finally, it is worth emphasizing perhaps, for geologists primarily concerned with stratigraphic studies of Earth's glacial record, that Milankovitch cycles cannot control either the activity of individual ice sheet margins nor deposition of stratigraphic successions at their margins. There is in effect, a "black box" between Milankovitch-forced global changes, the response of ice masses and the resulting stratigraphic record. Ice margin fluctuations are dynamic, strongly diachronous from one sector to another and controlled by many complexly-interrelated factors including local temperatures, precipitation, subglacial bed materials and topography (e.g. Boulton and Clark, 1990; MacAyeal, 1992). These factors create considerable "noise" in the glacial depositional record; the strongest Milankovitch signal is recorded in sediments in the deep-sea and along nonglaciated coastlines (e.g. Anderson, 1984; Ito and Katsura, 1992; Klein and Kupperman, 1992; Sections 16.13, 16.14). 21. THE ROLE OF ATMOSPHERIC CARBON DIOXIDE IN LATE CENOZOIC AND OLDER GLACIATIONS The pivotal role of atmospheric carbon dioxide in climate change is now widely recognised in relation to anthropogenic consumption of fossil fuels (Grubb et al., 1991 and refs. therein). The effects of changing CO 2 partial pressures have also been implicated in dictating the incidence of glaciation throughout Earth history; a suggestion first made by Arrhenius in 1896 and by Chamberlin (1899). Fleming (1992) provides a succint review of these early ideas. Fischer (1984) argued that the Earth's climates have oscillated between "greenhouse states", marked by elevated carbon dioxide values (Fig. 16.21), sluggish ocean circulation and moderately

N EYLES

182

high

16.21). In addition, Venus-like greenhouse conditions have been invoked to account for both the "missing" Archean glaciations and the long mid-Proterozoic non-glacial interval (Section 7.3). Possibly widespread Early Proterozoic glaciation has also been explained as a result of lowered atmospheric CO 2 levels (Section 8.2). The association of Proterozoic, Phanero-

average

global temperatures, and "icehouse states", characterized by continental ice sheets, vigorous ocean circulation and the widespread development of freezing temperatures at sea-level. According to Fischer, greenhouse conditions followed a Late Proterozoic "icehouse" state and lasted to the end of the Devonian with another warm interval from the Jurassic to the Eocene (Fig.

GLACIATIONS

A B

IIU I ICEHOUSE I

GREENHOUSE [

ICEHOUSE

I

GREENHOUSE I ICE ]

CLIMATESTATES ICOOLI WARM ICOOLI

I COOL [WARM I COOL IWARMICOOLI

WARM

CLIMATEMODES 0.71 O0

2O

18 ._J .< >

0.7095

~, 16

r~

C

~

--

SrSrla~Sr RATIOIN MARINEWATERS

--

coz CONTENTOF ATMOSPHERE

~ /~ I

I YI'--,

S_~-

ERROR

0.7090

RANGE

/

0.708S

12

0.7080

10

Z 0

8

0 u

6

0

4--

%

/k

~"

El 0 0 EE-Z,

\\~

o.7o7s

tt

tl

',

-

0.7070 0.7065

<

0---~ 600

Ii

I ~

500

0.7060

ISh D

C i-l" I i[ J 400 300 200 MILLIONSOF YEARSBEFOREPRESENT

IKi 100

I T

TYPE OF CARBONATECEMENT

D

t 'Aragonite Threshold' []High ~]

600

Mg Calcite and, less abundantly, Aragontte

Calcite; Mg content generally lower, increasing toward 'Threshold'

500

400 300 200 MILLIONSOF YEARSBEFOREPRESENT

100

0

Fig. 21.1. A. Glacial episodes during the Phanerozoic. B. Climate "states" (after Fischer, 1984) and "modes" (Alter Frakes et al., 1992). C. Atmospheric C O 2 levels (after Berner, 1990) and 87 Sr/ 86 Sr ratio in seawater (after Koepnick et al., 1988). D. Variations of carbonate cement (after Sandberg, 1983). E. Sea-level history (Vail et al., 1977; see Fig. 6.2)

183

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'VI'ING

zoic and Cenozoic glaciations with low levels of atmospheric CO 2 (Figs. 6.2, 21.1) is now widely used as evidence that "ice house" conditions, characterised by low partial pressures, alternate with "greenhouse" conditions (Roberts, 1976; Fischer, 1984; Veevers, 1990; Raymo, 1991). The short-term role of CO 2 in amplifying relatively weak "Milankovitch" forcing of Pleistocene glaciations has also been stressed (Section 20.3; Fig. 20.2). More recently, Frakes et al. (1993) have interpreted the Phanerozoic history of Earth's climate in terms of warm and cool "modes" not unlike Fischer's "states" but of shorter duration and containing a controversial cool element in the Mesozoic (Section 17; Fig. 21.1). 21.1 Sources o f atmospheric CO 2

The production of CO 2 is governed by long-term volcanism (Rampino, 1991) which can be approximated by the rate of ocean crust production. Examination of Fig. 6.2 shows a good agreement between accretion of oceanic crust and first-order fluctuations in global sea-level. Both these variables are driven by the global tectonic cycle (Wilson cycle) involving supercontinent rifting, dispersal of continents and amalgamation (Fig. 6.2). At times of supercontinent dispersal the length of active mid-ocean spreading centres is maximised and the rate of oceanic crust is highest; the volume of the ocean basins is thereby reduced and global sea-levels are correspondingly high. In contrast, tectonic quiescence prevails at times of supercontinent assembly when new ocean crust production is reduced and existing ocean crust cools and becomes more dense. At times such as these, the volume of ocean basins is maximised thereby lowering global sea-level (Worsley et al., 1986). There is a degree of circular reasoning involved in this model simply because the rate of ocean crust accretion is not derived independently but is calculated from the "Vail" sea-level curve (e.g. Gaffin, 1987). Controversy also surrounds many of

the assumptions on which the Vail curve is based (e.g. Miall, 1986, 1991, 1992; ChristieBlick, 1990). Despite these limitations, the changing rate of ocean crust accretion and associated magmatic outgassing provides a proxy record of p C O 2 levels in the atmosphere. It can be seen from Figs. 6.2 and 21.1 that glaciation in Earth history cannot be matched exactly with cycles in atmospheric p C O 2 pressure; indeed, Ordovician glaciation occurred at a time of very high pCO 2 pressures (Yapp and Poths, 1992). Nonetheless, as emphasised by Veevers (1990) and Berner (1992), Permo-Carboniferous and late Cenozoic glaciations in general, correlate with times of minimum CO 2 and warm periods, such as the Mesozoic, correspond to higher levels of atmospheric CO 2. 21.2 Sinks

The amount of carbon dioxide in the atmosphere is dependent not only upon the rate of CO 2 production but also depends on the efficiency of "scrubbing" processes involving the removal of CO 2 by biota, burial of carbonates and surficial weathering processes. Land plants greatly affect rates of silicate weathering and thus indirectly control atmospheric pCO 2. Possible invasion of Proterozoic landmasses by cyanophytes, followed by traceophyte radiation in the Siluro-Devonian and radiation of droughtadapted grasses after the Eocene have been argued to have facilitated Late Proterozoic, Late Palaeozoic and Quaternary glaciations as a result of enhanced continental weathering and pCO 2 drawdown (Robinson and Upchurch, 1991). Other workers have invoked a lowering of atmospheric carbon dioxide by accelerated burial of carbonates. Roberts (1971, 1976) proposed an "antigreenhouse" effect to explain global and allegedly synchronous Late Proterozoic glaciations. Glaciation was modelled to follow widespread deposition of dolomites ("Dolomite Event"). Atmospheric CO 2 levels are especially sensitive to orogenesis and the resulting sur-

184 ficial weathering of eroded sediment (Worsley et al., 1986; Berner, 1990; Brady, 1991). The CO 2 model for pre-Pleistocene glaciations as proposed by Chamberlin (1899) stresses the essential precondition of orogenic activity and the release of large volumes of crustally-derived sediment. Weathering of this sediment results in drawdown of atmospheric CO 2 level, anti-greenhouse cooling and a tendecy for widespread glaciation. This model has been recently tested by reference to strontium isotopes in seawater which provides an estimate of the degree of weathering on continental surfaces (Raymo, 1991). Long term variation in the ratio of 87Sr//86Sr in the world's oceans is determined by the relative production of high 87Sr//86Sr from continental erosion and global rivers compared with low 87Sr//86Sr derived from hydrothermal activity at mid-ocean ridges (Palmer and Edmond, 1989). Long term geological variation is therefore, a proxy record of enhanced orogenesis. Figures 6.2 and 21.1 suggest that periods of high 87Sr/S6Sr values during the Phanerozoic generally coincide with glaciation, with notable exceptions, suggesting a genetic association between glaciation, increased global orogenic activity and the rate of global chemical erosion. Increased seawater 875r//86Sr ratios in Late Proterozoic strata have also been related to glaciation by Burns and Matter (1991). The present day average value of the 875r/86Sr ratio in seawater is 0.7119 (Dia et al., 1992) and is controlled by run-off from large, low latitude rivers (e.g. Brahmaputra). There has been a general increase in this ratio during the late Cenozoic which reflects the increased average elevation of the continents at this time and uplift of the Tibetan Plateau (Fig. 16.21; Raymo and Ruddiman, 1992). In addition to this long term trend there have been short-term changes in the Sr ratio that show a 100 Ka periodicity corresponding with late Cenozoic glacial and interglacial cycles (Dia et al., 1992). Glacials are characterized by lowered CO 2 partial pressure, interglacials by elevated levels

N.EYLES (Shackleton and Pisias, 1985; Boyle, 1988; Lorius et al., 1990; Fig. 20.2). As related above, fluctuating CO 2 levels are thought to play an important role in amplifying the relative weak effects of "Milankovitch" orbital forcing (Section 20.3). Thus, there is a substantial body of data that shows a good relationship between global tectonism, atmospheric carbon dioxide and Phanerozoic glaciation. With regard to Late Proterozoic glaciations it is worth noting that Schermerhorn (1977) comprehensively refuted the "antigreenhouse" model of Roberts (1971, 1976) by showing that no unique "dolomite event " preceded Late Proterozoic glaciation and furthermore, that neither global extent nor global synchroneity of glaciation can be demonstrated. He reinforced arguments made earlier (Schermerhorn, 1974, 1975) that Late Proterozoic glaciation was regional in scope, diachronous in timing and accompanied tectonic differentiation of source areas and subsiding basins (e.g. Fig. 9.2). Clearly, a combined model is required where increased recognition is given to the effect of regional tectonism and uplift in creating adiabatic glaciers reinforced by the subsequent global drawdown of atmospheric CO 2 resulting from the weathering of crustal debris released from areas of active orogenesis (Table 21.1). Relatively well-understood late Tertiary global climatic changes may provide a key to understanding the role of atmospheric CO 2 in older glaciations (Table 21.1). Raymo and Ruddiman (1992) give a detailed account of the global climatic impact of late Tertiary uplift of the Tibetan Plateau. This single event has conditioned global change over the past 60 Ma by creating a substantial flux of weathered debris and lowered atmospheric CO 2 levels (Table 21.1). This in turn, acting in concert with other global tectonic changes such as widespread uplift around the margins of the North Atlantic Ocean (Figs. 19.2, 20.1), abrupt changes in ocean circulation (Fig. 20.3) and the uplift and thermal isolation of Antarctica (Sections, 18.1, 19.1) allowed am-

185

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C SE'lq'ING

TABLE 21.1 GENERAL

LATE CENOZOIC

INCREASED GLOBAL TECTONICACTIVITY (increased average continental elevation)

Collision of India + Eurasia

TIMETABLE

40 Ma

+ CONTINENTAL POSITIONING (oceanic circulation)

Migration of Australia, development of strong circum-Antarctic current

INCREASED CONTINENTAL WEATHERING + SEDIMENT SUPPLY

Initiation of Antarctic glaciation

36 Ma

Continued uplift of passive margin plateaux around North Atlantic REDUCTIONOF ATMOSPHERIC CO2 REDUCEDGLOBAL TEMPERATURES (increased effectiveness of Milankovitch forcing)

Adiabatic glaciation in Northern Hemisphere

10 Ma

Development of major Northern Hemisphere ice sheets

3Ma

'AL

G

~

PR E R V A A S T O W A R D. . . . . . . & E D MARINE SEDIMENTS ALONG PASSWE MARGINS AND IN RIFT BASINS + 20 Ma

plification of relatively weak Milankovitch variables and the formation of extensive ice sheets on passive margin plateaux (Figs. 19.7, 19.8, 19.9). It can be stated therefore that late Tertiary global change is tectonically driven and may provide a detailed model for other Phanerozoic glaciations. It must be pointed out that episodes of Phanerozoic glaciation have occurred with greatly elevated CO 2 levels such as during the Late Ordovician glaciation of North Africa when atmospheric pCO2 was 16 times higher than today (Berner, 1990; Yapp and Poths, 1992). Fischer (1984) suggested that this was an "overshoot" of the greenhouse effect whereby a heavy cloud cover raised the earth's albedo to a level where glaciers could form. This is a little bit like having one's cake (or ice cream) and eating it too (perhaps in a sauna). Nonetheless, Miller and Vernal (1992) show how a moderate increase in global temperatures created by CO2-induced greenhouse warming may, perversely it might seem, actually promote glaciation in highlatitudes as a result of higher than normal precipitation.

21.3 Capworlds and ringworlds; the climatic importance of palaeogeography and sea level As related above, the long term balance of CO z degassing and its removal as carbonate is governed by the weathering of crustal silicates (Walker et al., 1981). This process has been stable over geologic time and it has been estimated that global temperatures have not varied from modern conditions by more than 15°C (Kuhn et al., 1989). Climatic equability is maintained because under conditions of global warming, enhanced rates of weathering will lead to an increased rate of CO 2 removal from the atmosphere. If, on the other hand, global cooling obtains, then a lowered CO 2 removal rate will result in a temperature increase in response to a greenhouse condition. Long term sea level changes and the distribution of land and sea appears to be an important control on global temperatures and on any predisposition toward glaciation. A global rise in sea-level effectively reduces the area of emergent continents so atmospheric CO 2 levels would rise because less CO2 will be scrubbed from the atmosphere by subaerial weathering. Worsley and Kidder (1991) showed that halving the area of emergent land area could elevate CO 2 levels sufficient to raise surface temperatures over the continents by 10°C (Fig. 21.2). Furthermore, it was determined that global mean temperatures are highest when emergent or semiemergent continents are clustered around the pole ("capworld"). At this time, silicate weathering rates are reduced. The lowest global mean temperature occurs, counterintuitively, when continents form a "ringworld" clustered around the equator allowing the enhanced removal of atmospheric CO 2 by tropical weathering. Different configurations of continents under conditions of low and high sea level are shown in Fig. 21.2. Capworlds have the warmest global temperatures, ringworlds the coolest and "slice-worlds" perhaps representing conditions such as occur at the pre-

186

N. EYLES 3 =

tude glaciation in the Silurian and Devonian

,-,

O

~

~z

- 3°C

1Z*C

- 18°C

12°C

22°C

-

17°C

ZZ*C

1zOc

Z2°C

ZT*C

17°C

LOW SEA LEVEL

3°C

HIGH SEA LEVEL

LAND SURFACE

z7

c

GLOBAL MEAN ~

TEMPERATURE

~

[ ~ com~sEm o 22 C

Fig. 21.2. Global, polar and equatorial temperatures as a function of palaeogeography and sea level. Modified from Worsley and Kidder (1991).

sent day, have intermediate characteristics. Worsley and Kidder (1991) claim that land area and its average latitude alone control global climates on timescales greater than 1 Ma. In support of this argument they claim that the Late Proterozoic continental configuration closely resembled a ringworld with an average latitude of 21°. Such a palaeogeography would produce a low end mean global temperature ( ~ 17°C) which would allow glaciation along the poleward margins of the continents. Mean temperatures for tropical areas would still remain high (22°C; Fig. 21.2). Whilst there is no general agreement as to the configuration of Late Proterozoic continents (Fig. 9.2), all models do largely agree by depicting continental masses that straddle the low to mid-latitudes (Fig. 13.1). Late Ordovician glaciation is the product, according to Worsley and Kidder (1991), of short-term sea-level lowering (and therefore enhanced weathering and CO 2 scrubbing) in an otherwise warm Early Paleozoic cap-world with temperate poles. The model does not take into account well-documented high-lati-

(Section 15, Fig. 15.1). The model furthermore, cannot explain Permo-Carboniferous glaciation which is the longest single phase of Phanerozoic glaciation (c. 100 Ma; Fig. 16.1). The model rules out Cretaceous glaciation because of very warm polar temperatures (12°-17°C) created by high sea-levels in a "sliceworld" configuration. But this is inconsistent with growing evidence of' seasonal or perennial ice covers on high latitude landmasses during the Mesozoic (Section 17). The model is also unrealistic in that the present day is categorized as a "sliceworld" even though it clearly has a major land area (Antarctica) over the south polar region. Whilst it seems eminently reasonable that palaeogeography and sea level exert a strong influence on global climate, the assertion that land area and its average latitude is the first-order control on global climate is not substantiated if the model is tested against Earth's glacial record. The model does serve a very useful function of emphasising the complexity of palaeogeographic controls on long-term climate change. It is interesting to note in this regard that a common palaeogeographic denominator between Phanerozoic glaciations is the presence of a large continental landmass over the south pole during the Early and Late Palaeozoic and the late Cenozoic (Figs. 15.1., 15.3, 20.3).

21.4 The role of methane In addition to carbon dioxide, methane has also been implicated as an important regulator of glaciation. Ice-like compounds of water and natural gas (such as propane, ethane, methane, isobutane, carbon dioxide and hydrogen sulphide) form at high pressures and low temperatures and are present as gas-hydrate cements in many deep marine sediments (Fig. 21.3). The thickness of the hydrated zone below the sea-floor, which may be several hundreds of metres, is governed by geothermal heat flow, water temperature and pressure (i.e. water depth). Gas hydrates

187

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E T F I N G

A METHANEGAS &~ WATER

4

L i

0

10

20

30 0 10 20 TEMPERATURE (*C)

[]

....

s

[]

30

GAs

40

,R,sE,,

Fig. 21.3. Pressure/temperature conditions for the formation of gas hydrates in cold oceans. From Dillon and Paull (1983).

were first detected in Arctic marine sediments in the late 1960's and are widely associated with submarine permafrost (Max, 1991). Under low pressure, hydrates are only stable around the freezing point of water but at greater water depths and pressures, hydrates may be stable at 20°C (Fig. 21.3A). The typical relationship between the phase change of hydrate and free gas, and depth in a sediment column is shown in Fig. 21.3B. The lower limit of gas hydrate cemented sediment is governed by the geothermal heat gradient and is identified on seismic reflection profiles by well-defined bottom simulating reflectors. Gas hydrates occur over large areas of the eastern Canadian coastal margin where postglacial warming at the end of the last glaciation resulted in the release of large volumes of gas following hydrate degradation. Gas venting is recorded on the sea floor by conical depressions (pockmarks; Fig. 21.4) and has also been implicated in the triggering of large slumps on the sea-floor. Methane released from the floors of cold seas has recently been implicated as a factor in long-term climate change. The most common gas hydrate is that of methane (CH4); Paull et al. (1991) estimate that the global total of methane stored in submarine gas hydrates is equivalent to about 1 x 10 4 Gt of carbon. This is double the amount of carbon in all fossil fuels and some 3000 times the

amount of methane contained in the present day atmosphere (Max, 1991). Methane is a very important greenhouse gas so any exchange between the large volume stored in sediments and the atmosphere is potentially important in dictating long term climate change. Enhanced release of gas from offshore sediments could result from substantial hydrostatic pressure changes caused by glacially-lowered sea-levels. Sea-level induced pressure reductions would break down gas hydrates to produce large volumes of interstitial gas which is released by slumping of offshore sediments (Carpenter, 1981). Enhanced slumping and gas release during times of glacially-lowered sea level could result in greenhouse warming sufficient to play a significant role in terminating glaciation (Paull et al., 1991). This model is difficult to balance against data that shows low overall concentrations of methane in ice age atmospheres (Chappellaz, 1990; Lorius et al., 1990; Fig. 20.2). In a variant of the above model, Nisbet (1990) suggested that methane might be released by sea-level rise at the end of the glacial cycle. Arctic sediments that remain frozen year round (permafrost) are the most favourable sites for gas hydrate formation. During the sea-level fall that accompanies

DISTANCE (km) 00

015

1.0

1.5

E Z

0 I-<

I0 EAB~D(17Om) 2O

GASPLUME r~a~'~J~/~r~n

I-- 30 L.LJ Z LLI 40 n I--. 50 Oq

0

60 70

Fig. 21.4. High resolution seismic profile across gas charged Late Pleistocene sediments on a glacially-influenced continental shelf; eastern Canadian continental margin. Modified from Lewis and Keen (1990).

188 ice sheet growth on the continents large a m o u n t s of gas hydrate could develop in arctic continental shelf sediments newly exposed by falling sea-level to a cold climate. Subsequently, this gas would be released at the end of the glacial cycle when sediments were flooded and warmed by the postglacial rise in eustatic sea level. Methane returned to the atmosphere could then accelerate postglacial climate warming. Regardless of the differences between the models of Paull et al. (1991) and Nisbet (1990), ice core data from the Antarctic Ice Sheet show that fluctuations in atmospheric methane, carbon dioxide, atmospheric temperature and global ice volumes are in-phase (Fig. 20.2). The nature of this relationship begs the question as to the identity of a higher-order control. It is now widely recognised that astronomically-driven changes in solar irradiation of the Earth (Milankovitch cycles) are insufficient on their own to bring about glacial/interglacial climate cycles (Broecker and Denton, 1989; Broecker and Peng, 1989; Charles and Fairbanks, 1992). The role of dramatic changes in ocean circulation as an amplifier of Milankovitch cycles, is discussed in Section 20.3. It is possible that such changes in ocean circulation and the accompanying variation in bottom water temperatures could also control the extent and depth of hydrated sediments on the sea floor and the amount of gas vented to the atmosphere (see above). 22. TECTONO-STRATIGRAPHY OF LATE CENOZOIC GLACIATED BASINS Late Cenozoic glaciation occurred in a wide range of tectonic settings from young rift basins through mature passive margins to active margin forearc and backarc basins. It is the purpose of this section to briefly review the style of glacial sedimentation in each stressing the interrelationship between tectonics and sediment supply, depositional facies and stratigraphic successions and long term preservation potential.

N.EYLES Active subsidence is the key to long term preservation of late Cenozoic glacial deposits. Late Cenozoic glacial deposits of the Northern Hemisphere extend from the pole to about 40°N but the short-term ( < 1 Ma) preservation potential of extensive glacioterrestrial facies in North America and Europe is not great; most areas show deposits of the last glacial cycle ( < 125,000) resting erosively on pre-Pleistocene strata; older deposits were reworked by successive ice sheets. The long term post-Pleistocene potential for preservation ( > 10 Ma?) is not much better and most deposits will be reworked by rivers to the continental shelf and slope (e.g. Mississippi River and fan system). Neither is preservation assured for offshore marine glacial deposits unless they can be accomodated in the stratigraphic record as a result of tectonic subsidence. On the eastern Canadian continental shelf for example, no offshore deposits older than the last glaciation are preserved because the shelf is tectonically stable; the fate of such sediment in turn, is to be bulldozed to the shelf edge by the next ice sheet. In this case the base of slope and abyssal plain is the principal repository of Pleistocene glacial sediments in eastern North America. This situation can be contrasted with that of the North Sea Basin where at least 1 km of Quaternary continental shelf glacial strata are preserved; this area provides a very useful "modern" analog for glaciated basins in extensional tectonic settings (see below). Other good examples are found in Antractica along the Transantarctic Rift System where a thick late Cenozoic stratigraphy has been accomodated by rapid regional subsidence. The thickest (5 km) and longest ( < Miocene) record of late Cenozoic glaciation that is directly accessible for outcrop study is that preserved around the Gulf of Alaska in the northern Pacific Ocean. This area provides an excellent "modern" analog for ancient glaciated basins along active plate margins.

189

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

marked northward increase in the average elevation of the coastal margin of the development of a high-relief, cliffed coastline indented by glacially-overdeepened and structurally-controlled fiord basins (Fig. 19.8). The Tertiary evolution of this topography and its bearing on late Cenozoic climate change and glaciation has been reviewed in Section 19.2. The marine record shows that the mid-Pliocene (3-5 Ma) was the time of greatest environmental change when abrupt influxes of sub-polar planktonic foraminifera record the establishment of an Arctic to Antarctic cold current through the Labrador Sea (Gradstein et al., 1990). Ice-rafted debris first appears in the southern Labrador Sea and Central Atlantic just after 3 Ma. Extensive Northern Hemispheric ice covers were initiated at about 2.4 Ma (Section 18.3). Terrestrial records in Greenland and the Canadian Arctic Islands show climatic deterioration accompanied by a southward retreat of temperate vegetation between 5 and 1.7 Ma; the first ice sheet crossed the Scotian Shelf at about 0.8 Ma (Piper et al., 1990).

23. PASSIVE CONTINENTAL MARGINS

23.1 Eastern Canadian continental margin Arguably the most well studied example of a glaciated passive margin is that of eastern Canada, extending from the Scotian Shelf to Baffin Bay (Fig. 19.8). Rifting began in the south, progressed northwards and took about 100 Ma to complete. In the south, rifting between North America and Africa began in the Triassic and was essentially complete by the Middle Jurassic. The northern rifted margin, off Labrador and Baffin Island, results from later plate motions between North America, Europe and Greenland that were not completed until the Early Eocene at about 50 Ma. Along the Scotian margin, the earliest post-rift sediments are represented by carbonate banks (Jansa, 1981) similar to those of the early Iapetus Ocean in Europe which underlie extensive Late Proterozoic glacial deposits (Section 13). Whilst there is no agreement as to the precise kinematic model of rifting along the Western Atlantic margin, the Charlie Gibbs Fracture Zone at about 48°N separates the broad, shallow water shelves of the Grand Banks and Scotian Margin from the narrow, deeper water shelves of northeast Newfoundland, Labrador and Baffin Island (Fig. 19.8). There is also a UPLIFTED,PASSIVEMARGINMOUNTAINS& PLATEAU & FIORD- INDENTEDCOASTUNE

23.2 Glacial depositional record of the eastern Canadian continental shelf The Pleistocene depositional record on the continental shelf is meagre. The average EROOED,PENEPLAJNED

l-'+'-'1 CONTINENTALCRUST

[]

OCEANICCRUST



[]

EARLYRIFTSTRATA

P~.IEISTOCENE GLACICLASTICSEDIMENT

~

ICEBERGSCOURSAND

HUMMOCKY, SLIOETOPOGRAPHY

CUNOFOI~I'RIFTTO DRIFT' PASSIVEMARGINSTRATA

Fig. 23.1. General model of a glaciated passive plate margin based on eastern Canadian and northwest European continental margins. In the absence of marked subsidence, accumulation of a thick Pleistocene glacial record is prohibited (contrast with Figs. 24.1, 24.4 and 26.5).

190 thickness of Quaternary sediments on the Grand Banks is as little as 20 m, on the Scotian Shelf it is only 50 m. Maximum thicknesses on the northern deeper shelves seldom exceed 200 m and even these thicknesses are restricted to fiord basins along the inner coasts and their offshore extensions. Moreover, nearly all of this sediment belongs to the closing phases of the last glacial cycle (Wisconsin; < 100 Ka) when ice began to retreat from the marine environment after 18 Ka. Any pre-existing sediments were bulldozed by the ice sheet to the shelf edge and pushed down to the continental slope. The absence of older deposits on the continental shelf reflects the overall tectonic setting characterized by passive margin uplift inboard of a "hinge-line" which more of less coincides with the shelf edge (Fig. 23.1). This has prohibited the preservation of successive glacial and interglacial strata and has curtailed any "upbuilding" of the shelf. Instead, the shelf has been able to prograde by "outbuilding" of the continental slope. The shallow depth of the shelf results in widespread subaerial exposure at times of glacio-eustatically-lowered sea-level which prohibits any preservation from previous glacial cycles. The shelf can be regarded essentially, as an offshore extension of the uplifted plateaux surfaces (Fig. 23.1). Glacial erosion along the eastern Canadian continental margin has modified and embellished a pre-existing late Tertiary fluvial landscape. It has long been noted for example that many fiord basins are cut along pre-existing river valleys (Ives and Andrews, 1963; Pelletier, 1966). Piper et al. (1990) concluded that most of the shelf topography in eastern Canada is of fluvial origin. The eastern Canadian margin can be contrasted with other late Cenozoic glaciated margins which have experienced subsidence and which consequently, are characterised by high rates of "outbuilding" and "upbuilding" (e.g. West Antarctica; Larter and Barker, 1991; Section 18.1.3). Over most areas of the eastern Canadian

N.EYLES shelf, a single late Wisconsin basal till (probably a deformation till; Section 3.1.1c) overlies peneplained Paleozoic, Cenozoic or Tertiary strata. Networks of "tunnel" valleys that were cut subglacially, either by the focussed flow of meltwaters (Boyd et al., 1988) or by "streams" of deforming subglacial sediment (Boulton, 1987), form anastomosing networks incised as much as 550 m below the sea floor (Fig. 23.1). Thick glaciomarine silts occur on the deeper portions of the shelf and have been interpreted as having accumulated below an extensive floating ice shelf by King and Fader (1986). Small moraine ridges that corrugate the till surface were described as "lift-off" moraines recording uncoupling of a grounded ice sheet to form an ice shelf. Fluctuations of the grounding line of the ice shelf are allegedly recorded by "till tongues" that extend down the continental slope and down the slopes of overdeepened bedrock basins on the shelf. "Till tongues" are interbedded with thick successions of laminated and massive silts. It has recently been recognised that the glaciomarine deposits of the Scotian Shelf were not deposited below a floating ice shelf but are proglacial in origin. These sediments record the rapid retreat of an ice sheet margin terminating in deep water on a glacio-isostatically depressed shelf. "Lift-off" moraines are probably equivalent to " D e Geer" moraines that form annually along retreating ice margins either by winter readvance and sediment bulldozing, by squeezing into transverse crevasses or by sediment deformation accompanying the release of large tabular icebergs (M. Gipp, pers. commun., 1992). Laminated glaciomarine silt facies are characteristic of proglacial marine settings where large volumes of mud are released from tidally-influenced meltwater plumes (see Cowan and Powell, 1990; Section 4.5.1). Tongue-like intercalations of "till" within these facies are debris flow deposits recording downslope resedimentation of subglacial and proglacial sediment newly exposed by ice retreat. Late- and postglacial resedimenta-

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

tion of glacial sediment is a common characteristic of submarine slopes along glaciated continental margins irregardless of tectonic setting (Sections 4, 8, 10, 11, 12).

23.3 Glacial depositional record of the eastern Canadian continental slope It has been emphasised above that little glacial sediment is preserved on the continental shelf because of continued uplift. The major repository of Pleistocene glaciogenic sediment is the deeper water continental slope and rise adjacent to the large glaciated troughs that cross the shelf. Large submarine fans occur at the mouth of the Laurentian Channel (Laurentian Fan; Fig. 23.2) and adjacent to the Cartwright Trough in the Labrador Sea area. Elsewhere, deep sea fans are rare because of very effective slopeparallel sediment dispersal by contour currents (Piper et al., 1990). In these current

191

swept areas, glaciogenic sediment has accumulated within prograding slope prisms. Both submarine fans and slope prisms have experienced rapid progradation ("outbuilding") during successive glaciations. Resedimentation and erosional processes are dominant during times of ice expansion to the shelf edge and alternate with interglacial periods of reduced sediment supply dominated by hemipelagic deposition. It may, however, take many thousands of years for the slope to become stable after glaciation and mass flow processes can continue for some time before the slope reaches equilibrium (Piper et al., 1987). The Laurentian Fan lies at the mouth of the Laurentian Channel and comprises a tongue-like body up to 1000 km long and 300 km wide (Figs. 5.6, 23.2). It shows a dissected upper fan cut by several v-shaped tributary feeder channels, a large convex-upward mid fan with large u-shaped channels up to 30

A

60*

50c

40 c

Fig. 23.2. Laurentian Fan and Sohm Abyssal Plain with isopach of 1929 "Grand Banks" turbidite. Modified from Lewis and Keen (1990), Piper et al. (1988).

192 km wide and nearly 1 km deep, and a broad, concave-upward lower fan portion crossed by channels that extend to the Sohm Abyssal Plain. Thick (up to 7 m) gravel turbidite beds occur along the main feeder channels and pass downslope into thick silt beds within depositional lobes lying on the lower portion of the mid fan. The most extensive facies across the entire fan system are graded muds and silts which record "spilling" over the shelf edge of dilute muddy turbidites released from Pleistocene ice sheets (e.g. Hill, 1984; Swift, 1985). The uppermost part of the Laurentian Fan shows an irregular, hummocky relief created by the slumping of pebbly muds. These facies are the product of the rapid "rain-out" and slumping of muddy proglacial sediments when ice reached the shelf edge. The direct discharge of subglacially-transported deformation till being "squeezed" toward the ice front was probably an important process acting to deliver glacial sediment to the upper slope. The "till tongues" of King and Fader (1986) probably originated in this fashion. Postglacial slumping continues on the upper slope as shown by the sediment failures that accompanied the magnitude 7.2 Grand Banks earthquake of 18 November 1929 (Piper et al., 1988, 1990). Slumping triggered the release of a turbidity current which resulted in the successive downslope breakage of submarine cables; the timing of these breaks indicates current velocities between 30 and 70 k m / h . This flow deposited up to 100 cm of sediment on the Sohm Abyssal Plain (Fig. 23.2) and provides a good example of the effects of downslope reworking of glaciogenic sediment long after glaciation. It may take many tens of thousands or even millions of years for continental slopes to reach equilibrium after glaciation and highlights the hazards in making overly simple interpretations of climate and ice margin behaviour from ancient slope sucessions. Piper and Normark (1989) present an interpretation of over 1000 km of high resolution multichannel seismic reflection profiles

N.EYLES from the Scotian slope. Biostratigraphic age dating control suggests that the intial glacial influence along the slope was felt in the Late Pliocene and is coincident with the onset of Laurentian Fan growth. Average rates of Pleistocene sedimentation are about 30 c m / K a , which is about five times that for non-glacial sedimentation prior to the Late Pliocene. The extent of Pleistocene gullying on the slope, recording the delivery of coarse sediment by ice sheets, greatly constrains the correlation of seismic reflectors along the slope. The dominant glaciogenic sediment along the slope are graded sand and mud turbidites containing ice-rafted debris. These facies are contained either within submarine fan bodies or within prograding slope prisms that merge downslope with the abyssal plain. In eastern Canada, the thickness of Pleistocene glaciogenic sediments of this type approach 2 km on the Laurentian Fan and about 500 m on the Sohm Abyssal Plain. Subsidiary facies are diamictic or conglomeratic debris flows that are cut into the fan sediments or are interbedded within the upper slope portions of the slope prism. Flows have either a channeled cross-sectional geometry reflecting their release from "point sources" at the heads of feeder channels and gullies or a sheet-like form reflecting the spilling-over of glacigenic sediment from "line sources" along the shelf edge.

23. 4 Post-depositional deformation of shelf and slope glaciclastic sediments In addition to the effects of post-depositional slumping and downslope mass-flow glaciated continental margin sediments also undergo deformation as a result of high gas contents and the scour of floating ice masses. However, whilst these processes and their geological effects are extensively developed on modern day shelves that were subjected to Pleistocene glaciations the pre-Pleistocene record is surprisingly poor.

193

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E T T I N G

23.4.1 Gas venting Passive margin glaciogenic marine sediments can contain large volumes of biogenic or petrogenic gases in the form of ice-like crystalline solids (gas hydrates; e.g. Kvenholden, 1987; Fig. 21.3). The stability of gas hydrates and the possible climatic implications of gas venting has already been discussed (Section 21.4). Gas expulsion from degrading hydrate-cemented sediment is accompanied by pitting of the sea floor ("pockmarks"; King and MacLean, 1970; Horland and Judd, 1988; Fig. 21.4). On the eastern Canadian margin these conical depressions are widespread in water depths between 500 and 1100 m; widths vary from 1 to 300 m with depths of up to 15 m. They are particularly common in muddy sediments and extend over more than 20% of the deeper water areas on the Scotian Shelf (Lewis and Keen, 1990). They are commonly associated with iceberg scours suggesting an association between physical disturbance and gas release. Gas charged sediments are easily identified on seismic reflection profiles by acoustic masking and by bottom-simulating reflectors that mark the base of gas hydrate cemented sediments. They pose a potential threat to offshore development in terms of sea-bed foundation stability. The style of sediment deformation below surface pockmarks is not known and such structures await identification in pre-Pleistocene glaciogenic marine sediments. They are part of a group of glacially-related phenomena, such as iceberg scours and cold climate periglacial structures, that are widespread on the present day Earth's surface (e.g. Fig. 23.3) but which are not widely reported from the pre-Pleistocene rock record. 23.4.2 Iceberg scours A n o t h e r very c o m m o n deformational structure observed on the sea floor of glaciated passive margins are iceberg scours. As is also the case with "pockmarks", the wide extent and large number of scour struc-

90 °

70 °

30 °

40 °

1 BO °

160 °

1100

90°

0 I

2000 km J

Average Scale

Fig. 23.3. Extent of permafrost in Northern Hemisphere. After Williams and Smith (1989). Extent of offshore permafrost and gas hydrate-cemented sediment (Figs. 21.3, 21.4) not shown.

tures on modern shelves is not matched by descriptions from ancient strata where they appear to be very rare (Fischbein, 1987; Woodworth-Lynas and Guigne, 1990). This is particularly surprising given the predominance of marine sediments in the rock record and the wide extent of iceberg scours on modern day high-latitude shelves. In eastern Canada, virtually the entire continental margin lies within the area of iceberg influence (Fig. 23.4). The major source area is the Baffin Bay coastline of West Greenland, at about 70°N, where calving glaciers release an average annual flux of about 40,000 icebergs (Lewis and Woodworth-Lynas, 1990). About 400 of these survive their passage southwards to reach the Grand Banks (50°N) where they

194

N. EYLES

60*

zo*

60 °

\ \

\ \ I

J I / / /

/

20 o

Fig. 23.4. Extent of present day ice berg drift and seafioor disturbance along eastern Canadian continental margin. After Lewis and Keen (1990).

typically last about 2 weeks. The average berg appears to spend about 50% of its lifetime sitting on the sea floor and thus may leave some geological record of its existence. Iceberg scours are abundant on the muddy, and generally deeper, northern shelf off the coast of Labrador and Baffin Island but cover less than 5% of the southern, more shallow, sandy shelf. Observed scours are up to 20 m deep have flat floors, steep (60 °) side slopes in cohesive sediments and are flanked by berms of thrust sediment up to 6 m high. Scour widths may exceed 300 m and extend for more than 10 km. Large pits, commonly greater than 10 m in depth, record the break up and capsize of large tabular bergs and the subsequent passive failure of sea floor sediments (Clark and Landva, 1988; Fig. 23.5).

Current and wind-driven grounded bergs construct large scours which are known to exceed 300 m in width and extend for more than 10 km (Woodworth-Lynas and Guigne, 1990). Modern scours have been found at depths of 427 m in Baffin Bay and to 500 m in the Antarctic (Barnes and Lien, 1988); relict Pleistocene scours, reflecting glacio-eustatic sea-level lowering, occur to 750 m on the Labrador and Baffin slopes. Scours rapidly destroyed by sea-floor bottom currents on the southern shelves and in general are absent in water depths less than about 100 m. The repeated pummelling of sea-floor sediments by icebergs produces lag boulder gravels and more poorly-sorted diamicts termed "iceberg turbates" by Vorren et al. (1983). Lags are created by the repeated resuspension of mud during continued exposure to grounding events; diamicts by the deformational mixing of muddy sea-floor sediments and ice-rafted debris. These facies form a relatively coarse-grained "cap" resting on underlying sediments with an irregular contact that should show local penetrative deformation structures resulting from sediment loading below iceberg keels. On cohesive, muddy substrates characteristic of deeper water settings, the effects of repeated turbation may be to produce a chaotically-

~i!i~i~i~i~i~!~i~i~i~i~i~ii~ii~i~ii~!i!!ii~i~i~ii~i~i!i+ii~ii~i~i~i~i!ii!iiiiiiiii~!ii!i~i~i~iii~i~i!i!~ii!iii!i!i~i~ii!ii!iiii!i!iii!i~ii~ii~

~~

surfaces

WIND & CURRENT~,,'~ ,,, ~'X

i~f.,B ERM

i+~ " N Fig. 23.5. Simple passive failure model for formation of deep (approx. 10 m) pits and scours on glacially-influenced sea-floors. After Clark and Landva (1988).

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C SE'Iq'ING

structured diamict, composed of brecciated muds and ice-rafted clasts, resting on a faulted and perhaps folded substrate (Figs. 23.6, 23.7). These diamict facies may resemble deformation till resulting from subglacial shearing of overridden sediments (Figs. 3.2, 3.5, 3.6) but the overall depositional and sedimentary context of these sediments is very different. Pleistocene iceberg turbates and scours are commonly identified on seismic profiles from the Norwegian and Cana-

195

dian coastal margins (Vorren et al., 1983; Piper et al., 1990; Fig. 23.6) but not so far, from the pre-Pleistocene geological record. Thomas and Connell (1985) described structures exposed in outcrops of Pleistocene glaciolacustrine sediments that result from iceberg scour and dumping of ice-rafted debris. Eyles and Clark (1988) describe a large v-shaped structure in Pleistocene glaciolacustrine deltaic sediments thought to be the result of substrate deformation by a ground-

C

Fig. 23.6. A, B, C. High resolution seismic profiles of buried Late Pleistocene ( < 15 Ka) ice scours and turbated sediment; Emerald Basin, Eastern Canadian continental margin. Scale bar for A and B is 100 m long and 10 m high, that in C is 50 m long and 5 m high. Unpublished data courtesy of M. Gipp. Turbated sediments are draped by undisturbed postglacial strata.

196

N. EYLES

m

"~o

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETI'ING

197

Fig. 23.7 (continued). ing ice mass. In this case, it cannot be determined whether the ice mass was glacier ice or seasonal pack ice. This is an important consideration because on cold climate coastlines, intense compression between shore-fast pack ice and wind and current driven pack produces large pressure ridges ("starnukhi") that have deep keels. These produce sea bottom scour features in shallow water ( < 50 m) that are essentially indistinguishable from those produced by icebergs. Because of the potential difficulty of discriminating between the sedimentological effects of stamukhi and iceberg scour in the ancient record, Barnes and Lien (1988) argue that the all inclusive term "ice keel turbate" should be used in preference to the more specific iceberg turbate. Woodworth-Lynas and Guigne (1990) excavated two scours (c. 50 m wide, > 5 km

long) on the now drained floor of proglacial Lake Agassiz (1,000,000 kin 2) that developed toward the closing stages of the last glaciation in North America. Scours are incised into laminated and massive clays and show plastically-deformed and faulted clays below a well-defined scour incision surface. The orientation, dip and sense of movement of faults below the scour confirm that a simple two-dimensional foundation failure model can be used to predict the record of ice keel turbation in pre-Pleistocene strata (Fig. 23.4). On non-cohesive sandy substrates in shallower, more high energy settings, well-defined scours are unlikely to survive later reworking; ice keel turbation may be recorded by horizons of deformed sand showing evidence of exess pore water pressures, liquefaction, sediment loading and injections. These facies may resemble subaerial sedi-

198

N.EYLES

ments disturbed by freezing of interstitial ice and the growth of ice lenses in a cold periglacial climate. Again, however, the overall sedimentary context is very different and important diagnostic clues are provided by related facies. 24. RIFT BASINS

24.1 North Sea Basin North Sea rifting started in the Early Triassic and was most intense during the Middle Jurassic to earliest Cretaceous. The basin has since then developed into a thermal sag basin (Ziegler et al., 1989). Total subsidence over the past 65 Ma is about 3.25 km (Bjorslev Nielsen et al., 1986). Deep-seated and longlived structures in the basement have partially controlled rifting patterns leading to the development of complexly-linked basins floored by a major detachment system (Fig. 19.9; Gibbs, 1989). In the central North Sea, along the central graben and the AngloDutch basins, Pleistocene sediments approach 1000 m in thickness and record a sharp increase in subsidence rates (Ziegler and Van Hoorn, 1989). Isopachs of Pleistocene glaciclastic sediment follow the principal structural trends (Fig. 24.1). Accelerated subsidence along the central parts of the basin can probably be linked to complimentary uplift of the Scandinavian Shield and the Scottish Highlands and western Britain in response to the generation of a complex "upper plate" detachment system (Fig. 19.9). In the adjacent Irish Sea Basin the marine record locally reaches a maximum thickness of about 300 m. Basin-wide investigations of outcrops and seismic profiles shows the importance of sea-level and water depths in controlling behavior of ice sheet margins and the spatial distribution of facies and depositional systems. An important process has been the generation of high relative sealevels, as a result of crustal downwarping under the margins of the British ice sheet,

1

0

km

zoo

Fig. 24.1. Isopachs of Pleistocene glaciclasticsediment in the North Sea Rift system. Modified from Gibbs (1989) and Bjorslev Nielson et al. (1986). See Fig. 19.9 for underlying structure. Contours interval is 200 m. Thicknesses over 1000 m shown in black.

and rapid calving of the ice margin. This results in a short-lived phase of enhanced ice velocity as the ice disintegrates and is accompanied by the delivery of large volumes of meltwater and sediment to the ice margin (see Hughes, 1987; N. Eyles and McCabe, 1989a, b). The record of this flux is preserved across the inner continental shelf in the form of large morainal banks, built along the ice margin where it was able to stabilize temporarily in shallow water, and by tunnel valleys. The onshore subglacial record of fast ice flow is provided by drumlins formed by the deformation and selective erosion of overridden sediments (see Boulton and Hindmarsh, 1987). Space does not permit a detailed review of the sedimentology of morainal bank settings and the reader is referred to Powell (1981, 1990) for details. If these data are representative of other areas of marine-based ice sheets then the glacial stratigraphic record of any one glacial cycle is strongly biased to the final decay phase of the ice sheet. This should not be too surprising because it is at this time that most meltwater and sediment is released. (e.g. Ashley et al., 1991)

199

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETT1NG

]

/ C)r,¢:.=, )00 km

B,'

90 ° E --

--900W

INDIAN X--

[] )CEAN

[]

VLB VICTORIALANDBASIN

lUU

/

I

\

Fig. 24.2. Ross Sea/Weddell Sea Rift System and associated basins. The rift system is comparable in size with glaciated Late Proterozoic rifted margin basins (e.g. Figs. 9.2, 12.4) and the Cenozoic North Sea Rift system (Figs. 19.9 and 24.1). Cross-section X - X ' is shown on Fig. 24.4.

traced 3500 km across the entire continent (Kriftoffersen et al., 1991). Both basins share a similar history of structural development and sedimentation and are good analogs for older glaciated basins. The Ross Sea Rift system is comparable in size to the Basin and Range Province of the western North Amer-

24.2 The Ross Sea / Weddell Sea rift system of Antarctica The Ross Sea Rift and the Weddell Sea Basin form mirror-image counterparts either side of Antarctica (Fig. 24.2) and together comprise a glaciated rift system that can be 0o

D 90 °

W-

-

~

~

GLACIALLY-INFLUENCEDMARINESTRATA

90 o E

ROSSICESHELF

90* E ~ E Skin

180"

TRANSANTARCTIC MTNS.

1 80 °

ROSS SEA RIFT

AR~ClC

SEALEVEL

S ICESHELF

-5krn SOUTH POLE

Fig. 24.3. Oblique cross-section through the Ross Sea Rift (Fig. 24.2). Modified from Barrett et al. (1991).

200

N. EYLES

ica and the East African Rift system (Tessensohn et al., 1991). It is directly comparable to the North Sea Rift System (Fig. 24.1) and provides an excellent analogue for many of the glaciated Late Proterozoic sea ways that developed along the rifted margins of Laurentia in Australia, western North America and Europe (Figs. 9.2, 12.4). The 900 km wide Ross Sea embayment is a zone of crustal extension between West and East Antarctica and consists of a basement graben flanked by uplifted rift shoulders (Transantarctic Mountains; Figs. 24.3, 24.4). The graben was initiated during Late Mesozoic breakup of Gondwana and contains up to 14 km of sediment. These strata include a thick glaciomarine succession that post-dates Early Oligocene initiation of the Antarctic Ice Sheet at about 36 Ma in response to accelerated uplift of the Transantarctic Mountains. The mountains have risen about 5 km over the past 50 Ma but this figure disguises short-lived higher uplift rates (see Section 11.2). The Ross Sea contains three major graben systems (Victoria Land Basin, Central Trough and the Eastern Basin; Fig. 24.4). The thickest fill occurs within the Victoria Land Basin which is undergoing active rifting along the Terror Rift marked by calc-alkaline volcanoes. The Terror Rift is about 60 km wide and is delineated by faults that extend through the entire stratigraphic column (Fig. 23.4). Regional faults run parallel to the front of the Transantarctic Mountains and extend southTRANSANTARCTIC MOUNTAINS

VICTORIALAND BASIN

Oo,: x'

eastward into the Byrd Subglacial Basin which, along the Bentley Trough, extends to nearly 2000 m below sea-level (Fig. 23.2). A drill core in the Victoria Land Basin (CIROS-1; Barrett, 1989) in western McMurdo Sound reached the base of a 700 m thick glacial section in strata dated by diatoms and foraminifera at Early Oligocene. The upper half of the core is dominated by diamictites that rest on a lower mudstonedominated succession. This lower succession shows thin diamictites interbedded with turbidites typical of the infill of a deep (up to 600 m) basin close to a steep glaciated basin margin. Sedimentation rates averaged 200 m / M a (Barrett et al., 1989, 1991). The remainder of the core contains interbedded diamictites and fossil-rich mudstones that show pervasive soft sediment deformation again indicating the continued influence of a depositional slope and active tectonism. A prominent diamictite unit in mid-core is argued to coincide with the mid-Oligocene eustatic low of Haq et al. (1988) and is consistent with the growth of a large ice sheet across Antarctica. The magnitude of the recorded change in sea level is close to 200 m, and is associated with a 4 m.y. hiatus (Harwood et al., 1989) suggesting an underlying tectonic and not climatic control. The difficulty of identifying glacioeustatic sealevel variation from glacial successions preserved in areas of active tectonism has already been stressed (Sections 8, 10, 12, 14). More recent tectonic uplift along the Victo-

CENTRAL TROUGH

EASTERN BASIN

x

10

km~

~

km TERROR RIFT& VOLCANOES

Fig. 24.4. Cross-section through the Ross Sea Rift showing post-rift infill. The Victoria Land Basin contains a thick glaciclastic succession that post dates initiation the Antarctic Ice Sheet at c. 36 Ma. For location of section see Fig. 24.2.

201

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C SE'I'F1NG

ria Land Basin is recorded at the CIROS-1 site by sea floor truncation of Early Miocene sediments (c. 22 Ma). The averaged sedimentation rate in the western Ross Sea during the mid-Oligocene to Early Miocene was about 40 m / M a for the uppermost diamictite-dominated succession. Cooper et al. (1991a, b) suggest that sedimentation rates are doubled for the opposite (eastern) Ross Sea where seismic profiles show a thick, sea-ward prograding post-Oligocene glacial sequence. In the western Ross Sea, the rate of tectonic subsidence has outstripped the availability of sediment resulting in deep structural-bathymetric troughs. This tectonically-generated topography exerts a major influence on ice sheet dynamics. The considerable sub-ice topography below the Ross Ice Shelf and the Filchner Ice Shelf of the Weddell Sea dictate local flow directions and dynamics with the Antarctic Ice Sheet. There is a particularly interesting relationship between grabens and fast flowing (800 m/yr) ice streams within the ice sheet. The Cary Trough (CT), Byrd subglacial Basin (BSB) and Bentley Subglacial Trough (BST; Fig. 24.2) have a relief that exceeds 2000 m and whilst previously thought to be glacially-scoured "fiords" are now recognized as tectonically-controlled rifts (Storey, 1991). The most well-studied Antarctic ice stream is that of ice stream B within the Ross Ice Shelf (Fig. 24.2; Alley et al., 1987). This ice stream rests on a thin (average 6 m) unconsolidated diamict (deformation till; Section 3.1.1c) which is deforming throughout its thickness and is cutting an unconformity across Late Oligocene glaciomarine strata similar to that described from the adjacent Victoria Land Basin graben by Barrett (1989; see above). The deforming till layer is "dumped" en masse on a grounding line fan where the ice stream begins to float as the Ross Ice Shelf (Figs. 19.5, 24.2, 24.3). A structural control on the flow of the ice sheet, leading to the development of fast flowing ice streams, may have been a characteristic

I YUKON FLATS

(

2 TANANA

3 SUSITNA

~--~ ~N

~

ALASKA

~

4 COPPERR,VER ~ _ ~ " CENOZOIC

VOLCANOES /

• WHITE RIVER ~

/.~,S

- A ] ~

ANDESFFE / ~ -- FAULT

l

~

/2

~ J

~ ~

'

~

"~

~

CANADA

'

! I~"

~

I~%

c~t

o

Fig. 25.1. Glaciated "successor" basins developedalong strike-slip faults in the interior of Alaska. Thick Pleistocene glacial strata are preserved in these basins (see N. Eyles, 1987; C.H. Eyles and N. Eyles, 1989). After Schultz and Aydin (1990) and Little (1990).

of the rifted and often fiord-like margins of Late Paleozoic Gondwana glaciated basins (e.g. Figs. 16.2, 16.9). 25. CONTINENTAL STRIKE-SLIP AND INTERIOR RIFT BASINS 25.1 Alaska Significant thicknesses of Pleistocene glacial deposits are preserved in several non-marine, strike-slip basins of Alaska. These form curved, interior basins along the major fault systems that accomodate transpressional movements between the North American and Pacific Plates (Fig. 25.1) and can be regarded as "successor" basins since they overlie allochthonous/terranes accreted to North America by Mesozoic and Tertiary plate motions. The Chugach and Prince William terranes were in place by about 50 Ma (Plafker, 1987) and the Yukutat Block is currently in the process of being emplaced against the backstop of North American carrying with it the overlying thick, depositional record of late Cenozoic glaciation (Yakataga Formation; Fig. 26.3). The present day boundary between North America and the Pacific Plate is taken as the Aleutian Trench, and the Queen Charlotte/Fairweather fault

202 system which shows about 6 km of right lateral displacement during the Pleistocene. Modelling of the curved Alaskan strike-slip fault shows that uplift forces are produced on the inner concave margins and subsidence on the northern convex margins (e.g. Little, 1990; Schultz and Aydin, 1990). This is expressed as the Alaska Range and Chugach Mountain uplifts and their flanking basins (Fig. 25.1). In the Copper River Basin, as much as 300 m of Pleistocene glacial sediments are interbedded with andesite lavas and other volcaniclastics derived from the Wrangell Mountains volcanic complex. The basin contains a very large lake during successive Pleistocene glaciations when its southern outlet is plugged by glaciers on the Chugach Range; extensive exposures of glaciolacustrine sediments along the Copper River and its tributaries are described by N. Eyles (1987); volcanic tephra provide important marker horizons. Not unexpectedly, the glacial sedimentary record in the interior strike-slip basins of Alaska is dominated by lake sediments and mass-flow dominated fan-delta deposits. A major influence has been proximity to steep glaciated mountain slopes, episodic volcanism, downslope reworking of glacial and volcanic sediment and lake ponding (C.H. Eyles and N. Eyles, 1989). Primary in situ glacial facies have in most cases been reworked either by gravity or by meltstreams (Fig. 10.7). The restricted thickness ( < 300 m) of late Cenozoic glacial sediments in the interior strike-slip basins of Alaska contrasts with the very substantial record preserved offshore in the north Pacific Ocean (Section 26.1). The wet, oceanic coastal mountains support extensive ice fields and piedmont glaciers that release very large volumes of meltwater to the Pacific Ocean. In contrast, the interior is relatively arid and shut-off from oceanic influences. The dominant late Cenozoic sediment across much of Interior Alaska is aeolian cover sand and loess (e.g. Lea, 1990; Westgate et al., 1990).

N.EYLES

-~------~ ATE PLEISTOCENEGLACIAL & INTERGLACIALSTRATA

N

S

Fig. 25.2. The Kleszczow Graben system of southcentral Poland. After Brodzikowski et al. (1987).

25. 2 Poland

In Europe, Pleistocene glacial and interglacial strata are preserved in interior rift basins on the northern margins of the Alpine foreland that are analogous to that of the Alaskan basins described above. In Poland, glacial sediments occur within the 40 km long Kleszczow Graben which began to subside in the Late Oligocene (Brodzikowski et al., 1987; Fig. 25.2). Up to eight Pleistocene glacial cycles are preserved each separated by fluvial or lacustrine interglacial sediments. Local karstification of underlying Permian carbonates, halokinesis by salt diapirism, compaction of Miocene lignites and glacial reactivation of basement faults by ice sheet loading are, in combination, responsible for continued subsidence and preservation of a continental glacial record. Glacial sediments

203

EARTH'S GLACIAL RECORD AND ITS TECTONIC SETI'ING

Eocene

OIIgocene

IRio- [

Miocene

Paleogene

cane

Q

Neogene

L~

t

20 ~

4

Pacific

Northwest

10 ~

50

410

~

310

2FO

110 •

/~'

0

ka

Fig. 26.1. Terrestrial botanical record of climatic deterioration from western North America. Based on Wolfe (1978). zx: initiation of glaciation in the Antarctic (Zachos et al., 1992). • : initiation of glaciation in Gulf of Alaska (Lagoe et al., 1993). ~ : initiation of Northern Hemisphere continental ice sheets (Shackleton et al., 1984; Ruddiman et al., 1986).

are extensively deformed by differential subsidence across the graben (Brodzikowski et al., 1987). 26. ACTIVE M A R G I N BASINS

26.1 The Gulf of Alaska forearc basin In southern Alaska there is a well defined palaeobotanic record of Palaeogene and

Neogene climate deterioration (Fig. 26.1) but the timing of initial glaciomarine sedimentation in the North Pacific Ocean is controversial. The first arrival of glaciers to tidewater is recorded by dropstones and debris flows at the base of the Yakataga Formation. These strata subcrop across the Gulf of Alaska shelf where they reach thicknesses of almost 5 km and are spectacularly exposed along the flanks of rapidly-uplifting coastal mountains (Fig. 4.6). Ages based on mollusc biostratigraphic zones suggest that initial glaciation in the North Pacific occurred in the early Middle Miocene (15-16 Ma; Marincovich, 1990). In contrast, work on planktic foraminifera indicates that this event can be no older than Late Miocene ( < 11 Ma), and probably latest Miocene (c. 5-6 Ma; Lagoe et al., 1993; Section 18.3). The timing of these events is important because the Yakataga Formation represents the earliest, most complete and yet accessible record of Northern Hemisphere glaciation. Comparison with high-resolution oxygen isotopic records suggests a correlation of initial Yakataga glaciation to brief cold events in the Late Miocene/Early Pliocene in agreement with plankic microfossil dating of these events (see Lagoe et al., 1993 for a detailed review). The onset of glaciation around the Gulf of

WHITERIVERGLACIALS FIGS. 10.8,25.1

Mt. St. Elias

I

WRANGELL .TS

\

1

~r'~ .c~~

ALASKARANGE

FIG. 26.4

;

\ Chugach/St.EliasFault

MIDDLETON ISLAND

0 ..............

,'¢,

. ~ / 100km - - - -

~

,.~ /

/

/

/ / /

0

km

100

V

....

I

Fig. 26.2. Active plate margin setting of Late Cenozoic glacially-influenced Yakataga Formation (from C.H. Eyles et al.., 1991).

204

N. EYLES

et al. (1990), C.H. Eyles and Lagoe (1990) and C.H. Eyles et al. (1991) and N. Eyles et al. (1992b). The formation comprises a shallowing-upward succession from deep water, mass flow facies (debris flows and turbidites) to glacially-influenced shallow marine facies (blanket-like units of muds and diamictites of a rain-out/resedimented origin) deposited in a "ponded" forearc basin (Fig. 26.4). Thick (up to 300 m) postglacial muds are presently accumulating across the Gulf of Alaska shelf having been released from the Copper River delta and glaciated coastal embayments. The Yakataga Formation contains much evidence for the continued importance of downslope resedimentation triggered by large magnitude earthquakes (Fig. 4.3). Modern analogues are provided by the present day shelf/slope system around the Gulf of Alaska where extensive, earthquake-generated sediment slides can be mapped (Carlson and Molnia, 1977; Fig. 26.5). The slide on the upper portion of Kayak Trough is about 20 km long, has a maximum width of 15 km, a thickness of over 115 m and shows a hum-

Alaska is closely linked to tectonism associated with subduction of the Pacific Plate below North America and the uplift of high coastal mountain ranges over 5000 m in elevation (e.g. Mount St. Elias). The Pacific Plate is moving northwestward at 6.3 c m / y r and is subducting below North America along the line of the Aleutian Trench. To the east of the trench, movement between the two plates is accomodated by strike-slip movement. The Yakutat Block, an allochthonous terrane having originated from somewhere near to present day Oregon, is riding "piggy-back" on the Pacific plate and is currently docking against North America. A substantial thickness of late Cenozoic glaciclastic strata overlies the displaced terrane (Fig. 26.2). The Yakataga Formation records the rapid infiUing of a complex forearc basin on the landward side of an outer accretionary wedge (Fig. 26.3). Tectonic setting, stratigraphy, sedimentary facies, seismic facies and biofacies have been extensively reviewed by Plafker (1987), Lagoe et al. (1989), Carlson

/

o

-- -- -1 2 3 4 S

,,

/

#

BASE OF GLACIALSTRATA (c. 6 Ma; YAKATAGA FORMATION) ACCRETIONARYWEDGE SLOPEBASINSIN AREAS OF COMPRESSIONALRIDGING& 'TRELLISED' SLOPEGULLIES OUTERSHELF HIGH ABOVE ACCRETIONARYWEDGE AND SHALLOW WATER SUBMARINEBANKS SLIBMARINETROUGH CROSSINGSHELFAND DISCHARGINGDOWNSLOPEe.g. Yukutat Trench (Figs. 4.6, 26.7) DENDRITICGULLYON SMOOTHSLOPE;SEDIMENTLOST TO TRENCH

Fig. 26.3. Schematic model of modern Gulf of Alaska active margin basin. See Fig. 26.4 for vertical succession of glaciclastic facies.

205

EARTH'S GLACIAL RECORD AND ITS TECTONIC SETTING

O

thickness of 150 m and extends over 1200 km 2 of the shelf. These thicknesses are comparable to mass flow olistostromes exposed in outcrop (Fig. 4.3). In some areas of modern sliding gas cratering may have contributed to the development of a hummocky surface topography (Carlson and Molnia, 1977). The continental shelf edge in the Gulf of Alaska takes the form of a distinct " o u t e r shelf high" that lies above the accretionary wedge complex of the Aleutian T r e n c h (Fig. 26.4). The raised shelf edge is demarcated by submarine banks that are subject to active winnowing and are accumulating gravel lags and clastic carbonates (N. Eyles and Lagoe, 1989). Parts of the shelf are experiencing vigorous uplift accompanying the construction of the accretionary wedge. The adjacent continental slope shows a range of different morphologies in response to different structural settings (Fig. 26.6). West of Bering Trough it is traversed by a dendritic system of gulleys. These assume a distinctly trellised pattern to the west of Middleton Island reflecting active compressional "ridging" of the underlying slope (Carlson et al., 1990). Wide

S.L

®

S.L

PONDED,FOREARCBASINAND GLACIALLY-~IFLLIENC~D 'SHELF'FACIES

®

BIOLAC'ER ~ G L A C I A L L Y

INFLUENCEDMARINESTRATA

Fig. 26.4. Schematic model for glaciclastic sedimentation in a forearc basin. Glacial strata show vertical succession of deep water mass flow deposits (1), to shallow water "shelf" deposits (3), recording infilling and "ponding" of forearc basin behind accretionary wedge. Based on Yakataga Formation, Gulf of Alaska (C.H. Eyles et al., 1991); basin evolution shown here spans the last 6 Ma. mocky topography typified by disrupted seismic reflectors. A n o t h e r prominent slide area extends off the mouth of Icy Bay, has a

C/ylp'"

k

PRINCE WILLIAM SOUND

~c~c~

~ i,¢K,~'~/" S ~ . _ ~ A

/,/ /J~)

~

Are~ of a~ing

~

m

~~

--,OOm BathymetrJc contour

Y/~

~ ~

I

\

I

t ~~

I

ZOOm 't O0

"t-~

'~

~,

-u~,

Fig. 26.5. Areas of modern sediment instability along Gulf of Alaska continental margin. Affected sediments are last glaciation glacially-influenced marine diamicts and muds; similar slide movements are recorded in onshore outcrops of the Miocene/Pliocene Yahataga Formation (Fig. 4.3) suggesting sediment instability is a recurrent feature of the margin since the initiation of glaciation (c. 6 Ma). Offshore data modified from Carlson and Molnia (1977).

206

N. EYL~S

submarine troughs that cross the shelf from glaciated embayments along the inner shelf (e.g. Yakutat Trough; Carlson et al., 1982) discharge directly down the continental slope into the 700 km long Surveyor Channel system and Surveyor deep-sea fan (Figs. 5.3, 26.6). Exposed Yakataga Formation strata showed many examples of submarine feeder channels filled with debris flows, olistostromes and turbidites (Figs. 4.3, 4.6). The southern extent of the fan is defined by the Patton Seamounts and to the west by the Aleutian Trench. DSDP site 178 lies on the western flank of the fan; the onset of deposi-

tion across the fan coincides with the initiation of glaciation and the release of very large volumes of glacially-derived sediment to the Gulf (Fig. 5.3). Single channel seismic tracklines and drill site data across the lower reach of the Surveyor Channel show that the fan is composed of a lower turbidite sequence that is conformably overlain by fine-grained overbank deposits spilt from turbidity currents flowing down the Surveyor Channel (Stephenson and Embley, 1987). Sediment is are delivered directly to the eastern end of the Aleutian Trench (Fig. 26.6). Piper et al. (1973) esti-

CHUGACHMOUNTAINS ))

S7. ~LI~s R~NG~ Outcropof Yaklti{l| ~ Formation .--_ BathymetricContour~ in metrel Thicknelscontou~ of - - YakItagaFormation in kilometrel

RIVERDELTA .60 °

PENINSULA

~.LSEK

AMATULi TROUGH X

~ )%

,'j~

/f__;

/

J

-~-,,.~,~7/" ~

%%

CROSS SOUND

/'~1

~." " - ~

S ~-'~'a -~ ~c~

%

."

'

...

/.i

,I

/X/

O.D.P.

.

~"~ ............,~./f.. . , ....'.,.... ",.

..,', 5/'.:'_ _

.//¢.-

.,

//

~,T~,.~...~. . - ........

., ..

'~

A B Y S S A L ~

_

_

~

..

~.,,.:........

\

",'.,>._

--- ~ -

------.Y*....... . ,~

/

. /

.

/

"..'.--,~

¢

/

/

Fig. 26.6. Physiography of Gulf of Alaska continental margin and thickness of Late Cenozoic ( < 6 Ma) glaciclastic Yakataga Formation. A schematic model of margin is shown in Fig. 26.4. Total glaciclastic sediment volumes are shown on Fig. 5,3. Based on data in Plafker (1987), Carlson et al. (1982, 1990).

207

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C SE'Iq'ING

mate that sedimentation rates along the trench are as high as 3500 m / M a ; the equivalent figure for the abyssal plain is about 175 m / M a and that for the adjacent sea mounts is 40 m / M a which can be considered as the rate of "background" pelagic sedimentation. The rate of pelagic sedimentation is high and reflects the availability of large volumes of glaciclastic mud resuspended from submarine slides and by storms across the shelf. The situation across the Gulf of Alaska abyssal plain, where sediment derived from the shelf and slope is delivered over a substantial distance directly to the Aleutian Trench, contrasts with that seen along nonglaciated continental margins where submarine fans typically have shorter feeder channels that terminate in mid-fan. This morphological difference may reflect the much greater sediment loading of the Surveyor Fan. The fan is fed by glaciers in a temperate oceanic climate where disproportionately large volumes of sediment are released compared with other depositional settings (Fig. 5.3). The Aleutian Trench contains a much larger volume of sediment compared with non-glacial trenches as a result of the glacially-influenced setting (Piper et al., 1973). The primary role of glaciers along the Gulf of Alaska continental margin has been to "oversupply" sediment to the forearc basin; the rapid infilling of the basin is reflected in the overall shallowing-upward character of the Yakataga Formation (Fig. 26.4). This has also promoted the construction of a large accretionary wedge which acts as a "sill" defining the edge of the continental shelf thereby promoting the trapping of glaciclastic sediment along the margin. Ultimately, the volume of sediment may be too large to be digested by the subduction system which may eventually "choke"; the long term preservation potential of the late Cenozoic Yakataga Formation would appear to be rather good. These strata are being carried northwards onto the outer margin of the North American plate by the underlying

Yakutat Block (Fig. 26.2). The fate of these strata is to be smeared onto the North American plate and to be preserved as part of the currently forming Cordilleran "mobile belt". The late Cenozoic Gulf of Alaska provides an excellent tectono-stratigraphic analogue for Late Proterozoic Pan-African active margin basins (Section 10). The great thickness and extent of glacially-influenced marine strata in the Gulf of Alaska forearc basin testifies to the temperate maritime climatic setting of the Gulf. High coastal mountains intercept moistureladen air moving in from the Pacific Ocean and result in the discharge of enhanced volumes of glacial meltwater and mud to the Gulf (see Section 5.3). Inland, in the lee of the coastal mountains the record of late Cenozoic glaciation is very much more restricted. Interbedded debris flows, composed of reworked glaciclastic and volcaniclastic debris, fluvial and lacustrine sediments and lava flows are preserved along the valley of the White River (Fig. 25.1) and record glacially-influenced sedimentation within a small ( < 10 km 2) strike-slip basin close to an active volcano (C.H. Eyles and N. Eyles, 1989; Fig. 10.7). Glacially-influenced sedimentation may have begun as early as 10 Ma (Denton and Armstrong, 1969) but a total stratigraphic thickness of less than 1 km has been preserved.

26.2 Bransfield Strait backarc basin Bransfield Strait lies between the northern Antarctic Peninsula and the South Shetland Islands which lie inboard of the South Shetland Trench (Figs. 24.2, 26.7). Bransfield Basin records the transition from a rift basin to a marginal sea characterized by early sea floor spreading and voluminous piles of volcanic and glaciogenic sediment The South Shetland Islands are built around volcanic edifices of calc-alkaline lavas, recording eastward subduction of ancient Pacific Ocean crust (Aluk plate). Subduction ceased at about 6 Ma; extension of the Bransfield Basin

208

N. EYLES

occurred as the subducted slab cooled and began to sink (Jeffers et al., 1991). A broadly similar development occurred at the same time along the Andes with development of the Coastal Range, Central Valley and the Andean Range of Chile (Gonzalez-Ferran, 1991). The central rift of Bransfield Strait is delineated by oceanic shield volcanoes that are buried below considerable thicknesses of glacial and volcanically-derived marine sediments. Glaciation along the crests of the South Shetland Island began at or before 50 Ma (Eocene) and is recorded by thin (3.6 m) and isolated glaciomarine strata (Krakow Glaciation). These sediments are exposed on King George Island and contain Tertiary bivalves and scaphopods and redeposited Cretaceous coccoliths. Deposits relate to local alpine glaciation along the Antarctic Peninsula. Older, possibly Late Cretaceous lahar deposits on King George Island may record catastrophic melting of local ice caps during volcanic eruptions (Birkenmajer, 1991, 1992).

DRAKE PLATE"~. c.~'~

,,..,, ,4.7>.,

o

kl

SOUTHSHETLANDIS.

ANTARCTICPENINSULA

,.

,,oo

C'

10 km [ ~ INTRUSIVES ~ 1 OCEANICCRUST ~

~ ] It~SOZOIC/CENOZOICSTRATA& VO_CANICS

LATECENOZOICTO RECENTGLAClALLY-INFI.IJE NCED MARINE SEDIMENTS& ~ ' i ~ D VOLCANICS

Fig. 26.7. Physiography (Top) and cross-section (Bottom) of Bransfield Strait backarc basin, Antarctica. From various sources. See Fig. 24.2 for location. Compare with Fig. 10.6.

Younger (Oligocene) glaciomarine strata of the Polonez Cove Formation record the rapid progradation of submarine mass flows along the flanks of a glaciated volcanic arc (Porebski and Gradzinski, 1987). Andesitedacite lavas are interbedded with glaciomarine diamictites that pre-date the Early Miocene (22-20 Ma) and could be of mid- or Early Oligocene in age (Birkenmajer, 1991). Younger glaciations of Late Oligocene and Early Miocene age are recorded by glaciomarine deposits of the Legru and Melville glaciations, respectively (Birkenmajer, 1991). These strata contain a rich fossil and trace fossil record and were deposited in neritic to bathyal water depths (Gazdzicki et al., 1982; Gazdzicki and Pugaczewska, 1984); these strata are directly analagous with glaciomarine strata deposited on the margins of rapidly uplifting coastal mountains around the Late Miocene Gulf of Alaska (Yakataga Formation; Section 26.1). 27. E A R T H ' S G L A C I A L R E C O R D A N D ITS GEODYNAMIC SETTING; A SYNTHESIS

Figure 27.1 is a summary of the likely tectonic setting of Earth's glaciated basins. The bulk of the glacial record of the planet appears to be preferentially preserved in the fills of rifted, extensional basins where localised uplift of rift margins created the conditions necessary for regional glaciation. Uplift was complementary to strong basinal subsidence which allowed the accomodation of a glacially-influenced, predominantly marine sedimentary record. Excellent "modern" analogues can be found in the Transantarctic rift system (Fig. 24.2) and along the northwest European and eastern North America continental margins (Figs. 19.8, 19.9). Further detailed comparison could also be made with Early and Late Proterozoic rift settings in North America (Figs. 8.3, 12.2), Europe (Fig. 11.1, 11.2) and Australia (Fig. 12.4) and with Late Palaeozoic intracratonic basins (Fig. 16.2). The most common "glacial" facies types preserved in the rock record are not "tillites"

209

EARTH'S GLACIAL R E C O R D A N D ITS TECTONIC SETTING

and the other ice-contact terrestrial sediments deposited directly from glaciers but marine "rain out" and debris flow diamictites, and associated turbidite facies, that locally attain very considerable thicknesses. The volumetric importance of successions such as these in many glaciated basins testifies to the principal sedimentological role of glaciers in Earth history which is to deliver large volumes of poorly-sorted glaciclastic debris to basin margins where it is then reworked and redeposited by non-glacial depositional processes. Only a very small proportion of terrestrial ice contact sediments are preserved in the glacial record; probably much less than 10% of the total volume of glaciclastic strata. An observation that can be made directly from Fig. 27.1 is the relative scarcity in the rock record of glacial strata that were deposited in active margin basins. Certainly this

L. CENOZOIC

is the result of selective destruction of the record along subducting margins. However, as the example of the late Cenozoic Gulf of Alaska suggests (Section 26.1), the influx of substantial volumes of glaciclastic sediment can "choke" a subducting trench system and lead to the development of thick accretionary wedges which have excellent long term preservation potential. Nonetheless, it is also the case in such settings that because of active tectonism and steep slopes glaciclastic sediment stands a much greater chance of being completely reworked such that a glacial contribution may eventually be difficult to recognise in the basin fill. This situation characterises many Late Proterozoic basins where a glacial influence is commonly suggested by poorly-sorted mass flow deposits but where a glacial source is impossible to confirm (Section 10). The relative scarcity of active margin glacial deposits in KLESZCZOWBASIN. POLAND(FIG.25.2) NORTHSEA RIFT (FIGS. 19.9,24.1)

BRANSFIELDSTRAIT (FIG. 26.8)

GULF OF ALASKA(FIGS. 26.3,26.5)

i

ROSSBEA/WEDOELLSEAI

<3GMa

RIFT (FIG. Z4.2) J ALASKAN INTERIOR

EASTERNCANAI~AN CONTINENTALMARGIN (FIGS. 19.8,23.1) N.W.EUROPEAN CONTINENTALMARGIN (FIG. 19.9)

4FIG. zs.l~

L. PALAEOZOIC 350

-

2 5 0 Ma

KAROOBASIN,

PALAEO*P.~QFIC MARGINOF GONOWANA: EASTERNAUSTRAUA(FIG. 16.15, 16.19) ANTARCTICA(FIG. 16.18)

S. AFRICA

(FIGS. 16.9,16.1 Z)

ORDOVICIAN

WEST AFRICA ? CENTRAL SAUDI ARABIA

c. 4 0 0 Ma

L. PROTEROZOIC 8 0 0 - 5 5 0 Ma (FIG,

9.2)

KALAHARIBASIN (FIGS. 16.9,16.12) ARABIAN PENINSULA (FIGS. 16.19.16.20) PARANA BASIN. BRAZIL (FIGS. 16.8, 1B.10) INDIAN BASINS(FIG 16.4 AUSTRALIAN INTERIOR BASINS (FIGS. 16.~116.4

(FIG, I 5.2) DA/dARAMOBILEBELT (SECT. 10,2) ARABIANSHIELD (SECT. 10,3) PARAGUAyoARAGUAIAFOLD BELT (SECT. 10.3) TIDDILi~E BARN, NORTHAFRICA(SECT, 10.1)

GABI~ERS F'M, NR.D. 80~TON BAY GROUP (FIG. 10.6)

PALAEOATLANTIC MARGINOF LAURENTiA (FIGS, 11.1,11.2) PALAEO~A~'IFIC MARGINOF LAURENTIA (FIGS. 9.1t9.Z112.4~

BAKOYE GR., JBELIAT GP., W. AFRICA (FIGS. 10.1,10.2, 10.3)

E. PROTEROZOIC

HURONIAN SUPERGROU~ GO~A~ANOAFM (FIG. 8,3)

2 1 0 0 - 1 8 0 0 Ma

ARCHEAN

W~SWATEB3~O BA~,N

S, AFR~A (FIG. 7.1)

>2500Ma

? FOREARC TRE?

f)~/__~ ~

BACKARC

FORELAND

INTRACRATONIC/ PASSIVE MARGIN AULACOGENIC

.........

Fig. 27.1. Tectonic setting of earth's glacial record: Note preservational bias toward intracratonic and passive margin basins. It is in these settings that ice sheets are more likely to develop and glaciclastic strata be preserved (see text).

210 the ancient record may also reflect the generally smaller size of the cordilleran glaciers that typically develop in such settings. Rifted passive margins characterised by extensive area of uplifted plateaux, are clearly the preferred sites for the formation of larger ice masses; Figure 27.1 suggests that this tectonic precondition has been a recurrent feature of Earth's long glacial record and is very clearly a characteristic of Pleistocene glaciations. It has not been the intent of this paper to review all the many different possible models that have been proposed for the origin and timing of glaciations in Earth history. The primary purpose has been to emphasise the critical role of tectonic setting in creating the necessary conditions for glaciation and the preservation of a sedimentary record. It can be suggested that episodes of major glaciations in Earth's history are the result of the relationship between the initiation of regional ice covers in areas of active tectonism and subsequent global feedback mechanisms involving geochemical controls or extraterrestrial "astronomical" variables that allow (or restrict) glacier expansion (Table 21.1). Perennial snow fields or ice masses initiated adiabatically on elevated continental surfaces may evolve into regional ice sheet complexes in response to CO 2 induced and "Milankovitch" feedback mechanisms. Astronomical mechanisms can promote regional glaciation only if global temperatures have been lowered to threshold values by tectonicallydriven uplift; the so-called "insolation-topographic" theory (Emiliani and Geiss, 1959). It has long been recognised that Milankovitch astronomical forcing alone is insufficient to generate extensive glaciations; if it were otherwise then the record of ice sheets in Earth history would not be so sporadic (Fig. 6.2; Section 20.3). Section 20 has examined some of the processes that may amplify Milankovitch climate forcing. Prominent among these are changes in ocean circulation which in turn reflect long term plate tectonic reorganisations.

N.EYLES Recognition of the central importance of the regional tectonic setting to initiation of ice covers and preservation of glaciclastic strata carries important implications for the interpretation of depositional environments and controls on sedimentation. The traditional approach has been that of climatostratigraphy where successions of diamictites and associated facies are interpreted directly in terms of climatically-driven cycles of ice advance and retreat. This approach often presupposes direct sedimentation from glaciers in a terrestrial setting. A variant of the climatostraigraphic approach is that stratigraphic successions are the result of glacially-induced changes in global sea level. Many ancient glacial successions are part of thick shallowing-upward marine successions or alternatively, are capped by deeper water, fine-grained deposits. These are commonly interpreted with reference to glacio-isostatic cycles of crustal loading and rebound or glacio-eustatic changes in sea-level recording either postglacial climatic amelioration and rising sea levels or increasing ice volumes and sea-level drawdown (e.g. Eisbacher, 1985; Nystuen, 1985; Lopez Gamundi, 1989: Vaslet, 1990; Alvarez and Maurin, 1991; Deynoux, 1991; Young, 1992). However, given the strong bias of the glacial record to divergent plate margins, many such successions may be more satisfactorily interpreted in terms of a tectonostratigraphic approach emphasing source area uplift, an influx of poorly-sorted glaciclastic debris to a rapidly subsiding rifted basin followed by a phase of thermal subsidence and shale deposition (Fig. 8.3). A further implication arising from the tectonostratigraphic paradigm is that the age of the glaciclastic component in any one sedimentary basin does not axiomatically constrain the timing and duration of climate change merely the timing of sedimentpreserration. As is increasingly evident from the study of Late Palaeozoic basins in South America, southern Africa, India and Australia, a long phase of glacial erosion and non-deposition may precede basin subsi-

211

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SETTING

dence and accomodation of glacially-influenced strata (Section 16.10). This tectonic overprinting will complicate any effort at defining the precise timing of glaciation from one area to another and at establishing ice volumes. A similar problem, though on a very different time scale, emerges from investigation of Pleistocene basins where it can be shown that the depositional record is biased toward the shortlived closing phases of any one glacial cycle when sediment and water discharges are maximised (e.g. Section 24.1). These observations simply confirm the old saw about there being "more gaps than record". It is ironic that at a time of intense discussion of the role of global geochemical and astronomical controls on glaciations (e.g. Socci, 1992), combined with a much improved ability to model such influences, there is a steadily increasing body of data that implicate plate tectonic setting as a first-order control on the initiation of glaciation, on the preservation of a glacial record and on facies successions. The rock record, not computer models, is the primary source of data regarding ancient climates and as yet, the detailed stratigraphy, palaeogeography and geodynamic setting of most glaciated basins is poorly constrained. These basins offer great opportunities for field research aimed at identification of the relative roles of "local" tectonically-induced controls on glaciation from "global" and "astronomical" influences. The key lies in comprehensive sedimentological, stratigraphic, structural, palaeontological and geochemical investigations in concert with improved geochronological frameworks. ACKNOWLEDGEMENTS This paper is a contribution to International Geological Correlation Program #260 {Earth's Glacial Record}. Membership in the program has been a major stimulus to this project, the bulk of which was written up at the petroleum exploration laboratory of The

Brazilian State Oil Company, Petrobras, in Curitiba, Brazil and at Saudi Aramco in Dhahran, Saudi Arabia. Final writing was completed at Yaciementos Petroliferos Fiscales Bolivianos in Santa Cruz, Bolivia. The research on which this paper is based has been made possible by the generosity of the Natural Science and Engineering Research Council of Canada, the donors of the Petroleum Research Fund of the American Chemical Society, the International Geological Correlation Program and NATO. I wish to thank Almerio B. Franca, for providing a stimulating working environment in Brazil, and his family for their very generous hospitality. I enjoyed very useful discussions with Edison Emiliani and Sven Wolff at Petrobras, Tony Rocha-Campos and Paulo dos Santos at the University of Sao Paulo, Jan Visser at Bloemfontein in South Africa and Tom Stump, George Lyndts and Tom Connelly at Saudi Aramco. Discussions in Bolivia with Oscar Lopez Paulsen, Fernando Wiens and Tony Tankard were particularly informative as were those with Ian Fairchild, Max Deynoux, Grant Young, John Crowell, Vic Gostin, Ross Powell, Eugene Domack, Gus Bonorino and Julia Miller. A special vote of thanks goes to Mike Doughty who draughted the figures. Finally, I am indebted to my wife Carolyn for her toleration, ongoing discussions during the course of this and many other projects and for reading the first drafts of the manuscript. I thank her and our children, Christopher and Claire, for their patience. REFERENCES Aalto, R.K., 1971. Glacial marine sedimentation and stratigraphy of the Toby Conglomerate(Upper Proterozoic), southeastern British Columbia, northwestern Idaho and northeastern Washington. Can. J. Earth Sci., 8: 753-787. Abouzakhm, A.G. and Tarling, D.H., 1975. Magnetic anisotropy and susceptibility of Late Precambrian tillites form northwestern Scotland. J. Geol. Soc. London, 131: 647-652. Acharyya, S.K., 1975. Tectonic framework of sedimentation of the Gondwana of the Eastern Himalayas,

212 India. In: K.S.W. Campbell (Editor), Gondwana Geology. Aust. Natl. Univ. Press: pp. 663-674. Ad-Dabbagh, M.E. and Rogers, J.J., 1983. Depositional environments and tectonic significance of the Wajid Sandstone of Southern Saudi Arabia. J. Afr. Earth Sci., 1: 47-57. Adams, J.M., Faure, H. and Petit-Maire, N., 1992. Methane and Milankovitch cycles. Nature, 335: 214. Adh6mar, J.A., 1842. Revolutions de lamer. Privately published, Paris. Agassiz, L., 1840. Etudes sur les glaciers. Privately published, Neuchatel. Aggarval, H.R. and Oberbeck, V.R., 1992. Mathematical modeling of impact deposits and the origin of tillites. Eos, Abstr. Suppl., 73(43): 325. Aitken, J.D., 1991. Two late Proterozoic glaciations, MacKenzie Mountains, northwestern Canada. Geology, 19: 445-448. Aksenov, Y.M., Keller, B.M. and Sokolov, B.S., 1978. A general scheme for upper Precambrian stratigraphy of the Russian platform. Int. Geol. Rev., 22: 444-457. Aksu, A.E. and Hiscott, R.N., 1989. Slides and debris flows on the high-latitude continental slopes of Barfin Bay. Geology, 17: 885-888. Aksu, A.E. and Hiscott, R.N., 1992. Shingled Quaternary debris flow lenses on the north-east Newfoundland Slope. Sedimentology, 39: 193-206. Alonso, B., Anderson, J.B., Diaz, J.I. and Bartek, L.R, 1990. Pliocene-Pleistocene seismic stratigraphy of the Ross Sea: Evidence for multiple ice sheet grounding episodes. Antarctic Research Series, Am. Geophys. Union, 57, pp. 93-103. A1-Laboun, A.A., 1986. Stratigraphy and hydrocarbon potential of the Paleozoic succession in both Tabuk and Widyan Basins, Arabia. Am. Assoc. Pet. Geol. Bull., 40: 373-397. AI-Laboun, A.A., 1987. Unayzah Formation: A new Permian-Carboniferous unit in Saudi Arabia. Am. Assoc. Pet. Geol. Bull., 71: 29-38. Al-Laboun, A.A., 1988. The distribution of Carboniferous-Permian siliclastic rocks in the greater Arabian basin. Geol. Soc. Am. Bull., 100: 362-373. Alavi, M., 1991. Sedimentary and structural characteristics of the paleo-Tethys remnants in northeastern Iran. Geol. Soc. Am. Bull., 103: 938-992. Albritton, C.C., 1989. Catastrophic Episodes in Earth History. Chapman and Hall, 221 pp. Aleinikov, A.L. et al., 1980. Dynamics of the Russian and West Siberian platforms. In: A.W. Bally, P.L. Bender, T.R. McGetchin and R.I. Walcott (Editors), Dynamics of Plate Interiors. Am. Geophys. Union and Geol. Soc. Am. Geodynamic Series, 1, pp. 5371. Alley, R.B., Blankenship, D.D., Bentley, C.R. and Rooney, S.T., 1987. Till beneath Ice Stream B; 3,

N. EYLES Till deformation: Evidence and implications. J. Geophys. Res., 92: 8921-8929. Alvarez, P. and Maurin, J.-C., 1991. Sedimentation and tectonics in the Upper Proterozoic basin of Comba (Congo): sequence stratigraphy of the West-Congolian Supergroup and strike-slip damping model related to the Pan-African orogenesis. Precambrian Res., 50: 137-171. Ambrose, J.W., 1964. Exhumed paleoplains of the Precambrian shield of North America. Am. J. Sci., 262: 817-857. Amos, C.L., 1990. Modern sedimentary processes. In: J.J. Keen and G.L. Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geol. Surv. Can., 2: 611-673. Anderson, J.B., 1983. Ancient glacial-marine deposits: Their spatial and temporal distrubution. In: B.F. Molnia (Editor), Glacial-Marine Sedimentation. Plenum, New York, pp. 3-92. Anderson, J.B., Brake, C., Domack, E.W., Myers, N. and Wright, R., 1983. Development of a polar glacial-marine sedimentation model from Antarctic Quaternary deposits and glaciological information. In: B.F. Molnia (Editor), Glacial Marine Sedimentation. Plenum, New York, pp. 233-264. Anderson, J.B. and Molnia, B.F., 1989. Glacial-Marine Sedimentation. Short Course in Geology, 9. Am. Geophys. Union, Washington, D.C., 127 pp. Anderson, J.B. and Ashley, G.M. (Editors), 1991. Glacial Marine Sedimentation; Paleoclimatic Significance. Geol. Soc. Am. Spec. Pap., 261, 232 pp. Anderson, J.B., Kennedy, D.S., Smith, M.J. and Domack, E.W., 1991. Sedimentary facies associated with Antarctica's floating ice masses. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation; Paleoclimatic Significance. Geol. Soc. Am. Spec. Pap., 261. Anderson, J.B. and Bartek, L.R., 1992. Cenozoic glacial history of the Ross Sea revealed by intermediate resolution seismic reflection data combined with drill site information. Antarctic Research Series. Am. Geophys. Union, 56, pp. 231-263. Anderson, J.M., 1975. Turbidites and arthropod trackways in the Dwyka glacial deposits (Early Permian) of Southern Africa: Trans. Geol. Soc. S. Afr., 78: 265-273. Anderson, J.M., 1981. The Umfolozia arthropod trackways in the Permian Dwyka and Ecca Series of South Africa. J. Palaeontol., 55: 84-108. Anderson, R.Y., 1982. A long geoclimatic record from the Permian. J. Geophys. Res., 87: 7285-7290. Anderson, R.Y., 1984. Orbital forcing of evaporite sedimentation. In: A.L. Berger, J. Imbrie, J. Hays, G. Kukla and B. Salzman (Editors), Milankovitch and Climate, part 1. Reidel, Dordrecht, pp. 147-162. Anderton, R., 1982. Dalradian deposition and the late

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SETTING

Precambrian-Cambrian history of the N. Atlantic region: a review of the early evolution of the Iapetus Ocean. J. Geol. Soc. London, 139: 421-431. Anderton, R., 1985. Clastic facies models and facies analysis. In: P.J. Brenchley and B.P.J. Williams (Editors), Sedimentology--Recent Developments and Applied Aspects. Blackwell, Oxford, pp. 31-48. Andrews, J.T., 1978. Sea level history of Arctic coasts during the Upper Quaternary. Prog. Phys. Geol., 2: 375-407. Andrews, J.T., 1987. Late Wisconson glaciation and deglaciation of the Laurentide Ice Sheet. In: W.F. Ruddiman and H.E. Wright jr. (Editors), North America and Adjacent Oceans During the Last Deglaciation. Geol. Soc. Am. The Decade of North American Geology Project. The Geology of North America. K-3, pp. 13-37. Andrews, J.T., 1990. Fiord to deep sea sediment transfers along the northeastern Canadian continental margin: models and data. Geogr. Phys. Quaternaire, 44: 55-70. Andrews, J.T., 1992. A case of missing water. Nature, 358: 281. Antevs, E., 1925. Retreat of the last ice-sheet in eastern Canada. Dep. Mines Geol. Surv. Mem., 146, 141 pp. Antevs, E., 1953. Geochronology of the deglacial and neothermal ages. J. Geol., 61: 195-230. Armstrong, R.A., Compstone, W., Retief, E.A., Williams, I.S. and Welke, H.J., 1991. Zircon ion microprobe studies bearing on the age and evolution of the Witwatersrand triad. Precambrian Res., 53: 243-266. Ashley, G.M., 1975. Rhythmic sedimentation in glacial Lake Hitchcock, Massachusetts-Connecticut. In: A.V. Jopling and B.C. McDonald (Editors), Glaciofluvial and Glaciolacustrine Sedimentation. Soc. Econ. Paleontol. Mineral. Spec. Publ., 23: 304320. Ashley, G.M., Boothroyd, J.C. and Borns, H.W.Jr., 1991. Sedimentology of Late Pleistocene (Laureatide) deglacial-phase deposits, eastern Maine; an example of a temperate marine grounded ice-sheet margin. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation. Geol. Soc. Am. Spec. Pap., 261: 107-124. Ashley, G.M., Shaw, J. and Smith, N.D., 1985. Glacial Sedimentary Environments. Soc. Econ. Paleontol. Mineral. Short Course Notes, 16, 246 pp. Astakhov, V.I. and Isayeva, L.L., 1988. The "ice hill"; an example of retarded deglaciation in Siberia. Quat. Sci. Rev., 7: 29-40. Axelrod, D.I., 1984. An interpretation of Cretaceous and Tertiary biota in polar regions. Palaeogeogr. Palaeoclimatol. Palaeoecol., 45: 105-147. Bahcall, J.N. and Ulrich, R.K., 1988. Solar models,

213

neutrino experiments and helioseismology. Rev. Mod. Phys., 60: 297-372. Bailey, R.A., Huber, K.N. and Curry, R.R. 1990. The diamicton at deadman Pass, Central Sierra Nevada, California: A residual lag and colluvial deposit, not a 3 Ma glacial till. Bull. Geol. Soc. Am., 102: 11651173. Balkwill, H.R., 1987. Labrador Basin: Structural and Stratigraphic Style. In: C. Beaumont and A.J. Tankard (Editors), Basin Forming Mechanisms. Can. Soc. Pet. Geol. Mem., 12: 17-43. Balkwill, H.R., McMillan, N.J., MacLean, B., Williams, G.L. and Srivastava, S.P., 1990. Geology of the Labrador Shelf, Baffin Bay and Davis Strait. In: M.J. Keen and G.L. Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geology of Canada, part 2. Geol. Surv. Can., pp. 293-350. Ball, R., 1891. The Cause of an Ice Age. Kegan Paul, Trench, Trubner and Co., London, 180 pp. Bally, A.W., 1989. Phanerozoic basins of North America. In: A.W. Bally and A.R. Palmer (Editors), The Geology of North America: An Overview. Decade of North American Geology. Geol. Soc. Am., pp. 397-446. Banerjee, I., 1973. Sedimentology of Pleistocene glacial varves in Ontario, Canada. Geol. Surv. Can. Bull., 226, part A: 1-44. Banks, M.R. and Clarke, M.J., 1987. Changes in the geography of the Tasmania Basin in the Late Paleozoic. In: G.D. McKenzie (Editor), Gondwana 6. Am. Geophys. Union, Geophys. Monogr., 41: 1-14. Bannerjee, T., 1992. Talchir Sedimentation: A Study from Four Selected Gondwana Basins in Easter India. Ph.D. Thesis, Indian Inst. Technol., Kharagpur, 188 pp. Unpublished Bard, E., Hamelin, B. and Fairbanks, R.G., 1990. U-Th ages obtained by mass spectrometry in corals from Barbados; sea-level during the past 130,000 years. Nature, 346: 456-458. Barnes, C.R., 1986. The faunal extinction event near the Ordovician-Silurian boundary: a climatically-induced crisis. In: O. Walliser (Editor), Global BioEvents. Springer, Berlin, pp. 121-126. Barnes, C.R. and Williams, S.H., 1991. (Editors), Advances in Ordovician Geology. Geol. Surv. Can. Mem., 90-9. Barnes, P.W. and Lien, R., 1988. Icebergs rework shelf sediments to 500 m off Antarctica. Geology, 16: 1130-1133. Barrett, P.J., 1982. History of the Ross Sea region during the deposition of the Beacon Supergroup 400-180 million years ago. J.R. Soc. N.Z., 11: 447458. Barrett, P.J. (Editor), 1989. Antarctic Cenozoic history from the CIROS-1 drillhole, McMurdo Sound. Dep. Sci. Ind. Res. Bull., 245, 254 pp.

214 Barrett, P.J., Elston, D.P., Harwood, D.M., McKelvey, B.C. and Webb, P.N., 1987. Mid-Cenozoic record of glaciation and sea level change on the margin of the Victoria Land basin, Antarctica. Geology, 15: 634637. Barrett, P.J., Hambrey, M.J., Harwood, D.M., Pyne, A.R. and Webb, P.N., 1989. Synthesis. In: P.J. Barrett (Editor), Antarctic Cenozoic History from the CIROS-1 Drillhole, McMurdo Sound. Dep. Sci. Ind. Res. Bull., 245: 241-251. Barrett, P.J., Adams, C.J., Mclntosh, W.C., Swisher, C.C., III and G.S.Wilson, 1992. Geochronological evidence supporting Antarctic deglaciation three million years ago. Nature, 359: 816-818. Barrett, S.F. and Isaacson, P.E., 1987. Devonian paleogeography of South America. Can. Soc. Pet. Geol. Mem., 14: 655-667. Barron, E.J., 1985. Explanations of the Tertiary global cooling trend. Palaeogeogr. Palaeoclimatol. Palaeoecol., 50: 45-61. Barron, E.J., 1989. Studies of cretaceous climate. In: A. Berger, R.E. Dickinson and J.W. Kidson (Editors), Understanding Climate Change. Geophys. Monogr. Am. Geophys. Union, 52: 149-157. Barron, E.J., Thompson, S.L. and Schneider, S.H., 1981. An ice-free Cretaceous? Results from climate model simulations. Science, 212: 501-508. Barron, E.J., et al., 1989. Proceedings, Ocean Drilling Program, Initial Reports Ocean Drilling Program, 119. College Station, Texas, 942 pp. Barron, E.J., 1992. Lessons from past climates. Nature, 360: 533. Basu, T.N. and Shrivastava, B.B.P., 1981. Structure and tectonics of Gondwana basins of Peninsula India. In: M.M. Cresswell and P. Vella (Editors), Gondwana 5. A.A. Balkema, Rotterdam, pp. 177183. Bates, R.L. and Jackson, J.A., 1987. Glossary of Geology. Am. Geol. Inst., Alexandria, 754 pp. Battersby, D.G., 1976. Cooper Basin gas and oil fields, in Economic Geology of Australia and Papau New Guinea. Monogr. Ser # 5-8. Aust. Inst. Min. Metal., Melbourne, pp. 321-368. Beard, J.H., Sangree, J.B. and Smith, L.H., 1982. Quaternary chronology, paleoclimate, depositional sequences and eustatic cycles. Am. Assoc. Pet. Geol. Bull., 66: 158-169. Beckinsale, R.D., Reading, H.G. and Rex, D.C., 1976. Potassium-Argon ages for basic dykes from East Finnmark: stratigraphic and structural implications. Scott. J. Geol., 12: 51-65. Behrendt, J.C. and Cooper, A., 1991. Evidence of rapid Cenozoic uplift of the shoulder escarpment of the Cenozoic West Antarctic rift system and a speculation on possible climate forcing. Geology, 19: 360-363.

N. EYLES Bell, C.M., 1981. Soft-sediment deformation of sandstone related to the Dwyka Glaciation in S. Africa. Sedimentology, 28: 321-329. Bell, M. and Laine, E.P., 1985. Erosion of the Laurentide Region of North America by glacial and glaciofluvial processes. Quat. Res., 23: 154-174. Bell, R.T., 1968. Study of the Hurwitz Group, District of Keewatin. Geol. Surv. Can. Pap., 68-1A: 116-120. Benson, J.M., 1981. Late Carboniferous and Early Permian tillites north of Newcastle, New South Wales. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 480-484. Benton, M.J., 1991. Polar dinosaurs and ancient climates. Trends Ecol. Evol., 6: 28-30. Berger, A. and Loutre, M.F., 1989. Pre-Quaternary Milankovitch frequencies. Nature, 332: 133. Berger, A., Loutre, M.F. and Dehant, V., 1989. Influence of the changing lunar orbit on the astronomical frequencies of pre-Quaternary insolation patterns. Palaeoceanography, 4: 555-564. Berger, A. and Loutre, M.F., 1992. Astronomical solutions for paleoclimate studies over the last 3 million years. Earth Planet. Sci. Lett., 111: 369-382. Berner, R.A., 1990. Atmospheric carbon dioxide over Phanerozoic time. Science, 249: 1382-1386. Berner, R.A., 1992. Palaeo-CO 2 and climate. Nature, 358:114. Beuf, S., Biju-Duval, B., Stevaux, J. and Kulbicki, G., 1966. Ampleur des glaciations "siluriennes" au Sahara, leurs influences et leurs consequences sur la sedimentation. Rev. Inst. Fr. Pet., 11: 363-381. Beuf, S., Biju-Duval, B., de Charpal, O., Rognon, P., Gariel, O. and Bennacef, A., 1971. Les gres du Paleozoique inferieur au Sahara. Science et Technique du Petrole, 18. Inst. Fr. Pet., Editions Technip, 464 pp. Beukes, N.J. and Cairncross, B. 1991. A lithostratigraphic-sedimentological reference profile for the Late Archaean Mozaan Group, Pongola Sequence: application to sequence stratigraphy and correlation with the Witwatersrand Supergroup. S. Afr. J. Geol., 94: 44-69. Beydoun, Z.R., 1991. Arabian plate hydrocarbon geology and potential--A plate tectonic approach. Studies in Geology, 33. Am. Assoc. Pet. Geol., Tulsa, Oklahoma, 77 pp. Bidgood, D.E.T. and Harland, W.B., 1961. Palaeomagnetism in some East Greenland sedimentary rocks. Nature, 189: 633-634. Biju-Duval, B., Deynoux, M. and Rognon, P., 1981. Late Ordovician tillites of the Central Sahara. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 99-107. Binda, P.L. and Van Eden, J.G., 1972. Sedimentologi-

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETFING

cal evidence on the origin of the Precambrian great conglomerate (Kundelungu Tillite), Zambia. Palaeogeogr. Palaeoclimatoi. Palaeoecol., 12: 151-168. Birkenmajer, K., 1991. Tertiary glaciation in the South Shetland Islands, West Antarctica: Evaluation of data. Medd. Gronl. Geosci., 21: 629-632. Birkenmajer, K., 1992. Cenozoic glacial history of the South Shetland Islands and northern Antarctic Peninsula. In: J. Lopez-Martinez (Editor), III Congreso Geoiogico de Espana, Symposios T3, Salamanca, Espana, pp. 251-260. Bischof, J., Koch, J., Kubisch, M., Spielhagen, R.F. and Thiede, J., 1990. Nordic Seas surface ice drift reconstructions: evidence from ice rafted coal fragments furing oxygen isotope stage 6. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glacimarine Environments: Processes and Sediments. Geol. Soc. Spec. Publ., 53: 235-251. Bjorlykke, K., 1966. Sedimentary petrology of the sparagmites of the Rena District, S. Norway. Nor. Geol. Unders., 238: 5-53. Bjorlykke, K., 1967. The Eocambrian Reusch Moraine at bigganjargga and the geology around Varangerfjord Northern Norway. Nor. Geol. Unders., 251: 18-44. Bjorlykke, K., 1978. The eastern marginal zone of the Caledonide Orogeny in Norway, in Caledonian-Appalachian Orogeny of the North Atlantic Region. Geol. Surv. Can. Pap., 78-13. IGCP Project 27, Norwegian Contribution No. 5F, pp. 49-55. Bjorlykke, K., 1985. Glaciations, preservation of their sedimentary record and sea level changes. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 197-207. Bjorlykke, K., Elvsborg, A. and Hoy, T., 1976. Late Precambrian sedimentation in the central sparagmite basin of south Norway. Nor. Geol. Tidsskr., 56: 233-290. Bjorslev Nielsen, O., Sorensen, S., Thiede, J. and Skarbo, O., 1986. Cenozoic differential subsidence of North Sea. Am. Assoc. Pet. Geol. Bull., 70: 276-298. Blackwelder, E., 1930. Striated boulders as evidence of glacial action. Geol. Soc. Am. Bull., 41: 154. Blackwelder, E., 1931. Pleistocene glaciation in the Sierra Nevada and Basin Ranges. Bull. Geol. Soc. Am., 42: 865-922. Blandford, H.F. and Blandford, W.T., 1859. Indian Geological Survey Memoir, 1. Blondeau, K.M. and Lowe, D.R., 1972. Upper Precambrian glacial deposits of the Mount Rogers Formation, Central Appalachians, U.S.A. In: 24th Int. Geol. Congr. Proc. Section 1., Montreal, pp. 325332. Boardman, D.R. and Heckel, P.H., 1989. Glacial-eustatic sea-level curve for early Late Pennsylvanian sequences in north-central Texas and biostrati-

215

graphic correlation with curve for mid-continent North America. Geology, 17: 802-805. Bond, G., 1981. Late Paleozoic (Dwyka) glaciation in the Middle Zambezi region. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 55-57. Bond, G. et al., 1992. Evidence for massive discharges of ice bergs into the North Atlantic during the last glacial period. Nature, 360: 245-249. Bond, G.C., Nickeson, P.A. and Devlin, W.J., 1984. Breakup of a supercontinent between 625 Ma and 555 Ma: New evidence and implications for continental histories. Earth Planet. Sci. Lett., 70: 325345. Bond, G.C., Kominz, M.A. and Grotzinger, J.P., 1988. Cambro-Ordovician eustasy: Evidence from geophysical modelling of subsidence in Cordilleran and Appalachian passive margins, In: K.L. Kleinspehn and C. Paola (Editors), New Perspectives in Basin Analysis. Springer, New York, pp. 129-160. Boothroyd, J.C. and Nummedal, D., 1978. Proglacial braided outwash: a model for humid alluvial fan deposits. In: A.D. Miall (Editor), Fluvial Sedimentology. Can. Soc. Pet. Geol. Mem., 5: 641-668. Borns, H.W.J. and Hall, B.A., 1969. Mawson "Tillite" in Antarctica: preliminary report of a volcanic deposit of Jurassic age. Science, 166: 870-872. Boulton, G.S., 1972. Modern arctic glaciers as depositional models for former ice sheets. J. Geol. Soc. London, 128: 361-393. Boulton, G.S., 1975. Processes and patterns of subglacial sedimentation: a theoretical approach. In: A.E. Wright and F. Moseley (Editors), Ice Ages: Ancient and Modern. Seel House, Liverpool, pp. 7-42. Boulton, G.S., 1976. The origin of glacially fluted surfaces-observations and theory. J. Glaciol. 17: 287310. Boulton, G.S., 1978. Boulder shapes and grain-size distributions of debris as indicators of transport paths through a glacier and till genesis. Sedimentology, 25: 773-799. Boulton, G.S., 1979. Processes of glacier erosion on different substrata. J. Glaciol., 23: 15-38. Boulton, G.S., 1987. A theory of drumlin formation by subglacial sediment deformation. In: J. Menzies and J. Rose (Editors), Drumlin Symposium. Balkema, Rotterdam, pp. 25-80. Boulton, G.S., 1990. Sedimentary and sea level changes during glacial cycles and their control on glacimarine facies architecture. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glacimarine Environments: Processes and Sediments. Geol. Soc. Spec. Publ., 53: 15-52. Boulton, G.S., Dent, D.L. and Morris, E.M., 1974.

216 Subglacial shearing and crushing and the role of water pressures in tills from south-east Iceland. Geogr. Ann., 56: 153-145. Boulton, G.S. and Eyles, N., 1979. Sedimentation by valley glaciers; a model and genetic classification. In: C.H. Schluchter (Editor), Moraines and Varves. Balkema, Rotterdam, pp. 11-23. Boulton, G.S. and Deynoux, M., 1981. Sedimentation in glacial environments and the identification of tills and tillites in ancient sedimentary sequences. Precambrian Res., 15: 397-422. Boulton, G.S. and Hindmarsh, R.C.A., 1987. Sediment deformation beneath glaciers: Rheology and geological consequences. J. Geophys. Res., 92: 9059-9082. Boulton, G.S. and Clark, C.D., 1990. A highly mobile Laurentide ice sheet revealed by satellite images of glacial lineations. Nature, 346: 813-817. Bouma, A.H., 1964. Recent and ancient turbidites. Geol. Mijnbouw, 43: 375-379. Bouma, A.H., Normark, W.R. and Barnes, N.E. (Editors), 1984. Submarine Fans and Related Turbidite Systems. Frontiers in Sedimentary Geology. Springer, New York, 351 pp. Boyce, J. and Eyles, N., 1991. Drumlins carved by deforming till streams below the Laurentide Ice Sheet. Geology, 19: 787-790. Boyd, R., Scott, D.B. and Douma, M., 1988. Glacial tunnel valleys and Quaternary history of the outer Scotian Shelf. Nature, 333: 61-64. Boygle, J., 1993. The Swedish varve chronology--a review. Prog. Phys. Geogr., 17: 1-20. Boyle, E.A., 1988. Vertical oceanic nutrient fractionation and glacial/interglacial CO 2 cycles. Nature, 331: 55-56. Braakman, J.H., Levell, B.K., Martin, J.H., Potter, T.L. and Van Vliet, A., 1982. Late Palaeozoic Gondwana glaciation in Oman. Nature, 299: 48-50. Bradley, D.W. and Kusky, T.M., 1986. Geologic evidence for rate of plate convergence during the Taconic Arc-Continent collision. J. Geol., 94: 667681. Brady, P.V., 1991. The effect of silicate weathering on global temperature and atmospheric CO 2. J. Geophys. Res., 96: 18,101-18,106. Brathwaite, C.J.R., 1991. Dolomites, a review of origins, geometry and textures. Trans. R. Soc. Edinburgh Earth Sci., 82: 99-112. Brandt, K., 1986. Glacioeustatic cycles in the Early Jurassic?. Neues Jahrb. Geol. Palaontol. Monatsh., 5: 257-274. Brasier, M.D., 1982. Sea-level changes, facies changes and the late Precambrian-early Cambrian evolutionary explosion. Precambrian Res., 17: 105-123. Brasier, M.D., 1989. On mass extinction and the faunal turnover near the end of the Precambrian. In: S.K. Donovan (Editor), Mass Extinctions. Columbia University Press, New York, pp. 73-88.

N. EYLES Brathwaite, C.J.R., 1991. Dolomites, a review of origins, geometry and textures. Trans. R. Soc. Edinburgh Earth Sci., 82: 99-112. Brenchley, P.J., 1989. The Late Ordovician extinction. In: S.K. Donovan (Editor), Mass Extinctions. Columbia University Press, New York, pp. 104-132. Brenchley, P.J. Romano, M., Young, T.P. and Storch, P., 1991. Hinantian glaciomarine diamictites--evidence for the the spread of glaciation and its effect on Upper Ordovician faunas. Geol. Surv. Can. Pap., 90-9: 325-336. Broccoli, A.J. and Manabe, S., 1987. The influence of continental ice, atmospheric CO and land albedo on the climate of the last glacial maximum. Climate Dyn., 1:87-99 Brodzikowski, K., Gotowala, L., Kasza, L. and Van Loon, A.J., 1987. The Kleszcqow Graben (central Poland): reconstruction of deformational history and inventory of the resulting soft-sediment deformational structures. In: M.E. Jones and R.M.F. Preston (Editors), Deformation of Sediments and Sedimentary Rocks. Geol. Soc. London Spec. Publ., 29: 241-254. Brodzikowski, K. and Van Loon, A.J., 1991. Glacigenic Sediments. Developments in Sedimentology, 49. Elsevier, Amsterdam, 674 pp. Broecker, W.S. and Van Donk, J. 1970. Insolation changes, ice volumes and the O18 record in deep-sea cores. Rev. Geophys. Space Phys., 8: 169-197. Broecker, W.S. and Denton, G.H., 1989. The role of ocean-atmosphere reorganizations in glacial cycles. Geochim. Cosmochim. Acta, 53: 2465-2501. Broecker, W.S. and Peng, T.H., 1989. The cause of the glacial to interglacial atmospheric CO 2 change: A polar alkalinity hypothesis. Global Biogeochem. Cycles, 3: 215-239. Broecker, W.S., Bond, G. and Klas, M., 1990. A salt oscillator in the glacial Atlantic? 1. The concept. Paleoceanography, 5:469-477 Brooks, C.E.P., 1926. Climate Through the Ages. Ernest Benn Limited, New York, 395 pp. Brown, E.H., 1960. The Relief and Drainage of Wales. University of Wales Press, Cardiff. Bull, P.A., Culver, S.J. and Gardner, R., 1980. The nature of the Late Paleozoic glaciation in Gondwana as determined from an analysis of garnets and other heavy minerals. Discussion. Can. J. Earth Sci., 17: 282-283. Burbank, D.W., 1992. Causes of recent Himalayan uplift deduced from deposited patterns in the Ganges basin. Nature, 357: 680-683. Burke, K., Kidd, W.S.F. and Kusky, T.M., 1986. Archean foreland basin tectonics in the Witwatersrand, South Africa. Tectonics, 5: 439-456. Burns, S.J. and Matter, A., 1991. The strontium isotopic composition of Late Precambrian carbonates

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

(640-560 Ma) from Oman. Geol. Soc. Am. Annu. Meet. Abstr., 23: 97. Burret, C., Long, J. and Stait, B., 1990. Early-Middle Palaeozoic biogeography of Asian terranes derived from Gondwana. In: W.S. Mckerrow and C.R. Scotese (Editors), Palaeozoic Palaeogeography and Biogeography. Geol. Soc. Mem., 12: 163-174. Burton, R., Kendall, G.S.C. and Lerche, I., 1987. Out of our depth: on the impossibility of fathoming eustasy from the stratigraphic record. Earth-Sci. Rev., 24: 237-277. Butler, R.F., 1991. Paleomagnetism: Magnetic Domains To Geologic Terranes. Biackwell Scientific Press, Oxford, 319 pp. Caby, R. and Fabre, J., 1981. Tillites in the Latest Precambrian strata of the Touareg Shield (Central Sahara). In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 146-149. Cahen, L. and Lepersonne, J., 1981a. Late Paleozoic tillites of the Congo Basin in Zaire. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 43-47. Cahen, L. and Lepersonne, J., 1981b, Proterozoic diamictites of Lower Zaire. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 153-157. Caldeira, K. and Rampino, M.R., 1992. Mount Etna CO 2 may affect climate. Nature, 355: 401-402. Caldeira, K. and Kasting, J.F., 1992. Susceptibilty of the early Earth to irreversible glaciation caused by carbon dioxide clouds. Nature, 359: 226-228. Campana, B. and Wilson, R.B., 1955. Tillites and related glacial topography of South Australia. Eclogae Geol. Helv., 48: 1-30. Campbell, I.H. and Jarvis, G.T., 1984. Mantle convection and early crustal evolution. Precambrian Res., 26: 15-56. Caputo, M.V., 1985. Late Devonian glaciation in South America. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 291-317. Caputo, M.V. and Crowell, J.C., 1985. Migration of glacial centers across Gondwana during Paleozoic Era. Geol. Soc. Am. Bull., 96: 1020-1036. Caputo, M.V. and Silva, O.B., 1990. Sedimentacao e tectonica da Bacia do Solimoes. In: G.P.R. Gabaglia and E.J. Milani (Editors), Origem e Evolucao de Bacias Sedimentaires. Petrobras Petroleo Brasileiro, S.A., Rio de Janeiro, pp. 169-193. Car, D. and Ayres, L.D., 1991. A thick dacitic debris flow sequence, Lake of the Woods greenstone terrane, central Canada: resedimented products of Archean vulcanian, plinian and dome-building eruptions. Precambrian Res., 50: 239-260.

217

Card, K.D., 1978. Geology of the Sudbury-Manitoulin area, Districts of Sudbury and Manitoulin. Ontario Geol. Surv., pp. 238. Careaga, J.P., 1978. Prospeccion sismica en el noroeste de Bolivia. Rev. Tec. Yacimientos Pet. Fiscales Boliv., 7: 75-89. Carey, S.W. and Ahmad, N., 1960. Glacial marine sedimentation. In: G.O. Raasch (Editor), Geology of the Arctic, 2. Toronto University Press, pp. 865894. Carlson, P.R., 1989. Seismic reflection characteristics of glacial and glacimarine sediment in the Gulf of Alaska and adjacent fjords. Mar. Geol., 85: 391-416. Carlson, P.R. and Molnia, B.F., 1977. Submarine faults and slides on the continental shelf, northern Gulf of Alaska. Mar. Geotechnol., 2: 275-290. Carlson, P.R., Bruns, T.R., Molnia, B.F. and Schwab, W.C., 1982. Submarine valleys in the northeastern Gulf of Alaska: Characteristics and probable origin. Mar. Geol., 47: 217-242. Carlson, P.R., Bruns, T. and Fisher, M.A., 1990. Development of slope valleys in the glacimarine environment of a complex subduction zone, Northern Gulf of Alaska. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glacimarine Environments: Processes and Sediments. Geol. Soc. Spec. Publ., 53: 139-153. Carpenter, G., 1981. Coincident sediment slump/ clathrate complexes on the U.S. Atlantic continental slope. Geo-Mar. Lett., 1: 29-32. Carter, R.W.G., Orford, J.D., Forbes, D.L. and Taylor, R.B., 1990. Morphosedimentary development of drumlin flank barriers with rapidly rising sea level, Story Head, Nova Scotia. Sediment. Geol., 69: 117138. Casshyap, S.M., 1969. Petrology of the Bruce and Gowganda formations and its bearing on the evolution of Huronian sedimentation in the Espanola-Willisvile Area, Ontario (Canada). Palaeogeogr. Palaeoclimatol. Palaeoecol., 6: 5-36. Casshyap, S.M., 1977. Patterns of sedimentation in Gondwana Basins. In: 4th Int. Gondwana Symp. Calcutta. Hindustan Publishing Co., Delhi, pp. 527551. Cathles, L.M. and Hallam, A., 1991. Stress-induced changes in plate density, Vail sequences, epirogeny and short-lived global sea level fluctuations. Tectonics, 10: 659-671. Cecioni, 1981. Palaeozoic varve-like sediments in the Patagonian Archipelago, Chile. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 912. Chahine, M.T., 1992. The hydrological cycle and its effect on climate. Nature, 359: 373-380. Chamberlin, T.C., 1899. An attempt to frame a work-

218 ing hypothesis of the cause of glacial periods on an atmospheric basis. J. Geol., 7: 545-584. Chappell, J., 1983. A revised sea-level record for the last 300,000 years from Papua New Guinea. Search, 14: 99-101. Chappell, J. and Shackleton, N.J., 1986. Oxygen isotopes and sea level. Nature, 324: 137-138. Chappellaz, J., 1990. Ice core record of atmospheric methane and the past 160,000 years. Nature, 345: 127-131. Charles, C.D. and Fairbanks, R.G., 1992. Evidence from Southern Ocean sediments for the effect of North Atlantic deep-water flux on climate. Nature, 355: 416-422. Charrier, R., 1986. The Gondwana glaciation in Chile; description of alleged glacial deposits and palaeogeographic condintions bearing on the extent of the ice cover in southern South America. Palaeogeogr. Palaeoclimatol. Palaeoecol., 56: 151-175. Chen, J., 1991. Bathymetric biosignals and Ordovician chronology of eustatic variations. Geol. Surv. Can. Pap., 90-9: 299-311. Chen, Z., McA. Powell, C. and Balme, B.E., 1992. New Devono-Carboniferous palaeomagnetic results for Gondwanaland and the timing of the formation of Pangea. Eos, 73(43), Abstr. suppl., p. 150. Chowdhury, M.K.R., Laskar, B. and Mitra, N.D., 1975. Tectonic control of Lower Gondwana sedimentation in peninsular India. In: K.S.W. Campbell (Editor), Gondwana Geology. Australian National University Press, pp. 675-680. Christie-Blick, N., 1983. Glacial-marine and subglacial sedimentation, upper Proterozoic Mineral Fork Formation, Utah. In: B.F. Molnia (Editor), GlacialMarine Sedimentation. Plenum, New York, pp. 703-776. Christie-Blick, N., 1990. Sequence stratigraphy and sea-level changes in Cretaceous time. In: R.N. Ginsburg and B. Beaudoin (Editors), Cretaceous Resources, Events and Rhythms. Kluwer, Dordrecht, pp. 1-22. Christie-Blick, N., Mountain, G.S. and Miller, K.G., 1990. Seismic stratigraphic record of sea level change. In: Sea Level Change. Studies in Geophysics. National Research Council, National Academy Press, Washington, D.C., pp. 116-140. Chumakov, N.M., 1981a. Scattered stones in Mesozoic deposits of North Siberia. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 264. Chumakov, N.M., 1981b. Upper Proterozoic glaciogenic rocks and their stratigraphic significance. Precambrian Res., 15: 373-395. Chumakov, N.M., 1985. Glacial events of the past and their geological significance. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 319-346.

N. EVLES Chumakov, N.M. 1992. The problems of old glaciations. Sov. Sci. Rev. G. Geology, 1: 1-208. Chumakov, N.M. and Cailleux, A., 1971. Glaciation and eolisation dans L'est et le Nord de l'Europe a l'Eocambrien. Rev. Geomorph. Dyn., 20: 1-4. Chumakov, N.M. and Elston, D.P., 1989. The paradox of Late Proterozoic glaciations at low latitudes. Episodes, 12: 115-1120. Chumakov, N.M. and Semikhatov, M.A., 1981. Riphean and Vendian of the USSR. Precambrian Res., 15: 229-253. Chumakov, N.M. and Krasil'nikov, S.S., 1992. Lithology of Riphean tiiloids: Urinsk uplift area, Jena region. Lithol. Miner. Resour., 26: 249-264. Ciesielski, P.F., Kristofferson, Y. and others, 1991. Proceedings Of The Ocean Drilling Program, Scientific Results Leg 114. Ocean Drilling Program, 114. College Station, Texas, 826 pp. Clague, J.J., Evans, S.G. and Blown, I.G., 1985. A debris flow triggered by the breaching of a moraine-dammed lake, Klattasine Creek, British Columbia. Can. J. Earth Sci., 22: 1492-1502. Clark, D.L., 1982. Origin, nature and world climate effect of Arctic Ocean ice-cover. Nature, 300: 321325. Clark, J.I. and Landva, J., 1988. Geotechnical aspects of seabed pits in the Grand Banks area. Can. Geotech. J., 25: 448-454. Clark, P.U., 1989. Discussion: Relative differences between glacially crushed quartz transported by mountain and continental ice--some examples from North America and East Africa discussion. Am. J. Sci., 289: 1195-1198. Clauer, N. and Deynoux, M., 1987. New information on the probable isotopic age of the Late Proterozoic glaciation in West Africa. Precambrian Res., 37: 89-94. Clendenin, C.W., Charlesworth, E.G. and Maske, S., 1988. An early Proterozoic three-stage rift system, Kaapvaal Craton, South Africa. Tectonophysics, 145: 73-86. Cloetingh, S., 1988. Intraplate stresses: a tectonic cause for third-order cycles in apparent sea level?. In: C.K. Wilgus, B.S. Hastings, H.W. Posamentier, C.A. Ross and C.G.S. Kendall (Editors), Sea Level Changes: An Integrated Approach. Soc. Econ. Paleontol. Mineral. Spec. Publ., 42: 19-29. Coats, R.P., 1981. Late Proterozoic (Adelaidean) tillites of the Adelaide Geosyncline. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 537-548. Coats, R.P. and Preiss, W.V., 1980. Stratigraphic and geochronological reinterpretation of Late Proterozoic glaciogenic sequences in the Kinberley Region, Western Australia. Precambrian Res., 13: 181-208.

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SETTING

Coleman, A.P., 1907. A lower Huronian ice age. Am. J. Sci., 23: 187-192. Coleman, A.P., 1908. A lower Huronian ice age. J. Geol., 16: 149-158. Coleman, A.P., 1916. Dry land in geology. Geol. Soc. Am. Bull., 27: 175-192. Coleman, A.P., 1926. Ice Ages: Recent and Ancient. Macmillan, Lonmdon, 296 pp. Coleman, A.P., 1929. Long range correlation of varves. J. Geol., 37: 783-789. Coleman, A.P., 1941. The Last Million Years. University of Toronto Press (Reprinted 1976 by AMS Press, New York), 216 pp. Coleman, J.M. and Prior, D.B., 1988. Mass wasting on continental margins. Annu. Rev. Earth Planet. Sci., 16: 101-119. Colhoun, E.A., Mabin, M.C.G., Adamson, D.A. and Kirk, R.M., 1992. Antarctic ice volume and contribution to sea-level fall at 20,000 yr BP from raised beaches. Nature, 358: 316-319. Collinson, J.W., 1991. The palaeopacific margin as seen from East Antarctica. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 199-204. Collinson, J.D., Bevins, R.E. and Clemmensen, L.B., 1989. Postglacial mass flow and associated deposits preserved in palaeovalleys: the late Precambrian Moraeneso formation, North Greenland. Medd. Gronl. Geosci., 21, 26 pp. Condie, K.C., 1989. Plate Tectonics and Crustal Evolution. Pergamon Press, Oxford, 476 pp. Coombs, D.S., 1958. Zeolitised tuffs from the Kuttung glacial beds near Seaham, New South Wales. Aust. J. Sci., 21: 18. Cooper, A.K., Davey, F.J. and Behrendt, J.C., 1991a. Structural and depositional controls on Cenozoic and Mesozoic strata beneath the western Ross Sea. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 279-283. Cooper, A.K., Davey, F.J. and Hinz, K., 1991b. Crustal extension and origin of sedimentary basins beneath the Ross Sea and Ross Ice Shelf, Antarctica. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 285-291. Copper, P., 1977. Paleolatitudes in the Devonian of Brazil and the Frasnian-Famennian mass extinctions. Palaeogeogr. Palaeoclimatol. Palaeoecol., 21: 165-207. Copper, P., 1986. Frasnian-Fammenian mass extiction and cold water oceans. Geology, 14: 835-839. Corbitt, L.L. and Woodward, L.A., 1973. Upper Precambrian diamictite of Florida Mountains, southwestern New Mexico. Geol. Soc. Am. Bull., 84: 171-174.

219

Corey, C., 1991. Credit the Oceans? Nature, 352: 196197. Cowan, E.A. and Powell, R.D., 1990. Suspended sediment transport and deposition of cyclically interlaminated sediment in a temperate glacial fjord, Alaska, U.S.A.. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. Soc. Spec. Publ., 53: 75-90. Cowan, E.A., Powell, R.D. and N.D., S., 1988. Rainstorm-induced event sedimentation at the tidewater front of a temperate glacier. Geology, 16: 409-412. Crandell, D.R. and Waldron, H.H., 1956. A recent volcanic mudflow of exceptional dimenstions from Mt. Rainier, Washington. Am. J. Sci., 254: 349-362. Crawford, A.R. and Daily, B., 1971. Probable non-synchroneity of Late Precambrian glaciations. Nature, 230: 111-112. Crittenden, M.D., Christie-Blick, N. and Link, P.K., 1983. Evidence of two pulses of glaciation during the late Proterozoic in northern Utah and southeastern Idaho. Geol. Soc. Am. Bull., 94: 437-450. Croll, J., 1875. Climate and Time in their Geological Relations. Daldy, Isbister and Co., London. Croot, D.G. (Editor), 1988. Glaciotectonics--Forms and Processes. Balkema, Rotterdam, 212 pp. Crowell, J.C., 1957. Origin of pebbly mudstone. Bull. Geol. Soc. Am., 68: 993-1010. Crowell, J.C., 1964. Climatic significance of sedimentary deposits containing dispersed megaclasts. In: A.E.M. Nairn (Editor), Problems in Climatology. lnterscience, London, pp. 86-99. Crowell, J.C., 1978. Gondwana glaciation, cyclothems, continental positioning and climate change. Am. J. Sci., 278: 1345-1372. Crowell, J.C., 1981. Early Paleozoic glaciation and Gondwana drift: In: Paleoreconstruction of the Continents. Am. Geophys. Union, Geodynamic Series, 2: 45-49. Crowell, J.C., 1982. Continental glaciation through geologic time. In: Climate in Earth History. National Academy Press, Washington, D.C., pp. 77-82. Crowell, J.C., 1983a. Ice ages recorded on Gondwanan continents. Trans. Geol. Soc. S. Afr., 86: 237-261. Crowell, J.C., 1983b. The recognition of ancient glaciations. Geol. Soc. Am. Mem., 161: 289-297. Crowell, J.C. and Frakes, L.A., 1970. Phanerozoic glaciation and the causes of ice ages. Am. J. Sci., 268. 193-224. Crowell, J.C. and Frakes, L.A., 1971a. Late Palaeozoic glaciation of Australia. J. Geol. Soc. Aust., 17: 115155. Crowell, J.C. and Frakes, L.A., 1971b. Late Paleozoic Glaciation: Part IV, Australia. Geol. Soc. Am. Bull., 82: 2515-2540. Crowell, J.C. and Frakes, L.A., 1972. Late Paleozoic glaciation: Part V, Karoo Basin, South Africa. Geol. Soc. Am. Bull., 83: 2887-2912.

220

Crowell, J.C., Suarez-Soruco, R. and Rocha-Campos, A.C., 1980. Silurian glaciation in central South America. In: Fifth Int. Gondwana Symp. Wellington, New Zealand, pp. 105-110. Crowley, T.J., Mengel, J.G. and Short, D.A., 1987. Gondwanaland's seasonal cycle. Nature, 329: 803807. Crowley, T.J. and Baum, S.K., 1991a. Estimating Carboniferous sea-level fluctuations from Gondwana ice extent. Geology, 19: 975-977. Crowley, T.J. and Baum, S.K., 1991b. Towards reconciliation of Late Ordovician ( ~ 440 Ma) glaciation with very high CO 2 levels. J. Geophys. Res., 96: 597-622. Crowley, T.J. and Baum, S.K., 1992. Modeling late Paleozoic glaciation. Geology, 20: 507-510. Crowley, T.J., Baum, S.K. and Hyde, W.T., 1991. Climate model comparison of Gondwanan and Laurentide glaciations. J. Geophys. Res., 96: 9217-9226. Cuerda, A. and Azcuy, C.L., 1986. The Carboniferous System In The Argentine Republic (Synthesis). Subcommission On Carboniferous Stratigraphy, Cordoba, Argentina, 380 pp. Curray, J.R., 1965. Late Quaternary history, continental shelves of the United States. In: H.E. Wright and D.C. Frey (Editors), The Quaternary of the United States. Princeton University Press, pp. 723735. D'Orsay, A.M. and van de Poll, H.W., 1985. Quartz grain surface textures: Evidence for middle Carboniferous glacial sediment input to the Parrsboro Formation of Nova Scotia. Geology, 13: 285-287. Dalland, A., 1976. Erratic clasts in the Lower Tertiary deposits of Svalbard--evidence of transport by winter ice. Nor. Polarinst. Arbok, pp. 151-165. Dalmayrac, B., Laubacher, G., Marocco, R., Martinez, C. and Tomasi, P., 1980. La chaine hereynienne d'amerique du sud structure et evolution d'un orogene intracratonique. Geol. Rundsch., 69: 1-21. Daly, M.C., Chorowicz, J. and Fairhead, J.D., 1989. Rift basin evolution in Africa: the influence of reactivated steep basement shear zones. In: M.A. Cooper and G.D. Williams (Editors), Inversion Tectonics. Geol. Soc. Spec. Publ., 44: 309-334. Daly, M.C., Lawrence, S.R., Kimun'a, D. and Binga, M., 1991. Late Palaeozoic deformation in central Africa. A result of distant collision? Nature, 350: 605-607. Dalziel, I.W.D., 1992. On the organization of American plates in the Neoproterozoic and the breakout of Laurentia. GSA Today, 2: 237. Dalziel, I.W.D., Garrett, S.W., Grunow, A.M., Pankhurst, R.J., Storey, B.C. and Vennum, W.R., 1987. The Ellsworth-Whitmore Mountains crustal block: Its role in the tectonic evolution of West Antarctica. In: G.D. McKenzie (Editor), Gondwana

N. EYLES 6. Am. Geophys. Union, Geophys. Monogr., 41: 173-182. Dawson, W., 1872. Canadian Naturalist and Geologist, 1: 416. De Angelis, M., Barkov, N.I. and Petrov, V.N., 1987. Aerosol concentrations over the last climate cycle (160 kyr) from an Antarctic ice core. Nature, 325: 318-321. de Castro, J.C., 1989. The "Tiles": A glaciolacustrine facies from the Itarare Group. Bol. Geocienc. Petrobras, 3: 229-235. De Geer, G., 1912. A Geochronology of the Last 12,000 Years, in C.R. Cong. Geol. Int. Stockholm, 1910, pp. 241-253. De Geer, G., 1926. On the solar curve as dating the Ice Ages, the New York Moraine and Niagara Falls through the Swedish timescale. Geogr. Ann., 8: 253-283. de Wit, M., Jeffery, M., Bergh, H. and Nicolaysen, L., 1988. Geological map of sectors of Gondwana. Am. Assoc. Pet. Geol., Tulsa, Oklahoma. Scale 1 : 10,000,000. de Wit, M.J., Roering, C., Hart, R.J., Armstrong, R.A., de Rondc, C.E., Green, R.W.E., Tredoux, M., Peberdy, E. and Hart, R.A., 1992. Formation of an Archaean continent. Nature, 357: 553-562. Dennison, J., 1976. Appalachian Queenston delta related to eustatic sea-level drop accompanying late Ordovician glaciation centered in Africa. In: M.G. Bassett (Editor), The Ordovician System. University of Wales Press, Cardiff, pp. 107-120. Denton, G.H. and Armstrong, R.L., 1969. MiocenePiiocene glaciations in southern Alaska. Am. J. Sci., 267: 1121-1142. Denton, G.W. and Hughes, T.J. (Editors), The Last Great Ice Sheets. Wiley, New York, 484 pp. Derby, O.A., 1888. Spuren einer carboner eiszeit in Sudamerika. Neues Jahrb. Mineral., 2: 172-176. Destombes, J., Hollard, H. and Willefert, S., 1985. Lower Palaeozoic rocks of Morocco. In: C.H. Holland (Editor), Lower Palaeozoic of North-Western and West-Central Africa. Lower Palaeozoic Rocks of the World 4. Wiley, Chichester, pp. 91-336. Des Marais, D.J., Strauss, H., Summons, R.E and Hayes, J.M., 1992. Carbon isotope evidence for the stepwise oxidation of the Proterozoic environment. Nature, 359: 605-609. Deynoux, M., 1982. Periglacial polygonal structures and sand wedges in the Late Precambrian glacial formations of the Taoudeni Basin in Adrar of Mauretania (West Africa). Palaeogeogr. Palaeoclimatol. Palaeoecol., 39: 55-70. Deynoux, M., 1985a. Les glaciations du Sahara. La Reserche, 16: 986-987. Deynoux, M., 1985b. Terrestrial or waterlain glacial diamictites? Three case studies from the Late Pre-

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

cambrian and Late Ordovician glacial drifts in S. Africa. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 97-142. Deynoux, M., 1991. Field Trip Guide to Bamako, Mali. Int. Geol. Correl. Program 260. Int. Geol. Correl. Program, 73 pp. Deynoux, M. and Trompette, R., 1976. Discussion: Late Precambrian mixtites; glacial and/or non-glacial? Dealing especially with the mixtites of West Africa. Am. J. Sci., 276: 1302-1315. Deynoux, M., Trompette, R., Clauer, N. and Sougy, J. 1978. Upper Precambrian and lowermost Palaeozoic correlations in west Africa and in the western part of central Africa. Probable diachronism of the Late Precambrian tillite. Geol. Rundsch., 67: 615630. Deynoux, M. and Trompette, R., 1981. Late Ordovician tillites of the Taoudeni Basin, West Africa. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 89-96. Deynoux, M., Sougy, J. and Trompette, R., 1985. Lower Paleozoic of west Africa and the western part of central Africa. In: C.H. Holland (Editor), Lower Palaeozoic of North-Western and West-Central Africa. Lower Palaeozoic Rocks Of The World 4. Wiley, Chichester, pp. 337-495. Deynoux, M., Kocurek, G. and Proust, J.N., 1989. Late Proterozoic periglacial eolian deposits on the west African platform, Taoudeni Basin, western Mali. Sedimentology, 36: 531-549. Deynoux, M., Proust, J.N. and Simon, B., 1991. Late Proterozoic glacially-controlled shelf sequences in western Mali (West Africa). J. Air. Earth Sci., 12: 181-198. Dia, A.N., Cohen, A.S., O'Nions, R.K. and Shackleton, N.J., 1992. Seawater Sr isotope variation over the past 300 kyr and influence of global climate cycles. Nature, 356: 786-788. Dickins, J.M., 1985a. Late Palaeozoic and Early Mesozoic "orogeny" in eastern Australia. In: Advances in the Study of the Sydney Basin. Proc. Nineteenth Symp. Univ. Newcastle, pp. 8-9. Dickins, J.M., 1985b. Late Paleozoic glaciation. BMR J. Aust. Geol. Geophys., 9: 163-169. Dickins, J.M., 1993. Climate of the Late Devonian to Trassic. Palaeogeogr., Palaeoclimatol. Palaeoecol., 100:89-97 Dickins, J.M., Gostin, V.A. and Runnegar, B., 1968. Correlation and age of the Permian sequence in the southern part of the Sydney Basin. In: K.S.W. Campbell (Editor), Stratigraphy and Palaeontology: Essays in honour of Dorothy Hill. Australian National University Press, pp. 211-255. Diessel, C.F.K., 1980. Newcastle and Tomago Coal Measures. Bull. Geol. Surv. New South Wales, 26: 100-115.

221

Dillon, W.P. and Paull, C.K., 1983. Marine gas hydrates. 2. Geophysical evidence. In: J.L. Cox (Editor), Natural Gas Hydrates. Butterworth Publishers, Boston, pp. 73-90. Dillon, W.R. and Oldale, R.N., 1978. Late Quaternary sea-level curve; Reinterpretation based on glaciotectonic influence. Geology, 6: 56-60. Dionne, J.C., 1985. Forms, figures and glacial sedimentary facies of muddy tidal flats of cold regions. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 415451. Domack, E.W., 1983. Facies of Late Pleistocene Glacial-Marine Sediments on Whidbey Island, Washington: An Isostatic Glacial-Marine Sequence. In: B.F. Molnia (Editor), Glacial-Marine Sedimentation. Plenum, New York, pp. 535-570. Domack, E.W., 1984. Rhythmically bedded glaciomarine sediments on Whidbey Island, Washington. J. Sediment. Petrol., 54: 589-602. Domack, E.W., 1988. Biogenic facies in the Antarctic glacimarine environment: basis for a Polar glacimarine summary. Palaeogeogr. Palaeoclimatol. Palaeoecol., 63: 357-372. Domack, E.W., 1990. Laminated terrigenous sediments from the Antarctic Peninsula: the role of subglacial and marine processes. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glaciomarine Environments. Geol. Soc. Spec. Publ., 53: 91-103. Dong-li, S., 1993. On the Permian biogeographic boundary between Gondwana and Eurasia in Tibet, China as the eastern section of the Tethys. Palaeogeogr. Palaeoclimatol. Palaeoecol., 100: 59-77. Donnelly, T.W., 1982. Worldwide continental denudation and climatic deterioration during the late Tertiary: Evidence from deep-sea sediments. Geology, 10: 451-454. Dore, F., Dupret, L. and Le Gall, J., 1985. Tillites et tilloides du Massif armoricain. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 85-96. Dott, R.H., 1961. Squantum "Tillite", Massachusetts --Evidence of glaciation or subaqueous mass movements?. Geol. Soc. Am. Bull., 72: 1289-1306. Dott, R.H. and Batten, R.L., 1988. Evolution of the Earth. McGraw-Hill, New York, 512 pp. Dowdeswell, J.A., Hambrey, M.J. and Rutang, W., 1985. A comparison of clast fabric and shape in Late Precambrian and modern glacigenic sediments. J. Sediment. Petrol., 55: 691-704. Dowdeswell, J.D. and Scourse, J.D. (Editors), 1990. Glacimarine Environments: Processes and Sediments. Geol. Soc. London, Spec. Publ., 53. Draganic, I.G., Bjergbakke, E., Draganic, Z.D. and Sehested, K., 1991. Decompositon of ocean waters by potassium-40 radiation 3800 Ma ago as a source of oxygen and oxidizing species. Precambrian Res., 52: 337-345. Drake, J.J. and McCann, S.B., 1982. The movement of

222 isolated boulders on tidal flats by ice floes. Can. J. Earth Sci., 19: 748-754. Drewry, D.J., 1986. Glacial Geologic Processes. Edward Arnold, 276 pp. Dreyer, T., 1988. Late Proterozoic (Vendian) to Early Cambrian sedimentation in the Hedmark Group, southwestern part of the Sparagmite Region, southern Norway. Nor. Geol. Unders., 412: 1-27. Du Toit, A.L., 1927. A Geological Comparison of South America with South Africa. Carnegie Institute, Washington, 157 pp. Duplessy, J.C., Labeyrie, L., Arnold, M., Paterne, M., Duprat, J. and van Weering, T.C.E., 1992. Changes in surface salinity of the North Atlantic Ocean during the last deglaciation. Nature, 358: 485-488. Duringer, P., Paicheler, J.C. and Schneider, J.L., 1991. Un courant d'eau continu peut-ii generer des turbidites? Resultats d'experimentations analogiques. Mar. Geol., 99: 231-246. Edwards, M.B., 1975. Glacial retreat sedimentation in the Smalfjord Formation, Late Precambrianm, North Norway. Sedimentology, 22: 75-94. Edwards, M.B., 1978. Glacial Environments. In: H.G. Reading (Editor), Sedimentary Environments And Facies. Elsevier, New York, pp. 416-438. Edwards, M.B., 1979. Late Precambrian glacial loessites from North Norway and Svalbard. J. Sediment. Petrol., 49: 85-92. Edwards, M.B., 1984. Sedimentology of the Upper Proterozoic glacial record, Vestertana Group, Finmark, North Norway. Nor. Geol. Unders., 394: 1-76. Edwards, M.B. and Foyn, S., 1981. Late Precambrian tillites in Finnmark, North Norway. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 606-610. Eglinton, G., Bradshaw, A., Rosell, A., Sarnthein, M., Pflaumann, U. and Tiedemann, R., 1992. Molecular record of secular sea surface temperature changes on 100-year timescales for glacial terminations I,II and IV. Nature, 326: 423-426. Eisbacher, G.H., 1981. Sedimentary tectonics and glacial record in the Windermere Supergroup, Mackenzie Mountains, northwestern Canada. Geol. Surv. Can. Pap., 80-27, 40 pp. Eisbacher, G.H., 1985. Late Proterozoic rifting, glacial sedimentation and sedimentary cycles in the light of Windermere deposition, Western Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 231-254. E1-Nakhal, H.A., 1990. Glaciations in the Arabian Peninsula. Qatar Univ. Sci. Bull., 10: 287-295. Elverhoi, A., 1984. Glaciogenic and associated marine sediments in the Weddell Sea, Fjords of Spitsbergen and the Barents Sea: a review. Mar. Geol., 57: 53-88. Elverhoi, A., Lonne, O. and Seland, R., 1983. Glacima-

N. EYLES rine sedimentation in a modern fjord environment, Spitsbergen. Polar Res., 1: 127-149. Embleton, B.J.J. and Williams, G.E., 1986. Low palaeolatitude of deposition for late Precambrian periglacial varvites in South Australia: implications for palaeoclimatology. Earth Planet. Sci. Lett., 79: 419-430. Embleton, C., 1984. The Geomorphology of Europe. MacMillan, London, 465 pp. Embry, A., 1984. Upper Jurassic to lowermost Cretaceous straitgraphy, sedimentology and petroleum geology, Sverdrup Basin. Can. Soc. Pet. Geol. Abstr. Annu. Meet., pp. 49-50. Emiliani, C. and Geiss, J., 1959. On glaciations and their causes. Geol. Rundsch., 47: 576-601. Endal, A.S., 1981. Evolutionary variations of solar luminosity. Variations of the solar constant. NASA Conf. PUN., 2191: 175-183. Endal, A.S. and Sofia, S., 1981. Rotation in solar-type stars, I, Evolutionary models for the spin-down of the sun. Astrophys. J., 243: 625-640. England, P. and Molnar, P., 1990. Surface uplift, uplift of rocks and exhumation of rocks. Geology, 18: 1173-1177. England, P. and Molnar, P., 1991. Surface uplift, uplift of rocks and exhumation of rocks. Reply to comments. Geology, 19: 1053-1054. Epshteyn, O.G., 1978. Mesozoic-Cenozoic climates of Northern Asia and glacial-marine deposits. Int. Geol. Rev., 20: 49-58. Eriksson, K.A., 1983. Siliciclastic-hosted iron-formation in the early Archean Barberton and Pilbara sequences. J. Geol. Soc. Aust., 30: 473-482. Eyles, C.H., 1987. Glacially-influenced submarinechannel sedimentation in the Yakataga Formation, Middleton Island, Alaska. J. Sediment. Petrol., 57: 1004-1017. Eyles, C.H., 1988a. Glacially and tidally-influenced shallow marine sedimentation of the Late Precambrian Port Askaig Formation, Scotland. Palaeogeogr. Palaeoclimatol. Palaeoecoi., 68: 1-25. Eyles, C.H., 1988b. A model for striated boulder pavement formation on glaciated, shallow-marine shelves: An example from the Yakataga Formation, Alaska. J. Sediment. Petrol., 58: 62-71. Eyles, C.H. and Eyles, N., 1983a. A glaciomarine model for Upper Precambrian diamictites of the Port Askaig Formation, Scotland. Geology, 11: 692-696. Eyles, C.H. and Eyles, N., 1983b. Sedimentation in a large lake: a reinterpretation of the late Pleistocene stratigraphy at Scarborough Bluffs, Ontario, Canada. Geology, 11: 146-152. Eyles, C.H. and Eyles, N., 1984. Late Pleistocene glaciomarine sediments of the Isle of Man as a key to stratigraphic investigations in the Irish Sea basin. Geology, 12: 359-364.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

Eyles, C.H., Eyles, N. and Miall, A.D., 1985. Models of glaciomarine deposition and their applications to ancient glacial sequences. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 15-84. Eyles, C.H. and Eyles, N., 1989. The Late Cenozoic White River "Tillites" of southern Alaska: Subaerial slope and fan delta deposits in a strike-slip setting. Geol. Soc. Am. Bull., 101: 1091-1102. Eyles, C.H. and Lagoe, M.B., 1990. Sedimentation patterns and facies geometries on a temperate glacially-influenced continental shelf: the Yakataga Formation, Middleton Island, Alaska. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glacimarine Environments: Processes and Sediments. Geol. Soc. Spec. Publ., 53: 363-386. Eyles, C.H., Eyles, N. and Lagoe, M.B., 1991. The Yakataga Formation; A late Miocene to Pleistocene record of temperate glacial marine sedimentation in the Gulf of Alaska. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation; Paleoclimatic Significance. Geol. Soc. Am. Spec. Pap., 261. Boulder, Colorado, pp. 159-180. Eyles, C.H., Eyles, N. and Franca, A.B., 1993. Glaciation and tectonics in an active intracratonic basin; the late Paleozoic Itarare group, Parana Basin. Sedimentology, 40: 1-25. Eyles, N., 1987. Late Pleistocene debris flow deposits in large ice-contact lakes in British Columbia and Alaska. Sediment. Geol., 53: 33-71. Eyles, N., 1990. Late Precambrian "tillites" of the Avalonian-Cadomian belt; marine debris flows in an active tectonic setting. Palaeogeogr. Palaeoclimatol. Palaeoecol., 79: 73-98. Eyles, N. and Sladen, J.A., 1981. Stratigraphy and geotechnical properties of weathered lodgement till in Northumberland, England. Q.J. Eng. Geol., 14: 129-141. Eyles, N., Sladen, J.A. and Gilroy, S., 1982a. A depositional model for stratigraphic complexes and facies superimposition in lodgement tills. Boreas, 11: 317333. Eyles, N., Sasseville, D.R., Slatt, R.M. and Rogerson, R.J., 1982b. Geochemical denudation rates and solute transport mechanisms in a maritime temperate glacier basin. Can. J. Earth Sci., 19: 1570-1582. Eyles, N., Eyles, C.H. and Miall, A.D., 1983. Lithofacies types and vertical profile modes; an alternative approach to the description and environmental interpretation of glacial diamict sequences. Sedimentology, 30: 393-410. Eyles, N. and Miall, A.D., 1984. Glacial Facies. In: R.G. Walker (Editor), Facies Models. Geol. Assoc. Can., Toronto, pp. 15-38 Eyles, N. and Clark, B.M., 1985. Gravity induced softsediment deformation structures in glaciomarine se-

223 quences of the Late Proterozoic Port Askaig Formation, Scotland. Sedimentology, 32: 784-814. Eyles, N., Clark, B.M. and Clague, J.J., 1987a. Coarsegrained sediment gravity flow facies in a large supraglacial lake. Sedimentology, 34: 193-216. Eyles, N., Day, T.E. and Gavican, A., 1987b. Depositional influences upon the magnetic characteristics (NRM, AMS) of lodgement tills and glacial diamict facies. Can. J. Earth Sci., 24: 2436-2458. Eyles, N. and Clark, B.M., 1988. Storm-influenced deltas and ice-scouring in a Late Pleistocene glacial lake. Geol. Soc. Am. Bull., 100: 793-809. Eyles, N., Eyles, C.H. and McCabe, A.M., 1988a. Late Pleistocene subaeriai debris flow facies of the Bow Valley, near banff, Canada Rocky Mountains. Sedimentology, 35: 465-480. Eyles, N., Eyles, C.H. and McCabe, A.M., 1988b. Sedimentation in an ice-contact subaqueous setting: the Mid-Pleistocene "North Sea Drifts" of Norfolk, U.K.Q. Sci. Rev., 8: 57-74. Eyles, N. and Kocsis, S., 1988. Sedimentology and clast fabric of subaerial debris flows in a glacially-influenced alluvial fan. Sediment. Geol., 59: 15-28. Eyles, N. and Eyles, C.H., 1989. Glacially-influenced deep marine sedimentation of the Late Precambrian Gaskiers Formation, Newfoundland, Canada. Sedimentology, 36: 601-620. Eyles, N. and Kocsis, S.P., 1989. Sedimentological controis on gold in a late Pleistocene glacial placer deposit, Cariboo District, British Columbia, Canada. Sediment. Geol., 65: 45-68. Eyles, N. and Lagoe, M., 1989. Sedimentology of shellrich deposits (coquinas) in glaciomarine facies of the late Cenozoic Upper Yakataga Formation, Middleton Island, Alaska. Geol. Soc. Am. Bull., 101: 129-142. Eyles, N. and McCabe, A., 1989a. The Late Devensian ( < 22,000 ybp) Irish Sea basin: The sedimentary record of a collapsed ice sheet margin. Quat. Sci. Rev., 8: 307-351. Eyles, N. and McCabe, A.M., 1989b. Glaciomarine facies within subglacial tunnel valleys; the sedimentary record of glacioisostatic downwarping in the Irish Sea Basin. Sedimentology, 36: 431-448. Eyles, N., Mullins, H.T. and Hine, A.C., 1991. The seismic stratigraphy of Okanagan Lake, British Columbia; a record of rapid deglaciation in a deep "fiord-lake" basin. Sediment. Geol., 73: 13-41. Eyles, N. and Clague, J.J., 1991. Contrasting styles of glaciolacustrine sedimentation during ice sheet advance and retreat in central British Columbia. Geogr. Phys. Quat., 45: 317-331. Eyles, N., McCabe, A.M. and Bowen, D.Q., 1992a. Depositional record of Late Pleistocene surging glaciers, Eastern England. Quat. Sci. Rev. (in press).

224

Eyles, N., Logoe, M.B. and Vossler, S., 1992b. Ichnology of a glacially-influenced continental shelf and slope; the Late Cenozoic Gulf of Alaska (Yakataga Formation). Palaeogeogr. Palaeoclimatol. Palaeoecol., 94:193-221 Eyles, N. and Eyles, C.H., 1992, Glacial Depositional Systems. In: R.G. Walker and N.P. James (Editors), Facies Models: Response to Sea-level Change. Geol. Assoc. Can. Spec. Publ., pp. 73-100. Eyles, N. and Eyles, C.H., 1993. Glacial geologic confirmation of an intraplate boundary crossing the Parana Basin of Brazil. Geology, 21: 459-462. Fairbanks, R.G., 1989. A 17.000 year glacio-eustatic sea-level record; influence of glacial melting rates on the Younger Dryas event and deep ocean circulation. Nature, 342: 637-642. Fairchild, I.J., 1980. Sedimentationand origin of a Late Precambrian "dolomite" from Scotland. J. Sediment. Petrol., 50: 423-446. Fairchild, I.J., 1983. Effects of glacial transport and neomorphism on Precambrian dolomite crystal sizes. Nature, 304: 714-716. Fairchild, I.J., 1989. Doiomitic stromatolite-bearing units with storm deposits from the Vendian of East Greenland and Scotland: a case of facies equivalence. In: R.A. Gayer (Editor), The Caledonide Geology of Scandinavia. Graham and Trotman, London, pp. 275-283. Fairchild, I.J., 1992. Balmy shores and icy wastes: the paradox of carbonates associated with glacial deposits in Neoproteroqoic times. Sedimentol. Rev., 1 (in press). Fairchild, I.J. and Hambrey, M.J., 1984. The Vendian succession of northeastern Spitsbergen: petrogenesis of dolomite-tillite association. Precambrian Res., 26: 111-167. Fairchild, I.J., Hambrey, M.J., Spiro, B. and Jefferson, T.H., 1989. Late Proterozoic glacial carbonates in northeast Spitsbergen: new insights into the carbonate-tillite association. Geol. Mag., 126: 469-490. Fairchild, I.J. and Sp]ro, B., 1990. Carbonate minerals in glacial sediments: geochemical clues to palaeoenvironment. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. Soc. Spec. Publ., 53: 201-216. Feeley, M.H., Moore, T.C., Jr., Loutit, T.S. and Bryand, W.R., 1990. Sequence stratigraphy of Mississippi fan related to oxygen isotope sea level index. Am. Assoc. Pet. Geol. Bull., 74: 407-424. Fischbein, S.A., 1987. Analysis and interpretation of ice-deformed sediments from Harrison bay, Alaska. U.S. Geol. Surv. Open-file Rep., 87-262, pp. 73. Fischer, A.G., 1984. The two Phanerozoic supercycles. In: W.A. Berggren and J.A. Van Couvering (Editors), Catastrophes in Earth History; The New Uniformitarianism. Princeton University Press, pp. 129-150.

N. EYLES

Fischer, A.G., 1986, Climatic rhythms recorded in strata. Annu. Rev. Earth Planet. Sci., 14: 351-376. Fischer, A.G., Premoli Silva, I. and De Boer, P.L., 1990. Cyclostratigraphy. In: R.N. Ginsburg and B. Beaudoin (Editors), Cretaceous Resources, Events and Rhythms. Kluwer, Dordrecht, pp. 139-172. Fisher, R.V. and Schminke, H.V., 1984. Pyroclastic Rocks. Springer, Berlin, 472 pp. Fitzgerald, P.G., 1992. The Transantarctic Mountains of southern Victoria Land: The application of apatite fission track analysis of rift shoulder uplift. Tectonics, 11: 634-662. Fleming, J.R., 1992. T.C Chamberlin and H 2 0 climate feedbacks: A voice from the past. Eos, 73: 505-509. Flint, R.F., 1957. Glacial and Pleistocene Geology. Wiley, New York, 553 pp. Flint, R.F., 1971. Glacial and Quaternary Geology. New York, Wiley, 471 pp. Flint, R.F., Sanders, J.E. and Rodgers, J., 1960. Diamictitie, a substitute term for symmictite. Geol. Soc. Am. Bull., 71: 1809-1810. Forbes, D.L. and Taylor, R.B., 1987. Coarse-grained beach sedimentation under paraglacial conditions, Canadian Atlantic coast. In: D. Fitzgerald and P. Rosen (Editors), Glaciated Coasts. Academic Press, San Diego, pp. 52-86 Fortey, R.A., 1984. Global earlier Ordovician transgressions and regressions and their biological implications. In: D.L. Bruton (Editor), Aspects of the Ordovician System. Paleontol. Contrib. Univ. Oslo, 295: 37-50. Fortuin, A.R., 1984. Late Ordovician glaciomarine deposits (Orea Shale) in the Sierra De Albarracin, Spain. Palaeogeogr. Palaeoclimatot. Palaeoecol., 48: 245-261. Foster, C.B. and Waterhouse, J.B., 1988. The Granulatisporites confluens Oppel-zone and Early Permian marine faunas from the Grant Formation on the Barbwire Terrace, Canning Basin, Western Australia. Aust. J. Earth Sci., 35: 135-158. Foyn, S. and Siedlecki, S., 1980. Glacial stadials and interstadials of the Late Predambrian Smalfjord Tillite on Laksefjordvidda, Finnmark, North Norway. Nor. Geol. Unders., 358: 31-45. Frakes, L.A., 1979. Climates Throughout Geologic Time. Elsevier, Amsterdam, 310 pp. Frakes, L.A. and Crowell, J.C., 1969. Late Paleozoic glaciation: I, South America. Geol. Soc. Am. Bull., 80: 1007-1042. Frakes, L.A., Amos, A.J. and Crowell, J.C., 1969. Origin and stratigraphy of Late Paleozoic diamictites in Argentina and Bolivia. In: IUGS Symposium, Gondwana Stratigraphy, Buenos Aires, 1967. Earth Sci., 2: 821-843. Frakes, L.A. and Crowell, J.C., 1970. Late Paleozoic glaciation: II, Africa exclusive of the Karroo basin. Geol. Soc. Am. Bull., 81: 2261-2286.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

Frakes, L.A., Matthews, J.L. and Crowell, J.C., 1971. Late Paleozoic glaciation: Part III, Antarctica. Geol. Soc. Am. Bull., 82: 1581-1604. Frakes, L.A. and Crowell, J.C., 1975. Characteristics of modern glacial marine sediments: Application to Gondwana glacials. In: K.S.W. Campbell (Editor), Gondwana 3. Australian National University Press, pp. 373-380. Frakes, L.A. and Francis, J.E., 1988. A guide to Phanerozoic cold polar climates from high latitude ice-rafting in the Cretaceous. Nature, 333: 547-549. Frakes, L.A., Francis, J.E. and Sytkus, J.I., 1993. Climate Modes of the Phanerozoic. Cambridge University Press, 274 pp. Franca, A.B. and Potter, P.E., 1991. Stratigraphy and reservoir potential of glacial deposits of the Itarar6 Group (Carboniferous-Permian), Paranfi Basin, Brazil. Am. Assoc. Pet. Geol. Bull., 75: 62-85. Franseen, E.K., Watney, W.L., St.C. Kendall, C.G. and Ross, W., 1991. (Editors) Sedimentary Modelling. Kansas Geological Survey Bulletin 233, 524 pp. French, H. and Harry, D.G., 1990. Observations on buried glacier ice and massive segregated ice, western Arctic coast, Canada. Permafrost Periglacial Processes, 1: 31-43. Gabaglia, G.P.R. and Milani, E.J., 1990. Origem e Evolucao de Bacias Sedimentaires. Petrobras Petroleo Brasileiro S.A., Rio de Janeiro, 414 pp. Gaffin, S., 1987. Ridge volume dependence on sea floor generation rate and inversion using long term sea-level change. Am. J. Sci., 287: 596-611. Gair, J.E., 1981. Lower Proterozoic glacial deposits of northern Michigan. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 803-806. Galloway, W.E., 1976. Sediments and stratigraphic frameworks of the Copper River Fan Delta, Alaska. J. Sediment. Petrol., 46: 726-737. Gardner, S. and Hiscott, R.N., 1988. Deep-water facies and depositional setting of the lower Conception Group (Hadrynian), southern Avalon Peninsula, Newfoundland. Can. J. Earth Sci., 25: 1579-1594. Gayer, R.A. and Rice, A.H.N., 1989. Palaeogeographic reconstruction of the pre- to syn-Iapetus rifting sediments in the Caledonides of Finnmark, N. Norway. In: R.A. Gayer (Editor), The Caledonide Geology of Scandinavia. Graham and Trotman, London, pp. 127-142. Gazdzicki, A., Gradzinski, R., Porebski, S.J. and Wrona, R., 1982. Pholadid Penitella borings in glaciomarine sediments (Pliocene) of King George Island, Antarctica. Neues Jahrb. Geol. Palaentol. Monatsh., 12: 723-735. Gazdzicki, A. and Pugaczewska, H., 1984. Biota of the "Pecten conglomerate" (Polonez Cove Formation, Pliocene) of King George Island (South Shetland Islands, Antarctica). Stud. Geol. Pol., 79: 59-120.

225 Geikie, J., 1863. On the phenomena of the glacial drift of Scotland. Transactions, Geological Soc. Glasgow. 1: 1-190. Geikie, J., 1896. The Great Ice Age. Appleton, New York, 850 pp. George, T.N., 1966. Geomorphic evolution in Hebridean Scotland. Scott. J. Geol., 2: 1-14. Gerard, J.-C., Hauglustaine, D.A. and Francois, L.M., 1992. The faint young sun climatic paradox: A simulation with an interactive seasonal climate-sea ice model. Global Planet. Change, 5: 133-150. Ghibaudo, G., 1992. Subaqueous sediment gravity flow deposits; practical criteria for their field description and classification. Sedimentology, 39, 423-454. Ghosh, P.K., 1975. The environment of coal formation in the Peninsular Gondwana basins of India. In: K.S.W. Campbell (Editor), Gondwana Geology. Australian National University Press, pp. 221-231. Gibbard, P.D. and Stuart, A.J., 1974. Trace fossils from proglacial lake sediments. Boreas, 3: 69-74. Gibbs, A.D., 1984. Structural evolution of extensional basin margins. J. Geol. Soc. London, 141: 609-620. Gibbs, A.D., 1989. A model for linked basin development around the British Isles. In: A.J. Tankard and H.R. Balkwill (Editors), Extensional Tectonics and Stratigraphy of the North Atlantic Margins. Am. Assoc. Pet. Geol. Mem., 46: 501-510. Ginsburg, R.N. and Beaudoin, B. (Editors), 1990. Cretaceous Resources, Events and Rhythms. Kluwer, Dordrecht, 352 pp. Gjessing, J., 1967. Norway's paleic surface. Nor. Geogr. Tiddskr., 21: 69-132. Goldhammer, R.K., Dunn, P.A. and Hardie, L.A., 1987. High frequency glacio-eustatic sea-level oscillations with Milankovitch characteristics recorded in Middle Triassic platform carbonates in nothern Italy. Am. J. Sci., 287: 853-892. Goldstein, B.A., 1989. Waxing and Waning in stratigraphy, play concepts and prospectivity in the Canning Basin. APEA Journal, 29: 466-508. Gomez, B., 1990. Comment on "Microfabrics and quartz microstructures confirm glacial origin of the Sunnybrook drift in the Lake Ontario basin". Geology, 18: 1032-1033. Gomez, B., Dowdeswell, J.A. and Sharp, M.J., 1988. Microstructural control of quartz sand grain shape and texture: Implications for the discrimination of debris transport pathways through glaciers. Sediment. Geol., 57: 119-129. Gonzalez, C.R., 1982. Evidence for the neopaleozoic glaciation in Argentina. INQUA Symposia on the Genesis and Lithology of Quaternary Deposits. Balkema, Rotterdam, pp. 271-276. Gonzalez, C.R., 1990. Development of the Late Palaeozoic glaciations of the south American Gondwana in western Argentina. Palaeogeogr. Palaeoclimatol. Palaeoecol., 79: 257-287.

226 Gonzalez, C.R. and Bossi, G.E., 1987. Descubrimiento del Carbonico inferior marino de la Precordillera Argentina, in Actas 4th Congr. Latinoamer. Paleontol., Sta. Cruz de la Sierra, Bolivia, 2: 713-729. Gonzalez-Bonorino, G., 1992. Carboniferous glaciation in Gondwana. Evidence for grounded marine ice and continental glaciation in southwestern Argentina. Palaeogeogr. Palaeoclimatol. Palaeoecol., 91: 363-375. Gonzalez-Ferran, O., 1991. The Bransfield rift and its active volcanism. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 505509. Goodfeliow, W.D., Geldsetzer, H., McLaren, D.J., Orchard, M.J. and Klapper, G., 1988. The FrasnianFammenian extiction: Current results and possible causes. In: N.J. McMillan, A.F. Embry and D.J. Glass (Editors), Devonian of the World. Can. Soc. Pet. Geol., 3: 9-22. Goodwin, A.M., 1991. Precambrian Geology. Academic Press, New York, 666 pp. Gostin, V.A., 1986. Proterozoic Clastic Sedimentation in relation to Glaciations and Tectonism: Flinders Ranges, South Australia. Twelth Int. Sedimentol. Congr. Canberra, Australia, Excursion 27B. Department of Geology and Geophysics, Adelaide Gostin, V.A. and Herbert, C., 1973. Stratigraphy of the Upper Carboniferous and Lower Permian sequence, southern Sydney Basin. J. Geol. Soc. Aust., 20: 49-70. Gradstein, F.M., Jansa, L.F., Srivastava, S.P., Williamson, M.A., Bonham Carter, G. and Stam, B., 1990. Aspects of North Atlantic Paleo-Oceanography. In: M.J. Keen and G.L. Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geol. Surv. Can., 2: 351-390. Graedel, T.E, Sackmann, I.J. and Boothroyd, A.I., 1991. Early solar mass loss: A potential solution to the weak sun paradox. Geophys. Res. Lett., 18: 1881-1884. Grahn, Y. and Caputo, M.V., 1992. Early Silurian glaciations in Brazil. Palaeogeogr. Palaeoclimatol. Palaeoecol., 99: 9-15. Graindor, M.J., 1954. Note preliminaire sur la glaciation intracambrienne dans Le Massif Armoricain. Bull. Soc. Geol. Fr., 6: 17-24. Grandville, B.F., 1982. Appraisal and development of a structural and stratigraphic trap oil field with reservoirs in glacial to periglacial clastics. In: M.T. Halbouty (Editor), The Deliberate Search for the Subtle Trap. Am. Assoc. Pet. Geol. Mem., 37: 267-286. Gravenor, C.P., 1975. Erosion by continental ice sheets. Am. J. Earth Sci., 275: 594-604. Gravenor, C.P., 1979. The nature of the Late Paleozoic

N. EYLES glaciation of Gondwana as determined from an analysis of garnets and other heavy minerals. Can. J. Earth Sci., 16: 1137-1153. Gravenor, C.P., 1980. Chattermarked garnets and heavy minerals from the Late Paleozoic glacial deposits of southeastern Brazil. J. Earth Sci., 17: 156-159. Gravenor, C.P. and Gostin, V.A., 1979. Mechanism to explain the loss of heavy minerals from Upper Palaeozoic tillites of South Africa and Australia and the Late Precambrian tillites of Australia. Sedimentology, 26: 707-717. Gravenor, C.P., yon Brunn, V. and Dreimanis, A., 1984. Nature and classification of waterlain glaciogenic sediment exemplified by Pleistocene, Late Palaeozoic and Late Precambrian deposits. EarthSci. Rev., 20: 105-166. Grosvald, M.G., Vtyurin, B.I., Sukhodrovskiy, V.L. and Shishorina, Z.G., 1986. Underground ice in western Siberia: origin and geological significance. Polar Geogr. Geol., 10:173-183 Grubb, M.J., Victor, D.G. and Hope, C.W., 1991. Pragmatics in the greenhouse. Nature, 354: 348-350. Guan, B., Wu, R., Hambrey, M.J. and Geng, W., 1986. Glacial sediments and erosional pavements near the Cambrian-Precambrian boundary in western Henan Province, China. J. Geol. Soc. London, 143: 311-323. Gupta, S.K. and Sharma, P., 1992. On the nature of the ice cap on Tibetan Plateau during the Late Quatemary. Global Planet. Change, 5: 339-344. Gust, D.A., Biddle, K.T., Phelps, D.W. and Uliana, M.A., 1985. Associated Middle to Late Jurassic volcanism and extension in southern South America. Tectonophysics, 116: 223-253. Gustavson, T.C., 1975. Sedimentation and physical limnology in proglacial Malaspina Lake, southeastern Alaska. In: A.V. Jopling and B.C. McDonald (Editors), Glaciofluvial and Glaciolacustrine Sedimentation. Soc. Econ. Paleontol. Mineral. Spec. Publ., 23: 249-263. Hadley, D.G. and Schmidt, D.L., 1975. Non-glacial origin for conglomerate beds in the Wajid Sandstone of Saudi Arabia. In: K.S.W. Campbell (Editor), Gondwana Geology. Australian National University Press, pp. 357-371. Hambrey, M.J., 1982. Late Precambrian diamictites of northeastern Svalbard. Geol. Mag., 119: 527-551. Hambrey, M.J., 1983. Correlation of Late Proterozoic tillites in the North Atlantic region and Europe. Geol. Mag., 120: 209-232. Hambrey, M.J. and Harland, W.B. (Editors), 1981. Earth's Pre-Pleistocene Glacial Record. Cambridge University Press. Hambrey, M.J. and Harland, W.B., 1985. The Late Proterozoic glacial era. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 255-272.

E A R T H ' S G L A C IA L R E C O R D A N D ITS T E C T O N I C SE'VI'ING

Hambrey, M.J. and Spencer, A.M., 1987. Late Precambrian glaciation of central East Greenland. Medd. Gronl. Geosci., 19, 50 pp. Haq, B.U., Hardenbol, J. and Vail, P.R., 1988. Mesozoic and Cenozoic chronostratigraphy and cycles of sea level change. In: C.K. Wilgus, B.S. Hastings, H.W. Posamentier, C.A. Ross and C.G.S. Kendall (Editors), Sea Level Changes: An Integrated Approach. Soc. Econ. Paleontol. Mineral. Spec. Publ., 42: 71-108. Hargraves, R.B., 1986. Faster spreading or greater ridge length in the Archean? Geology, 14: 750-752. Harland, W.B., 1964. Critical evidence for a great infra-Cambrian glaciation. Geol. Rundsch., 54: 4561. Harland, W.B., 1981. Chronology of Earth's glacial and tectonic record. J. Geol. Soc. London, 138: 197-203. Harland, W.B. and Bidgood, D.E.T., 1959. Palaeomagnetism in some Norwegian Sparagmites and the Late pre-Cambrian Ice Age. Nature, 184: 18601862. Harland, W.B., Herod, K.N. and Krinsley, D.H., 1966. The definition and identification of tills and tillites. Earth-Sci. Rev., 2: 225-256. Harland, W.B. and Herod, K.N., 1975. Glaciations through time. In: A.E. Wright and F. Moseley (Editors), Ice Ages: Ancient and Modern. Seel House Press, Liverpool, pp. 189-216. Harland, W.B. and Wright, N.J.R., 1979. Alternate hypothesis for the pre-Carboniferous evolution of Svalbard. Skr. Nor. Polarinst., 167: 90-117. Harland, W.B., Armstrong, R.L., Cox, A.V., Craig, L.E., Smith, A.G. and Smith, D.G., 1989. A Geologic Time Scale. Cambridge University Press, 263 pp. Harrington, H.J., 1971. Glacial-like striated floor originated by debris laden torrential water flows. Am. Assoc. Pet. Geol. Bull., 55: 1344-1347. Harrison, C.G.A., 1990. Long-term eustacy and epirogeny in continents. In: Sea Level Change. Studies in Geophysics. National Research Council. National Academy Press, Washington, D.C., pp. 141-155. Hart, J.K., 1992. Sedimentary environments associated with Glacial Lake Trimingham, Norfolk, U.K. Boreas, 21: 119-136. Harvey, L.D., 1988. Climatic impact of ice-age aerosols. Nature, 334: 333-335. Harwood, D.M., 1985. Late Neogene climatic fluctuations in the southern high latitudes: implications of a warm Pliocene and deglaciated Antarctic continent. S. Afr. J. Sci., 81: 239-241. Harwood, D.M., 1991. The changing style of Ceonozoic Antarctic glaciations. In: Annual Meeting, Program with Abstracts. Geol. Assoc. Can., p. A52. Harwood, D.M., Barrett, P.J., Edwards, A.R., Rieck, H.J. and Webb, P.N., 1989. Biostratigraphy and

227

chronology. In: P.J. Barrett (Editor), Antarctic Cenozoic History from the CIROS-1 drillhole, McMurdo Sound. DSIR Bull., 245: 231-239. Haworth, R.T., 1982. Geology of the continental margin: Eastern Canada. In: Regional Geological Synthesis. Decade of North American Geology. Geol. Soc. Am., pp. 133-143. Hay, W.W., Barron, E.J., Sloan, J.L.I. and Southam, J.R., 1981. Continental drift and the global pattern of sedimentation. Geol. Rundsch., 70: 302-315. Hay, W.W. and Leslie, M.A., 1990. Could possible changes in global groundwater reservoir cause eustatic sea level fluctuations?, in Sea Level Change. Washington, D.C., National Academy Press, National Research Council, Studies in Geophysics pp. 161-170. Heckel, P.H., 1986. Sea-level curve for Pennsylvanian eustatic marine transgressive-regressive depositionai cycles along mid-continent outcrop belt, North America. Geology, 14: 330-334. Hefferan, K.P., Karson, J.A. and Saquaque, A., 1992. Proterozoic collisional basins in a Pan-African suture zone, Anti-Atlas Mountains, Morocco. Precambrian Res., 54: 295-319. Helal, A.H., 1964. On the occurrence and stratigraphic position of Permo-Carboniferous tillites In SaudiArabia. Geol. Rundsch., 54: 193-207. Heiwig, J., 1972. Stratigraphy, sedimentation, palaeogeography and palaeoclimates of Carboniferous ("Gondwana") and Permian of Bolivia. Am. Assoc. Pet. Geol. Bull., 56: 1008-1033. Henrich, R.,1991. Cycles, rhythms and events on high input and low input glaciated continental margins. In: H. Einsele (Editor), Cycles and Events in Stratigraphy. Springer, Berlin, pp. 751-780. Herbert, C., 1981. Late Palaeozoic glacigenic sediments of the southern Sydney Basin, New South Wales. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 488-491. Herrington, P.M. and Fairchild, I.J., 1989. Carbonate shelf and slope facies evolution prior to Vendian glaciation, central East Greenland. In: R.A. Gayer (Editor), The Caledonide Geology of Scandinavia. Graham and Trotman, London, pp. 263-273. Herve, F., Godoy, E., Parada, M.A., Ramos, V., Rapela, C., Mpodozis, C. and Davidson, J., 1987. A general view on the Chilean-Argentine Andes, with emphasis on their early history. In: J.W.H. Monger and J. Francheteau (Editors), Circum-Pacific Orogenic Belts and Evolution of the Pacific Ocean Basin. Geodynamics Series 18. Am. Geophys. Union, Washington, D.C., pp. 97-113. Heubeck, C., 1992. Sedimentology of large olistoliths, southern cordillera central, Hispaniola. J. Sediment. Petrol., 62: 474-482.

228 Higgins, A.K., 1981. The Late Precambrian Tillite Group of the Kong Oscars Fjord and Kejser Franz Josefs Fjord region of east Greenland. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 778-781. Hill, P., 1984. Sedimentary facies of the Nova Scotian upper and middle continental slope, offshore eastern Canada. Sedimentology, 31: 293-311. Hillaire-Marcel, C. and Causse, C., 1989. The Late Pleistocene Laurentide Glacier: T h / U dating of its major fluctuations and 180 range of the ice. Quat. Res., 32: 125-135. Hine, A.C., Locker, S.D., Tedesco, L.P., Mullins, H.T., Hallock, P., Belknap, D.F., Gonzales, J.L., Neumann, A.C. and Snyder, S.W., 1992. Megabreccia shedding from modern low-relief carbonate platforms, Nicaraguan Rise. Geol. Soc. Am. Bull., 104: 928-943. Hodel, K.L., Reimnitz, E. and Barnes, P.W., 1988. Microtextures of quartz grains from modern terrestrial and subaqueous environments, north slope of Alaska. J. Sediment. Petrol., 58: 24-32. Hodell, D.A., Elmstrom, K.M. and Kennett, J.P., 1986. Latest Miocene benthic 018 changes, global ice volume, sea-level and "Messinian salinity crisis". Nature, 320: 411-414. Hodell, D.A., Benson, R.H., Kennett, J.P. and Bied, K.R., 1989. Stable isotope stratigraphy of Latest Miocene sequences in northwest Morocco. The Bou Regreg section. Paleoceanography, 4: 467-482. Hoffman, P.F., 1989. Precambrian geology and tectonic history of North America. In: A.W. Bally and A.R. Palmer (Editors), The Geology of North America: An Overview. Decade of North American Geology. Geol. Soc. Am., pp. 447-512. Hoffman, P.F., 1991. Did the break-out of Laurentia turn Gondwanaland inside-out? Science, 252: 14091412. Hofmann, H.J., Narbonne, G.M. and Aitken, J.D., 1990. Ediacaran remains from intertillite beds in northwestern Canada. Geology, 18: 1199-1202. Holland, H.D., 1984. The chemistry of the earlier atmosphere and oceans. In: The Chemical Evolution of the Atmosphere and Oceans. Princeton University Press, pp. 89-127. Holland, H.D. and Beukes, N.J., 1990. A paleoweathering profile from Grigualand West, South Africa: evidence for a dramatic rise in atmospheric oxgyen between 2.2 and 1.9 BYBP. Am. J. Sci., 290: 1-34. Holtedahl, O., 1918. Bidrag til finmarkens geologi. Nor. Geol. Unders., 84: 1-314. Horland, M. and Judd, A.G., 1988. Seabed Pockmarks and Seepages. Graham and Trotman, London, 293 PP. Horsthemke, E., Ledendecker, S. and Porada, M., 1990.

N. EYLES Depositional environments and stratigraphic correlation of the Karoo Sequence in northwestern Damaraland. Commun. Geol. Surv. Namibia, 6: 6373. Houston, R.S., Lanthier, L.R., Karstrom, K.K. and Sylvester, G.G., 1981. Late Proterozoic diamictite of southern Wyoming. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 795-799. Howell, P.D. and van der Pluijm, B.A., 1990. Early history of the Michigan basin: Subsidence and Appalachian tectonics. Geology, 18: 1195-1198. Hoyle, F., 1981. Ice. New English Library, 209 pp. Hsu, K.J., Li, J., Chen, H., Wang, Q., Sun, S. and Sengor, A.M.C., 1990. Tectonics of South China: Key to understanding West Pacific geology. Tectonophysics, 183: 9-39. Huff, W.D., Bergstrom, S.M. and Kolata, D.R., 1992. Gigantic Ordovician ash fall in North America and Europe: Biological, tectonomagmatic and eventstratigraphic significance. Geology, 20: 875-878. Hughes, T., 1987. Ice dynamics and deglaciation models when ice sheets collapsed. In: W.F. Ruddiman and H.E., Wright Jr. (Editors), North America and Adjacent Oceans during the Last Deglaciation. Decade of North American Geology K-3. Geol. Soc. Am., Boulder, Colorado, pp. 183-220. Husseini, M.I., 1988. The Arabian Infracambrian extensional system. Tectonophysics, 143: 93-103. Husseini, M.I., 1989. Tectonic and depositional model of Late Precambrian-Cambrian Arabian and adjoining plates. Am. Assoc. Pet. Geol. Bull., 73: 1117-1131. Hutchinson, R.W., 1992. Late Proterozoic stratigraphy and the Canada-Australia connection: Comment. Geology, 20: 765-766. Idnurm, M. and Giddings, J.W., 1988. Australian Precambrian polar wander: A review. Precambrian Res., 40: 61-88. Ilyin, A.V., 1990. Proterozoic supercontinent, its latest Precambrian rifting, breadup, dispersal into smaller continents and subsidence of their margins: Evidence from Asia. Geology, 18: 1231-1234. Imbrie, J. and Imbrie, K.P., 1979. Ice Ages: Solving the Mystery. McMillan Press, New York, 224 pp. Isotta, C.A.L., Rocha-Campos, A.C. and Yoshida, R., 1969. Striated pavement of the upper Precambrian glaciation in Brazil. Nature, 222: 466-468. Ito, M. and Katsura, Y., 1992. Inferred glacio-eustatic control for high-frequency depositional sequences of the Plio-Pleistocene Kazusa Group, a forearc basin fill in Boso Peninsula, Japan. Sediment. Geol., 80: 67-75. Ives, J.D. and Andrews, J.T., 1963. Studies in the physical geography of north-central Baffin Island, Northwest Territories. Geogr. Bull., 19: 5-48.

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SE'FFING

Jablonski, D., 1985. Marine regressions and mass extinctions: A test using the modern biota. In: J.W. Valentine (Editor), Phanerozoic Diversity Patterns. Princeton University Press, pp. 335-354. Jackson, L.E., 1979. A catastrophic glacial outburst flood (jokulhlavp) mechanism for debris flow generation at the SPiral Tunnels, Kicking Horse basin, British Columbia. Can. Geotech. J., 16: 806-813. Jackson, M.T. and Van De Graaff, W.J.E., 1981. Geology of the Officer Basin, Western Australia. Department of National Development and Energy, Bureau of Mineral Resources, Geology and Geophysics, Bull., 206, 97 pp. Jackson, T.A., 1965. Power-spectrum analysis of two "varved" argillites in the Huronian Cobalt Series (Precambrian) of Canada. J. Sediment. Petrol., 35: 877-886. Jamieson, T.F., 1865. On the history of the last glacial changes in Scotland. Q.J. Geol. Soc., 21: 161-203. Jansa, L.F., 1981. Mesozoic carbonate platforms and banks of the eastern North American margin. Mar. Geol., 44: 97-117. Jansen, E. and Sjoholm, J., 1990. Reconstruction of glaciation over the past 6 Myr from ice-borne deposits in the Norwegian Sea. Nature, 349: 600-603. Jansen, J.H.F., Woensdegt, C.F., Kooistra, M.J. and van der Gaast, S.J., 1987. Ikaite pseudomorphs in the Zaire deep-sea fan: An intermediate between calcite and porous calcite. Geology, 15:245-248 Jeffers, J.D. anderson, J.B. and Lawver, L.A., 1991. Evolution of the Bransfield basin, Antarctic Peninsula. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 481485. John, B.S., ed., 1979. The Winters of the World. Earth Under the Ice Ages. David and Charles, London, 256 pp. Johnsen, S.J., Clausen, H.B., Dansgaard, W., Fuhrer, K., Gundestrup, N., Hammer, C.U., Iversen, P., Jouzel, J., Stauffer, B. and Steffensen, J.P., 1992. Irregular glacial interstadials recorded in a new Greenland ice core. Nature, 359: 311-313. Jones, B.G., Gostin, V.A. and Dickins, J.M., 1986. Sydney Basin Field Guide. 12th Int. Sedimentol. Congr. Canberra, Australia. Int. Assoc. Sedimentol. Excursion IC41 pp. Jones, J.P., 1985. The southern border of the Guapore Shield in western Brazil and Brazil: An interpretation of its geologic evolution. Precambrian Res., 28: 111-135. Jones, R.L. and Mitchell, J.F.B., 1991. Is water vapour understood? Nature, 353: 210. Jorgensen, G.J. and Bosworth, W., 1989. Gravity modeling in the Central African Rift System, Sudan: rift geometries and tectonic significance. J. Afr. Earth Sci., 8: 283-306.

229

Jouzel, J., Lorius, C., Petit, J.R., Genthon, C., Barkov, N.I., Kotlyankov, V.M. and Petrov, V.M., 1987. Vostok ice core: a continuous isotope temperature record over the last climatic cycle (160,000 years). Nature, 329: 403-408. Karlstrom, K.K., Flurkey, A.J. and Houston, R.S., 1983. Stratigraphy and depositional setting of Proterozoic rocks of southereastern Wyoming: record of an Early Proterozoic Atlantic-type cratonic margin. Geol. Soc. Am. Bull., 94: 1287-1294. Kasting, J.F., 1987. Theoretical constraints on oxygen and carbon dioxide concentrations in the Precambrian atmosphere. Precambrian Res., 34: 205-229. Kasting, J.F., 1991. Box models for the evolution of atmospheric oxygen: an update. Global Planet. Change, 97: 125-131. Kasting, J.F., Pollack, J.B. and Ackerrnan, T.P. 1984. Response of Earth's surface temperature to increases in solar flux and implications for loss of water from Venus. Icarus, 57: 335-355. Kasting, J.F. and Toon, O.B., 1989. Climate evolution on the terrestrial planets. In: S.K. Atreya, J.B.Pollack and M.S. Matthews (Editors), Origin and Evolution of Planetary and Satellite Atmospheres. University of Arizona Press, Tucson, pp. 423-449. Kaszycki, C.A., 1987. A model for glacial and proglacial sedimentation in the shield terrane of southern Ontario. Can. J. Earth Sci., 24: 2373-2391. Kaszycki, C.A. and Shilts, W.W., 1980. Glacial erosion of the Canadian Shield: calculation of average depths. Atomic Energy of Canada Ltd., Tech. Rep., TR-106. Kaufman, A.J., Knoll, A.H. and Awramik, S.M., 1992. Biostratigraphic and chemostratigraphic correlation of Neoproterozoic sedimentary successions: Upper Tindir Group, northwestern Canada, as a test case. Geology, 20: 181-185. Keen, C.E. and Beaumont, C., 1990. Geodynamics of rifted continental margins. In: M.J. Keen and G.L. Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geology of Canada 2. Geol. Surv. Can., pp. 391-472. Keigwin, L.D., 1987. Toward a high-resolution chronology for latest Miocene palaeoceanographic events. Palaeoceanography, 2: 639-660. Keith, B.D., 1989. Regional facies of the Upper Ordovician Series of eastern North America. In: B.D. Keith (Editor), The Trenton Group (Upper Ordovician Series) of Eastern North America. Studies in Geology, 29. Am. Assoc. Pet. Geol., pp. 1-7. Keller, B.M., 1973. Great glaciations in history of the earth. Int. Geol. Rev., 15: 1067-1074. Kemp, E.M., 1975. The palynology of Late Palaeozoic glacial deposits. In: K.S.W. Campbell (Editor), Gondwana 3. Australian National University, Canberra, pp. 397-413. Kemp, J. and Young, G.M., 1981. Upper Proterozoic

230 diamictites in northwestern Saudi Arabia. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 344-347. Kennett, J.P., 1982. Marine Geology. Prentice-Hall, Englewood Cliffs, 812 pp. Kennett, J.P. and Shackleton, N.J., 1976. Oxygen isotopic evidence for the development of the psychrosphere 38 M.y. ago. Nature, 260: 513-515. Keppie, J.D. and Dostal, J., 1991. Late Proterozoic tectonic model for the Avalon Terrane in Maritime Canada. Tectonics, 10: 842-850. Kerrich, R., 1992. Continents ring the changes. Nature, 359: 16-17. King, G. and Ellis, M., 1990. The origin of large local uplift in extensional regions. Nature, 348: 689-693. King, L.F. and Fader, G.B., 1986. Wisconsinan glaciation of the continental shelf, southeastern Atlantic Ocean. Geol. Surv. Can. Bull., 363, 72 pp. King, L.F. and MacLean, B., 1970. Pockmarks on the Scotian Shelf. Geol. Soc. Am. Bull., 81: 3141-3148. Kingsley, C.S., 1984. Dagbreek fan-delta: An alluvial placer to prodelta sequence in the Proterozoic Welkom goldfield, Witwatersrand, South Africa. In: E.H. Koster and R.J. Steel (Editors), Sedimentoiogy of Gravels and Conglomerates. Can. Soc. Pet. Geol. Mem., 10: 321-330. Klein, G., de V. and Kupperman, J.B., 1992. Pennsylvanian cyclothems: Methods of distinguishing tectonically induced changes in sea level from climatically-induced changes. Bull. Geol. Soc. Am., 104: 166-175. Klein, G., de V. and Willard, D.A., 1989. Origin of the Pennsylvanian coal-bearing cyclothems of North America. Geology, 17: 152-155. Klein, G., de V. and Hsui, A.T., 1987. Origin of cratonic basins. Geology, 19: 330-342. Klitgard, K.D., Hutchinson, D.R. and Schouten, H., 1988. U.S. Atlantic continental margin: Structural and tectonic framework. In: R.E. Sheridan and J.A. Graw (Editors), The Atlantic Continental Margin, U.S. The Geology of North America 1-2. Geol. Soc. Am., pp. 19-55. Knoll, A.H., 1991. End of Proterozoic Eon. Sci. Am., 265: 64-73. Knoll, A.H. and Walters, 1992. Latest Proterozoic stratigraphy and Earth history. Nature, 356: 673678. Kobluk, D.R., 1984. Coastal palaeokarst near the Ordovician-Silurian boundary, Manitoulin Island, Ontario. Can. Bull. Pet. Geol., 32: 398-407. Koepnick, R.B., Denison, R.E. and Dahl, D.A., 1988. The Cenozoic seawater 87Sr/86Sr curve: Data review and implications for correlation of marine strata. Paleoceanography, 3: 743-756. Koster, E.H. and Steel, R.J. (Editors), 1984. Sedimen-

N. EVLES tology of Gravels and Conglomerates. Can. Soc. Pet. Geol. Mem., 10, 441 pp. Kriftoffersen, Y. and Hinz, K., 1991. Crustal development: Weddell Sea-Ross Sea region. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 225-230. Krogh, T.E., Davis, D.W. and Cirfu, F. 1984. Precise U-Pb zircon and badddeleyite ages for the Sudbury area. In: E.G. Pye, A.J. Naldrett and P.E. Giblin (Editors), The Geology and Ore Deposits of the Sudbury Structure. Ontario Geol. Surv. Spec. Vol., 1: 443- 446. Kruck, W. and Thiele, J., 1983. Late Palaeozoic glacial deposits in the Yemen Arab Republic. Geol. Jahrb. Reihe B, 46: 3-29. Kuhle, M., 1987. Subtropical mountain and highland glaciation as ice age triggers and the waning of the glacial periods in the Pleistocene. GeoJournal, 14: 393-421. Kuhn, W.R., 1992. Avoiding a permanent ice age. Nature, 359: 196-197. Kuhn, W.R., Walker, J.C.G. and Marshall, H.G., 1989. The effect on Earth's surface temperature from variations in rotation rate, continent formation, solar luminosity and carbon dioxide. J. Geophys. Res., 94: 11,129-11,136. Kulka, P.A. and Stanistreet, I.G., 1991. Record of the Damaran Khomas Hochland accretionary prism in central Namibia: Refutation of an "ensialic" origin of a Late Proterozoic orogenic belt. Geology, 19: 473 -476. Kumpulainen, R. and Nystuen, J.P., 1985. Late Proterozoic basin evolution and sedimentation in the westernmost part of Baltoscandia. In: D.G. Gee and B.A. Sturt (Editors), The Caledonide Orogen-Scandinavia and Related Areas. Wiley, New York, pp. 213-232. Kurtz, D.D., 1981. Early Proterozoic diamictites of the Black Hills, South Dakota. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 800-802. Kutzbach, J.E., Guetter, P.J., Ruddiman, W.F. and Prell, W.L., 1989. Sensitivity of climate to Late Cenozoic uplift in Southern Asia and the American West: Numberical Experiments. J. Geophys. Res., 94: 18,393-18,407. Lagoe, M.B., Eyles, C.H. and Eyles, N., 1989. Paleoenvironmental and paleoclimatic significance of foraminiferal biofacies in the glaciomarine Yakataga Formation, Middleton Island, Gulf of Alaska. J. Foraminiferal Res., 19: 194-204. Lagoe, M.B., Eyles, C.H., Eyles, N. and Hale, C., 1993. Dating the onset of Late Ceonzoic glaciation in the north Pacific Ocean. Geol. Soc. Am. Bull.: In press.

E A R T H ' S G L A C I A L R E C O R D A N D ITS T E C T O N I C S E T T I N G

Laird, 1972. Stratigraphy and sedimentology of the Laksefjord Group, Finnmark. Nor. Geol. Unders., 278: 13-40. Lajtai, E.Z., 1967. The origin of some varves in Toronto, Canada. Can. J. Earth Sci., 4: 633-639. Lambert, A.M. and Hsu, K.J., 1979a. Non-annual cycles of varve-like sedimentation in Walensee, Switzerland. Sedimentology, 26: 453-461. Lambert, A.M. and Hsu, K.J., 1979b. Varve-like sediments of the Walensee, Switzerland. In" C. Schluchter (Editor), Moraines and Varves: Origin/Genesis/Classification. Balkema, Rotterdam, pp. 287294. Lambert, I.B. and Donnelly, T.H., 1991. Atmospheric oxygen levels in the Precambrian: a review of isotopic and geological evidence. Global Planet. Change, 97: 83-91. Larter, R.D. and Barker, P.F., 1989. Seismic stratigraphy of the Antarctic Peninsula Pacific margin: A record of Pliocene-Pleistocene ice volume and paleoclimate. Geology, 17: 731-734. Larter, R.D. and Barker, P.F., 1991. Neogene interaction of tectonic and glacial processes at the Pacific margin of the Antarctic Peninsula. In: D.I.M. MacDonald (Editor), Sedimentation, Tectonics, Eustasy: Sea-Level Changes At Active Margins. Int. Assoc. Sedimentol. Spec. Publ., 12: 165-186. Larter, R.D. and Cunningham, A.P., 1993. The depositional pattern and distribution of glacial-interglacial sequences on the Antarctic Peninsula Pacific margin. Mar. Geol., 109: 203-219. Lash, G.G., 1987. Sedimentology and possible palaeoceanographic significance of mudstone turbidites and associated deposits of the Pen Argyl Member, Martinsburg Formation (Upper Ordovician), Eastern Pennsylvania. Sediment. Geol., 54: 113-135. Laskar, J., 1989. A numerical experiment on the chaotic behaviour of the solar system. Nature, 338: 237-238. Laskar, J., Joutel, F. and Robutel, P., 1993. Stabilization of the Earth's obliquity by the Moon. Nature, 361: 615-617. Laznicka, P., 1988. Breccias and Coarse Fragments. Developments in Economic Geology, 25. Elsevier, Amsterdam, 832 pp. Lea, P.D., 1990. Pleistocene periglacial aeolian deposits in southwestern Alaska: Sedimentary facies and depositional processes. J. Sediment. Petrol., 60: 582-591. Le Blanc-Smith, G. and Eriksson, K.A., 1979. A fluvioglacial and glaciolacustrine deltaic depositional model for Permo-Carboniferous coals of the northeastern Karoo Basin, South Africa. Palaeogeogr. Palaeoclimatol. Palaeoecol., 27: 67-84. Leblanc, M., 1975. Ophiolites precambriennes et gris arsenies de cobalt (Bou Azzer, Maroc). Ph.D. Thesis, Univ. Paris VI, Paris, 329 pp. Unpublished Leblanc, M., 1981. The Late Precambrian Tiddiline Tilloid of the Anti-Atlas, Morocco. In: M.J. Ham-

231

brey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 120-122. Legget, R.F. (Editor), 1976. Glacial Till: An Inter-disciplinary Study. R. Soc. Can. Spec. Publ., 12, Ottawa, 412 pp. Leggett, J.K., McKerrow, W.S., Cooks, L.R.M. and Rickards, R.B., 1981. Periodicity in the early Paleozoic realm. J. Geol. Soc. London, 138: 167-176. Legrand, P., 1985. Lower Palaeozoic rocks of Tunisia. In: Holland, C.H. (Editor), Lower Palaeozoic of North-Western and West-Central Africa. Lower Palaeozoic Rocks of the World, 4. Wiley, Chichester, pp. 1-4. Lehman, S.J. and Keigwin, L.D., 1992. Deep circulation revisited. Nature, 358: 197-198. Leighton, M.W., Kolata, D.R., Oltz, D.F. and Eidel, J.J. (Editors), 1990. Interior Cratonic Basins. Am. Assoc. Pet. Geol. Mem., 51, 819 pp. Leighton, M.W. and Kolata, D.R., 1990. Selected interior cratonic basins and their place in the scheme of global tectonics; a synthesis. In: M.W. Leighton, D.R. Kolata, D.F. Oltz and J.J. Eidel (Editors), Interior Cratonic Basins. Am. Assoc. Pet. Geol. Memoir, 51: 729-799. Lemos, R.S., Strachan, R.A. and Topley, C.G., 1990. The Cadomian Orogeny. Geol. Soc. Spec. Publ., 51, 423 pp. Levell, B.K., Braakman, J.H. and Rutlen, K.W., 1988. Oil-bearing sediments of Gondwana glaciation in Oman. Am. Assoc. Pet. Geol. Bull., 72: 775-796. Lewis, C.F.M. and Keen, J.J., 1990. Constraints to development. In: J.J. Keen and G.L. Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geol. Surv. Can., 2: 793-823. Lewis, C.F.M. and Woodworth-Lynas, C.M.T., 1990. Ice scour. In: Keen and Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geol. Surv. Canada, 2: 785-793. Lin, J. and Watts, D.R., 1988. Paleomagnetic constraints on Himalayan Tibetan tectonic evolution. In: R.M., Shackleton, J.F. Dewey and B.F. Windley (Editors), Tectonic Evolution of the Himalayas and Tibet. R. Soc. London, pp. 172-188. Lindsey, D.A., 1969. Glacigenic rocks in the Early Proterozoic Chibougamau Formation of northern Quebec. Geol. Soc. Am. Bull., 80: 1685-1702. Lindsey, D.A., 1971. Glacial marine sediments in the Precambrian Gowganda Formation of Whitefish Falls, Ontario, Canada. Palaeogeogr. Palaeoclimatol. Palaeoecol., 9: 7-25. Lindsay, J.F., 1966. Carboniferous subaqueous massmovement in the Manning-Macleay Basin, Kempsey, New South Wales. J. Sediment. Petrol., 36: 719-732. Lindsay, J.F., 1968. The development of clast fabric in mud flows. J. Sediment. Petrol., 38: 1242-1253. Lindsay, J.F., 1989. Depositional controls on glacial

232 facies associations in a basinal setting, Late Proterozoic, Amadeus Basin, Central Australia. Palaeogeogr. Palaeoclimatol. Palaeoecol., 73: 205-232. Lindsay, J.F., Korsch, R.J. and Wilford, J.R., 1987. Timing and breakup of a Proterozoic supercontinent: Evidence from Australian intracratonic basins. Geology, 15: 1061-1064. Lister, G.S., Etheridge, M.A. and Symonds, P.A., 1986. Detachment faulting and the evolution of passive continental margins. Geology, 14: 246-250. Lister, G.S., Etheridge, M.A. and Symonds, P.A., 1991. Detachment models for the formation of passive continental margins. Tectonics, 10: 1038-1064. Little, T., Holcombe, R.J., Gibson, G.M., Ottler, R., Gans, P.B. and McWilliams, M.O., 1992. Exhumation of late Paleozoic blueschists in Queensland, Australia, by extensional faulting. Geology, 20: 231234. Little, T.A., 1990. Kinematics of wrench and divergent-wrench deformation along a central part of the Border Ranges Fault system, Northern Chugach Mountains, Alaska. Tectonics, 9: 585-611. Long, D.G.F., 1974. Glacial and paraglacial genesis of conglomeratic rocks of the Chibougamau Formation (Aphebian), Chibougamau, Quebec. Can. J. Earth Sci., 11: 1236-1252. Long, D.G.F., 1981. Glacigenic rocks in the Early Proterozoic Chibougamau Formation of northern Quebec. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 817-820. Long, D.G.F., 1991. A non-glacial origin for the Ordovician (Middle Caradocian) Cosquer Formation, Vesyarc'h, Crozon Peninsula, Brittany, France. Geol. J., 26: 279-294. Long, J.A. and Suarez, J.M., 1978. Informacion sismica indicativa de erosion en el Carbonico de Bolivia, relacionada con te tecnica de anomalias de amplitud. Rev. Tec. Yacimientos Petroliferos Fiscales Boliv., 7: 51-74. Lonne, I. and Mangerud, J., 1991. An Early or Middle Weichselian sequence of proglacial, shallow marine sediments on western Svalbard. Boreas, 20: 85-104. Lopez Gamundi, O.R., 1989. Postglacial transgressions in Late Paleozoic basins of western Argentina; A record of glacioeustatic sea level rise. Palaeogeogr. Palaeoclimatol. Palaeoecol., 71: 257-270. Lorius, C., Jouzel, J., Raynaud, D., Hansen, J. and Le Treut, H.E., 1990. The ice-core record: climate sensitivity and future greenhouse warming. Nature, 347: 139-145. Lovelock, J., 1979. Gaia: A New Look at Life on Earth. Oxford Univ. Press. Lowe, D.R., 1976. Subaqueous liquefied and fluidized sediment flows and their deposits. Sedimentoiogy, 23: 285-308.

N. EYLES Lowe, D.R., 1982. Sediment gravity flows, II. Depositional models with special reference to the deposits of high density turbidity currents. J. Sediment. Petrol., 52: 279-297. Mackay, J.R.,1979. Pingos of the Tuktoyaktuk Peninsula area, Northwest territories. Geogr. Phys. Quaternaire, 33: 3-61. MacAyeal, D.R., 1992. Irregular oscillations of the West Antarctic ice sheet. Nature, 359: 29-32. MacCarthy, G.R. 1958. Glacial boulders on the Arctic coast of Alaska. Arctic, 11: 71-85. MacDonald, D.I.M. (Editor), 1991. Sedimentation, Tectonics, Eustasy: Sea-Level Changes at Active Margins. Int. Assoc. Sedimentol. Spec. Publ., 12. Mahaney, W.C., 1990. Microfabrics and quartz microstructures confirm glacial origin of the Sunnybrook drift in the Lake Ontario basin. Geology, 18: 145-148. Mahaney, W.C., Vortisch, W. and Julig, P., 1988. Relative differences between glacially crushed quartz transported by mountain and continental ice-some examples from North America and East Africa. Am. J. Sci., 288: 810-826. Mahmoud, M.D., Vaslet, D. and Husseini, M.I., 1992. The Lower Silurian Qalibah Formation of Saudi Arabia: An important hydrocarbon source rock. Am. Assoc. Pet. Geol. Bull., 76: 1491-1506. Malcuit, R.J. and Winters, R.R., 1986. The Late Proterozoic glaciations: Possible product of the evolution of the Earth-Moon system. Int. Geol. Congr. Abstr., 26: 600. Marincovich, L., 1990. Molluscan evidence for early middle Miocene marine glaciation in southern Alaska. Geol. Soc. Am. Bull., 102: 1591-1599. Marmo, J.S. and Ojakangas, R.W., 1984. Lower proterozoic glaciogenic deposits, eastern Finland. Geol. Soc. Am. Bull., 95: 1118-1128. Marshall, J.R. and Overbeck, V.R., 1992. Textures of impact deposits and the origin of tillites. Eos Abstr. Suppl., 73(43): 324. Martin, H., 1975. Structural and palaeogeographical evidence for an Upper Palaeozoic sea between southern Africa and South America. In: K.S.W. Campbell (Editor), Gondwana Geology. Australian National University Press, Canberra, pp. 37-51. Martin, H., 1981. The Late Paleozoic Dwyka Group of the South Kalahari Basin in Namibia and Botswana and the subglacial valleys of the Kaokoveld in Namibia. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 61-65. Martin, H. and Wilczewski, N., 1970. Palaeoecology, conditions of deposition and the paleogeography of the marine Dwyka beds of S.W. Africa. In: Second Gondwana Symposium. Geol. Soc. S. Afr., pp. 225232.

E A R T H ' S G L A C I A L R E C O R D AND ITS T E C T O N I C S E T T I N G

Martin, H., Porada, H. and Walliser, O.H., 1985. Mixtite deposits of the Damara sequence, Namibia: Problem of interpretation. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 159-196. Martin, M.D., Stanistreet, I.G. and Camden-Smith, P.M., 1989. The interaction between tectonics and mudflow deposits within the Main Conglomerate Formation in 2.8 to 2.7 Ga Witwatersrand Basin. Precambrian Res., 44: 19-38. Martinez, E.D., 1991. Litoestratigrafia del Carbonifero del Altiplano de Bolivia. Rev. Tec. Yaciementos Petroliferos Fiscales Boliv., 12: 295-302. Martini, I.P. and Glooschenko, W.A., 1985. Cold climate peat formation in Canada and its relevance to Lower Permian coal measures of Australia. EarthSci. Rev., 22: 107-140. Martini, I.P. and Johnson, D.P., 1987. Cold-climate, fluvial to paralic coal-forming environments in the Permian Collinsville Coal Measures, Bowen Basin, Australia. Int. J. Coal Geol., 7: 365-388. Martinson, D.G., Pisias, N.G., Hays, J.D., Imbrie, J., Moore, T.C. and Shackleton, N.J., 1987. Age dating and the orbital theory of the Ice Ages. Development of a high-resolution 0 to 300,000 year chronostratigraphy. Quat. Res., 27: 1-29. Masetti, D., Neri, C. and Bosellini, A., 1991. Deepwater asymmetric cycles and progradation of carbonate platforms governed by high frequency eustatic oscillations (Triassic of the dolomites, Italy). Geology, 19: 336-339. Mather, K.F. and Wengerd, S.A., 1965. Pleistocene age of the "Eocene" Ridgway Till, Colorado. Geol. Soc. Am. Bull., 76. Mathews, R.K., 1986. Oxygen isotope record of icevolume history: 100 million years of glacio-eustatic sea level flucuation. In: J.S. Schlee (Editor), Interregional Unconformities and Hydrocarbon Accumulation. Am. Assoc. Pet. Geol. Mem., 36: 97-107. Matsch, C.L. and Ojakangas, R.W., 1991. Comparisons in depositional style of "polar" and "temperate" glacial ice: Late Paleozoic Whiteout Conglomerate (West Antarctica) and late Proterozoic Mineral Fork Formation (Utah). In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation; Paleoclimatic Significance. Geol. Soc. Am. Spec. Publ., 261: 191-206. Max, M., 1991. National workshop on gas hydrates. Eos, 72: 476-477. McCann, A.M. and Kennedy, M.J., 1974. A probable glaciomarine deposit of Late Ordovician-Early Silurian age from the north central Newfoundland Appalachian belt. Geol. Mag., 111: 549-564. McClaren, C., 1842. The glacial theory of Professor Agassiz. Am. J. Sci., 42: 346-365. McClure, H.A., 1978. Early Paleozoic glaciation in Arabia: Palaeogeogr. Palaeoclimatol. Palaeoecol., 25, 315-326.

233

McClure, H.A., 1980. Permian-Carboniferous glaciation in the Arabian Peninsula. Geol. Soc. Am. Bull., 91: 707-712. McClure, H.A., Hussey, E. and Kaill, I., 1988. Permian-Carboniferous glacial deposits in southern Saudi Arabia. Geol. Jahrb. Reihe B, 68: 3-31. McElhinny, M.W., Giddings, J.W. and Embleton, B.J.J., 1974. Palaeomagnetic results and late Precambrian glaciations. Nature, 248: 557-561. McGillivray, J.G. and Husseini, M.I., 1992. The Paleozoic Petroleum geology of Central Arabia. Am. Assoc. Pet. Geol. Bull., 76: 1473-1490. McKelvey, B.C., 1981. Carboniferous tillites in the New England area of New South Wales. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 476-479. McKelvey, B.C., Webb, P.N., Harwood, D.M. and Mabin, M.C.G., 1991. The Dominion Range Sirius Group: a record of the late Pliocene-early Pleistocene Beardmore Glacier. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 675-682. McPhie, J., 1987. Andean analogue for Late Carboniferous volcanic arc and arc flank environments of the western New England Orogen, New South Wales, Australia. Tectonophysics, 138: 269-288. Meert, J.G. and Van der Voo, R.,1992. Evidence for a high paleolatitude of North America in the latest Precambrian and rapid drift during the Cambrian. Eos Abstr. Suppl., 73(43): 150. Mercer, J.H., 1986. Southernmost Chile; a modern analog of the southern shores of the Ross embayment during Pliocene warm intervals. Antarct. J.U. S., 21: 103-105. Mercer, J.H. and Sutler, J.F., 1982. Late MioceneEarliest Pliocene glaciation in southern Argentina. Implications for global ice sheet history. Palaeogeogr. Palaeoclimatol. Palaeoecol., 38: 185-206. Metcalfe, I., 1988. Origin and assembly of south-east Asian continental terranes. In: M.G. Audley-Charles and A. Hallam (Editors), Gondwana and Tethys. Geol. Soc. Spec. Publ., 37: 101-118. Miall, A.D., 1983a. Glaciomarine sedimentation in the Gowganda Formation (Huronian), northern Ontario. J. Sediment. Petrol., 53: 477-491. Miall, A.D., 1983b. Basin analysis of fluvial sediments. In: J.D. Collinson and J. Lewin (Editors), Modern and Ancient Fluvial Systems. Int. Assoc. Sedimentol. Spec. Publ., 6: 279-286. MiaU, A.D., 1985. Sedimentation on an early Proterozoic continental margin under glacial influence: the Gowganda Formation (Huronian), Elliot Lake area, Ontario, Canada. Sedimentology, 32: 763-788. Miall, A.D., 1986. Eustatic sea level changes interpreted from seismic stratigraphy: a critique of the

234 methodology with particular reference to the North Sea Jurassic record. Am. Assoc. Pet. Geol. Bull., 70: 131-137. Miall, A.D., 1990. Principles of Sedimentary Basin Analysis. Springer, New York, 668 pp. Miall, A.D., 1991. Stratigraphic sequences and their chronostratigraphic correlation. J. Sediment. Petrol., 61: 497-505. Miall, A.D., 1992. Exxon global cycle chart: An event for every occasion? Geology, 20: 787-790. Middleton, G.V., 1993. Sediment deposition from turbidity currents. Annu. Rev. Earth Planet. Sci., 21: 89-114. Middleton P.D., Marshall, J.D. and Brenchley, P.J.,1991. Evidence for isotopic changes associated with Late Ordovician glaciation from brachiopods and marine cements from central Sweden. Geol. Surv. Can. Pap., 90-9: 313-324. Miller, G.H. and Vernal, A., 1992. Will greenhouse warming lead to Northern Hemisphere ice sheet growth? Nature, 355: 244-246. Miller, H., 1984. Orogenic development of the Argentinian/Chilean Andes during the Palaeozoic. J. Geol. Soc. London, 141: 885-892. Miller, J.M.G., 1985. Glacial and syntectonic sedimentation: The Upper Proterozoic Kingston Peak Formation, southern Panamint Range, eastern California. Geol. Soc. Am. Bull., 96: 1537-1553. Miller, J.M.G. and Waugh, B., 1991. Permo-Carboniferous glacial sedimentation in the central Transantarctic Mountains and its palaeotectonic significance. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 205208. Miller, K.G., Fairbanks, R.G. and Mountain, G.S., 1987. Tertiary oxygen isotope synthesis, sea level history and continental margin erosion. Paleoceanography, 2: 1-19. Miller, R.M., 1983. The Pan-African Damara Orogen of South West Africa/Namibia. Geol. Soc. S. Afr. Spec. Publ., 11: 431-515. Mills, H.H., 1984. Clast orientation in Mount St. Helens debris flow deposits, North Fork, Toutle River, Washington. J. Sediment. Petrol., 54: 626-634. Milnes, A.R. and Bourman, R.P., 1972. A Late Palaeozoic glaciated granite surface at Port Elliot, South Australia. Trans. R. Soc. South Aust., 96: 149-155. Molnar, P., 1990. The rise of mountain ranges and the evolution of humans: A causal relation? Ir. J. Earth Sci., 10: 199-207. Molnar, P. and England, P., 1990. Late Cenozoic uplift of mountain ranges and global climate change: chicken or egg? Nature, 346: 29-34. Moncrieff, A.C.M., 1989. The Tillite Group and related rocks of East Greenland: implications for Late Proterozoic palaeogeography. In: R.A. Gayer (Edi-

N. EYLES tor), The Caledonide Geology of Scandinavia. Graham and Trotman, London, pp. 285-297. Moncrieff, A.C.M. and Hambrey, M.J., 1988. Late Precambrian glacially-related grooved and striated surfaces in the Tillite Group of central East Greenland. Palaeogeogr. Palaeoclimatol. Palaeoecol., 65: 183-200. Moores, E.M., 1991. Southwest U.S.-East Antarctic (SWEAT) connection: A hypothesis. Geology, 19: 425 -428. Morel, P. and Irwing, E., 1978. Tentative Palaeocontinental maps for the early Phanerozoic and Proterozoic. J. Geol., 86: 535-561. Morner, N.A., 1980. Earth movements, paleoceanography, paleoclimatology and eustasy: major cenozoic events in the North Atlantic. Geol. Foren. Stockholm Forh., 102: 261-268. Morner, N.A., 1987. Pre-Quaternary long-term changes in sea-level. In: R.J.N. Devoy (Editor), Sea Surface Studies: A Global View. Croom Helm, New York, pp. 233-241. Morris, W.A., 1977. Paleolatitude of glaciogenic upper Precambrian Rapitan Group and the use of tillites as chronostratigraphic marker horizons. Geology, 5: 85-88. Mpodozis, C. and Kay, S.M., 1992. Late Paleozoic to Triassic evolution of the Gondwana margin: Evidence from Chilean Frontal Cordilleran batholiths (28°S to 31°S). Geol. Soc. Am. Bull., 104: 999-1014. Murphy, J.B. and Nance, R.D., 1991. Supercontinent model for the contrasting character of Late Proterozoic orogenic belts. Geology, 19: 469-472. Murray, C.D., 1993. Seasoned travellers. Nature, 361: 586. Mustard, P.S., 1991. Normal faulting and alluvial-fan deposition, basal Windermere Tectonic Assemblage, Yukon, Canada. Geol. Soc. Am. Bull., 103: 1346-1364. Mustard, P.S. and Donaldson, J.A., 1987a. Early Proterozoic ice-proximal glaciomarine deposition: The lower Gowganda Formation at Cobalt, Ontario, Canada. Geol. Soc. Am. Bull., 98: 373-387. Mustard, P.S. and Donaldson, J.A., 1987b. Substrate guarrying and subglacial till deposition by an early Proterozoic ice sheet: Evidence from the Gowganda Formation at Cobalt, Ontario, Canada. Precambrian Res., 34: 347-368. Mustard, P.S. and Donaldson, J.A., 1990. Paleokarst breccias, calcretes, silcretes and fault talus breccias at the base of Upper Proterozoic Windermere strata, Northern Canadian Cordillera. J. Sediment. Petrol., 60: 525-539. Mutter, J.C., Talwani, M. and Stoffa, P.L., 1982. Origin of seaward dipping reflectors in oceanic crust off the Norwegian margin by "subaerial sea-floor spreading". Geology, 10: 353-357. Myers, J.S., 1990. Precambrian tectonic evolution of

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T T I N G

part of Gondwana, southwestern Australia. Geology, 18: 537-540. Myers, J., 1993. Precambrian history of the West Australian Craton and adjacent orogens. Annu. Rev. Earth Planet. Sci., 21: 453-486. Nance, R.D., Murphy, J.B., Strachan, R.A., D'Lemos, R.S. and Taylor, G.K., 1991. Late Proterozoic tectonostratigraphic evolution of the Avalonian and Cadomian terranes. Precambrian Res., 53: 41-78. Negrutsa, T.F. and Negrutsa, V.Z., 1981. Archean Pebozero tilloids of Karelia, U.S.S.R.. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 687-688. Nemec, W. and Steel, R.J. (Editors), 1988. Fan Deltas: Sedimentology and Tectonic Settings. Blackie, London, 444 pp. Nemec, W., Steel, R.J., Porebski, S.J. and Spinnagr, A., 1984. Domba Conglomerate Devonian, Norway: Process and lateral variability in a mass flowdominated lacustrine fan-delta. In: E.H. Koster and R.J. Steel (Editors), Sedimentology of Gravels and Conglomerates. Mem. Can. Soc. Petrol. Geol., 10: 295-320. Nesbitt, H.W. and Young, G.M., 1982. Early Proterozoic climates and plate motions inferred from major element chemistry of lutites. Nature: 715-717. Newell, N.D., 1957. Supposed Permian tillites in northern Mexico are submarine slide deposits. Bull. Geol. Soc. Am., 68: 1569-1576. Newman, M.J. and Rood, R.T., 1977. Implications of solar evolution for the Earth's early atmosphere. Science, 198: 1035-1035. Nisbet, E.G., 1990. The end of the ice age. Can. J. Earth Sci., 27: 148-157. Nummedal, D. et al., 1974. Recent migration of the Skeidarasandur shoreline, southeast Iceland. Final Rep., NG0921-73-6-0258. Naval Ordinance Lab. Dep. Geol. Univ. South Carolina, 183 pp. Nystuen, J.P., 1976a. Facies and sedimentation of the Late Precambrian Moelv Tillite in the eastern part of the Sparagmite Region, southern Norway. Nor. Geol. Unders., 329: 1-70. Nystuen, J.P., 1976b. Late Precambrian Moelv tillite deposited on a discontinuity surface associated with a fossil ice wedge, Rendalen, southern Norway. Nor. Geol. Tidsskr., 56: 29-56. Nystuen, J.P., 1982. Late Proterozoic Basin evolution on the Baltoscandian Craton: the Hedmark Group, Southern Norway. Nor. Geol. Unders., 375: 1-74. Nystuen, J.P., 1985. Facies and preservation of glaciogenic sequences from the Varanger Ice Age in Scandinavia and other parts of the North Atlantic region. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 209-229. Nystuen, J.P., 1987. Synthesis of the tectonic and sedi-

235

mentological evolution of the late Proterozoic-early Cambrian Hedmark Basin, the Caledonian Thrust Belt, southern Norway. Nor. Geol. Tidsskr., 67: 395-418. Nystuen, J.P. and Saether, T., 1979. Clast studies in the Late Precambrian Moelv Tillite and Osdal Conglomerate, Sparagmite Region, south Norway. Nor. Geogr. Tidsskr., 59: 239-251. Nystuen, J.P. and Kumpulainen, R., 1981. Upper Proterozoic basin evolution and sedimentation on the western margin of the Baltoscandian Craton, Sothern and Central Scandinavia. Terra Cognita, 1: 62. O'Brien, P.E. and Christie-Blick, N., 1992. Glacially grooved surfaces in the Grant Group, Grant Range, Canning Basin and the extent of Late Palaeozoic Pilbara ice sheets. J. Aust. Geol. Geophys., 13: 87-92. O'Brien, P.E., Lindsay, J.F., Southgate, P.N., Jackson, M.J., Rennard, J.M. and Sexton, M.J., 1992. Sequence stratigraphy of glacial sediments in an intracratonic basin--Grant Group, Canning Basin, Western Australia. Am. Assoc. Pet. Geol. Bull., 76: 1120. Oberbeck, V.R., Marshall, J.R. and Aggarwal, H., 1993. Impacts, tillites and the breakup of Gondwanaland. J. Geol., 101: 1-20. Oca, I.M., 1989. Geografia y Recursos Naturales de Bolivia. Acad. Nac. Cienc. Bolivia, La Paz, 574 pp. Oeiofsen, B.W., 1987. The biostratigraphy and fossils of the Whitehill and Irati Shale Formations of the Karoo and Parana Basins. In: G.D. McKenzie (Editor), Gondwana 6. Geophys. Monogr. Am. Geophys. Union, pp. 131-138. Oerlemans, J., 1982. A model of the Antarctic Ice Sheet. Nature, 297: 550-553. Oglesby, R.J., 1989. A GCM study of Antarctic glaciation. Clim. Dyn., 3: 135-156. Ojakangas, R.W., 1988. Glaciation: An uncommon "mega-event" as a key to intracontinental and intercontinental correlation of Early Proterozoic basin fill, North American and Baltic Cratons. In: K.L. Kleinspehn and C. Paola (Editors), New Perspectives in Basin Analysis. Springer, New York, pp. 431-444. Ojakangas, R.W. and Matsch, C.L., 1980. Uper Precambrian (Eocambrian) Mineral Fork Tiilite of Utah: a continental glacial and glaciomarine sequence. Geol. Soc. Am. Bull., 91: 495-501. Oliveira, L.O.A., 1989. Aspectos da evolucao termomecanica da bacia do Parana no Brasil. Rev. Bras. Geocienc., 19: 330-342. Osleger, D. and Read, J.F., 1992. Relation of eustasy to stacking patterns of meter-scale carbonate cycles, Late Cambrian, U.S.A.. J. Sediment. Petrol., 61: 1225-1252. Ovenshine, A.T., 1970. Observations of iceberg rafting

236 in Glacier Bay, Alaska and the identification of ancient ice rafted deposits. Geol. Soc. Am. Bull., 81: 891-894. Page, N.J., 1981. The Precambrian diamictite below the base of the Stillwater Complex, Montana. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 821-825. Page, N.J., 1971. Subglacial limestone deposition in the Canadian Rocky Mountains. Nature, 229: 42-43. Palmer, M.R. and Edmond, J.M., 1989. The strontium isotopic budget of the modern ocean. Earth Planet. Sci. Lett., 92: 11-26. Park, J. and Oglesby, R.J., 1991. Milankovitch rhythms in the Cretaceous: A GCM modelling study. Palaeogeogr. Palaeoclimatol. Palaeoecol., 90: 329-355. Patterson, J.G. and Heaman, L.M. 1991. New geochronologic limits on the depositional age of the Hurwitz Group, Trans-Hudson hinterland, Canada. Geology, 19: 1137-1140. Patterson, W.S.B., 1972. Laurentide ice sheet; Estimated volumes during the late Wisconsin. Rev. Geophys. Space Phys., 10: 885-917. Paul, M.A. and Eyles, N., 1990. Constraints on the preservation of diamict facies (melt-out tills) at the margins of stagnant glaciers. Quat. Sci. Rev., 9: 51-69. Paull, C.K., Ussler, W., III and Dillon, W.P., 1991. Is the extent of glaciation limited by marine gas-hydrates? Geophys. Res. Lett., 18: 432-434. Peach, B.N. and Horne, J., 1930. Chapters on the Geology of Scotland. Oxford University Press. Pelletier, B.R., 1966. Development of submarine physiography in the Canadian Arctic and its relation to crustal movements. In: G.D. Garland (Editor), Continental Drift. R. Soc. Can. Spec. Publ., 9: 77-101. Pesonen, L.J., 1990. The drift of Fennoscandia during the Proterozoic with special reference to the Bergslagen province, south-central Sweden; a reply. Geol. Foren. Stockholm Forh., 113: 251-253. Pesonen, L.J., Torsvik, T.H., Elming, S.A. and Bylund, G., 1989. Crustal evolution of Fennoscandia-palaeomagnetic constraints. Tectonophysics, 162: 27-49. Pettijohn, F., 1957. Sedimentary Rocks. Harper, New York, 718 pp. Pfirman, S., Gascard, J.C., Wollenburg, I., Mudie, P. and Abelmann, A., 1989. Particle laden Eurasian Arctic sea ice: observations from July and August 1987. Polar Res., 7: 59-66. Phillips, A.C., Smith, N.D. and Poweil, R.D. (Editors), 1991. Laminated sediments in prodelta deposits, Glacier Bay, Alaska. Glacial Marine Sedimentation. Geol. Soc. Am. Spec. Pap., 261: 51-60. Pickerell, R.K., Pajari, G.E. and Currie, K.L., 1979. Evidence of Caradocian glaciation in the Davidsville

N. EVLES group of northeastern Newfoundland. Geol. Surv. Can. Pap., 79-1c: 67-72. Pickering, K.T., Hiscott, R.N. and Hein, F.J., 1989. Deep-Marine Sedimentation. Unwin Hyman, London, 416 pp. Pickrill, R.A. and Irwin, J., 1983. Sedimentation in a deep glacier-fed lake--Lake Tekapo, New Zealand. Sedimentoiogy, 30:63-75 Pimentel, M.M. and Fuck, R.A., 1992. Neoproterozoic crustal accretion in central Brazil. Geology, 20: 375-379. Piper, D.J.W., Normark, W.R. and Sparkes, R., 1987. Late Cenozoic stratigraphy of the Central Scotian Slope, eastern Canada. Bull. Can. Pet. Geol., 35: 1-11. Piper, D.J.W., Shor, A.N. and Hughes-Clarke, A.E., 1988. The 1929 "Grand Banks" earthquake, slump and turbidity current. Geol. Soc. Am. Spec. Pap., 229: 77-92. Piper, D.J.W. and Normark, W.R., 1989. Late Cenozoic sea-level changes and the onset of glaciation: impact on continental slope progradation off eastern Canada. Mar. Pet. Geol., 6: 336-348. Piper, D.J.W., Mudie, P.J., Fader, G.B., Josenhans, H.W., MacLean, B. and Vilks, G., 1990. Quaternary History. In: M.J. Keen and G.L. Williams (Editors), Geology of the Continental Margin of Eastern Canada. Geol. Surv. Can., 2: 475-607. Piper, D.J.W. and Stow, D.A.V., 1991. Fine-grained turbidites. In: G. Einsele, W. Richen and A. Seilacher (Editors), Cycles and Events in Stratigraphy. Springer, Berlin, pp. 360-376 Piper, D.J.W., van Huene, R. and Duncan, J.R. 1973. Late Quaternary sedimentation in the active eastern Aleutian Trench. Geology, 1: 19-22. Piper, J.D.A., 1981. Palaeomagnetic study of the (Late Precambrian) West Greenland Kimberlife-Lamprophyre suite: definition of the Hadrynian Track. Phys. Earth Planet. Inter., 27: 164-186. Piper, J.D.A., 1983. Dynamics of the continental crust in Proterozoic times. In: L.G.J. Medaris, C.W. Byers, D.M. Mickelson and W.C. Shanks (Editors), Proterozoic Geology. Geol. Soc. Am. Mem., 161: 11-34. Piper, J.D.A., 1985. Continental breakup and dispersal in Late Precambrian-Early Cambrian times: prelude to Caledonian orogenesis. In: D.B. Gee and G.A. Sturt (Editors), The Caledonide Orogen-Scandinavia and Related Areas. Wiley, New York, pp. 19-34. Pique, A. and Skehan, J.W., 1992. Late Paleozoic orogenies in Western Africa and Eastern North America: The diachronous closure of the Theic Ocean. Tectonics, 11: 392-404. Plafker, G., 1987. Regional geology and petroleum potential of the northern Gulf of Alaska continental

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T T I N G

margin. In: D.W. Scholl, A. Grantz and J.G. Vedder (Editors), Geology and Resource Potential of the Continental Margin of Western North America and Ajacent Ocean Basin. Earth Science Series, 7. Circum-Pacific Council for Energy and Mineral Resources, Houston, pp. 229-268. Plafker, G., Richter, D.H. and Hudson, T., 1977. Reinterpretation of the origin of inferred Tertiary tillite in the northern Wrangell Mountains, Alaska. In: K.M. Blean (Editor), The United States Geological Survey in Alaska: Accomplishments during 1976. U.S. Geol. Surv. Circ., 751-B: 52-54. Planke, S., Skogseid, J. and Eldholm, O., 1991. Crustal structure off Norway, 62° to 70°N north. Tectonophysics, 189: 91-107. Plint, A.G., 1991. High frequency relative sea-level oscillations in Upper Cretaceous shelf clastics of the Alberta foreland basin: Possible evidence for a glacio-eustatic control? In: D.I.M. MacDonald (Editor), Sedimentation, Tectonics and Eustasy. Int. Assoc. Sedimentol. Spec. Publ., 12: 409-428. Plint, A.G., Eyles, N., Eyles, C.H. and Walker, R.G., 1992. Control of sea-level change. In: R.G. Walker and N.P.James (Editors), Facies Models: Response to Sea-level Change. Geol. Assoc. Can., pp. 15-25 Plumb, K.A., 1981. Late Proterozoic (Adelaidean) tillites of the Kimberley-Victoria River region, Western Australia and Northern Territory. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 504-536. Plumb, K.A., 1991. New Precambrian time scale. Episodes, 14: 139-140. Porebski, S.J. and Gradzinski, R., 1987. Depositional history of the Polonez Cove Formation (Oligocene), King George Island, West Antarctica: a record of continental glaciation, shallow-marine sedimentation and contemporaneous volcanism. Stud. Geol. Pol., XCIII: 7-62. Porebski, S.J., Meischner, D. and Gorlich, K., 1991. Quaternary mud turbidites from the South Shetland Trench (West Antarctica): recognition and implications for turbidite facies modelling. Sedimentology, 38: 691-715. Porter, S.E. and Orombelli, G., 1981. Alpine rockfall hazards. Science, 69: 67-75. Posamentier, H.W., Jervey, M.T. and Vail, P.R., 1988. Eustatic controls on clastic deposition i-conceptual framework. In: C.K. Wilgus, B.S. Hastings, H.W. Posamentier, C.A. Ross and C.G.S. Kendall (Editors), Sea Level Changes: An Integrated Approach. Soc. Econ. Paleontol. Mineral. Spec. Publ., 42: 109124. Postma, G., 1986. Classification for sediment gravity flow deposits based on flow conditions during sedimentation. Geology, 14: 291-294.

237

Postma, G., Roep, T.R. and Ruegg, G.H.J., 1983. Sandy gravelly mass flow deposits in an ice marginal lake (Saalian, Leuvenumsche Beek Valley, Veluwe, the Netherlands) with emphasis on plug-flow deposits. Sediment. Geol., 34: 59-82. Postma, G., Nemec, W. and Kleinspehn, K.L., 1988. Large floating clasts in turbidites: a mechanism for their emplacement. Sediment. Geol., 58: 47-61. Powell, C.M. and Veevers, J.J., 1987. Namurian uplift in Australia and South America triggered the main Gondwanan glaciation. Nature, 326: 177-179. Powell, R.D., 1981. A model for sedimentation by tidewater glaciers. Ann. Giaciol., 2: 129-134. Powell, R.D., 1984. Glacimarine processes and inductive lithofacies modelling of ice shelf and tidewater glacier sediments based on Quaternary examples. Mar. Geol., 57: 1-52. Powell, R.D., 1990. Glacimarine processes at grounding-line fans and their growth to ice-contact deltas. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glaciomarine Environments. Geol. Soc. Spec. Publ., 53: 53-74. Powell, R.D. and Elverhoi, A., 1989. Modern glacimarine environments: glacial and marine controls of modern lithofacies and biofacies. Mar. Geol., 85: 101-418. Prentice, M.L. and Matthews, R.K., 1988. Cenozoic ice-volume history. Development of a composite oxygen isotope record. Geology, 16: 963-966. Proust, J.N., Deynoux, M. and Guillocheav, F., 1990. Effets conjugues de l'estatisme et de risostasie sur les plates-formes stables en periode glaciaire: Exemple des depots glaciares du Proterozoic superior de L'Afrique de l'Ouest au Mali occidental. Bull. Soc. Geol. Fr., 6: 637-681. Quinlan, G.M. and Beaumont, C., 1984. Appalachian thrusting, lilthospheric flexure and the Paleozoic stratigraphy of the eastern interior of North America. Can. J. Earth Sci., 21: 973-996. Ramaswamy, V., 1992. Explosive start to last ice age. Nature, 359: 14. Ramli and Crook, K.A.W., 1978. Early Permian depositional environments, southern Sydney Basin. APEA J., 18: 70-76. Ramos, V.A., 1989. The birth of southern South America. Am. Sci., 77: 444-450. Rampino, M.R., 1991. Volcanism, climate change and the geologic record. In: R.V. Fisher and G.A. Smith (Editors), Volcanism, Tectonics and Sedimentation. Soc. Econ. Palaeontol. Mineral. Spec. Publ., 45: 9-18. Rampino, M.R. and Self, S., 1992. Volcanic winter and accelerated glaciation following the Toba supereruption. Nature, 359: 50-52. Ramsay, A.E., 1855. On the occurrence of angular, subangular, polished and stratified fragments and

238 boulders in the Permian breccia of Shropshire. Q.J. Geol. Soc. London, 11: 185-205. Rao, C.P., 1981. Geochemical differences between tropical (Ordovician) and subpolar (Permain) carbonates, Tasmania, Australia. Geology, 89: 205-209. Rao, P.C. and Green, D.C., 1982. Oxygen and carbon isotopes of early Permian cold-water carbonates, Tasmania, Australia. J. Sediment. Petrol., 52: 11111125. Rappol, M., 1985. Clast fabric strength in tills and debris flows compared for different environments. Geol. Mijnbouw, 64: 327-332. Raup, D.M. and Sepkovsky, J.J., 1982. Mass extinctions in the marine fossil record. Science, 215: 1501-1503. Raymo, M.E., 1991. Geochemical evidence supporting T.C. Chamberlin's theory of glaciation. Geology, 19: 344-348. Raymo, M.E., Ruddiman, W.F., Backman, J., Clement, B.M. and Martinson, D.G., 1989. Late Pliocene variation in northern hemisphere ice sheets and north Atlantic deep water circulation. Paleoceanography, 4: 413-446. Raymo, M.E. and Ruddiman, W.F., 1992. Tectonic forcing of Late Cenozoic climate. Nature, 359: 117122. Raymo, M.E. and Ruddiman, W.F., 1993. Tectonic forcing of Late Cenozoic climate: Reply. Nature, 361: 124. Raymond, A., 1987. Comment and reply on "Frasnian and Fammenian mass extinctions and cold water oceans." Geology, 15:777-778 Reading, H.G. and Walker, R.G., 1966. Sedimentation of Eocambrian tillites and associated sediments in Finnmark, Northern Norway. Palaeogeogr. Palaeoclimatol. Palaeoecol., 2: 177-212. Redfern, J., 1991. Subsurface facies analysis of Permocarboniferous glaciogenic sediments, Canning Basin, Western Australia. In: H. Ulbrich and A.C. RochaCampos (Editors), Gondwana Seven Proceedings. Inst. Geoscienc. Univ. Sao Paulo, pp. 349-363. Reimnitz, E., Marinkovich, L., McCormick M. amd Briggs, W.M., 1992. Suspension freezing of bottom sediment and biota in the Northwest Passge and implications for Arctic Ocean sedimentation. Can. J. Earth Sci., 29: 693-703. Rendu, M.L.C., 1840. Theorie des Glaciers de la Savoie. Chez Puthod, Chambery. Reprinted by Macmillan and Co., London, 1874. Retallack, G.J., 1990. Soils of the Past. Unwin Hyman, Boston, 520 pp. Reusch, H., 1891. Skruingsmerker og moraenegrus eftervist i finmarken fra en period megetaeldre end "istiden". Nor. Geol. Unders., 1: 97-100. Rich, P.V., Rich, T.H., Wagstaff, B.E., Mason, J.M., Douthitt, C.B., Gregory, R.T. and Felton, E.A., 1988. Evidence for low temperatures and biologic

N. EYLES diversity in Cretaceous high latitudes of Australia. Science, 242: 1403-1406. Riding, R., 1992. The algal breath of life. Nature, 359: 13-14. Robardet, M. and Dore, F., 1988. The Late Ordovician diamictic formations from southwestern Europe: North Gondwana glaciomarine deposits. Palaeogeogr. Palaeoclimatol. Palaeoecol., 66: 19-31. Roberts, J.D., 1971. Later Precambrian glaciation: an anti-greenhouse effect? Nature, 234: 216-217. Roberts, J.D., 1976. Late Precambrian dolomites, Vendian glaciation and synchroneity of Vendian glaciations. J. Geol., 84: 47-63. Roberts, J.D., 1977. Discussion of Late Precambrian dolomites, Vendian glaciation and synchroneity of Vendian glaciations: A reply. J. Geol., 85: 251-252. Roberts, J.D. and Engel, B.A., 1980. Carboniferous palaeogeography of the Yarrol and New England Orogens, eastern Australia. J. Geol. Soc. Aust., 27: 167-186. Robineau, B. and Ritz, M., 1990. Geoelectrical signature of the central Mauritanides deep structure, Mauritania, West Africa. Tectonics, 9: 1649-1661. Robinson, J.M., 1991. Phanerozoic atmospheric reconstructions: a terrestrial perspective. Global Planet. Change, 97: 51-62. Robinson, J.M. and Upchurch, G., 1991. Silicate weathering and land plant evolution. Angiosperms vs. grasses as agents of change. In: Annual Meeting, Abstracts with Programs: A456. Geol. Soc. Am. San Diego, Rocha-Campos, A.C., 1981a, Early Palaeozoic Iapo Formation of Paranfi, Brazil. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 908-910. Rocha-Campos, A.C., 1981b, Late Ordovician-Early Silurian Trombetas Formation, Amazon Basin, Brazil. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 896-898. Rocha-Campos, A.C. and Rosier, O., 1978. Late Paleozoic Faunal and Floral Successions in the Paranfi basin, Southeastern Brazil. Bol. IG Inst. Geocienc. USP, 9: 1-16. Rocha-Campos, A.C. and dos Santos, P.R., 1981. Late Paleozoic tillites of the Lutoe Series, Angola. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 48-51. Rocha-Campos, A.C. and Hasui, Y., 1981. Tillites of the Macaubas Group (Proterozoic) in central Minas Gerais and southern Bahia, Brazil. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 933-939. Rocha-Campos, A.C. and Sundaram, D., 1981. Geolog-

E A R T H ' S G L A C IA L R E C O R D AND ITS T E C T O N I C S E T F I N G

ical and palynological observations on Late Paleozoic varvites from the Itarare Subgroup, Parana Basin, Brazil. Anais II Congr. Latino-Americo Paleontol. Porto Alegre, Brazil, pp. 257-275. Rodolfo, K.S. and Arguden, A.T., 1991. Rain-lahar generation and sediment delivery systems at Mayon Volcano, Philippines. In: R.V. Fisher and G.A. Smith (Editors), Sedimentation in Volcanic Settings. Soc. Econ. Palaeontol. Mineral. Spec. Publ., 45: 71-87. Roop, S., Mullins, H.T., Gartner, S., Huang, T.C., Joyce, E., Prutzman, J. and Tjalmsa, L., 1992. Climatic forcing of cyclic carbonate sedimentation during the last 5.4 million years along the west Florida continental margin. J. Sediment. Petrol., 61: 10701088. Roscoe, S.M., 1973. The Huronian Supergroup, a Paleoaphebian succession showing evidence of atmospheric evolution. In: G.M. Young (Editor), Huronian Stratigraphy and Sedimentation. Geol. Assoc. Can. Spec. Publ., 12: 31-47. Ross, C.A. and Ross, J.R.P., 1985. Late Paleozoic depositionai sequences are synchronous and worldwide. Geology, 13: 27-30. Ross, C.A. and Ross, J.R.P., 1988. Late Paleozoic transgressive-regressive deposition. In: C.K. Wilgus, B.S. Hastings, H.W. Posamentier, C.A. Ross and C.G.S. Kendall (Editors), Sea Level Changes: An Integrated Approach. Soc. Econ. Paleontol. Mineral. Spec. Publ., 42: 227-247. Ross, G.M., 1991. Tectonic setting of the Windermere Supergroup revisited. Geology, 19: 1125-1128. Ruddiman, W.F., Raymo, M. and Mclntyre, A., 1986. Matuyama 41,000-year cycles; North Atlantic Ocean and northern hemisphere ice sheets. Earth Planet. Sci. Lett., 80: 117-129. Ruddiman, W.F. and Kutzbach, J.E., 1989. Forcing of Late Cenozoic northern hemisphere climate by plateau uplift in Southern Asia and the American West. J. Geophys. Res., 94: 18,409-18,427. Ruddiman, W.F., Prell, W.L. and Raymo, M.E., 1989. Late Cenozoic uplift in Southern Asia and the American West: Rationale for general circulation modeling experiments. J. Geophys. Res., 94: 18,379-18,391. Runnegar, B., 1984. The Permian of Gondwanaland. In: Proc. 27th Int. Geol. Congr. VNU Science Press, 1: 305-339. Runnegar, B., 1991. Precambrian oxygen levels estimated from the biochemistry and physiology of early eukaryotes. Global Planet. Change, 97: 97-111. Rust, I.C., 1981. Early Palaeozoic Pakhuis Tillite, South Africa. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 113-117. Rutka, M. and Eyles, N., 1989. Ostracod faunas in Late Pleistocene glaciolacustrine diamict facies,

239 southern Ontario, Canada and their paleolimnoiogical significance. Palaeogeogr. Palaeoclimatol. Palaeoecol., 73: 61-76. Ryder, J.M., 1971. The stratigraphy and morphology of paraglacial alluvial fans in south-central British Columbia. Can. J. Earth Sci., 8: 279-298. Saether, T. and Nystuen, J.P., 1981. Tectonic framework, stratigraphy, sedimentation and volcanism of the Late Precambrian Hedmark Group, Osterdalen, south Norway. Nor. Geogr. Tidsskr., 61: 193-211. Saettern, J., 1992. Glacial erosion processes need improved understanding. Oil Gas J., June 22: 82-83. Sagan, C. and Mullen, G., 1972. Earth and Mars: Evolution of atmospheres and surface temperatures. Science, 177: 52-56. Saito, R., 1969. Glacier problems of late Pre-Cambrian eon. Kumamoto J. Sci., 8: 7-44. Salinas, C., Oblitas, J. and Vargas, C., 1978. Exploracion del sistema Carbonifero en la cuenca oriental de Bolivia. Rev. Tec. Yacimientos Petroliferos Fiscales Boliv., 7: 5-50. Salop, L.J., 1977. Glaciations, rapid changes in organic evolution and their relationship with cosmic phenomena. Int. Geol. Rev., 19: 1271-1291. Sanford, B.V., 1987. Palaeozoic geology of the Hudson Platform. In: C. Beaumont and A.J. Tankard (Editors), Sedimentary Basins and Basin Forming Mechanisms. Can. Soc. Pet. Geol. Mem., 12: 483-505. Sanford, B.V., Thompson, F.J. and McFall, E., 1985. Plate tectonics--A possible controlling mechanism in the development of hydrocarbon traps in southwestern Ontario. Bull. Can. Pet. Geol., 33: 52-11. Sayles, R.W., 1919. Seasonal deposition in aqueo-glacial sediments.. Mem. Mus. Comp. Zool. Harvard College, XLVII, 67 pp. Schieber, J., 1990. Significance of styles of epicontinental shale sedimentation in the Belt Basin, mid-Proterozoic of Montana, U.S.A. Sedimet. Geol., 69: 297-312. Schenk, P.E., 1965. Depositionat Environment of the Gowganda Formation (Precambrian) at the south end of Lake Timagami, Ontario. J. Sediment. Petrol., 35: 309-318. Schenk, P.E., 1972. Possible Late Ordovician glaciation of Nova Scotia. Can. J. Earth Sci., 9: 95-107. Schenk, P.E. and Lane, T.E., 1981. Early Palaeozoic tillites of Nova Scotia, Canada. In: H.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 707-710. Scherer, R.P., 1991. Quaternary and Tertiary microfossils from beneath ice stream B: Evidence for a dynamic West Antarctic Ice Sheet history. Palaeogeogr. Palaeoclimatol. Palaeoecol., 90: 395-412. Schermerhorn, L.J.G., 1966. Terminology of mixed coarse-fine sediments. J. Sediment. Petrol., 36: 831836.

240 Schermerhorn, L.J.G., 1974a. No evidence for glacial origin of Late Precambrian tilloids in Angola. Nature, 252: 114-116. Schermerhorn, L.J.G., 1974b. Late Precambrian mixtites: Glacial and/or non-glacial. Am. J. Sci., 673824. Schermerhorn, L.J.G., 1975. Tectonic framework of Late Precambrian supposed glacials. In: A.E. Wright and F. Moseley (Editors), Ice Ages: Ancient and Modern. Geol. J. Spec. Publ., 6: 241-274. Schermerhorn, L.J.G., 1977. Late Precambrian dolomites, Vendian glaciation and synchroneity of Vendian glaciations: A discussion. J. Geol., 85: 247-250. Schermerhorn, L.J.G., 1983. Late Proterozoic glaciation in the light of CO 2 depletion in the atmosphere. Geol. Soc. Am. Mem., 161: 309-315. Schermerhorn, L.J.G. and Stanton, W.I., 1966. Tilloids in the West Congo geosyncline. Q.J. Geol. Soc. London, 119: 201-241. Schlanger, W. and Philip, J., 1990. Cretaceous carbonate platforms. In: R.N. Ginsburg and B. Beaudoin (Editors), Cretaceous Resources, Events and Rhythms. Kluwer, Netherlands, pp. 173-195. Schmidt, P.W. and Morris, W.A., 1977. An alternative view of the Gondwana Palaeozoic apparent polar wander path. Can. J. Earth Sci., 14: 2674-2678. Schmidt, P.W., Williams, G.E. and Embleton, B.J.J., 1991. Low palaeolatitude of Late Proterozoic glaciation: early timing of remanence in haematite of the Elatina Formation, South Australia. Earth Planet. Sci. Lett., 105: 355-367. Schmok, J.P. and Clarke, G.K.C., 1989. Lacustrine sedimentary record of ice-damned Neoglaciai Lake Alsek. Can. J. Earth Sci., 26: 2092-2105. Scholl, D.W., Grantz, A. and Vedder, J.G. (Editors), 1987. Geology and Resource Potential of the Western North America and Adjacent Ocean Basins-Beaufort Sea to Baja California. Circum-Pacific Council for Energy and Mineral Resources, 6, Houston, 799 pp. Schubert, G., 1991. "Fhe lost continents. Nature, 354: 358-359. Schultz, A.W., 1984. Subaerial debris flow deposition in the Upper Palaeozoic Cutler Formation, Western Colorado. J. Sediment. Petrol., 54: 759-772. Schultz, R.A. and Aydin, A., 1990. Formation of interior basins associated with curved faults in Alaska. Tectonics, 9: 1387-1407. Schwab, F.L., 1976. Depositional environments, provenance and tectonic framework: upper part of the Late Precambrian Mount Rogers Formation, Blue Ridge Province, Southwestern Virginia. J. Sediment. Petrol., 46: 3-13. Schwab, F.L., 1977. Grandfather Mountain Formation. J. Sediment. Petrol., 47: 800-810. Schwarcz, H. and Eyles, N., 1991. Laurentide Ice Sheet

N. EYLES extent inferred from stable isotopic composition (O,C) of lacustrine ostracodes at Toronto, Canada. Quat. Res., 35: 305-320. Schwarzbach, M., 1958. Die "tillite" von Menorca und das problem devonischer Vereisungen. Sonderverottentlichungen Geol. Inst. Univ. Koln, 3: 4-19. Schwarzbach, M., 1975. Discussion: the terminology and stratigraphic nomenclature of proven and possible glaciogenic sediments. J. Geol. Soc. Aust., 22: 255-256. Scotese, C.R. and Barrett, S.F., 1990. Gondwana's movement over the south Pole during the Palaeozoic: Evidence from lithological indicators of climate. In: W.S. McKerrow and C.R. Scotese (Editors), Palaeozoic Palaeogeography and Biogeography. Geol. Soc. London Mem. 12: 75-85. Scotese, C.R. and McKerrow, W.S., 1991. Ordovician plate tectonic reconstructions. Geol. Surv. Can. Pap., 90-9: 271-282. Scott, D.B., Boyd, R. and Medioli, F.S., 1987. Relative sea-level changes in Atlantic Canada: Observed level and sedimentological changes vs. theoretical models. In: D. Nummedal, O.H. Pilkey and J.D. Howard (Editors), Sea-Level Fluctuations and Coastal Evolution. Soc. Econ. Paleontol. Mineral., Spec. Publ., 41: 87-96. Sellers, W.D., 1990., The genesis of energy balance modelling and the cool sun paradox. Palaeogeogr. Palaeoclimatol. Palaeoecol., 82: 217-224. Selwyn, A.R.C., 1859. Geological Notes of a Journey in S. Australia from Cape Jarris to Mt. Serle. Adelaide. Sen, D.P and Banerji, T., 1991. Permo-Carboniferous proglacial lake sedimentation in the Sahajuri Gondwana basin, India. Sediment. Geol., 71: 47-58. Sengor, A.M.C., Altiner, D., Cin, A., Ustaomer, T. and Hsu, K.J., 1988. Origin and assembly of the Tethyside orogenic collage at the expense of Gondwana Land. In: M.G. Audley-Charles and A. Hallam (Editors), Gondwana and Tethys. Geol. Soc. Spec. Publ., 37: 119-181. Shackleton, N.J., 1987. The carbon isotope record of the Cenozoic: history of organic carbon burial and of oxygen in the ocean and atmosphere. In: J. Brooks and A.J. Fleet (Editors), Marine Petroleum Source Rocks. Geol. Soc. Spec. Publ., 26: 423-434. Shackleton, N.J. and Opdyke, N.D., 1973. Oxygen isotope and paleomagnetic stratigraphy of equatorial Pacific core V28-238: oxygen isotope temperatures and ice volumes on a 105 and 106 year scale. Quat. Res., 3: 39-55. Shackleton, N.J., Backman, J., Zimmerman, H., Kent, D.V., Hall, M.A., Roberts, D.G., Schnitker, D., Baldauf, J.G., Desprairies, A., Homrighausen, R., Huddlestun, P., Keene, J.B., Kaltenback, A.J., Krumsiek, K.A.O., Morton, A.C., Murray, J.W. and

EARTH'S G L A C I A L R E C O R D AND ITS TECTONIC SETTING

Westberg-Smith, J., 1984. Oxygen isotope calibration of the onset of ice-rafting and history of glaciation in the North Atlantic region. Nature, 307: 620-623. Shackleton, N.J. and Pisias, N.G., 1985. Atmospheric carbon dioxide, orbital forcing and climate. In: E.T. Sundquist and W.S. Broecker (Editors), The Carbon Cycle. Am. Geophys. Union Geophys. Monogr., 32: 303-317. Shaw, J., 1977. Sedimentation in an alpine lake during deglaciation, Okanagan Valley, British Columbia, Canada. Geogr. Ann., 59A: 221-240. Shaw, J., 1979. Genesis of the Sveg tills and Rogen moraines of central Sweden: a model of basal melt out. Boreas, 8: 409-426. Shaw, R.D., Etheridge, M.A. and Lambeck, K., 1991. Development of the Late Proterozoic to midPaleozoic intracratonic Amadeus Basin in central Australia: A key to understanding tectonic forces in plate interiors. Tectonics, 10: 688-721. Sheehan, P.M., 1973. The relation of late Ordovician glaciation to the Ordovician-Silurian changeover in North American brachiopod faunas. Lethaia, 6: 147-154. Sheldon, R.P., 1984a, Ice-ring origin of the Earth's atmosphere and hydrosphere and Late ProterozoicCambrian phosphogenesis. Geol. Surv. India Spec. Publ., 17: 17-21. Sheldon, R.P., 1984b, Precambrian ice-ring model to account for changes in exogenic regimes from Proterozoic to Phanerozoic eras. In: Proc. 5th Int. Field Workshop on Phosphorite. Kumming, China, 2: 227-243. Shilts, W.W., 1986. Geological models for the configuration, history and style of disintegration of the Laurentide Ice Sheet. In: M.J. Woldenberg (Editor), Models in Geomorphology. Allen and Unwin, Boston, pp. 73-91. Sibrava, V., Bowen, D.Q. and Richmond, G.M. (Editors), 1986. Quaternary glaciations in the Northern Hemisphere. Quat. Sci. Rev., 5, 514 pp. Siedlecka, A. and Roberts, D., 1972. A late Precambrian tilloid from Varangerhalvoya: evidence of both glaciation and subaqueous mass movement. Nor. Geol. Tidsskr., 52: 135-141. Signor, P.W. and Moores, E.M., 1991. Breakup of the Proterozoic supercontinent: evidence from early Cambrian biogeography. Annu. Meet. Abstr. Programs: A456. Geol. Soc. Am. San Diego. Sigurdsson, H., Sparks, R.S.J., Carey, S. and Huang, T.C., 1980. Volcanogenic sedimentation in the Lesser Antilles Arc. J. Geol., 88: 523-540. Simpson, G.C., 1930. The climate during the Pleistocene Period. Proc. R. Soc. Edinburgh, 50: 262296. Sivell, W.J. and McCulloch, M.T., 1991. Neodymium

241 isotope evidence for ultra-depleted mantle in the early Proterozoic. Nature, 354: 384-387. Skinner, B.J. and Porter, S., 1987. Physical Geology. Wiley, New York. Sloss, L.L, 1963. Sequences in the cratonic interior of North America. Geol. Soc. Am. Bull., 74: 93-114. Sloss, L.L., 1988. Conclusions. In: L.L. Sloss (Editor), Sedimentary cover--North American craton. Geology of North America Series, D-2. Geol. Soc. Am., pp. 25-51. Sloss, EL., 1990. Epilog. In: M.W. Leighton, D.R. Kolata, D.F. Oltz and J.J. Eidel (Editors), Interior Cratonic Basins. Am. Assoc. Pet. Geol. Mem., 51: 799-805. Slowey, N.C. and Curry, W.B., 1992. Enhanced ventilation of the North Atlantic subtropical gyre thermocline during the last glaciation. Nature, 358: 665668. Smith, A.B., 1988. Late Paleozoic biogeography of East Asia and paleontological constraints on plate tectonic reconstructions. In: R.M. Shackleton, J.F. Deewey and B.F. Windley (Editors), Tectonic Evolution of the Himalayas and Tibet. R. Soc. London: 189-226. Smith, G.A. and Lowe, D.R., 1991. Lahars: Volcanohydrologic events and deposition in the debris flowhyperconcentrated flow continuum. In: R.V. Fisher and G.A. Smith (Editors), Sedimentation in Volcanic Settings. Soc. Econ. Palaeontol. Mineral. Spec. Publ., 45: 59-70. Smith, N.D., 1978. Sedimentation processes and patterns in a glacier-fed lake with low sediment input. Can. J. Earth Sci., 15: 741-756. Socci, A.D., 1992. Climate, glaciation and deglaciation, controls, pathways, feedbacks, rates and frequencies. Mod. Geol., 16: 279-316. Socci, A.D. and Smith, G.W., 1987. Evolution of the Boston Basin: A sedimentological perspective. In: C. Beaumont and A.J. Tankard (Editors), Basin Forming Mechanisms. Can. Soc. Pet. Geol. Mem., 12: 87-99. Sognian, L., Guogan, M., Zhenjia, G. and Weixing, L., 1985. Sinian ice ages and glacial sedimentary facies-areas in China. Precambrian Res., 29: 53-63. Solheim, A. and Pfirman, S.L., 1985. Sea-floor morphology outside a grounded, surging glacier; Brasvellbreen, Svalbard. Mar. Geol., 65: 127-143. Solokov, B.S. and Fedonkin, M.A., 1986. Global biological events in the Late Precambrian. In: O. Walliser (Editor), Global Bio-Events. Springer, Berlin, pp. 105-108. Solomon, S., 1986. Comment and reply on "Quartzgrain surface textures: Evidence for middle Carboniferous glacial sediment input to the Parrsboro Formation of Nova Scotia". Geology, 14: 92-93. Sonderholm, M. and Jepsen, H.F., 1991. Proterozoic

242 basins of North Greenland. Bull. Gronl. Geol. Unders., 160: 49-69. Soper, N.J. and Anderton, R., 1984. Did the Dalradian slides originate as extensional faults? Nature, 307: 357-360. Souchez, R.A. and Lemmens, M., 1985. Subglacial carbonate deposition: an isotopic study of a presentday case. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 357-364. Spencer, A.M., 1971. Late Precambrian glaciation in Scotland. Mem. Geol. Soc. London, 6: 1-48. Spencer, A.M., 1975. Late Precambrian glaciation in the North Atlantic region. In: A.E. Wright and F. Moseley (Editors), Ice Ages: Ancient and Modern. J. Geol. Spec. Issue, 6: 217-240. Spencer, A.M., 1985. Mechanisms and environments of deposition of late Precambrian geosynclinal tillites: Scotland and East Greenland. Palaeogeogr. Palaeoclimatol. Palaeoecol., 51: 143-158. Spicer, R.A., 1987. The significance of the Cretaceous flora of northern Alaska for the reconstruction of the climate of the Cretaceous. Geol. Jahrb. Reihe A, 96: 265-291. Spicer, R.A. and Parrish, J.T., 1986. Paleobotanical evidence for cool north polar climates in middle Cretaceous (Albian-Cenomanian) time. Geology, 14: 703-706. Spjeldnaes, N., 1964. The Eocambrian glaciation in Norway. Geol. Rundsch., 54: 24-45. Spjeldnaes, N., 1976. Ordovician Climates. In: M.G. Bassett (Editor), The Ordovician System. University of Wales, Cardiff, pp. 67-79. Stanistreet, I.G., Martin, McB.D., Spencer, R. and Beneke, D., 1988. The importance of diamictites in an understanding of the tectonics and sedimentation of the Witwatersrand Basin. In: Proc. 22nd Congr., Geol. Soc. S. Afr., Durban, pp. 607-610. Stanistreet, I.G., Kukla, P.H. and Henry, G., 1991. Sedimentary basinal responses to a Late Precambrian Wilson cycle; the Damara orogen and Nama foreland, Namibia. J. Afr. Earth Sci., 13: 141-156. Stanley, S.M., 1988. Paleozoic mass extinctions: shared patterns suggest global cooling as a common cause. Am. J. Sci., 288: 334-352. Stauffer, P.H. and Mantajit, 1981. Late Palaeozoic tilloids of Malaya, Thailand and Burma. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 331-335. Stauffer, P.H. and Peng, L.C., 1986. Late Paleozoic glacial marine facies in Southeast Asia and its implications. Geol. Soc. Malays. Bull., 20: 363-397. Steidtmann, J.R., et al., 1986. Geometry, distribution and provenance of teconogenic conglomerates along the southern margin of the Wind River Range, Wyoming. In: J.A. Peterson (Editor), Paleotectonics

N. EYLES and Sedimentation in the Rocky Mountain Region, United States. Am. Assoc. Pet. Geol., Tulsa, Oklahoma. Steiner, J. and Grillmair, E., 1973. Possible galactic causes of periodic and episodic glaciations. Bull. Geol. Soc. Am., 84: 1003-1018. Steiner, J. and Falk, F., 1981. The Ordovician Lederschiefer of Thuringia. In: M.J Hambrey and W.B., Harland (Editor), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 579-581. Stephenson, A.J. and Embley, R., 1987. Deep-sea fan bodies, terrigenous turbidite sedimentation and petroleum potential, Gulf of Alaska. In: D.W. Scholl, A. Grantz and J.G. Vedder (Editors), Geology and Resource Potential of the Continental Margin of Western North America and Adjacent Ocean Basins. Circum-Pacific Council for Energy and Mineral Resources Earth Science Series, 6. Houston, Texas, pp. 503-522. Stewart, J.H., 1972. Initial deposits in the Cordilleran Geosycline: evidence of a Late Precambrian ( < 850 M.y) continental separation. Geol. Soc. Am. Bull., 83: 1345-1360. Stihler, S.D., Stone, D.B. and Beget, J.E., 1992. "Varve" counting versus tephrochronology and 137Cs and 210 Pb dating: A comparative test at Skilak Lake, Alaska. Geology, 20: 1019-1022. Stoeser, D.B. and Camp, V.E., 1985. Pan-African microplate accretion of the Arabian Shield. Geol. Soc. Am. Bull., 96: 817-826. Stoker, M.S., Harland, R. and Graham, D.K., 1991. Glacially-influenced basin plain sedimentation in the southern Faeroe-Shetland Channel, northwest United Kingdom continental margin. Mar. Geol., 100: 185-200. Storey, B.C., Dalziel, I.W.D., Garrett, S.W., Grunow, A.M., Pankhurst, R.J. and Vennum, W.R., 1988. West Antarctica in Gondwanaland: crustal blocks, reconstruction and breakup processes. Tectonophysics, 155: 381-390. Storey, B.C., 1991. The crustal blocks of West Antarctica within Gondwana: reconstruction and break-up model. In: M.R.A. Thomson, J.A. Crame and J.W. Thomson (Editors), Geological Evolution of Antarctica. World and Regional Geology 1. Cambridge University Press, pp. 587-592. Stow, D.A.V., 1985. Deep-sea clastics: Where are we and where are we going? In: P.J. Brenchley and B.P.J. Williams (Editors), Sedimentology--Recent Developments and Applied Aspects. Blackwell, Oxford, pp. 67-94. Stow, D.A.V. and Piper, D.J.W., 1984. Fine-grained sediments: Deep water Processes and Facies. Geol. Soc. London Spec. Publ., 15. Street-Perrott, F.A., 1992. Tropical wetland sources. Nature, 355: 23-24.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SE'ITING

Stridsberg, S., 1980. Sedimentology of Upper Ordovician regressive strata in Vastergotland. Geol. Foren. Stockholm Forh., 102: 213-221. Stromberg, B., 1983. The Swedish varve chronology. In: J. Ehlers, (Editor), Glacial Deposits In North-West Europe. Balkema, Rotterdam, pp. 97-106. Stump, E., Miller, J.M.G., Korsch, R.J. and Edgerton, D.G., 1988. Diamictite from Nimrod Glacier area, Antarctica; possible Proterozoic glaciation on the seventh continent. Geology, 16: 225-228. Stump, E. and Fitzgerald, P.G., 1992. Episodic uplift of the Transantarctic Mountains. Geology, 20: 161164. Stupavsky, M., Symons, D.T.A. and Gravenor, C.P., 1982. Evidence for metamorphic remagnetisation of upper Precambrian tillite in the Dalradian Supergroup of Scotland. Trans. R. Soc. Edinburgh Earth Sci., 73: 59-65. Sugden, D., 1992. Antarctic ice sheets at risk? Nature, 359: 775-776. Sun Dong-li, 1993. On the Permian biogeographic boundary between Gondwana and Eurasia in Tibet, China as the eastern section of the Tethys. Palaeogeogr. Palaeoclimatol. Palaeoecol., 100: 59-77. Surlyk, F., 1991. Tectonostratigraphy of North Greenland. Bull. Gronl. Geol. Unders., 160: 25-47. Sutherland, 1870. Notes on an ancient boulder clay of Natal. Q.J. Geol. Soc. London, 26: 514-517. Swift, S.A., 1985. Late Pleistocene sedimentation on the continental slope and rise off western Nova Scotia. Geol. Soc. Am. Bull., 96: 832-841. Syvitski, J.P.M., Burreli, D.C. and Skei, J.M. (Editors), 1987. Fiords: Processes and Products. Springer, Berlin, 379 pp. Tainton, S. and Meyer, F.M., 1990. The stratigraphy and sedimentology of the Promise Formation of the Witwatersrand Supergroup in the Western Transvaal. S. Afr. J. Geol., 93: 103-117. Tajika, E. and Matsui, T., 1992. Evolution of terrestrial proto-CO 2 atmosphere coupled with thermal history of the earth. Earth Planet. Sci. Lett., 113: 251-266. Talbot, C.J. and Von Brunn, V., 1989. Melanges, intrusive and extrusive sediments and hydraulic arcs. Geology, 17: 446-448. Tankard, A.J., 1986. On the depositional response to thrusting and lithosphere flexure: Examples from the Appalachian and Rocky Mountain basins. In: P.A. Allen and P. Homewood (Editors), Foreland Basins. Int. Assoc. Sedimentol. Spec. Pubi., 8: 369392. Tankard, A.J., Jackson, M.P.A., Eriksson, K.A., Hobday, D.K., Hunter, D.R. and Minter, W.E.L., 1982. Crustal Evolution of Southern Africa. 523 pp. Tankard, A.J. and Balkwill, H.R. (Editors), 1989. Extensional Tectonics and Stratigraphy of the North

243 Atlantic Margins. Am. Assoc. Pet. Geol. Mem., 46, 641 pp. Tarling, D.H., 1974. A palaeomagnetic study of Eocambrian tillites in Scotland. J. Geol. Soc. London, 130: 163-177. Tate, R., 1878. Proceedings, Ordinary meeting of 5th February. Trans. Philos. Soc. Adelaide, 1. Taylor, S.R. and McLennan, S.M., 1985. The Continental Crust: Its Composition and Evolution. Oxford Scientific Publications, 312 pp. Teller, J.T. and Clayton, L. (Editors), 1983. Glacial Lake Agassiz. Geol. Assoc. Can. Spec. Pap., 26. Tessensohn, F. and Worner, G., 1991. The Ross Sea rift system, Antarctica: structure, evolution and analogues. In: M.R.A. Thomson, J.A. Crame and J.W. Thompson (Editors), Geological Evolution of Antarctica. Cambridge University Press, pp. 273277. Tewari, R.C. and Casshyap, S.M., 1982. Palaeoflow analysis of the late Palaeozoic Gondwana deposits of Giridih and adjoining basins of Bihar and palaeogeographic implications. J. Geol. Soc. India, 23: 67-79. Tewari, R.C. and Casshyap, S.M., 1983. Cyclicity in the Early Permian fluviatile Gondwana coal measures - - a n example from Giridih and Saharjuri basins of Bihar. Sediment. Geol., 35: 297-312. Thickpenny, A. and Leggett, J.K., 1987. Stratigraphic distribution and palaeo-oceanographic significance of European early Paleozoic organic-rich sediments. In: J. Brooks and A.J. Fleet (Editors), Marine Petroleum Source Rocks. Geol. Soc. Spec. Publ., 26: 231-248. Thomas, G.S.P. and Connell, R.J., 1985. Iceberg drop, dump and grounding structures from Pleistocene glacio-lacustrine sediments, Scotland. J. Sediment. Petrol., 55: 243-250. Thompson, J., 1871. On the occurrence of pebbles and boulders of granite in schistose rocks in Islay. In: 40th Annu. Meet. British Assoc. Liverpool Transactions: 88. Thompson, J., 1877. On the geology of the island of Islay. Trans. Geol. Soc. Glasgow, 5: 200-222. Thompson, J.B. and Newton, C.R., 1988. Late Devonian mass-extinction: Episodic climatic cooling or warming? In: N.J. McMillan, A.F. Embry and D.J. Glass (Editors), Devonian of the World. Can. Soc. Pet. Geol., 3: 29-34. Thornton, R.C.N., 1974. Hydrocarbon Potential of Western Murray Basin and Infrabasins. Rep. Invest., 41. Dep. Mines Geol. Surv. South Aust. Tietzsch-Tyles, D., 1989. Evidence of intracratonic Finnmarkian orogeny in central Norway. In: R.A. Gayer (Editor), The Caledonide Geology of Scandinavia. Graham and Trotman, London, pp. 47-62. Torsvik, T.H., Olessen, O., Ryan, P.D. and Trench, A.,

244 1990. On the palaeogeography of Baltica during the Palaeozoic: new palaeomagnetic data from the Scandinavian caledonides. Geophys. J. Int., 103: 261-279. Torsvik, T.H., Ryan, P.D., Trench, A. and Harper, D.A.T., 1991. Cambrian-Orovician paleogeography of Baltica. Geology, 19: 7-10. Trendall, A.F., 1981. The Lower Proterozoic Meteorite Bore Member, Hamersley Basin, Western Australia. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 555-559. Trettin, H., 1989. The Arctic Islands. In: A.W. Bally and A.R. Palmer (Editors), The Geology of North America: An Overview. Decade of North American Geology. Geol. Soc. Am., pp. 349-370. Trompette, R., 1981. Late Precambrian tillites of the Volta basin and the Dahomeyides Orogenic Belt, Benin, Ghana, Niger and Upper-Volta. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 135-139. Tucker, M. and Reid, P.C., 1981a. Late Precambrian glacial sediments, Sierra Leone. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 132-134. Tucker, M. and Reid, P.C., 1981b. Late Ordovician glaciomarine sediments, Sierra Leone. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 97-98. Tucker, R.D. and Pharoah, T.C., 1991. U-Pb zircon ages of late Precambrian rocks in southern Britain. J. Geol. Soc. London, 148: 435-443. Tuckwell, K.D., 1981. Adelaidean diamictites of the Broken Hill District of New South Wales. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 531-536. Twist, D. and Cheney, E.S., 1986. Evidence for the transition to an oxygen-rich atmosphere in the Roosberg Group, South Africa--A Note. Precambrian Res., 33: 255-264. Uliana, M.A. and Biddle, K.T., 1987. Permian to Late Cenozoic evolution of northern Patagonia: main tectonic events, magmatic activity and depositional trends. Am. Geophys. Union Mem., 40: 271-286. Uiiana, M.A., Biddle, K.T. and Cerdan, J., 1989. Mesozoic extension and the formation of Argentine sedimentary basins. In: A.J. Tankard and H.R. Balkwill (Editors), Extensional Tectonics and Stratigraphy of the North Atlantic Margins. Am. Assoc. Pet. Geol. Mere., 46: 599-614. Urrutia-Fucugauchi, J. and Tarling, D.H., 1983. Palae-

N. EYLES omagnetic properties of Eocambrian sediments in northwestern Scotland: implications for world-wide glaciation in the Late Precambrian. Palaeogeogr. Palaeoclimatol. Palaeoecol., 41: 325-344. Vagnes, E. and Amundsen, H.E.F., 1993. Late Cenozoic uplift and volcanism on Spitsbergen: Caused by mantle convection? Geology, 21:251-254 Vail, P.R., Mitchum, R.M. and Thompson, S., 1977. Seismic stratigraphy and global changes in sea-level. Am. Assoc. Pet. Geol. Mem., 26: 83-97. Vail, P.R., Hardenbol, J. and Todd, R.G., 1984. Jurassic unconformities, chronostratigraphy and sea-level changes from seismic stratigraphy and biostratigraphy. In: J.S. Schlee (Editor), Interregional Unconformities and Hydrocarbon Accumulations. Am. Assoc. Pet. Geol. Mem., 36: 129-144. Van de Graaf, W.J.E., 1981. Early Permian Lyons Formation, Carnarvon Basin, Western Australia. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 453-458. Van der Meer, J.J.M., 1987. Tills and Glacitectonics. Balkema, Rotterdam, 270 pp. Van der Voo, R., 1988. Paleozoic palaeogeography of North America, Gondwana and intervening displaced terranes: comparisons of paleomagnetism with paleoclimatology and biogeographic patterns. Geol. Soc. Am. Bull., 100: 311-324. Van der Voo, R. and Meert, J.G., 1991. Late Proterozoic paleomagnetism and tectonic models: a critical appraisal. Precambrian Res., 53: 149-163. Van Houten, F.B., 1957. Appraisal of Ridgway and Gunnison "tillites", southwestern Colorado. Geol. Soc. Am. Bull., 68: 383-388. Van Wagoner, J.C., Posamentier, H.W., Mitchum, R.M., Vail, P.R., Sarg, J.F., Loutit, T.S. and Hardenbol, J., 1988. An overview of the fundamentals of sequence stratigraphy. In: C.K. Wilgus, B.S. Hastins, C.G.C. Kendall, H. Posamentier, C.A. Ross and J. van Wagoner (Editors), Sea-level Changes: an Integrated Approach. Soc. Econ. Paleontol. Mineral. Spec. Publ., 42. Tulsa, pp. 39-45. Vaslet, D., 1990. Upper Ordovician glacial deposits in Saudi Arabia. Episodes, 13: 147-161. Vaslet, D., Berthiaux, A., Le Strat, P., Kellogg, K.S. and Vincent, P.L., 1987. Geologic map of the Baq'a quadrangle, sheet 27F, Kingdom of Saudi Arabia. Saudi Arabian Deputy Ministry for Mineral Resources. Scale 1:250,000. Geoscience Map GM-116 A. Vavrdova, M., Isaacson, P.E., Martinez, E.D. and Bek, J., 1991. Palinologia del limite Devonico-Carbonifero en torno a Lago Titikaka, Bolivia: Resultados prelimares. Rev. Tec. Yaciementos Petroliferos Fiscales Boliv., 12: 303-313.

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

Veeh, H.H. and Veevers, J.J., 1970. Sea level at - 175 m off the Great Barrier Reef 12,600 to 17,000 years ago. Nature, 226: 536-537. Veevers, J.J., 1984 (Editor). Phanerozoic Earth History of Australia. Clarendon Press, Oxford, 418 pp. Veevers, J.J., 1990. Tectono-climate supercycle in the billion year plate tectonic eon: Permian Pangean icehouse alternates with Cretaceous dispersed continents greenhouse. Sediment. Geol., 68: 1-16. Veevers, J.J. and Powell, C.M., 1987. Late Paleozoic glacial episodes in Gondwanaland reflected in transgressive-regressive depositional sequences in Euramerica. Geol. Soc. Am. Bull., 98: 475-487. Villeneuve, M., 1989. The geology of the MadinaKouta Basin (Guinea-Senegal) and its significance for the geodynamic evolution of the western part of the West African Craton during the Upper Proterozoic period. Precambrian Res., 44: 305-322. Villeneuve, M. and Dallmeyer, R.D., 1987. Geodynamic evolution of the Mauritanide, Bassaride and Rokelide Orogens (West Africa). Precambrian Res., 37: 19-28. Visser, J.N.J., 1981. The mid-Precambrian tillite in the Grigualand West and Transvaal Basins, South Africa. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 180-184. Visser, J.N.J., 1983. Submarine debris flow deposits from the Upper Carboniferous Dwyka Tillite Formation in the Kalahari Basin, South Africa. Sedimentology, 30: 511-523. Visser, J.N.J., 1987. Influence of topography on the Permo-Carboniferous glaciation in the Karoo Basin and adjoining areas, southern Africa. In: G.D. McKenzie (Editor), Gondwana 6. Geophys. Monogr., 41. Am. Geophys. Union, pp. 123-129. Visser, J.N.J., 1989a. Episodic Palaeozoic glaciation in the Cape-Karoo Basin, South Africa. In: J. Oerlemans (Editor), Glacier Fluctuations and Climatic Change. Kluwer, Boston, pp. 1-12. Visser, J.N.J., 1989b. The Permo-Carboniferous Dwyka formation of southern Africa: Deposition by a predominantly subpolar marine ice sheet. Palaeogeogr. Palaeoclimatol. Palaeoecol., 70: 377-391. Visser, J.N.J., 1991. The paleoclimatic setting of the Late Paleozoic marine ice sheet in the Karoo Basin of southern Africa. In: J.B. Anderson and G.M. Ashley (Editors), Glacial Marine Sedimentation: Paleoclimatic Significance. Geol. Soc. Am. Spec. Pap., 261: 181-190. Visser, J.N.J. and Loock, J.C., 1987. Ice margin influence on glaciomarine sedimentation in the PermoCarboniferous Dwyka Formation from the southwestern Karoo, South Africa. Sedimentology, 34: 929-941. Visser, J.N.J. and Loock, J.C., 1988. Sedimentary facies of the Dwyka Formation association with the

245 Nooitgedacht glacial pavements, Barkly West District. S. Afr. J. Geol., 91: 38-48. Vogt, P.R. and Tucholke, B.E., 1989. North Atlantic Ocean basin: Aspects of geologic structure and evolution. In: A.W. Bally and A.R. Palmer (Editors), The Geology of North America: An Overview. Decade of North American Geology. Geol. Soc. Am., pp. 53-80. Volk, T., 1993. Cooling in the late Cenozoic. Nature, 361: 123. von Brunn, V. and Gold, D.J.C., 1992. Diamictite in the Archaean Pongola sequence southeast of Piet Retief. Geol. Soc. S. Afr. Congr., 24. Bloemfontein, pp. 458-459. von Brunn, V. and Marshall, C.G.A., 1989. Glaciated surfaces and the base of the Dwyka Formation near Pietermaritzburg, Natal. S. Afr. J. Geol., 92: 420426. Vorren, T.O., Hald, M., Edvardsen, M. and Lind-Hansen, O.W., 1983. Glacigenic sediments and sedimentary environments on continental shelves: General principles with a case study from the Norwegian shelf. In: J. Ehlers (Editor), Glacial Deposits in Northwest Europe. Balkema, Rotterdam, pp. 61-76. Vorren, T.O., Kristoffersen, Y. and Andreassen, K., 1986. Geology of the inner shelf west of North Cape, Norway. Nor. Geol. Tidsskr., 66: 99-105. Vorren, T.O., Lebesbye, E., Andreasses, K. and Larsen, K.-B., 1989. Glacigenic sediments on a passive continental margin as exemplified by the Barents Sea. Mar. Geol., 85. Walker, J.G.C., Hays, P.B. and Kasting, J.K., 1981. A negative feedback mechanism for the long-term stabilization of Earth's surface temperature. J. Geophys. Res., 86: 9776-9782. Walker, J.G.C., Klein, G., Schidlowski, M., Schopf, J.W., Stevenson, D.J. and Walter, M.R., 1983. Environmental evolution of the Archean-Early Proterozoic Earth. In: J.W. Schopf (Editor), The Earth's Earliest Biosphere: Its Origin and Evolution. Princeton University Press, pp. 260-289. Walker, R.G., 1979. Facies and facies models. In: R.G. Walker (Editor), Facies Models. Geol. Assoc. Can., pp. 1-7. Walker, R.G., 1992. Facies, facies models and modern straigraphic concepts. In: R.G. Walker and N.P. James (Editors), Facies Models: Response to SeaLevel Change. Geol. Assoc. Can., pp. 1-15. Walker, R.G., 1992. Turbidites and submarine fans. In: R.G. Walker and N.P. James (Editors), Facies Models: Response to Sea-level Change. Geol. Assoc. Can., pp. 239-264. Walker, R.G. and Eyles, C.H., 1988. Geometry and facies of stacked shallow-marine sandier upward sequences dissected by erosion surface, Cardium Formation, Willesden Green, Alberta. Am. Assoc. Pet. Geol. Bull., 72: 1469-1494.

246 Walker, R.G. and James, N.P. (Editors), 1992. Facies Models: Response To Sea-Level Change. Geol. Assoc. Can., 409 pp. Walter, M.R., 1981. Late Proterozoic tillites of the southwestern Georgina Basin, Australia. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 525-530. Walter, M.R. and Bauld, J., 1983. The association of sulphate evaporites, stromatolitic carbonates and glacial sediments; examples from the Proterozoic of Australia and the Cenozoic of Antarctica. Precambrian Res., 21: 129-148. Wang, H. and Qiao, X., 1984. Proterozoic stratigraphy and tectonic framework of China. Geol. Mag., 121: 599-614. Wang, Y., Lu, S., Gao, Z., Lin, W. and Ma, G., 1981. Sinian tillites in China. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 386-401. Washburn, A.L., 1979. Geocryology. A Survey of Periglacial Processes and Environments. Edward Arnold, London. Waterhouse, J.B., 1982. An early Permian cool-water fauna from pebbly mudstones in south Thailand. Geol. Mag., 119: 337-355. Webb, P.N., Harwood, D.M., McKelvey, B.C., Mercer, J.H. and Stoff, L.D., 1984. Cenozoic marine sedimentation and ice-volume variation on the East Antarctic craton. Geology, 12: 287-291. Webb, P.N., Harwood, D.M., McKelvey, B.C., Mabin, M.C.G. and Mercer, J.H., 1986. Late Cenozoic tectonic and glacial history of the Transantarctic Mountains. Antarct. J.U.S., 21: 99-100. Wegener, A., 1929. The Origin of Continents and Oceans. Dover Publications Inc., New York. Wehr, F., 1986. A proglacial origin for the Upper Proterozoic Rockfish conglomerate, Central Virginia, U.S.A. Precambrian Res., 34: 157-174. Weimer, P., 1991. Sequence stratigraphy of the Mississippi fan related to oxygen isotope sea level index: Discussion. Am. Assoc. Pet. Geol. Bull., 75: 15001507. Weimer, R.J., 1992. Developments in sequence stratigraphy: Foreland and cratonic basins. Am. Assoc. Pet. Geol. Bull., 76: 965-982. Weller, J.M., 1930. Cyclic sedimentation of the Pennsylvanian Period and its significance. J. Geol., 38: 97-135. Wells, A.T., 1981. Late Proterozoic diamictites of the Amadeus and Ngalia Basins, central Australia. In: M.J. Hambrey and W.B. Hartand (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 515-524. Wernicke, B., 1985. Uniform-sense normal simple shear of the continental lithosphere. Can. J. Earth Sci., 22: 108-125.

N. EYLES Westgate, J.A., 1968. Linear sole markings in Pleistocene till. Geol. Mag., 106: 501-505. Westgate, J.A., Stemper, B.A. and Pewe, T.L., 1990. A 3 m.y. record of Pliocene-Pleistocene loess in interior Alaska. Geology, 18: 858-861. Whalley, W.B. and Krinsley, D.H., 1974. A scanning electron microscope study of surface textures of quartz grains from glacial environments. Sedimentology, 21: 87-105. White, R.S., 1989. Initiation of the Iceland plume and opening of the North Atlantic. In: A.J. Tankard and H.R. Balkwill (Editors), Extensional Tectonics and Stratigraphy of the North Atlantic Margins. Am. Assoc. Pet. Geol. Mem., 46:149-154 (559-566). Wiebols, J.H., 1955. A suggested glacial origin for the Witwatersrand conglomerates. Trans. Geol. Soc. S. Afr., 48: 367-387. Wilde, P. and Berry, W.B.N., 1984. Destabilization of the oceanic density structure and its significance to marine "extinction" events. Palaeogeogr. Palaeoclimatol. Palaeoecol., 48: 143-162. Wilgus, C.K., Hastings, B.S., Posamentier, H.W., Ross, C.A. and Kendall, C.G.S. (Editors), 1988. Sea Level Changes: An Integrated Approach. Soc. Econ. Paleontol. Mineral. Spec. Publ., 42, 407 pp. Williams, B.P.J., Wild, E.K. and Suttill, R.J., 1985. Paraglacial aeolianites: Potential new hydrocarbon reservoirs, Gidgealpa Group, Southern Cooper Basin. APEA J., 291-310. Williams, G.E., 1975. Late Precambrian glacial climate and the Earth's obliquity. Geol. Mag., 112: 441-465. Williams, G.E., 1979. Sedimentology, stable isotope geochemistry and palaeoenvironment of dolostones capping late Precambrian glacial sequences in Australia. J. Geol. Soc. Aust., 26: 377-386. Williams, G.E., 1986. Precambrian permafrost horizons as indicators of palaeoclimate. Precambrian Res., 32: 233-242. Williams, G.E., 1988. Cyclicity in the Late Precambriam Elatina formation, South Australia, solar or tidal signature? Clim. Change, 13: 117-128. Williams, G.E., 1989. Late Precambrian tidal rhythmites in South Australia and the history of the Earth's rotation. J. Geol. Soc. London, 146: 97-111. Williams, P.L., Vaslet, D., Johnson, P.R., Berthiaux, A., Le Strat, P. and Fourniguet, J., 1986. Geologic map of the Jabal Habashi quadrangle, sheet 26F, Kingdom of Saudi Arabia. Saudi Arabian Deputy Ministry for Mineral Resources. Scale 1:250,000., Geoscience Map GM-98 A. Williams, P.J. and Smith, M.W., 1989. The Frozen Earth. Cambridge University Press, 306 pp. Winterer, E., 1963. Late Precambrian pebbly mudstone in Normandy, France. In: A.E.M. Nairn (Editor), Problems in Palaeoclimatology. Interscience, London, pp.159-178. Winterer, E.L. and Von der Borch, C.C., 1968. Striated

EARTH'S G L A C I A L RECORD AND ITS TECTONIC SETTING

pebbles in a mudflow deposit, South Australia. Palaeogeogr. Palaeoclimatol. Palaeoecol., 5: 205211. Wolfe, J.A., 1978. A paleobotanical interpretation of Tertiary climates in the northern hemisphere. Am. Sci., 66: 694-703. Wood, R.J., Edrich, S.P. and Hutchinson, I., 1989. Influence of North Atlantic tectonics on the largescale uplift of the Stappen High and Loppa High, Western Barents Shelf. In: A.J. Tankard and H.R. Balkwill (Editors), Extensional Tectonics and Stratigraphy of the North Atlantic Margins. Am. Assoc. Pet. Geol. Mem., 46: 559-566. Woodruff, F., Savin, S.M. and Douglas, R.G., 1981. Miocene stable isotope record: A detailed deep Pacific Ocean study and its paleoclimatic implications. Science, 212: 665-668. Woodworth-Lynas, C.M.T. and Landva, J., 1988. Sediment deformation by ice scour: Centre for Cold Ocean Resources, Memorial Univ. Newfoundland, Eng. Rep., 27 pp. Woodworth-Lynas, C.M.T. and Guigne, J.Y., 1990. Iceberg scours in the geological record: examples from glacial Lake Agassiz. In: J.A. Dowdeswell and J.D. Scourse (Editors), Glaciomarine Environments: Processes and Sediments. Geol. Soc. London Spec. Publ., 53: 217-234. Wopfner, H., 1972. Depositional History and Tectonics of South Australian Sedimentary Basins. Miner. Resour. Rev. South Aust., 133: 32-50. Wopfner, H., 1981. Development of Permian intracratonic basins in Australia. In: M.M. Creswell and P. Vella (Editors), Gondwana 5. Balkema, Rotterdam, pp. 185-190. Wopfner, H. and Kreuser, T., 1986. Evidence for Late Palaeozoic glaciation in southern Tanzania. Palaeogeogr. Palaeoclimatol. Palaeoecol., 56: 259-275. Worsley, T.R. and Kidder, D.L., 1991. First-order coupling of paleogeography and CO2, with global surface temperature and its latitudinal contrast. Geology, 19: 1161-1164. Worsley, T.R., Nance, R.D. and Moody, J.B., 1986. Tectonic cycles and the history of the Earth's bogeochemical and paleoceanographic record. Palaeoceanography, 1: 233-263. Wright, J.D., 1993. Deglaciation triggered by the resumption of North Atlantic deep water. EOS, January 12: 24. Wundt, W., 1944. Die Mitwirkung der Erdbahnelemente bei der Enstehung der Eiszeiten. Geol. Rundsch., 34: 713-747. Yapp, C.J. and Poths, H., 1992. Ancient atmospheric CO 2 pressures inferred from natural goethites. Nature, 355: 342-344. Yardley, B.W.D., Vine, F.J. and Baldwin, C.T., 1982. The plate tectonic setting of NW Britain and Ire-

247 land in late Cambrian and early Ordovician times. J. Geol. Soc. London, 139: 455-463. Yeo, G.M., 1981. The Late Proterozoic Rapitan glaciation in the northern Cordillera. In: F.H.A. Campbell (Editor), Proterozoic Basins of Canada. Geol. Surv. Can. Pap., 81-10: 25-46. Yianping, L., Yongan, L., Sharps, R., McWilliams, M. and Ghao, Z., 1991. Sinian paleomagnetic results from the Tarim block, Western China. Precambrian Res., 49: 61-71. Young, G.M., 1970. An extensive Early Proterozoic glaciation in North America. Palaeogeogr. Palaeoclimatol. Palaeoecol., 7: 85-101. Young, G.M., 1975. Geochronology of Archean and Proterozoic rocks in the southern district of Keewatin: Discussion. Can. J. Earth Sci., 12: 1250-1254. Young, G.M., 1976. Iron formation and glaciogenic rocks of the Rapitan Group, Northwest Territories, Canada. Precambrian Res., 3: 137-158. Young, G.M., 1981. Diamictites of the Early Proterozoic Ramsay Lake and Bruce Formations, north shore of Lake Huron, Ontario, Canada. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge University Press, pp. 813-816. Young, G.M., 1983. Tectono-sedimentary history of Early Proterozoic rocks of the northern Great Lakes region. In: L.G. Medaris Jr. (Editor), Early Proterozoic Geology of the Great Lakes Region. Geol. Soc. Am. Mem., 160: 15-34. Young, G.M., 1988. Proterozoic plate tectonics, glaciation and iron formations. Sediment. Geol., 58, 127144. Young, G.M., 1991. The geologic record of glaciation: relevance to the climatic history of Earth. Geosci. Can., 18: 100-108. Young, G.M., 1992a, Late Proterozoic stratigraphy and the Canada-Australia connection. Geology, 20: 215-218. Young, G.M., 1992b, Neoproterozoic glaciation in the Broken Hill area, New South Wales, Australia. Bull. Geol. Soc. Am., 104: 840-850. Young, G.M. and McLennan, S.M., 1981. Early Proterozoic Padlei Formation, Northwest Territories, Canada. In: M.J. Hambrey and W.B. Harland (Editors), Earth's Pre-Pleistocene Glacial Record. Cambridge University Press, pp. 790-794. Young, G.M. and Nesbitt, H.W., 1985. The Gowganda Formation in the southern part of the Huronian outcrop belt, Ontario, Canada. Stratigraphy, depositional environments and regional tectonic significance. Precambrian Res., 29: 265-301. Young, G.M. and Gostin, V.A., 1989. An exceptionally thick upper Proterozoic (Sturtian) glacial succession in the Mount Painter area, South Australia. Geol. Soc. Am. Bull., 101: 834-845.

248 Young, G.M. and Gostin, V.A., 1991. Late Proterozoic (Sturtian) succession of the North Fiinders Basin, South Australia: an example of temperate glaciation in an active rift setting. In: J.B. Anderson and G.M. Ashley (Editors), Glacial marine sedimentation; paleoclimatic significance. Geol. Soc. Am. Spec. Pap., 261: 207-223. Youngs, B.C.; 1975. The Geology and Hydrocarbon Potential of the Pedirka Basin. Rep. Investigations, 44. Dep. Mines, Geol. Surv. South Australia. Yuelen, W., Soqnian, L., Zhengjia, G., Weixing, L. and Guognan, M., 1981. Sinian tillites of China. In: M.J. Hambrey and W.B. Harland (Editors), Earth's PrePleistocene Glacial Record. Cambridge, Cambridge University Press, pp. 386-401. Zachos, J., Breza, J.R. and Wise, S.W., 1992. Early Oiigocene ice-sheet expansion on Antarctica: Stable isotope and sedimentological evidence from Kerquelen Plateau, southern Indian Ocean. Geology, 20: 569-573. Zalan, P.V., 1991. Influence of pre-Andean orogenies on the Paleozoic intracratonic basins of South America. In: IV Simp. Bolivariano, Memorias, 1. Zalan, P.V., Wolff, S., Conceicao, J.C.J., Marques, A., Astolfi, M.A.M., Vieira, I.S., Appi, V.T. and Zarotto, O.A., 1990. Bacia do Parana. In: G.P.R. Gabagha and E.J. Milani (Editors), Origem e Evolucao de Bacias Sedimentares. Petrobras: 135168. Zalan, P.V., et al., 1990. The Parana Basin, Brazil. In:

N. EYLES M.W. Leighton, D.R. Kolata, D.F. Oltz and J.J. Eidel (Editors), Interior Cratonic Basins. Am. Assoc. Pet. Geol. Mem., 51: 681-708. Zhang, H. and Zhang, W., 1985. Palaeomagnetic data, Late Precambrian magnetostratigraphy and tectonic evolution of eastern China. Precambrian Res., 29: 65-75. Zhao, X., Coe, R.S., Liu, C. and Zhou, Y., 1992. New Cambrian and Ordovician paleomagnetic poles for the North China Block and their paleogeographic implications. J. Geophys. Res., 97: 1767-1788. Zhaochang, Z., Yuhzen, L., Songnian, L. and Huakun, 1993. Lithology, sedimentology and genesis of the Zhengmuguan Formation of Ningxia, China. In: M Deynoux and J. Miller (Editors), Glaciation in Earth History. Cambridge University Press, In Press. Ziegler, P.A. and Van Hoorn, B., 1989. Evolution of North Sea Rift System. In: A.J. Tankard and H.R. Balkwill (Editors), Extensional Tectonics and Stratigraphy of the North Atlantic Margins. Am. Assoc. Pet. Geol. Mem., 46: 471-500. Zolnai, A.I., Price, R.A. and Helmstaedt, H., 1984. Regional cross-section of the Southern Province adjacent to Lake Huron, Ontario: Implications for the tectonic significance of the Murray Fault Zone. Can. J. Earth Sci., 21: 447-456. Zubtsov, E.I., 1972. Precambrian tillites of Tien-Shan and their stratigraphic value. Byull. Mosk. Ova. Ispyt. Otd. Geol., 1: 42-56.