zyxwvutsrq zyxwvutsrq JOURNAL OF QUATERNARY SCIENCE 1986 1 (1) 45-56 @ 1986 Longman Group UK Ltd The Quaternary glacial sequence in Ecuador: a reinterpretation of the work of Walter Sauer zyxwvu zyxw CHALMERS M. CLAPPERTON Department of Geography, University of Aberdeen, Scotland, UK RAMON VERA Departamento de Geologia, Escuela Politkcnica Nacional, Quito, Ecuador Clapperton, Chalmers M. and Vera, Ramon 1986.The Quaternary glacial sequence in Ecuador: a reinterpretation of the work of Walter Sauer. journal of Quaternary Science, Vol. 1, pp. 45-56. ISSN 0143-2826 ABSTRACT: The Quaternary glacial sequence proposedfor the EcuadorianAndes by Walter Sauer is critically reviewed. Examination of his field evidence at sections exposing Quaternary sediments east of Quito has led to a fundamental reinterpretation. Deposits which Sauer considered as glacial, glacio-fluvial, glacio-lacustrine and pluvio-glacial in origin appear to have been formed mainly by volcanic, volcano-loessic, laharic, fluvial, colluvial and pedogenic processes. KEYWORDS: Andes, stratigraphical diagrams, cangahua, volcano-loessic, lahar. Background By the early 20th century a number of European geologists, climbers and travellers had found evidence that glaciers in the tropical mountains of Ecuador had formerly been more extensive (Reiss and Stubel, 1892; Whymper, 1892; Wolf, 1879; Meyer, 1907). The first suggested sequence of Quaternary glaciation was by Meyer (1 907) who, on the basis of morphological evidence, proposed two glacial-pluvial periods equivalent to the European Riss and Wurm ages; he believed that these were separated by a period of dry interglaciation. Sauer (1950, 1965, 1971) evolved a more elaborate scheme of Quaternary events from stratigraphical evidence exposed in deep stream-cut sections immediately east of Quito. He identified a pluvio-glacial and three glacial periods separated by three interglaciations. When first proposed (1950) this sequence appeared to correlate closely with that long accepted for the European Alps (cf. Penck and Bruckner, 1909). Although Kennerly and Bromley (1971 have briefly described glacial features observed in the Llanganati mountains of the eastern cordillera and Hastenrath (1981) has discussed the glaciation of the Ecuadorian Andes in general, the scheme proposed by Sauer in 1950 has remained unchallenged as the model for Quaternary glaciation in Ecuador. This paper reviews the evidence used by Sauer to determine the former existence of glacial and interglacial periods and suggests that alternative interpretations are more plausible. Journal of Quaternary Science canyon, in Quebrada Guarangupugru and at the Guangopolo electric power station (Fig. 1). His interpretation of the sediments, together with cursory observations of what he believed were glacial deposits at other localities in Ecuador, formed the basis for his Quaternary sequence. The two main diagrams are reproduced in this paper as Figure 2. The following discussion summarises Sauer’s scheme and draws attention to some of the assumptions he made. The ‘Pluvio-glacial’: The lower 34 m of sediments exposed in the Rio Chiche canyon consist of alternating beds of conglomerate and fine volcanic material, containing inclusions of unconsolidated gravels and coarse sand. Sauer (1965 p.265) believed that these were the products of intense denudation of ‘Neo-Tertiary’ beds of lava and tephra under pluvio-glacial climatic conditions. He assumed a moist environment to account for the coarse sediments and also that the Ecuadorian Andes had not yet been uplifted to altitudes of more than 3000m. The last assumption permitted him to argue that small glaciers confined to the higher summits were present during this period. Sauer concluded that any morainic deposits were removed by larger glaciers during subsequent glaciations but interpreted the coarse sands and gravels as fluvio-glacial products from the small glaciers. The 1st Interglacial: Quaternary sequence of Sauer (1950,1965) Sauer published three stratigraphical diagrams depicting sedirnentsexposed east of Quito in the intermontanedepression known as the Sangolqui basin; the sites are in the Rio Chiche The identification of an interglacial period seems to be based on Sauer’s interpretation of an unconformity at the top of his ‘pluvio-glacial’ beds in the Rio Chiche section. He argued that the interglacial period i s recognised from an interruption (caused by renewed uplift of the cordillera) to what he called ‘fluvio-lacustrine’ sedimentation and by the appearance of true alluvial sediments. JOURNALOF QUATERNARY SCIENCE 46 zyxwvutsrqpon Figure 1 Map showing location of sites discussed in the text. The 2nd Glaciation: Figure 2a shows that 17.8m of sediment interpreted as a ‘glacio-lacustrine’ deposit is Sauer‘s evidence for a second glacial period. He believed this interval to have been chayacterised by large ‘Malaspina-like’ outlet glaciers advancing from the eastern cordillera to terminate in lakes occupying the intermontane basins. Heenvisagedthat melting icebergs calved from floating glaciers and dropped mounds of coarse sediment onto the lake floors. The repetition of ‘glacio-lacustrine’ beds in the sequence suggested many readvances of the glaciers, but at no point did Sauer discuss evidence for the former existence of large lakes or explain his interpretation of the sediments. The 2nd Interglacial: This was characterised by the first appearance in the stratigraphical column of sediment interpreted by Sauer as ’eolian cangagua‘. He believed this to be a sediment of volcanic origin, but did not precisely explain the mechanism of its deposition. He distinguished the material from underlying ‘lacustrine cangagua’, a member of the previously-formed ’glacio-lacustrine’ formation. In a discussion of ’cangagua‘ Sauer (1950, p.23-26) drew attention to its similarity with loess (we presume he was thinking of European loess), but because he thought that cangagua lacked the calcium carbonate content which he believed to be typical of loess deposits, he concluded that the sediment was not of loessal origin. In view of a mineral assemblage similar to that of andesites and dacites, Sauer decided that cangagua was an eolian sediment produced during warm and dry interglacial conditions (unlike the European loess which i s the product of cold glacial conditions). Sauer subsequently used ‘eolian cangagua’ as a marker bed for interglacial periods and made the fundamental and unsubstantiated assumption that the main period of Pleistocene volcanism began during the 2nd interglacial, thereby providing a source for the sediment. He argued without any supporting evidence that volcanoes such as Rucu Pichincha, llalo and part of Chimborazo developed at this time. Sauer’s observation that cangagua lacks calcium carbonate is not strictly true since several horizons in the Rio Chiche canyon contain abundant nodules of cangagua cemented by interstitial calcite. Furthermore, microscopic examination of thin sections of cangagua show that it is highly siliceous (70% SiOz), corresponding more with an origin from dacitic - rhyolitic magma than from andesite. Also present i n the ‘eolian cangagua’ are ball-like concretions (bolas de cangagua), which are the fossilised nests of dung beetles (Coprinisphaera ecuadoriensis). Sauer identified the origin of these features and also noted the presence of other fossils incorporated within the cangagua -the bones of large grazing mammals such as Mastodon, Giant Llama and Horse (Hoffstetter, 1952). On the basis of such evidence he envisaged a semi-arid savannah or steppe-like grassland landscape in the intermontane basins, thereby supporting his view that the ‘eolian cangagua’ must be an interglacial deposit. zyxwv zyxw The 3rd Glaciation: Sauer (1950 p.26-27) concluded that the 3rd glaciation was the largest of all because renewed tectonic activity had created higher land. Volcanoes which, he believed, had been constructed during the preceding interglacial period became glacially sculptured for the first time, and ‘Malaspina‘ outlet glaciers from ice caps on the eastern cordillera again reached the intermontane basins. He considered that this powerful glaciation created most of the ‘U‘-shaped glacial valleys present zyx zyxwvutsrq zyxwvuts THE QUATERNARY GLACIAL SEQUENCE IN ECUADOR: A REINTERPRETATION OF THE WORK OF WALTER SAUER 1 m Aeolian cangagua Cangagua balls 10.0 Limestone concretions 1.o 1.5, 0.7 2.0 Pumice - 0.3 3 r d Interglaciation 47 Volcanic ash Lacustrinb cangagua 2.0 0.4 1.1 1.o Fine sand 0.3 1.o 1.2 Lacustrine cangagua and pumice Palaeosol 6.0 Lacustrine cangagua, sandy with pumice 0.1 1.1 Ice stillstand layer Sandy cangagua Sandy-loamy cangagua with fossils Limonite concretions Lacustrine, loamy cangagua Sand layer Palaeosol Moraine Lacustrine cangagua with limonite Erosional discordance Aeolian cangagua with balls Lava boulders Sand and gravel 4.0 1.o 1.5 3rd Glaciation 3.0 2nd Interglaciation 2.0 2.0 1.5 1.5 1.5 1.5 2.0 2nd Glaciation zyxwvutsr Glacio-lacustrine deposits 2.0 2.0 1.8 1st Interglaciation Pluvio-glacia 1-1 310 14.0 0 O .".". .".". ." 0 0 25. Fluvial erosion and sedimentation F l Beds of various volcanic material 26. with boulders, gravels and coarse sand in monotonic repetition Figure 2a Interpretationof sediments exposed in the Rio Chiche gorge (Fig. 4) by W. Sauer (1950, 1971). D JOURNALOF QUATERNARY SCIENCE 48 1 2 3 4 5 zyxw zyxw Moraine Fossll soils Eolian cangagua L a c u s t r i n e aedtmen t s Lacustrine cangagua 6 Fine send 7 Dark g r a v e l s B Dark g r a v e l s 9 Lacustrine c a n g a g u a 10 Eolian c a n g a g u a Figure 2b Interpretation of sediments exposed in the Quebrada Cuarangupugru (Fig. 7) by W. Sauer (1 950). in the cordilleras of Ecuador. The main sedimentary and stratigraphical evidence for this glacial period consists of ’dark gravels’ in Quebrada Guarangupugru, ‘fluvioglacial gravels’ at Guangopolo and ‘moraine’ in the Rio Chiche section. The last deposit required glaciers to extend 1Okm from the eastern cordillera, and on the basis of this Sauer had to assume that tectonic uplift created land sufficiently high to generate such large glaciers. Reinterpretationof the Rio Chiche and Quebrada Guarangupugru Sections zyxwvutsrq The 3rd Interglacial In the Rio Chiche section Sauer‘s marker beds of ‘eolian cangagua’ overlie sediments of the 3rd glaciation, and on this evidence he assumed a return to warm and dry interglacial conditions. Renewed uplift, faulting and volcanic activity were also suggested for this period, leading to the formation of the younger parts of Pichincha and Chimborazo volcanoes. Sauer also proposed (1950 p.27) that tectonic dislocation had caused drainage incision and the drying up of lakes in the intermontane basins; this assumption was necessary to account for the absence of ‘lacustrine cangagua’ during and after the ’fourth glaciation‘. The 4th Glaciation: The 4th Glaciation was considered to be of relatively small extent although glaciers became large enough to remove all traces of terminal moraines constructed during the 3rd glaciation. According to Sauer (1 950, p.291, renewed uplift of the cordillera permitted sufficient nourishment on the higher land for glaciers to re-occupy the ’U’ valleys created during the previous glaciation and thus reach similar limits. Sauer observed fresh terminal and recessional moraines of the 4th glaciation in parts of the cordillera, and since these moraines appeared to contain a matrix of ’eolian cangagua’, he concluded that volcanic activity had continued during this glacial period. The Postglacial: This period is characterised by eolian cangagua overlying deposits of the 4th glaciation. The inference is drawn that warm and dry conditions accompanied by volcanic activity have affected the intermontane basins during the postglacial period. Figures 3a, 3b and 3c illustrate the present writers‘ interpretation of the sedimentsand stratigraphy exposed in the RioChiche canyon and in the Quebrada Guarangupugru at the same sites described by Sauer. The characteristics of each section will be discussed briefly before considering their implications for the Quaternary sequence proposed by Sauer. The Rio Chiche Section Approximately 120m of sediments are exposed by road- and river-cut sections where the Tumbaco-Pifo road crosses the Rio Chiche at an altitude of 2400m (Fig. 1). The sediments fall naturally into two major formations separated by a distinct discontinuity (Fig. 4); these are the Chiche Formation and the *Cangahua Formation (Vera 1983). The following discussion provides descriptions and new interpretations of the two formations. The Chiche Formation: This has three sedimentary components: a conglomerate interbedded with cemented fine tephra and sporadic beds and pockets of unconsolidated sand, gravel and cobbles. The Chiche Formation can be subdivided into lower and upper units that are 29m and 48m thick respectively. The basis for subdivision is merely the more frequent occurrence in the upper part of the formation of interbedded tephra layers (mean thickness 30cm) and more sand and gravel pockets. Conglomerate beds in the lower unit are up to 1.5 m thick and less than 1.0m thick in the upper unit. Discontinuous lenses of sand, gravel and cobbles as well as conspicuous layers of black ashes occasionally occur in the lower unit. Pockets of gravels zyxwvu *Sauer (1950) used the spelling ‘cangagua’ whereas Vera (1983) used ‘cangahua‘. Cangagua is the Indian word for “hard soil”, whereas cangahua is the modern spelling adopted by INEMIN (Ecuadorian Geological Survey). THE QUATERNARY GLACIAL SEQUENCE IN ECUADOR: A REINTERPRETATION OF THE WORK OF WALTER SAUER Qepth below ground surface Thickness Conglomerate with sub-rounded zyx 49 - sub-angular volcanic clasts (up to lOcm diameter) : gradual decrease in clast size to fine sandy beds 20cm thick. This sedimentary cycle is repeated 8 times Towards the top a sand lens shows cross - stratification. The section is typified by dark coloured beds (conglomerates) alternating with light coloured beds (sandy and fine tephra with occasional beds zyxwvutsr zyxw P of volcanic clasts 30cm thick). zyxwv zyxwv zyxwv zyxwv zyxw Light coloured beds predominate.. Tephra with normal stratification containing volcanic clasts towards the middle. 6 { o.*.b.. .be. b\ ............... : ........ '.I ::.:::::...:.I.,:: :.'. : ... ....... Grey-black conglomerate with dark sub-angular volcanic clasts. This section consists of cemented conglomerate beds interstratified with light coloured tephra beds. . . . . The clasts consist mainly of dark andesites + + but vitric basalts are also present ( 1 %). . . ........ .,: ............. .o''...0'0....ij ....b... . . .. .. .. . . :........... Figure 3a Interpretation of sediments composing the Chiche Formation in the Rio Chiche gorge. and well-developed fluvial channel-fills of 4-5 m width and depth are present in the upper unit. The conglomerate has been analysed in detail by Vera (1983). It is composed predominantly of sub-angular shaped clasts of volcanic origin, the most common lithologies being andesites (pyroxene-olivine dominant), trachyandesite (with biotite) and basaltic andesite; also present are traces of rhyolite, chert, obsidian and graywacke. The sandy matrix consists of primary crystals, including plagioclase (60%), amphibole (5-7%), pyroxene (6-1O%), biotite (3-5%)and magnetite (5%), and fine lithic volcanic fragments; it i s cemented by secondary calcium carbonate and iron oxide. Measurements of the long axes of clasts indicate a S-N preferred alignment with angles of imbrication dipping towards their provenance area in the south. The conglomerate is poorly sorted in general; grading structures vary from normal to inverted within the same bed, to beds lacking any recognisable internal stratification. it is most commonly a dark-coloured clast-supported deposit, clast size ranging from 2 cm-8cm with occasionally larger cobbles and blocks present; interstices between clasts are filled with fine, silt-like material. Such characteristics indicate that the conglomerate i s more like a coarse mud flow deposit than a fluvial or glacio-fluvial sediment. The interbedded tephra i s a cream-coloured, structureless sediment conspicuously different from the conglomerate. Although it i s rnineralogically similar to the latter, it contains no large clasts and has a mean grain size of O.l20mm, the range of coarse-medium silt to fine sand. Lower beds of tephra are characterised by poor development of silica diagenesis and pedogenesis, suggesting relatively rapid deposition of the sedimentary units; tephras in the upper part of the formation appear to have undergone more alteration through pedogenic processes before the deposition of interbedded conglomerates, suggesting a slower rate of sedimentation. The Chiche Formation was periodically incised by streams which cut channels ca 4-6m wide and 3-5rn deep into the aggrading surface of the basin (Fig. 4). The channels are filled with loose accumulations of gravels and cobbles that are predominantly sub-angular to sub-rounded in shape; the clean smoothed surfaces of most clasts suggest that the deposit is mainly a water-worn stream sediment. Cobbles exceeding 50cm are common and in general these deposits are much coarser than the cemented conglomerate. North of llalo a wider range of lithologies is present in the gravels and cobbles and although andesitic rocks dominate, the presence of vitric clasts indicates a more easterly source for these sediments. The characteristics of the Chiche Formation suggest the following interpretation. The conglomerates appear to be consolidated volcanic mud flows (lahars) from a southerly source. The presence of some structures and imbrication indicates that a substantial water content was present in some of the lahars, particularly in the upper unit of the formation. The tephra layers most probably represent fall-out material from the explosive volcanism which generated the lahars. Because THICKNESS THICKNESS Accumulated Acarmlated Parti 0.65 0.85 Cangahua.. yellow-fawn 28.98 29.39 29.79 2.85 2.50 0.40 0.40 Pumice. 5mm size Volcanic sand Cangahua. soft 3 1.79 2.00 Cangagua with pumice clasts. 4-6cm 33.7s 33.89 2.00 0.10 34.9s 1.10 35.3s 0.40 36.54 1.15 37.54 1.oo Fluvial sand and gravel Cangahua. light fawn, partially bedded Qritty aluvium Cangahua. prismatic joints Cangagua. cubic joints 11.35 ,n Cangagua-soil. black-brown 0.45 Cangaeua with pumice clasb 39.44 1.90 40.44 1.00 41.04 0.60 Cangagua. cubic joints 15.50 Soil. rubified Cangahua 19.50 45.04 4.00 45.34 0.30 20.70 22.25 Pumice-sand with iron pan Cangahua. cubic joints 1.50 0.20 Hard white sediment 48.24 1.20 Cangahua. cubic joints, plant remains 48.54 0.30 Stream gravel, 15cm size 50.14 50.24 1.60 0.10 50.84 0.80 Stream gravel Soil, rubified zyxwvutsrqpon 25.45 26.48 52.44 Figure 3b Interpretationof the Cangahua Formation in the Rio Chiche gorge. 1.60 Cangahua with pumice clasts 5cm size, lenses of stream pebbles THICKNESS Partial A C C,umulated 0.60 0.03-0.08- zyxwvutsrqponmlkji zyxwvu zyxwvutsrqponm + THICKNESS Accumulated Partial Present soil: stoney Fine tephra Coarse lahar deposit Weathered, light-brown, fine tephra : light greyorange mottling Organic rich soil : developed on fine tephra Pumiceous tephra layer and l c m size particles Coarse tephra 2.10-0.18-- Soil, weathered. light chocolate 28.72 0.90- - 2847 0.75- - Fine tephra. weathered Fine tephra. weathered with poor organic soil on top : contalns stones up to 4cm size 27.84 27.82 GAP +Gap 41.47 44.47 44.97 Weathered, clayey tephra-based soil 47.17 White fine grained tephra Weathered pumiceous tephra 48.22 48.23 4863 48.87 49.47 Fine tephra. lens Grey-brown soil Fine tephra with pumice Mid-brown soil, clayey Sandy tephra Brown clayey soil Alluvial sand and silt with 11 laminae Alluvial sand, mottled grey and brown Weathered fine tephra Yellow-fawn mottled : upper 10-12cm is grey-brown clayey soil Fine tephra. weathered greylbrown : u p p e r 1 2 - 1 5 c m i s y e l l o w l b r o w n Soil 50.07 51.09 53.1 9 54.93 55.87 Tephra, weathered with lapilli-like aggregates of quartz and pumice : top 45cm is reddishlchocolate soil 58.67 60.17 Light grey tephra containing pumice particles 1-4cm size Figure 3c Interpretation of sediments exposed in the Quebrada Guarangupugru. Stream 52 zyxwvutsrqponm JOURNAL OF QUATERNARY SCIENCE zyxwvutsrqpo Figure 4 The Rio Chiche gorge site described in Figures 2a, 3a and 3b. The Cangahua Formation lies above bridge level, the Chiche Formation lies below. One of the ’cut-and-fill’ pockets of gravel and cobbles crops out in the Chiche Formation at the centre bottom of the photograph. lahars require a source of water, the likely origin of those in the Sangolqui basin is from snow- and ice-coveredvolcanoes lying to the south. Since recent eruptions of Cotopaxi have discharged lahars northwards through the basin, this volcano is an obvious source for at least some of the Chiche conglomerates. The lithological characteristicsof the conglomerate are consistent with an origin from this volcano and from adjacent extinct volcanoes which may have been active in glacial times. The lenses and pockets of unconsolidated fluvial sand, gravel and cobbles appear to be cut-and-fill sediments related to the occasional discharge of stream floods across the former land surfaces underlain by the aggrading sequence of laharic conglomerate and tephra beds. All or part of these fluvial deposits seem to have originated in valleys draining to the Sangolqui basin mainly from the eastern cordillera. It is not yet clear whether the deposits are the produce of meltwaters from former glaciers, or merely relate to heavy rainstorms. The Cangahua Formation: A distinct sedimentological discontinuity separates the Chiche Formation from the overlying Cangahua Formation which continues the section a further 52 m to the present land surface. The principal differences between the two are that the Cangahua Formation does not contain beds of conglomerate and is mainly unconsolidated. The Cangahua Formation is so-called (Vera 1983) because it is dominated by the fawn-coloured fine volcaniclastic sediment known in Ecuador as cangahua; it i s interbedded with three other sedimentary facies. Fluvial deposits at the base of the formation mark the last occasion when either fluvial or volcanic floods crossed the Sangolqui basin. Three palaeosols mark intervals when pedogenesis on the land surface was not interrupted by heavy fall-out from volcanic activity or by slope movements and stream action. Eruptions from adjacent volcanoes produced distinctive beds of pumiceous lapilli and sand-sized tephra on several occasions. The beds of very conspicuouswhite pumice are two pyroclastic flow deposits separated by a unit of cangahua; the flows appear to have come from eruptions in the eastern cordillera, but the source has not yet been determined. In some horizons the cangahua has evolved diagenetically through the development of amorphous silica and contains scattered ‘relic’ pumice fragments. Pieces of broken beetle nests and reworked cangahua fragments also occur in places. These may be colluvial deposits originating from the downslope movement and mixing of materials. Particular features of the Cangahua Formation which may have palaeoclimatic implications are the cangahua nodules, the fossilised beetle nests and the buried soils. The nodules are composed of cangahua particles cemented together with interstitial secondary calcite and vary in shape from smooth ovoid to rough irregular masses commonly from 10-13 cm in size (Fig.5). A road-cut on the east side of the Chiche canyon exposes the nodule horizons most clearly. They occur in significant numbers only above the white pumice horizons and appear to be irregularly disseminated through 4m of homogenous cangahua. Similar nodules contained in wind-blown silts have been observed and studied in northeast Argentina where they are believed to have formed during cool semi-arid conditions that coincided with glacial periods (Tonni and Fidalgo 1983). Those in Ecuador may have formed in similar circumstances: that is, from the diagenesis of calcium carbonate contained in the cangahua (also an air-born silt) as a zy zyxw zyx THE QUATERNARY GLACIAL SEQUENCE IN ECUADOR: A REINTERPRETATION OF THE WORK OF WALTER SAUER 53 interpreted as marker beds corresponding to times of greater moisture and vegetation cover, possibly during interstadial or interglacial periods. The Quebrada Guarangupugru Section Examples of cangahua nodules (top)and fossil beetle nests (bottom)from the Cangahua Formation. Figure 5 result of particular climatic conditions. As in Argentina, these could have been cool and semi-arid, coinciding with glacier expansion in the adjacent cordillera. At high altitude in equatorial latitudes strong evaporation over a bare semi-arid landscape could have influenced groundwater movement in such a way as to achieve the necessary diagenesis of the calcite. A variety of materials overlie the main nodule-bearing horizon, including layers of hard and soft cangahua (Fig. 3). The hard layers appear to be partially cemented and contain many noduleswhereas the soft layers are relatively free from them and more commonly contain fossil beetle nests. Sauer (1950) observed the possible link between dung beetles and large grazing animals, the former depending on excretions of the latter for their procreation; the fossils of both occur in the Cangahua Formation. Dung beetles are no longer present in Ecuador and exist mainly in semi-arid parts of the world populated by grazing animals. It is believed that dung beetles excavate spherical caves close to the surface of silt deposits; they fill these with balls of dung as a food supply for the larvae the instant they hatch from eggs laid in a hollow in the dung-ball. Some types of dung beetle burrow to a common level of c. 30cm to bury their’ nests; others develop dendritic networksof burrow tunnels to depths of 2-4 m (Crowson, 19811. The latter lead to a more random dispersal of nest balls throughout 2-4m of sediment. After the young beetle has consumed its food supply and has left the nest, fine sediment surrounding the ball subsequently creeps into the space inside. Because the shell is more consolidated and harder than surroundingsediment it survives as a relic feature (Fig. 5). Since dung beetles at present exist only in semi-arid parts of South America, Sauer interpreted the relic balls in Ecuador as evidence that such conditions formerly occurred in the intermontane basins at a time when grazing animals roamed across a steppe-like landscape; he implied, however, that this was during interglacial periods. Since climatic conditions in the Amazon basin during glacial build-up were drier than at present (Goudie 1983, pp. 96-97, Ochsenius, 1985), it may be inferred that climate in the adjacent Andes was also more arid because their predominant rain-bearing air masses were derived from the same sources. Thus the steppe-like conditions suggested by thebeetleand animal remainsaremore likely to havecoincided with the cooler and drier glacial periods rather than with interglacial periods as Sauer envisaged. The buried soil horizons are conspicuous because of their reddish-dark brown colour and clearly-developed profiles. Their palaeoclimatic significance is that they indicate distinct periods of soil development interrupting more prolonged episodes of cangahua sedimentation when conditions were too arid or cold to permit comparable pedogenesis. The buried soils, therefore, may be The Quebrada Guarangupugru is a narrow canyon cut into slopes descending from the llumbisi horst to the Sangolqui graben. The canyon grades to the Rio San Pedro, which drains the depression between llalo volcano and the llumbisi horst before joining the Rio Chiche to become the Rio Guayllabamba (Fig. 1). The valley, at an altitude of 25001-11, is well outside the limits of the last glaciation as defined by terminal moraines. The section described by Sauer (1950 p.33) is easily found because of the distinctive dark band in the upper part labelled on his diagram as a fossil soil. Figure 3c illustrates the present writers’ interpretation of the sediments exposed at this site; in their opinion, the only horizon correctly interpreted by Sauer is the ‘fossil soil’. It is clear that Sauer mistook laharic deposits generated by eruptions of Cotopaxi for morainic deposits of his so-called 3rd glaciation. Laharic deposits discharged down the Rio San Pedro depression have been mapped by Miller et a/. (1978), the most recent event occurred in 1877. Beneath the superficial laharic deposit (Fig. 6) the well-stratified sediments, interpreted by Sauer as glacio-lacustrine sediments, consist of at least 1 7 palaeosols interbedded with fine-grained volcanic materials (Fig. 7); the latter are composed mainly of silt- and sand-sizedtephras mixed with occasional layers and inclusions of pumice lapilli. The mean organic content of the palaeosols is 3.14%; the highest value of 4.6% is in layer 5 (Figs. 3c and 7). Stream sands and gravels predominate at the base of the section, suggesting that the site experienced more fluvial activity when the land surface was c. 65 m lower. Sediments exposed in the Quebrada Guarangupugru are different from those composing the Cangahua Formation in the Rio Chiche in that they lack concretionary nodules and compacted layers; fossilised beetle nests are also rare. Such contrasts could be explained by site differences; the Rio Chiche area is a lower, relatively dry basin floor site whereas the Cuarangupugru site is on a higher slope, possibly with more abundant moisture and a more dense vegetation cover. The absence of the white pumice layers so conspicuous at the Rio Chiche site (and in various sections exposed along the QuitoCayambe road) is puzzling if the ages of sediments at the two zy zy zyxwvutsrqpo Figure 6 Lahar deposits from Cotopaxi exposed near the Cuangopolo electric power station; interpreted by W. Sauer as a moraine. 54 zyxwvutsrqponm JOURNAL OF QUATERNARY SCIENCE of the river; the bedding and imbrication of these suggest that they were derived from valleys cut into contiguous slopes of llalo volcano. As at the other two sites, it was not possible to interpret any particular series of glacial-interglacial events from the sediments exposed at Guangopolo. Chronology zyxw Only three dates relating to the age of cangahua deposits in the Sangolqui basin are known. One is from Guangopolo where a piece of wood obtained from above the sand and gravel horizons and at the base of the cangahua gave a radiocarbon age of >48800 yr BP (Bristow et a/., 1980). Another is from Bonifaz (1972) who concluded from archaeological evidence and obsidian hydration-dating of microliths that the main deposition of cangahua had ended by 12900 yr BP or possibly by 21 600 yr BP. The third is from a sample taken 4 m from the surface of the Cangahua Formation (at the base of horizon 22 of Estrada, 1941) in the vicinity of Quito; it has given an age of c. 35000 BP by thermoluminescent dating (A.M.D. Gemmell, pers comm). The formation in this locality is c.8m thick, and although there are no data on its rate of accumulation, these dates encourage speculation that the Cangahua Formation may have accumulated entirely within the last 100000 yr. zyxwvutsrqp zyxwvutsrq zyxwvutsr Figure 7 Palaeosols interbedded with alluvial sediments in the Quebrada Cuarangupugru. Soil layers 5-14 of Figure 3c are visible (top to bottom). sites are broadly corresponding. One possible explanation is that the Guarangupugru site in the lee of llalo was protected from the pumice ash flow. The stratigraphical equivalent of the pumice eruption is possibly the 33cm layer of white, finegrained tephra at a depth of c. 12m. This interpretation is supported by the observation that immediately above this tephra is a 4 m layer of homogeneous, apparently weathered, cangahua-like material, possibly the equivalent of the 4 m cangahua layer overlying the white pumice in the Rio Chiche section. The Guangopolo Section Sediments exposed on both banks of the Rio San Pedro immediately downstream from the electric power station (Fig. 1) were not sampled or analysed in detail. However, field examination of the deposits indicated that they were primarily of volcanic, laharic, fluvial and colluvial origin. For example, large boulders up to 2m size are scattered widely over the ground surface in this area above the stream canyon and were probably derived from the same lahar event observed at Quebrada Guarangupugru; beds of tephra and fine pumice lapilli commonly occur in the cangahua-like horizons, and at least two buried soils are present. A major difference between the sediments exposed at Guangopolo and the other two sites is the greater abundance of water-laid materials. Over 20m of bedded sands, gravels and cobbles are exposed on the east bank Conclusions We conclude that the sediments interpreted by Sauer (1950, 1971) as glacial, glacio-fluvial and glacial-lacustrine are not of such origins and that the foundation for his Quaternary sequence must therefore be considered untenable. It is not yet possible to replace Sauer’s scheme with an equally elegant model of Quaternary glacial and interglacial periods but Table 1 illustrates a tentative sequence based on more recent fieldwork (Hastenrath, 1981; Clapperton, 1983; Clapperton and McEwan, 1985; Clapperton, 1986). The table shows that a reasonably clear pattern of glacier fluctuations is known for the last c. 40000 yr. Most morainic landforms seem to be associated with this period. Some morphologically subdued ridges may have been deposited earlier in the last glaciation, losing their fresh form through denudation and a covering of cangahua. Although theentire sequence suggested in Table 1 needs to be thoroughly tested and refined with more dating, two outstanding problems require early attention. One is to determine whether or not the belt of 3-4 closely-spaced moraines assigned to the last glaciation represent glacier fluctuations widely separated in time, but falling within the last 100000 yr. For example, interpretation of the oxygen-isotope record from ice sheet and ocean bed cores suggests global periodsof cooling separated by warmer (interstadial) conditions peaking around 70000, 50000 and 20000 BP. It i s possiblethat glacier systems more sensitiveto climatic fluctuations than the large continental ice sheets may have respondedto the cold peaks by advancing. Thus parts of the moraine complexes in Ecuador and the partially oxidised till overlain by peat with a radiocarbon age >40000 BP may have been formed by glacier advances coinciding with cold peaks in the middle and early parts of the last glaciation. Evidence from the adjacent Andes in Colombia suggests that glaciers were more extensive before c. 35000 BP zyxwvut zy zyxwvutsrqpo THE QUATERNARY GLACIAL SEQUENCE IN ECUADOR: A REINTERPRETATIONOF THE WORK OF WALTER SAUER 55 Table 1 Quaternary glacial sequence in the Ecuadorian Andes. The range of valuesfor altitudinal limits of each formation is due to the asymmetry in distribution, the lower values occurring on the eastern (windward) side of the mountains Formation Characteristics Altitudinal Limits(m) Age Determination Little Ice Age Moraines, trim-lines. 4100-4800 ~15th-early~ 2 0 t h . Neoglacial Moraines (superposed in places). Moraines. 3900-4600 pre-2000 BP 3800-4400 12000-10000 BP 3000-3900 C35000 BP Late-glac ia I Last glaciation (late) Last glaciation (earlier) Pre-last glaciation Moraines, erosion forms. Subdued moraines, erosion forms, oxidised ti IIc2m. Ox idised ti II>2m, deep1y weat hered diamictite, erosion forms. Source Hastenrath, 1981 Clapperton 1986 Clapperton 1986 Clapperton & McEwan 1985 Clapperton & McEwan 1985 Clapperton & McEwan 1985 zyxwvutsrq 2900-3800 >40000 BP 2750-3250 no data than during the period of maximum cold which occurred after c. 20000 BP when the climate became drier (van der Hammen et al., 1981). The problem in Ecuador may be resolved if volcanic sediments interbedded with glacial deposits can be found and dated with thermoluminescence and fission track methods and if biostratigraphicalanalyses, such as those by van der Hammen and associates in Colombia, are applied to deep cores from sedimentary basins. The second problem is how to interpret deeply weathered diamicites such as those exposed in parts of the western and southern cordilleras. In terms of their clast characteristics shape, size-range, varied lithology - the diamictites are similar to adjacent or overlying glacial deposits, but since the clasts are weathered to sand or clay it is not possible to detect evidence of glacial polish or striations; the diamictites could therefore be of mudflow, debris avalanche or alluvial origin. However, because the sediments closely resemble adjacent glacial deposits and are not found beyond the limits of the glacially eroded parts of the valleys in which they are situated, they have been tentatively interpreted as deeply weathered tills (Clapperton 1983). The degree of weathering implies an age greater than the last glaciation and, by analogy with weathered tills in the Chilean lake district (also a warm, humid environment), they could be older than the penultimate glaciation. It is known that glacier fluctuations have affected parts of the Andes periodically over the last 3 My (Clapperton 1983). The penultimate glaciation seems to have been more extensive than the last throughout the Andes. In the southern Andes at least four major glaciations have occurred during the last 800000 yr, each one less extensive than the previous. Old, deeply weathered tills may therefore be expected in the Ecuadorian Andes, assuming that these mountains were high enough (through tectonic uplift) earlier in. the Quaternary. Since this paper concludes that the sedimentary sequence exposed by deep sections in the intermontane basins of Ecuador is largely of non-glacial origin, partly because of the accumulation of volcanic deposits, it is suggested that the longest record of Quaternary glaciation and interglaciation is more likely to be preserved in the nonvolcanic southern cordillera. Clapperton 1983 References zyxw BONIFAZ, E. 1972. Microlitos arqueologicos. Quito. BRISTOW, R., CEVALLOS, L., LONGO, R. and MASIN, M. 1980. Mapa geoldgico del Ecuador, Escala 1 :50,000,Hoja 84-SW, Sangolqui, Quito. CLAPPERTON, C. M. 1983. The Glaciation of the Andes: Quaternary Science Reviews, 2, 83-155. CLAPPERTON, C. M. 1986. Glacial geomorphology, Quaternary glacial sequence and palaeoclimatic inferences in the Ecuadorian Andes. First lnternational Conference on Geomorphology Proceedings, Manchester 7985. (In press). CLAPPERTON, C. M. and McEWAN, C. 1985. Late Quaternary moraines in the Chimborazo area, Ecuador. Arctic and Alpine Research, 17, 135-1 42. CROWSON, R. A. 1981. The biology of the Coleoptera, Academic Press, London, 802 pp. ESTRADA, A. 1941. Contribucion geologica para el conocimiento de la Cangagua de la region interandina y del Cuaternario en general en el Ecuador. Anales de la Universidad Central Quito, 66,405-488. GOUDIE, A. 1983. Environmental Change. 2nd Edition, Clarendon Press, Oxford, 258 pp. HASTENRATH, S . 1981. Glaciation of the Ecuadorian Andes. A.A. Balkema, Rotterdam, 159 pp. HOFFSTETTER, R. 1952. Les mammifieres pleistocenes de la Republique de I'Equateur. Mem. SOC. Geol. Fr. (NSI 31,66,391 pp. KENNERLY, J. B. and BROMLEY, R. J. 1971. Geologyand geomorphology of the Llanganati Mountains, Ecuador. lnstituto Ecuatoriano de Ciencias Naturales (Quito), 73, 3-1 6. MEYER, H. 1907. In den Hochanden von Ecuador. D, Reimer-Ernst Vohsen, Berlin, vol 1, 522 pp. MILLER, C. D., MULLINEAUX, D. R. and HALL, M. L. 1978. Reconnaissance map of potential volcanic hazards from Cotopaxi Volcano Ecuador: U.S. Geological Survey Miscellaneous Investigations Series, Map 1-1072. OCHSENIUS, C. 1985. Pleniglacial desertization, large-animal mass extinction and Pleistocene-Holocene boundary in South America. Revista de Geografi'a Norte Grande, 12,35-47. PENCK, A. and BRUCKNER, E. 1909. Die Alpen im Eiszeitalter: Tauchmitz, Leipzig, 1 199 pp. REISS, W. and STUBEL, A. 1892. Reisen in Sudamerika. Das Hochgebirge der Republik Ecuador. Petrographixhe Untersuchungen, Vol. 1: West-Cordillere; Vol. 2: Ost-Cordillere. Asher & Co., Berlin, 358 and 356 pp. SAUER, W. 1950 Contribuciones para el conocimiento del Cuaternirioen El Ecuador. Analesdela UniversidadCentral, Quito, 77, (328): 327-364. zyxwvu zyxwv zyxwvutsrq AcknowledgmentsThanks are due to the Royal Society, The Carnegie Trust for the Universities of Scotland, and Aberdeen University for supporting field work in Ecuador in 1983 and 1985. 56 zyxwvutsrqpon zyxwvutsrqp zyxwvutsr zyxwvutsrq jOURNAL OF QUATERNARY SCIENCE SAUER, W. 1965. Geologia del Ecaudor. Editorial de Ministerio de Educacion, Quito, 383 pp. SAUER, W. 1971. Geologie von Ecuador. (Beitrage zur Regionalen Geologie der Erde, Vol. 1 1 ) Gebrueder Borntraeger, Berlin-Stuttgart, 316 pp. TONNI, E. P. and FIDALCO, F. 1983. Geology and palaeontology of Pleistocene sediments at Punta Hermengo area (Miramar, Provinceof Buenos Aires, Argentina): Some palaeoclimatic aspects. In, Rabassa, I. (Ed). Quaternary ofSouth America andAntarctica Peninsula Vol 1, 23-53. Balkema, Rotterdam. VAN DER HAMMEN, T., BARELDS, 1. DE jONG, H. and DE VEER, A. A. 1981. Glacial sequence and environmental, history in the Sierra Nevada del Cocuy (Colombia). Paleogeography, Palaeoclimatology, Palaeoecology, 32, 247-340. VERA, R. 1983. Caracteristicas petrogrdficas y palaeogeogrdficas del conglomerado Chiche en 10s alredadores del Ilalo. Unpublished report. Politecnica Nacional, Quito, 4pp. WHYMPER, E. 1892. Travels amongst the Great Andes ofthe Equator. Charles Scribner's Sons, New York, 456 pp. WOLF, T. 1879. Geografia y geologia del Ecuador Erockhaus, Leipzig, 671 pp.