Ii
Antarctic Geology and Gondwanaland
CAMPBELL CRADDOCK
Department of Geology & Geophysics University of Wisconsin, Madison
Although ideas of drifting continents had been advanced earlier, it was not until late in the nine- teenth century that geologists first suggested the existence of a large ancestral continent in the South- ern Hemisphere in the geologic past. This hypo- t etical protocontinent—termed Gondwanaland—in- Figure 1. Reassembly of Gondwanaland (after du Tot, 1937). luded present-day South America, Africa, Arabia, Madagascar, Ceylon, peninsular India, Antarctica, and Australia (see Fig. 1). Because of its position in surprisingly like Australia, only larger—around t e reassembly, Antarctica must clearly play an im- which, with wonderful correspondences in outline, portant role in determining whether Gondwanaland the remaining puzzle-pieces of Gondwanaland can actually ever existed. Despite its great importance with remarkable precision be fitted." tD the concept, however, the geology of Antarctica Because so little was then known about the geol- as almost unknown when Gondwanaland was first ogy of Antarctica, du Toit s Gondwanaland reas- postulated. Only in the past few years has our knowl- sembly in effect predicted the geologic patterns to be dge of Antarctica, the last fragment of the supposed expected in the antarctic interior. It is a tribute to protocontinent to be explored, advanced to a point the genius and foresight of du Toit that subsequent ihere meaningful geologic tests of the Gondwana- geologic work in Antarctica has largely confirmed land hypothesis are possible. these predictions. Antarctica was first sighted 150 years ago from Preparations for the International Geophysical ships sailing near the tip of the Antarctic Peninsula. Year (1957-1958) marked the beginning of the cur- Later expeditions collected continental rocks such rent phase of scientific exploration. Twelve nations as granite and sandstone from icebergs, while the established more than 40 bases in Antarctica to sup- first geologic specimens obtained on the continent port their IGY programs, and scientific activity has were collected in 1894 at Cape Adare, south of New continued at a high level since that time. Ski-equip- Zealand. The ensuing 20 years saw vigorous explora- ped aircraft and tracked vehicles have opened the tion by expeditions from several nations, and the entire continent to geologic study, and only a few broad pattern of antarctic geology began to emerge. areas of significant rock exposure remain unvisited Private and national expeditions before and after by geologists today. It is safe to say that our knowl- World War II added further geologic information, edge of the geology of this remote continent has but large areas of the continent still remained unex- more than doubled during the past 15 years. plored and unknown up to 1955. Most of these early ologic observations were confined to the coastal tinge of Antarctica; except for the areas covered by Outline of Antarctic Geology he South Pole parties of Shackleton, Amundsen, and cott, and the Gould party in support of Byrd s polar Early in this century, after geologic data became flight, the vast interior of the continent remained available from the Antarctic Peninsula, the Ross Sea eologically unknown. region, and other coastal localities, it was recognized In 1937, the great South African geologist Alex- that Antarctica could be divided into two major ánder du Toit published a book which must be rated geologic provinces. The first comprises the larger part among the most important ever written in the field of the continent that faces mainly upon the Atlantic of geology. In this brilliant synthesis, he set forth and Indian Oceans; since most of this province lies in detail the geologic evidence then available for in the area of east longitudes, it is commonly known continental drift and the existence of Gondwanaland. as East Antarctica (see Fig. 2). The second province Regarding Antarctica, he stated: consists of the smaller part of the continent that faces "The role of the Antarctic is a vital one. As will mainly upon the Pacific Ocean and is commonly be observed from Fig. 7 [present Fig. 1], the shield of known as West Antarctica. East Antarctica is a typi- East Antarctica constitutes the key-piece --shaped cal continental shield or stable platform, consisting
May-June 1970 53
0• Era I Period I Epoch I Age in my.
Quaternary -- 1-2
Cenozoic Pliocene - I Miocene - 25 Tertiary Oligocene - 40 Eocene -- 60 1. Paleocene --70- In Cretaceous - 135 EAST Mesozoic Jurassic 180 2 ANTARCTICA triassic 225- SOUth POlO 90°E - Premise 270 Carbooiferoos - 350 P Devonian WEST I Paleozoic 400 ANTARCTICA Silurian 440 Ordovician - 500 Cambrian --600- Precambrian pa S8 EA •Jco Figure 3. Geologic time scale.
180° brian and Cambrian strata, generally folded and somewhat metamorphosed, that occur in the Trans. Figure 2. Geologic provinces of Antarctica antarctic Mountains and at a few coastal localities. Overlying the East Antarctic basement complex i of a foundation of older igneous and metamorphic a succession of mainly flat-lying sedimentary an rocks overlain by a sequence of younger, flat-lying volcanic rocks. -These beds have been called the stratified rocks. By contrast, West Antarctica is Beacon Group (sedimentary) and the Ferrar Group composed of generally younger rocks that are widely (igneous), and they represent the Gondwana se deformed and metamorphosed. Intrusive and extru- quence—as described from the other southern con sive igneous rocks are abundant, and volcanic activ- tinents—in Antarctica. The Beacon Group contain ity continues there today. rocks as old as Devonian, and the Ferrar Group Rock exposures comprise less than 5 percent of rocks as young as Jurassic. These groups are exten- the area of the continent. Those in East Antarctica sively exposed in the Transantarctic Mountains, but occur in an oval belt consisting of the coastal region they have been found in place at only a few localities and the Transantarctic Mountains. The nunataks along the coast of East Antarctica. and mountain ranges within this belt reveal a base- The geologic history of West Antarctica is complex ment complex composed mainly of high-grade meta- and poorly known. All rocks whose ages are known morphic rocks and intrusive igneous rocks. Gneisses appear to have formed during the last 600 million of the granulite facies are the most abundant rocks, years, and no definitely Precambrian rocks have yet but lower-grade metamorphic rocks of the amphi- been discovered. In much of West Antarctica, the bolite and green schist facies also occur. A wide oldest rocks are igneous intrusive and metamorphic range of igneous rocks has been reported, but felsic varieties that form a basement complex believed to varieties such as granite are most common. These be Paleozoic in age. Sedimentary and volcanic se- crystalline basement rocks record a complex geologic quences of probable Paleozoic and Mesozoic age are history involving several cycles of deformation, meta- widely distributed, and most of these rocks art, morphism, and emplacement of igneous intrusives. strongly folded and somewhat metamorphosed. In In contrast to other continents, Antarctica has so trusive igneous rocks were emplaced across much o far revealed no rocks with apparent ages greater than West Antarctica during the Mesozoic, and perhaps 2 billion years, and indeed few older than 1.5 by. the early Cenozoic. Upper Cenozoic volcanic an This fact may imply that 1) the oldest rocks in the sedimentary layers on the Antarctic Peninsula ar basement of East Antarctica are truly younger than flat-lying and undisturbed. Volcanism which begax those of other continental shields, or 2) insufficient in the Miocene has continued into recent times in sampling has yet taken place to discover the oldest much of coastal West Antarctica. rocks present, or 3) later metamorphic events, es- pecially in the early Paleozoic (see Fig. 3), have reset the mineral clocks, causing apparent radio- Major Discoveries Bearing on Gondwanaland metric ages younger than the true ages of the rocks. The youngest known rocks in the East Antarctic base- In fitting Antarctica into his Gondwanaland re- ment complex are the fossiliferous Upper Precam- assembly, du Toit had to depend largely on the shape
54 ANTARCTIC JOURNAL of the continent because so little was then known suppose that Permian and modern geography are about its geology. Since that time, progress in ant- identical. Can a reasonable dispersal mechanism, arctic geology has yielded a number of discoveries such as wind or water currents, be found to connect that bear on the Gondwanaland problem. Eight lines these distant lands? And, is it possible that these two of evidence that appear significant are discussed widely separated lands, one polar and the other 1elow: tropical, could support nearly identical floras when 1. The basement rocks of coastal East Antarctica their climates would have been so different? re similar, both in a general way and in some de- Thus, the fossil animal and plant record in Ant- $ails, to those along the matching coasts of the other arctica strongly suggests that the present geographic Iondwanaland fragments in du Toit s reassembly. isolation of the continent did not exist during Paleo- Work by geologists of several nations has shown that zoic and at least early Mesozoic time. he structural grain in these ancient antarctic rocks 4. The Ellsworth Mountains fold belt formed in s compatible with that in like rocks of the suggested early Mesozoic time and probably represents the ri atching coast. The basement rocks of all these continuation of the Cape fold belt of South Africa areas are compositionally similar, consisting of high- and a part of du Toit s Samfrau geosyncline. Be- rade metamorphic rocks such as granulite gneisses, tween the Transantarctic Mountains of East Antarc- long with igneous intrusives such as granite. In tica and the coastal belt of West Antarctica lies a particular, an unusual hypersthene-bearing granitic large region that has been explored only during the rock, termed charnockite, is widespread in eastern last 15 years. Bedrock exposures in this area are Africa, Ceylon, eastern India, and coastal East Ant- found in the Ellsworth Mountains and in nunatak arctica. groups and small ranges to the south and west. On 2. The rocks of the Beacon Group in Antarctica the basis of lithologic similarity and structural con- are generally similar to the Paleozoic and Mesozoic tinuity, most of these outcrops have been assigned to Gondwana sedimentary successions on the other a new tectonic province, the Ellsworth Mountains southern continents and continental islands. The fold belt. The thick sedimentary sequence is mainly lower Beacon consists of detrital sedimentary rocks Paleozoic in age (some Precambrian strata may be as old as Devonian in some localities. Beds of ancient present) and has undergone strong post-Permian tillite were first discovered in the Transantarctic folding. Some of these formations resemble the Bea- Mountains in 1958, and many other tillite localities con Group formations in the Transantarctic Moun- are now known in Antarctica. These tillites occur tains, but differ in that they are considerably thicker in the Beacon Group and are considered Carbonif- and distinctly folded. In Part of the province, these erous or Permian in age. Overlying these glacial deformed Paleozoic strait are invaded by granitic beds are younger Permian strata which commonly batholiths that were empLced during late Triassic include coal beds and bear the Glossopteris flora. to early Jurassic time. In its present setting, the This succession of distinctive rock types can be Ellsworth Mountains fold belt is a puzzling tectonic matched, at least in part, in Australia, India, Mada- fragment resting between East Antarctica and coastal gascar, Africa, the Falkland Islands, and South West Antarctica. Both in stratigraphy and in struc- America. tural style, however, it bears a strong resemblance to 3. The Paleozoic and Mesozoic fossil record in the Cape fold belt of southern Africa and the fold rA ntarctica, only now emerging in detail, bears a mountains of eastern Argentina. In the Gondwana- strong resemblance to that found in the other south- land reassembly, it represents the natural continua- ern continents. Marine faunas are as old as Cambrian tion of a fold belt that begins in Argentina and con- and include archaeocyathids, trilobites, gastropods, tinues across southern Africa into Antarctica. bryozoans, fish, echinoderms, brachiopods and pelecy- 5. Jurassic igneous rocks, mainly basaltic in com- pods. These animals flourished in shallow waters position, are widespread in Antarctica. These rocks close to land under conditions similar to those on occur both as volcanic deposits and as shallow in- modern continental shelves, and it is unlikely that trusive bodies such as sills and dikes. They are com- they migrated across deep ocean basins. Land ani- mon throughout the length of the Transantarctic mals, less abundant, also show close affinities to the Mountains, where they have been termed the Ferrar faunas of the other southern continents. The recent Group. These Jurassic mafic rocks may be compared discoveries of Triassic reptiles and amphibians in to rocks of similar age and composition that occur the Transantarctic Mountains are of great impor- over large areas of Brazil, southern Africa, and Tas- tance. Antarctic fossil floras, especially the Permian- mania. In the Gondwanaland reassembly, it is rea- Carboniferous Glossopteris flora, bear a strong simi- sonable to interpret the rocks of all these areas as larity to other southern floras. The pronounced belonging to a single igneous province, perhaps re- overlap between the Glossopteris floras of Antarctica lated to the initial fragmentation of the protocon- and India, for instance, poses two questions if we tinent. Jurassic volcanic rocks of more varied corn-
May-June 1970 55 position are abundant in the Antarctic Peninsula have been identified in glacial erratics in the Ross and common along the coast of West Antarctica. Sea area, but these rocks have not been found in These latter rocks, along with counterparts in Ar- place. Volcanic rocks at least 7 million years old gentina and Australia, may be the products of activ- overlie a glaciated surface in the Jones Mountains ity in a mobile belt along the margin of Gondwana- of coastal West Antarctica, and late Cenozoic vol- land. canoes to the west in the same province contain der 6. Late Cretaceous to early Tertiary igneous plu- posits that suggest eruption of the lavas through th tons, mainly granitic in composition, are widespread ice sheet. An interesting record of glacial and vol+. in the Antarctic Peninsula and westward along the canic events of the last few million years is preserve4 coastal sector of West Antarctica. Similar intrusive in some of the deglaciated valleys of the Transantarc rocks are typical of the western margin of the Amer- tic Mountains. Thus, although there still remain icas from Canada to Tierra del Fuego. The presence much to be learned about the Cenozoic history o of such plutons, along with the geologically young Antarctica, it seems clear that by 7 million years ago deformation and abundant Cenozoic volcanism, sug- the continent was isolated from the other Gondwana gest that coastal West Antarctica may be properly land fragments and was in a geographic position fa considered a part of the circum-Pacific mobile belt vorable to the growth of an ice sheet. I of Mesozoic-Cenozoic time. Igneous and tectonic events in the Southern Hemisphere part of this mo- bile belt may be related to the breakup of Gondwana- Antarctica as a Fragment of Gondwanaland land, since the belt lies along the leading edges of the drifting fragments. One anomaly is the presently aseismic character of Antarctica; elsewhere, the Du Toit s reassembly of Gondwanaland, if correct, circum-Pacific belt is typified by numerous earth- requires that certain geologic elements exist in Ant- quakes. arctica. The basement rocks of coastal East Antarc- tica must resemble in composition and structure those 7. Magnetic-anomaly belts parallel to and sym- of the matching coasts of the other Gondwanaland metrical about the mid-ocean ridges strongly suggest fragments. The Beacon Group should be present that sea-floor spreading and continental drift have throughout the length of the Transantarctic Moun- occurred. Within the last 15 years, oceanographic tains and should resemble the Gondwana sequences surveys have revealed this unexpected pattern of of the other southern continents. Upper Paleozoic alternating belts of high and low magnetic intensity, glacial beds should occur in Antarctica. The Paleo- a pattern which initially defied explanation. Con- zoic and Mesozoic fossil record should consist of current studies of paleomagnetism from both lava forms resembling those elsewhere in Gondwanaland flow and marine sediment sequences, however, pro- rather than unique forms developed in an isolated duced evidence that the earth s magnetic field ap- continent. The eastward extension of the Samfrau pears to undergo periodic reversals of its polarity. If geosyncline and the Cape fold belt from southern the alternating high- and low-intensity belts are Africa should be found in Antarctica. Jurassic mafic ascribed to these polarity reversals, then the slowly igneous rocks can be expected to be present in th spreading sea-floor can be thought of as a magnetic tape that "freezes in" the existent magnetic field as Transantarctic Mountains along with the strata o the Beacon Group. new crust is formed at the mid-ocean ridges by the cooling of silicate melts from the earth s interior. If The rapidly accumulating geologic data from Ant- this interpretation proves correct, we may hope to arctica strongly suggest that each of these prediction learn both the rate of sea-floor spreading and the will prove to be correct. The accuracy of du Toit time when the various fragments of Gondwanaland predictions, the striking geologic similarities between began to separate. Antarctica and the other southern lands, and the new evidence for sea-floor spreading combine t 8. The Antarctic Ice Sheet appears to have formed provide a very strong case for the reality of Gond- at least 7 million years ago, suggesting that Antarc- wanaland. tica was a separate, polar continent by that date. The late Cenozoic history of the continent is obscure be- cause few deposits of this age are known. A limited number of paleomagnetic measurements suggest that Future Research Antarctica was in its present latitude even at the beginning of the Cenozoic. Tertiary floras from the Although the Gondwanaland concept is winning Antarctic Peninsula area suggest that moderate tem- wide acceptance, further testing of the hypothesis peratures prevailed there during part of the era, but against the emerging facts of antarctic geology is fossil penguins have been recovered from beds con- necessary. The geologic similarities established in a sidered Miocene in age. Early Tertiary microfloras general way must be critically examined through
56 ANTARCTIC JOURNAL local, detailed studies. While most of Antarctica has been geologically mapped on a reconnaissance scale, The Fossil Tetrapods of the remaining unknown areas must be investigated. At the same time, more detailed mapping of certain Coalsack Bluff areas is necessary, both to clarify the structural framework of the continent and to test fully the EDWIN H. COLBERT correctness of du Toit s predictions. Museum of Northern Arizona The basement rocks of East Antarctica require nuch more investigation. Those in the Transantarc- ic Mountains appear to be younger than those in Tetrapods and Continental Drift oastal East Antarctica; if correct, this fact is im- ortant in understanding the tectonic evolution of The evidence for an ancient continent of Gond- he East Antarctic shield. Detailed comparative wanaland, the fragments of which now constitute tudies of the ancient rocks of coastal East Antarctica the several Southern Hemisphere land masses and their assumed counterparts on the matching the peninsula of India, is various. Much of it is very Zoasts of Africa, India, and Australia are needed. convincing. Moreover, the numerous lines of inde Radiometric age determinations on these igneous and pendent data would seem to fit together reasonably metamorphic rocks will be a great help in testing the well, giving us an integrated concept of the supposed original continuity of basement rock terranes in Ant- Gondwana continent and its fragmentation. Yet, arctica with those in the other southern lands. almost all of the evidence is equivocal in one way or Detailed studies of the Beacon Group are just be- another, resting to a greater or lesser degree on cer- ginning, and it deserves examination along the length tain basic assumptions, the validity of which some- of the Transantarctic Mountains to establish lateral times may be called into question. changes in composition, age, and origin of these The theory of Gondwanaland and continental strata. Thorough knowledge of the coastal localities drift must be in accord with all of the evidence if it is especially important to allow careful comparison is to be valid. In this regard, the records of the rocks with the Gondwana successions in Africa, India, and and fossils are particularly important. As the re- Australia. Additional fossil discoveries can be ex- nowned geologist Walter Bucher remarked in 1964: pected in the Beacon Group and in the stratified "Ultimately the proof [for drift] must come from the rocks of West Antarctica. geologic and paleontologic record." Two kinds of studies of the geologic structure in One of the problems that has faced the paleontol- Antarctica will enable further testing of the Gond- ogists, particularly the vertebrate paleontologists, is wanaland reassembly. On the one hand, additional that drift has not been essential to explain the dis- knowledge of the areal geology will lead to better tribution of fossils on the southern continents. This definition of the tectonic provinces of the continent, is certainly true for the land-living vertebrates of such as the Ellsworth Mountains fold belt, and to Cenozoic age: their distributions, from the beginning more accurate comparison of these provincial bound- of the Cenozoic era to the present, are on the whole aries in Antarctica and the other Gondwana frag- very readily explained by the present intercontinen- ments. On the other hand, thorough study of the tal land connections, notably the isthmian link be- structural style within each province is necessary to tween the two Americas and the undoubted Bering. test the supposed continuity between that province Strait "bridge" that periodically connected the New n Antarctica and a tectonic province in another World with the Old throughout the past 60 million outhern land. years or more of geologic history. For older fossil Finally, continued investigation of paleomnagiietisrn vertebrate faunas, however, the concept of a former 11 Antarctica should provide valuable insights into (;ondvarsa1and that subsequently fragmented, its its geologic history. Although this technique is corn- several fragments—now the Southern Hemisphere nonly a complicated one in l)1 a(t1( e, it (an, under conti nents---dri fting to their present positions, has avorahie circumstances, he used to establish paleo- become increasingly attractive within the past two latitudes. Thus, the paleomagnetism locked in ant- decades. The very close resemblances between the arctic rocks, together with the mnagnetic-anorrialv Triassic reptiles of South Africa and South America, belts frorn the ocean floor, offer hope of direct con- for example, would seem to indicate a close conti- firmation, not only of the reality, but also of the guity of these continents, with a single, unified range chronology of continental drift. for the land-living animals then inhabiting what are now widely separated land masses. Yet, although Reference du Toit, Alexander L. .1937. Our Wandering Continents Cura tor Emeritus, The American Museum of Natural Edinburgh, Oliver and Boyd. 366 p. History, and Professor Emeritus, Columbia University.
May-June 1970 57