The eastern part of the Western Cordillera of Ecua-dor comprises thick buoyant oceanic plateaus asso-ciated to island arc-tholeiites and subduction relatedcalc-alkaline series, accreted to the continental mar-gin of Ecuador from late Cretaceous to Eocene times(Kerr et al. 2002, Mamberti et al. 2003). Two oceanicplateau sequences have been identified: the SanJuan oceanic plateau dated to 123 Ma (Lapierre etal. 2000) and the Guaranda oceanic plateau (90-86Ma, Mamberti et al. 2001) considered as remnant ofthe Cretaceous Caribbean Oceanic Province(CCOP). Mamberti et al. (2003) suggest that thisplateau is radiogenically enriched in 206Pb/204Pb and207Pb/204Pb and contains a HIMU component similarto that observed in the Gorgona basalts and Gala-pagos lavas. Guaranda plateau Pb isotopes plot in arestricted field overlapping those of Pacific MORB.

Samples studied here were taken from four UpperCretaceous arc-sequences in the northern part of theWestern Cordillera of Ecuador (Rio Cala, Macuchi)and in the Chogòn-Colonche Cordillera (Las Or-quideas and Cayo). These four island arcs rest onthe CCOP. They consist predominantly of clinopy-roxene-bearing basalts and andesites. The completepetrological and geochemical study of these rocksreveals that some have a primitive island-arc nature(MgO values range from 6 to 11 wt%). Their arc-affinity is shown by the Nb, Ta and Ti negativeanomalies. These rocks are LREE-enriched andtheir bearing clinopyroxenes show a tholeiitic affinity(FeOt-TiO2 enrichment and CaO depletion from coreto rim within a single crystal and the whole sample).

Initial Nd, and Pb isotope ratios are very homoge-neous and suggest that these rocks result from mix-ing of three components: an E-Pacific MORB mantle(Hauff et al. 2003), an EM2 enriched component,and a HIMU (Zindler & Hart. 1986) component.Characterising the EM2 and HIMU components isimportant to constrain the genesis of these arc-

volcanics. The subduction zone that generated thelate Cretaceous arcs occurred far from the conti-nental margin, in an oceanic environment (Jaillard etal., 1995). This implies that no terrigenous detritalsediments have interacted with the source at this pe-riod. Thus, the EM2 component can only result fromthe melting of subducted pelagic sediments. TheCCOP that supports these arc sequences is char-acterised by a HIMU component (Révillon et al.1999, Mamberti et al. 2001) that could have beenassimilated by the island arc lavas.

Mixing models reveal that proportions of 20% ofthe HIMU component and 20% of the pelagic sedi-ment end-member are needed to explain samplechemistry. These surprisingly high proportions canbe explained by the young age of the CCOP (5 to 10Ma) when the Late Cretaceous arcs developed. TheCCOP, basement of these arc sequences, wasprobably still hot and easily assimilated by the islandarc lavas.

REFERENCES

Hauff, F., Hoernle, K. and Schmidt Angelika. (2003). Sr-Nd-Pbcomposition of Mesozoic Pacific oceanic crust (site 1149and 801, ODP Leg 185): Implications for alteration of oceancrust and the input into the Izu-Bonin-Mariana subductionsystem. Geochem., Geophys., Geosyst. V.4, N°8.

Jaillard, E., Ordoñez, O., Suárez, J., Toro, J., Iza, D., Lugo, W.,2004. Stratigraphy of the Late Cretaceous-Paleogene de-posits of the Western Cordillera of Central Ecuador: Geo-dynamic implications. J. South Am. Earth Sci., 17, 49-58.

Kerr, A.C., Aspden, J.A., Tarney, J. and Pilatasig, L.F., 2002.The nature and provenance of accreted terranes in West-ern Ecuador: Geochemical and tectonic constraints. Jour-nal of the Geological Society, London, 159, 577-594.

Mamberti, M. (2001). Origin and evolution of two Cretaceousoceanic plateaus accreted in Western Ecuador (SouthAmerica), evidenced by petrology, geochemistry and iso-topic chemistry. Thèse univ. Lausanne-Grenoble,

The role of the Cretaceous Caribbean Oceanic Plateauin the genesis of late cretaceous arc magmatism in Ecuador

*Allibon, J., **Monjoie, P., **Lapierre, H., **Jaillard, E., *Bussy, F. & ***Bosch, D.

* Institut de Mineralogie et de Geochimie, Université de Lausanne, Suisse

** Laboratoire de Géodynamique des Chaines Alpines, UJF-Grenoble1

*** Laboratoire Tectonophysique, UMR-CNRS 5568, cc049 Université Montpellier II

Mamberti, M., Lapierre, H., Bosch, D., Ethien, R., Jaillard, É.,Hernandez, J., Polvé, M. (2003). Accreted fragments of theLate Cretaceous Caribbean-Colombian Plateau in Ecuador.Lithos, 66, 173-199.

Révillon, S., Arndt, N.T., Hallot, E., Kerr, A.C., Tarney, J., 1999.Petrogenesis of picrites from the Caribbean Plateau andthe North Atlantic magmatic province. Lithos, 49, 1-21.

Zindler, a. & Hart, S.R. (1986).Chemical geodynamics. Ann.Rev. Earth Planet. Sci. Lett., 14: 493-571.

In this part of Chile and Argentina located at 40-42°S, the active subduction setting of the Nazcaplate beneath the South America plate, the meltingof the Patagonian Ice Sheet during the Late glacialand the concomitant growth of large stratoconesover the most active volcanoes of the Americas, re-sulted in a complex Late Quaternary geomorphol-ogy.

Northern Patagonia is characterized by strongprecipitations driven by the Westerlies over the SEPacific and is including many sub-aqueous environ-ments that are sensitive natural archives of past en-vironmental changes: lakes of glacial, tectonic orvolcanic origin, but also fjords and bays that wereflooded by the postglacial sea level rise. Strongwesterly winds and intense volcanic activity in thestudy area result in the formation of thick and unsta-ble andosoils (soils developed on volcanic ash)draping the steep morphologies of the Andes. Suchclimate and soils favours the development of a verydense vegetation cover consisting of a temperateevergreen rain forest.

In this study, the recent sedimentation processesin four contrasting lacustrine and marine basins ofNorthern Patagonia are documented by high-resolution seismic reflection profiling (3.5 kHz) andshort cores at selected sites in deep lacustrine ba-sins. The regional correlation of the cores is providedby the combination of 137Cs dating in lakes Puyehue(Chile) and Frías (Argentina), and by the identifica-tion of the Cordon Caulle 1921-22 and 1960 tephrasin lakes Puyehue and Nahuel Huapi (Argentina) andin their catchment areas. This event stratigraphy al-lows correlating across the Andes the formation ofstriking sedimentary events in these basins with theconsequences of May-June 1960 earthquakes andthe well-documented induced Cordon Caulle erup-tion next to the Puyehue volcano; only 38 hours afterthe main shock offshore Valdivia. This rhyodacitic

fissure eruption was triggered along the Liquiñe-Ofqui Fault Zone (LOFZ), a major active intra-arcshear zone in this part of the Andes. Thousands oflandslides were also triggered in the Andes along theLOFZ, during the multiple seismic shocks of themajor 21-22 Mai 1960 subduction earthquakes (Mw9.5).

Some of these well-documented landslides werecatastrophic and deeply affected the drainage basinsof most of the glacial lakes from the Chilean LakeDistrict. In Argentina earthquake-triggered landslideswere also documented, but the most striking histori-cal chronicles are the descriptions of violent lakewater oscillations or destructive waves triggered bythe main seismic shock, in some of the largest Ar-gentinean glacial lakes between 39.5 and 46°S.

Our study shows that while this catastrophe in-duced a major hyperpycnal flood deposit of ca. 3.106

m3 in the proximal basin of Lago Puyehue, it onlytriggered and unusual organic rich layer in theproximal basin of Lago Frías, but destructive wavesand a large sub aqueous slide in the distal basin ofLago Nahuel Huapi. A very recent megaturbidite inthe two distal basins of Reloncavi fjord (Chile) lo-cated close to the LOFZ is also suggesting that 1960co-seismic movements in this area may have trig-gered the remobilization of ca. 187.106 m3 of marinesediments. Sub bottom profiling in these contrastinglacustrine and marine basins at both sides of theAndes are highlighting the recurrent incidence ofmajor sedimentary events during the Late Holocene.In order to confirm these reconstructions on the im-pact of 1960 earthquakes in Northern Patagonia,further studies should include dense grids of high-resolution seismic profiling and detailed studies ofsediment cores. Moreover, these deep lacutrine andmarine basins have the potential to document therecurrence of major subduction earthquakes in thispart of South America over several millennia.

Consequences of Mai 1960 major subduction earthquake in theAndes and on lacustrine and marine sedimentation of Northern

Patagonia (Chile, Argentina)

a Chapron, E., b Ariztegui, D., c Mulsow, S., d Villarosa, G., c Pino, M.,d Outes, V., e Charlet, F. & f Juviginié, E.

a Geological Institute, ETH Zentrum, Zurich; Switzerland, b Institute F.A. Forel and Department of Geology andPaleontology, University of Geneva, Switzerland; c Instituto de Geociencias, Universidad Austral de Chile, Valdivia,Chile; d Centro Regional Universitario Bariloche, Universidad Nacional del Comahue, Bariloche, Argentina; e Renard

Centre of Marine Geology, Gent University, Belgium; f Physical Geography, Université de Liège, Belgium.

bGlaciology and Geomorphodynamics Group, Department of Geography, University of Zurich, Switzerland

*Remote Sensing Laboratories, Department of Geography, University of Zurich, Switzerland

The Quindío-Risaralda basin lies west of the CentralCordillera. It is limited westward by the Tertiarysediments of the Serranía de Sta Barbara, whichwere folded during the late Tertiary Andean tectonicphase and separate the the basin from the CaucaValley (Fig. 1). Several SSW-NNE trending majorfault lineaments dissect the area, particularly theRomeral Fault System (Fig. 1B). The latter is associ-ated with the activity of the Ruiz-Tolima volcanicsystem, which produced the material deposited inthe fans. Following the dramatic earthquake of Ar-menia in 1999, sedimentological, geomorphologicaland tectonic studies have been undertaken to betterunderstand the geology of this area and ultimatelyunravel the mechanisms associated with the Pleisto-cene geological history of this zone (Gorin et al., thissymposium).

Geographically, the sedimentary infill of theQuindío-Risaralda basin can be subdivided fromsouth to north into the Quindío, Pereira and CartagoFans (Fig. 1C). Several individual units can be dis-tinguished within these deposits according to theirstratigraphic succession, lateral continuity, genesisand sedimentological and petrographic parameters,(Guarin et al., 2005). Sedimentology shows a transi-tion with the increasing water/sediment ratio fromdebris avalanches to debris flows, transitional flows,hyperconcentrated flows and finally normal stream-flows. This gradation permits the subdivision of thefans into proximal, intermediate and distal parts (Fig.1C). The imbrication of the different units permitstheir relative dating. Thicknesses vary from morethan 200 meters in the proximal part to less than 50meters in the most distal parts. In the distal part ofthe fans, particularly along the La Vieja River and inthe Cartago Fan in the north (Fig. 1C), volcaniclasticmass flows are interbedded with fluvio-lacustrinesediments of the Zarzal Formation (Neuwerth et al.,2005). Palynological data indicate an age of less

than one million years for the latter formation.Therefore, the largest part of the mass-flow fans areof Pleistocene age (Suter et al., this symposium).

So far, the origin of subsidence in the Quindío-Risaralda basin cannot be clearly explained. Never-theless, a clear relation exists between the activefault pattern, present-day drainage pattern and dis-tribution of volcaniclastic units. The basin bears theexpression of three major transpressional faulttrends observed at a more regional scale: the N-Strending Cauca-Romeral System, the E-W trendingSalento System and the SW-NE trending PalestinaSystem (Fig. 2). The present-day drainage pattern ofthe fans has been heavily influenced by the verticalthrows along the fault systems. This multiple activefault system led to the formation of localized pull-apart basins that became depositional lows for thevolcaniclastic units. The study of the Quindío-Risaralda basin is integrated with that of the ZarzalFormation at the western edge of the volcaniclasticfans and in the Cauca Valley (Suter et al., this Sym-posium). It is hoped to come up with an integratedinterpretation of the dynamics of this interandean re-gion.

This research is supported by the Swiss NationalScience Foundation (grant no. 21-67080.01).

REFERENCES

Gorin, G., Guarin, F., Neuwerth, R., Suter, F., Espinosa, A. &Guzman, C. (2005): Contribution of Quaternary sedimentsto the understanding of the tectonic history in Central Co-lombia: the volcaniclastic fans in Quindío-Risaralda and theZarzal Formation in the Cauca Valley. 3rd Swiss Geo-science Meeting, Zurich 2005.

Guarin, F., Gorin, G.& Espinosa, A. (2005): A Pleistocenestacked succession of volcanic mass flows in Central Co-lombia: the Quindío-Risaralda Fan. Acta Vulcanologica (inpress).

The interandean Quindío-Risaralda basin in Central Colombia andits Pleistocene infill by stacked volcaniclastic mass flows derived

from the Central Cordillera

*Guarin, F., *Gorin, G. & **Espinosa, A.

* Department of Geology-Paleontology, University of Geneva, Switzerland, Fernando.Guarin@terre.unige.ch;Georges.Gorin@terre.unige.ch

** Faculdad de Ingeneria, University of Quindío, Armenia, Colombia, espinosaar@epm.net.co

.Neuwerth, R., Suter, F., Guzman, C. & Gorin, G. (2005): Soft-sediment deformations in a tectonically active area : thePleistocene Zarzal Formation in the Cauca Valley (WesternColombia). Sedimentary Geology (in press).

Suter, F., Neuwerth, R., Guzman, C. & Gorin, G. (2005): Depo-sitional model for the Quaternary Zarzal Formation (Colom-

bia) and its stratigraphic relationship with the volcaniclasticmass flows derived from the Central Cordillera. 3rd SwissGeoscience Meeting, Zurich 2005

.

Figure 1. A) Location of study area; B) Geological cross-section AA’ through the Quindío-Risaralda basin; see Fig. 1C for location. C)Simplified geological map. The fluvio-volcanic infill of the Quindío-Risaralda basin is illustrated by a Digital Elevation Model (DEM).The zone framed in white underwent detailed field studies. The profile below the map shows the present-day surface slope of theQuindío Fan up to the Quindío volcano.

Figure 2. Interpretation of active faults and drainage patterns in the Quindío Fan (i.e., the lower part of Figure 1C). The limits ofdrainage compartments are shown as dotted lines on the interpretation at the right handside and as white lines on the DEM. Thethree major fault trends observed both at local and regional scales are shown at the right handside.

Healed microfractures and mineral or fluid inclusions in olivinexenocrysts and in xenoliths from Tatara San Pedro complex, Central

Chile: Evidence for subsolidus and remobilisation history

Ginibre, C. & Dungan, M. A.

Department of Mineralogy, University of Geneva, Switzerland

catherine.ginibre@terre.unige.ch, Michael.Dungan@terre.unige.ch

Phenocryst proportions in many mafic lavas of theQuaternary Tatara-San Pedro complex (TSPC;36ES, Chilean Andes) are minor compared to modalabundances of coarse olivine, clinopyroxene, andplagioclase xenocrysts derived from disaggregatedmafic-ultramafic xenoliths. The xenocrystic nature ofthe olivine crystals is indicated by the presence ofembayments, melt channels with mineral composi-tion different from the groundmass, and an abun-dance of healed microfractures (HMF, Dungan &Davidson, 2004). Modelling of iron diffusion in olivineshows that the residence time of these crystals in thebasalt is on the order of a few years (Costa & Dun-gan, 2005).

Other units, in particular a Holocene dacite flowfrom Volcan San Pedro contain abundant xenolithsof amphibole- and phlogopite-bearing gabbros(Costa et al., 2002.). Most minerals in these xeno-liths, and in particular the olivines, also containabundant HMF. The dacitic composition of the hostlava implies moderate temperatures (<980°C), andtherefore less thermal impact following incorporationin the host magma on the observed features. This isconsistent with observations suggesting limited re-action / melting of the xenoliths with the hostmagma.

We are investigating these HMF and their inclu-sions in both types of rocks. Because these featuresare found in volcanic rocks but originally formed inplutonic rocks, they provide two types of information:1) The original fractures are remnants from a sub-solidus history, indicating the extent of remobilisationof plutonic roots of the arc. 2) The history of remobi-lisation has reheated the xenocrysts in a basaltic anddacitic environment for various amounts of time.

The HMF’s are show two types of characteristicfeatures: 1) many of the HMF’s are marked by thinlinear low-Fo zones (_Fo = 1 to 5; width= 3 to 10µm) along the healed microfracture. 2) various kindsof fluid and mineral inclusions were trapped afterhealing of the fracture. We have investigate thechemical variations in olivine adjacent to the HMFusing back scattered electron images and electronmicroprobe In order to assess the extent of modifi-cation that the HMF have undergone following incor-poration of host in magma. Some of the HMF par-tially retain subsolidus histories in the form of fluidand or mineral inclusions. We use Raman spec-trometry to identify these phases.

Back-scattered electron images (Fig 1a) andelectron microprobe analyses delineate narrowzones of high Fe contents along the HMF. Becauseof the rapid diffusion of iron in olivine, the fact thatthese features have not been erased shows thatthey are relatively late and the shape of the Fe pro-file can be modelled in terms of residence time, as-suming the temperature known.

In the olivine xenocrysts from the basalts, the iron-rich zone is usually narrow (<5 µm) with a sharpboundary, and (isothermal) iron diffusion modellinggives time since crystallisation of a few hours atmagmatic temperatures, indicating that these frac-tures were probably open upon eruption and that thecrystallisation of the iron rich zone is mainly post-eruptive. Some longer diffusion times (a few days)suggest possible crystallisation during ascent. Inboth cases most information (fluids) from the sub-solidus stage will have been lost. The large effect ofremobilisation is consistent with the high tempera-ture of the host lavas and seems to be correlatedwith the residence time of the olivine in the lava.

Figure 1. HMF in olivine xenocrysts from TSPC basalt. a)Backscattered electron image showing Fe rich zone (lightergrey); b) Transmitted light microscopy.

In contrast the olivines from the xenoliths en-trained in the Holocene dacite, rarely show the ironrich zone along the HMF axis, although such healedfractures are especially abundant. When present, theiron-rich zones have a much more diffuse profilesimilar to those near olivine rims. The time given byiron diffusion at temperatures between 925 and525°C is a few years to over 1 Myrs. This seems toolong for post-eruptive crystallisation event at mag-matic temperatures. Therefore, these iron variationsprobably represent either a crystallisation in the re-opened HMF immediately after xenolith incorporationor a low temperature subsolidus history.

First results of Raman spectroscopy analyses forthe identification of fluid and mineral inclusions showfollowing features :Whereas most of the HMF’s that show recent re-opening, in the form of Fe-rich zones, seem, as ex-pected, to have lost their trapped fluids (or containlow pressure fluids), some others still retain some ofthe original fluids / minerals, modified to an unknownextent following incorporation.

Figure 2. HMF in olivine: HMF1 (showing Fe rich line in BSEimage) is empty whereas HMF2 inclusions, partially decripi-tated, still contain CO2, and N2 as well as magnesite.

The moist common fluid inclusions are rich inCO2 (gaseous state) with or without significantamount of N2. (Fig 2). The inclusions exhibit partialdecrepitation features, and, in some cases, magne-site as a reaction product of the CO2 with olivine. Nofluid water was observed, but a few inclusions seemto contain antigorite, a reaction product of water witholivine. We also found anhydrite crystals in some in-clusions suggesting very oxidizing conditions.

A more detailed Raman study is under way, in or-der to characterise the fluids phases. The xenolithsand xenocrysts are thought to come from the plu-tonic root of the arc at relatively shallow depth. Theinclusions present in the olivines have, therefore, re-corded fluids flowing through the crust. Althoughmodified by the post incorporation heating, the inclu-sions contained in xenocysts and xenoliths have thepotential to reveal the nature of the crust sampledand assimilated by the magma, especially in terms ofvolatile elements and fluid phases.

REFERENCES

Costa, F. Dungan, M.A. & Dinger B.S.(2002): Hornblende- andphlogopite-bearing gabbroic xenoliths from Volcan SanPedro (36°S), Chilean Andes: Evidence for melt and fluidmigration and reactions in subduction-related plutons.Journal of Petrology, 43(2) : 219-241.

Dungan, M.A. & Davidson J. (2004): Partial assimilative recy-cling of the mafic plutonic roots of arc volcanoes: An exam-ple from the Chilean Andes. Geology 32(9): 773-776.

Costa, F. & Dungan, M. A. (2005): Short timer scales of mag-matic assimilation from diffusion modelling of multiple ele-ments in olivine Geology, in press.

The subduction of the buoyant aseismic CarnegieRidge beneath the active margin of Ecuador, sinceat least 6 Ma, results in several modifications in themagmatic activity of this region, namely the wideningof the volcanic arc and noticeable changes of thegeochemical composition of the erupted lavas, dueto sources and P-T conditions variations. Till today,Ecuadorian rear-arc volcanism was known throughprevious studies focused on the quaternary Sumacovolcano (0.5°S-77.6°W), which erupted hauynephenocrysts-bearing undersaturated basalts, withalkaline affinities (Colony et Sinclair 1928; Barraganet al. 1998; Bourdon et al. 2003).

Figure 1. Geodynamical simplified map of the Northern Andes(modified from Gutscher et al. 1999). Grey arrows correspondto the direction of subduction. Filled triangles represent therear-arc edifices.

The recent characterization of two new volcanicformations, the Puyo scoria cones and the Mera lavaflows, wich outcrop around the Puyo city (latitude1°S), allow to extend the Ecuadorian rear-arcprovince some 100 km south of the three previouslyknown large stratovolcanoes (Sumaco, Pan deAzucar, Yanaurcu).

Petrographical and geochemical data obtained onthese poorly known edifices provide new constraintson the volcanism of this area and preliminarymodellings about the petrogenetical origin of rear-arcmagmas are also presented.

On one hand, the Puyo cones represent ninescoria cones, with several associated lava flows,aligned along a NW-SE fissure. They sit over theMera-Upano detritic formation, whose erosionsurface was dated at about 40 kaBP (Bès de Berc etal. 2005). Consequently, these young scoria coneswere probably emplaced during a single fissuraleruptive event of late Pleistocene to possibleHolocene age. The scoriae and lavas coming fromthese small formations consist of olivine +clinopyroxene ± phlogopite phenocrysts-bearingabsarokites.

On the other hand, the Mera lavas arerepresented by several isolated thick lava flows,whose emission centre as well as their age (probablyupper Pleistocene) still remain unknown. Theselavas range in composition from medium-K basaltsto high-K andesites, bearing olivine andclinopyroxene phenocrysts.

The enrichment of the rear-arc lavas in allincompatible elements, particularly in niobium and inpotassium, like their low content in SiO2, give them

Phlogopite-bearing absarokites and basic andesites from theEcuadorian rear-arc area: Genesis and evolution

1Hoffer, G., 1,2Eissen, J.-P., 3Beate, B., 4Bourdon, E., 5Fornari, M.,1Laporte, D., 1Martin, H., 6Samaniego, P. & 7Cotten, J.

1 Laboratoire Magmas et Volcans, UMR 6524, Université Blaise Pascal, Clermont-Ferrand, France2 Institut de Recherche pour le Développement (IRD), URM 163, Clermont-Ferrand, France

3 Departamento de Geologia, Escuela Politecnica Nacional, Quito, Ecuador4 Institut de Géologie, Université de Neuchâtel, Suisse

5 Instituto Geofisico, Escuela Politecnica Nacional, Quito, Ecuador6 Institut de Recherche pour le Développement (IRD), UMR 6526, Géosciences Azur, Université de Nice, France

7 Domaines Océaniques, UMR 6538, Université de Bretagne Occidentale, France

an atypical character. Indeed, these compositionsdiffer noticeably from fore arc or main arc lavas, wichare mainly andesitic. Although only few isotopicaldata are available, the 87Sr/86Sr (0.7041-0.7042) and143Nd/144Nd (0.5128) isotopic ratios, obtained onthree Mera samples, show that extremely limitedcontinental contamination might have affected thesemagmas during their ascent towards the surface,despite the presence of an old cratonic basement(Guyana shield) and their similarities with the forearc lavas.

Figure 2. K2O vs. SiO2 diagram showing Puyo cones and Meralavas. Field for Sumaco (rear-arc stratovolcano) and Altarvolcano (edifice belonging to the main arc) come from Bourdonet al. 2003 and from IRD geochemical data base.

The genesis of the rear-arc magma is alsoinvestigated, as for the chemical composition andthe mineralogy of the mantellic source, the residualassemblage, the degree of partial melting and the P-T conditions of the mantle wedge during the melting.Petrogenetical modellings, constrained with majorand trace elements, allow a preliminary estimation ofthese parameters. The peridotitic mantle, previouslyenriched by about 3% of slab-melts, undergoes avery low partial melting event (1-1.5%). The residualassemblage contains always olivine, orthopyroxeneand phlogopite. Nevertheless, depending on themodellings, clinopyroxene and garnet might bepresent in the residue. All the lavas of the Puyocones series and most of the Mera lavas can beexplained only by a slight increase of the degree ofpartial melting (from 1% to 2%), linked with thedisappearance of the residual phlogopite. However,the two most differentiated Mera lavas (>55% SiO2)require additional fractional crystallization, leaving acumulate with olivine, clinopyroxene, plagioclase,titano-magnetite and some apatite.

Subsequently, an experimental approach isenvisaged, in order to improve these preliminaryresults. Low degree partial melting experiments of aperidotite, enriched by an adakitic liquid(Quimsacocha adakite, Ecuador), will be carried outin a piston-cylinder apparatus, for the purpose ofevidence the importance of the hydrated phases(e.g. phlogopite or pargasite) during the melting.Effectively, the presence of phlogopite in the mantlesource, as well as its residual behaviour duringpartial melting, is a key point of the genesis of suchundersaturated incompatible rich melts in the rear-arc area of the Ecuadorian subduction zone.

REFERENCES

Bès de Berc, S., Soula, J.C., Baby , P., Souris, M.,Christophoul, F. & Rosero, J. (2005): Geomorphic evidenceof active deformation and uplift in a modern continentalwedge-top-foredeep transition: example of the easternEcuadorian Andes. Tectonophysics 399: 351-380.

Barragan, R., Geist, D., Hall., M.L., Larson, P. & Kurz, M.(1998): Subduction controls on the compositions of lavasfrom the Ecuadorian Andes. Earth and Planetary ScienceLetters 154: 153-166.

Bourdon, E., Eissen, J.-P., Gutscher, M.-A., Monzier, M., Hall,M.L. & Cotton, J. (2003): Magmatic response to earlyaseismic ridge subduction: the Ecuadorian margin case(South America). Earth and Planetary Science Letters 205:12-138.

Colony, R.J. & Sinclair J.H. (1928): The lavas of the volcanoSumaco, Eastern Ecuador, South America. AmericanJournal of Science 216: 299-312.

Hesse, M. & Grove, T.L. (2003): Absarokites from the westernMexican Volcanic Belt: constraints on mantle wedgeconditions. Contributions to Mineralogy and Petrology 146:10-27.

The desert parts of the Andes of northern Chile areregarded as being one of the oldest landscapes onEarth. Therefore, landscape forming processes mustact at very slow rates. These slow rates have pro-moted controversial ideas on the evolution of thecentral Andean mountain chain and discussionswhether climatic or tectonic forces predominate thegeodynamic evolution of the Andes (see: Lamb andDavis, 2003; Hartley, 2005).

In order to quantify the rates of various landscapeforming processes we analyzed erosion rates of hill-slope interfluves across the slope of the westernCentral Andes (Arica area, northern Chile) in a tran-sect from the Coastal Cordillera to the Western Es-carpment into the Western Cordillera. The data con-sist of the analysis of several long-lived terrestrialcosmogenic nuclides (10Be, 21Ne, 26Al - mostly inquartz of the Oxaya-ignimbrites) forming bedrockand preliminary data from catchment wide denuda-tion rates derived from cosmogenic 21Ne in riversediments. These long-lived cosmogenic nuclidesare associated with timescales of millions of years,depending on the erosion rate. Furthermore, weanalysed sediment yield data from river gauging sta-tions representative for the last decade. Erosionrates determined by the cosmogenic nuclide analysisare estimated back into the middle to late Mioceneand rates are on the order of 10-100cm/My at thehyperarid Western Escarpment (Atacama Desert)and the Costal Cordillera. In contrast, bedrock ero-sion rates for the semiarid Western Cordillera are upto >3000cm/My, at least back into the Holocene/latePleistocene. Likewise, catchment wide erosion ratesof the Lluta-drainage system yield similar orders ofmagnitudes. Sediment yield data obtained on adecadal scale indicate denudation rates of - again - asimilar order of magnitude. These landscapes form-ing processes rates are one to two orders of magni-tude higher compared to the desert parts (Western

Escarpment) but suggest a coupling between hill-slope and channel processes.

Erosion and denudation rates positively correlatewith elevation and the historical precipitation record,suggesting a coupling between climate and erosion.In addition, it is suggested that the very old land-scapes could be preserved in the western CentralAndes thanks to low tectonic activity and the pre-vailing dry climate during the late Cenozoic.

The analyses of multiple terrestrial cosmogenicnuclides and the use of various “erosion-island” dia-grams allowed the identification of system states(disequilibrium, transient, steady state) of the nuclidesystem as well as possible complex exposure histo-ries. Complex exposure histories for samples ana-lysed by cosmogenic nuclides were identified fornon-bedrock samples, such as boulders or amalga-mated clast samples (disequilibrium state). Cos-mogenic nuclide concentrations from bedrock sam-ples of the lower Western Escarpment, however,imply near steady-state or transient states over mil-lion year timescales likely caused by processes suchas episodic bedrock spalling in the cm-scale.

Landscape processes studied by morphometricanalysis suggest near-steady state conditions formost of the western slope of the Andes (Kober etal., 2005, in press; Kober et al., subm. 2005). Al-though the surfaces have demonstrably exposedsince the Miocene, timescales to achieve cos-mogenic nuclide saturation (dynamic equilibrium)and landscape steady states may not be necessarilythe same. Nevertheless, landscape forming proc-esses during the late Cenozoic act with very lowrates and relief modification is therefore almost neg-ligible.

Surface processes and associated timescales: Cosmogenic nuclideand sediment yield data from the Central Andes of northern Chile

a Kober, F., b Ivy-Ochs, S., c Schlunegger, F., d Zeilinger, G., e Kubik, P.W., f Baur, H. & f Wieler, R.

a Institute of Geology & Institute of Isotope Geology, ETH Zurich,, CH-8092 Zurich, Switzerland

kober@erdw.ethz.ch, Tel.: ++41-44-6323637b Institute of Particle Physics, ETH, CH-8093 Zurich & Institute of Geography, University of Zurich, CH-8057 Zurich

Switzerland; c Institute of Geology, University of Bern, CH-3012 Bern, Switzerland; d Institute of Geosciences, Universityof Potsdam,, D-14476 Golm/Potsdam, Germany; e PSI/c/o Institute of Particle Physics, ETH Hoenggerberg,CH-8093

Zurich, Switzerland; f Institute of Isotope Geology, ETH Zurich,, CH-8092 Zurich, Switzerland

REFERENCES

Hartley, A.J., 2005. What caused Andean uplift? In: 6th ISAG.IRD, Barcelona, pp. 824-827.

Kober, F., Ivy-Ochs, S., Schlunegger, F., Baur, H., Kubik, P.W.and Wieler, R., subm. 2005. Denudation rates and a topog-raphy-driven precipitation threshold in northern Chile: mul-tiple cosmogenic nuclide data and sediment yield budgets.Geomorphology.

Kober, F., Schlunegger, F., Zeilinger, G. and Schneider, H.,2005, in press. Surface uplift and climate change: Thegeomorphic evolution of at the Western Escarpment of theAndes of northern Chile between the Miocene and present.In: S. Willet, N. Hovius, D. Fisher and M. Brandon (Editors),Tectonics, Climate and Landscape evolution. GSA SpecialPaper.

Lamb, S. and Davis, P., 2003. Cenozoic climate change aspossible cause of the rise of the Andes. Nature, 425: 792-797.

The Torres del Paine Laccolith (TPL) in Patago-nia/Chile is part of a chain of isolated Miocene intru-sions, which intruded into the eastern foothills of thesouthernmost Andes of Chile and Argentina. TheTPL has been dated by Halpern (1973) at 12±2 Ma(Rb/Sr model) and13±1 Ma (K-Ar biotite), respec-tively.

The laccolith intruded at a shallow level (2-4 km)between mudstones, sandstones and conglomeratesof the Cretaceous Punta Barrosa and Cerro TorreFormation. It consists of a basal part with layeredgabbroic and minor dioritic and granitic rocks (Paine-Mafic-Complex PMC, Michael 1984).

The main i-type granite is peraluminous and canbe subdivided into an alkali-feldspar porphyritic me-dium grained granite. A fluid-saturated miarolitic gra-nophyric alkali-granite is found towards the host-rock. Miaroles contains mainly quartz and feldspar,with some biotite, tourmaline, sphene, pyrite andfayalite. Granitic phases with miaroles, inside themain-granite body, bordered by biotite-Schlieren in-dicate transport and ascent of fluid-saturated magmathrough the crystallizing mush.

Major element chemistry indicates a developmenttowards evolved granites from centre to the rim ofthe intrusion.

Contact between granite and host-rock is sharpand with minor or no stoping in the roof areas. Strik-ing features are multi-phase dykes, oriented perpen-dicular to the host rock contact. They continue forseveral tens of meters into the host rock. Late basal-tic, rhyolitic and composite dykes crosscut the intru-sion and the host rocks.

Field evidence shows the intrusion of the graniteafter the PMC and a feeder-zone is located at thewestern end, in the Lago Grey area. Here PMC-rocks are vertically cut by the granite. Here no fluid-saturation is evident at the contact of PMC andgranite. Further east granite overlays the PMC withmostly sharp contacts. Edges of the PMC are par-tially broken and intruded by the granite.

Sediments south of the intrusion have been re-gionally deformed showing horizontal N-S trendingfold-axes, being tighter in the western part. Ap-proaching the intrusion the fold-axes begin to dipsouthwards with increasing angle closer to the gran-ite. In the west structures indicate the ascending limbof an anticline, which is underlain by the PMC andthe granite. Similar features, with inverse dip rela-tions, can be observed at the western contact.

Vertical emplacement of the granite most likelyoccurred at the level of PMC resulting in uplift ofsedimentary strata.

We present models using the discrete elementmodel (DEM),(Malthe-Sorenssen et al. 2004), fordifferent scenarios of emplacement for the graniticlaccolith to reproduce the field observations.

REFERENCES

Halpern, M. (1973): Regional geochronology of Chile South of50°S Latitude. Geological Society of America, Bulletin, 84,p. 2407-2422.

Michael, P.J. (1984): Chemical differentiation of the Cordilleradel Paine granite (southern Chile) by in situ fractional crys-tallization. Contributions to Mineralogy and Petrology, Vol.87, p. 179-195.

A. Malte-Sørenssen et.al. (2004): Formation of saucer-shapedsills, in: Physical Geology of High-Level Magmatic Systems(F. Breitkreuz, C. Petford, eds), Geological Society, Lon-don, Special Publications, 234, 215-227.

The Torres del Paine laccolith, S-Chile

*Michel, J., *Baumgartner, L.P, **Malthe-Sørenssen, A., *Darbelly, B.,

***Oberhänsli, R., *Putlitz, B., & *Robyr, M.

* Institute of Mineralogy and Geochemistry, University of Lausanne, Switzerland

** Physics of Geological Processes, University of Oslo, Norway

*** Institute of Geoscience, University of Potsdam, Germany

The Eastern Cordillera of Peru represents a major,yet relatively unstudied part of the proto-Andeancontinental margin. Paleozoic to early Mesozoicbatholiths that span its length exhibit profound andsystematic variations in the chemistry and timing ofemplacement from north to south (Mégard, 1978;Soler, 1991; Vidal et al., 1995; Jacay et al., 1999).As products of long-lived magmatic episodes, theseplutonic belts mark loci of active lithosphericboundaries between the western Amazonian Cratonand variable Neoproterozoic to Paleozoic crustaldomains during the final assembly and ultimatebreak-up of Pangea. Recognizing variations in theirgeochemical signature through time and spaceplaces constraints on the type of tectonism along thepaleo-margin, the composition and provenance ofcrustal members involved, as well as the nature ofthe underlying lithospheric mantle.

Here, a new data set from plutonic rocks of theEastern Cordillera is integrated with the existinggeochemical, chronometric and isotopiccharacterizations of the Peruvian landmass and aprovisional geodynamic model is proposed for theLate Devonian - Early Jurassic evolution of thissegment of the western Gondwana.

A striking relationship exists between the threeprincipal plutonic belts of eastern PeruvianCordillera:

(1) Mississippian to Pennsylvanian I-typemetaluminous to peraluminous, hornblende-dominated granitoids are restricted to the segmentnorth of 11oS (dominantly north of 9oS), and displaycalc-alkaline evolutionary trends with elevatedLILE/HFSE ratios characteristic of continentalsubduction zones;

(2) Mid-Permian to Early Triassic peraluminous, S toI-type, mica-rich granitoids of the (south) centralPeru, are comagmatic with the compositionally

bimodal, calc-alkaline to tholeiitic lavas of the MituGroup and are characterized by restricted bimodalcompositional range (66-72 wt. % SiO2), Feenrichment, lack of Nb anomalies, Ba depletionsrelative to Th and Rb and higher Ga/Al ratios, all ofwhich are associated with the transitional post-orogenic to within-plate granitoid suites;

(3) Late Triassic-Early Jurassic peralkaline, A-typeplutons of the southern Cordillera de Carabayaintrude alkaline Mitu Gr. basalts. They are nephelinenormative, and characterized by highly elevated HFSelements (ZR, Ti, and P).

Combined 87Sr/86Sr, 143Nd/144Nd isotopic ratios aswell as various Pb isotope systematics from thethree intrusive provinces however lack systematicvariations, and suggest uniformly large degrees ofassimilation of the Proterozoic Amazonian basementthroughout, thus constraining their paleo-geographicposition proximal to or within the Gondwana craton.

Interestingly, high precision U/Pb (zircon) and39Ar/40Ar (mica, hbd.) geochronometry reveal ageneral younging-southward trend. A ~ 20 Ma longmagmatism associated with the formation of aMississippian continental arc in the north-centralCordillera Oriental culminated between 336-325 Ma,and was followed by c.a. 40 Ma hiatus, brieflypunctuated during a 314-312 Ma episode of orogenicAu-Ag mineralization and a 307-305 Ma, S-typemagmatic pulse, both interpreted to reflect anepisode of tectonic uplift of the convergent margin(Haeberlin et al., 2002).

Resumption of Permo-Triassic magmatism (279-230 Ma) initially saw deposition of the bimodal calc-alkaline to tholeiitic volcanics of the Mitu Groupcontemporaneously with the emplacement of thepost-collisional S-type plutons in the south-centralEastern Cordillera (Soler, 1991). The magmaticactivity throughout Triassic was marked by eruption

The Paleozoic-Mesozoic geodynamic transition along the WesternGondwanan margin – Geochemical and chronometric constraints

from the Eastern Peruvian Cordillera

Miskovic, A., Schaltegger, U. & Chew, D.

Département de Minéralogie ; Université de Genève

13 Rue des Maraîchers, 1205 Genève, Switzerland

miskovic@terre.unige.ch

of progressively more mafic and alkalic Mitu lavasand initiation of the A-type plutonism (sensu stricto)that peaked between 216-205 Ma in thesouthernmost Carabaya Batholith (Kontak et al.,1990).

Complementarity of the arc and rift-relatedplutonic belts in the eastern Peruvian Andes pointsto a major tectono-magmatic change that took placealong this segment of the proto-Andean margin ofGondwana during the late Paleozoic.

Any self-consistent tectonic model for the regionmust take into account the following:

(1) an apparent absence of the cratonic crust undermost of the Western Peruvian Cordillera north of 13o

S as inferred from isotopic (Mukasa and Tilton,1984) and gravimetric surveys (Polliand et al, 2005);

(2) Existence of a constructive continental margin asinferred from the subduction-related plutonismrestricted to the northern Eastern Cordillera of Peruduring mid-to-late Mississippian. The activityresumed 25 Ma later along the Chilean FrontalCordillera (Mpodozis and Kay, 1992);

(3) Purely Gondwanan Pb isotopic signature of boththe Carboniferous and Permo-Triassic plutonic rocks(Macfarlane, 1999);

(4) A north-to-south transition from subduction-related I-type through the S-type, post-orogenicleucogranitoids into the rift-associated A-typeplutons, and

(5) A diachronous onset of the Permo-Triassic rift-related magmatism in the central and southern Peruwith a younging-southward trend (Sempere et al.,2002).

The aforementioned geochemical and tectonicevidence can be integrated in a geodynamic modelin which an originally orthogonal eastwardsubduction of the paleo-Pacific crust below thewestern Gondwana during the Late Devonian toEarly Carboniferous became strongly obliquetowards south-east thus imposing a sinistral strike-slip stress regime on the Gondwanan margin andinduced a counter-clockwise rotation of the northernedge of the Arequipa terrane (Figure 4).

The proposed change in strike of the subductioncould have resulted in transport of a buoyantsegment of oceanic crust (island arc root / plateau),which plugged the subduction zone and resulted inan ocean-ward trench migration coupled with aninitial margin uplift and subsequent fore-arcextension. This scenario explains the “craton- free”basement underlying the Western Cordillera ofnorthern Peru as well as development of ubiquitousensialic basins filled with the Mitu Gr. Molasses andbimodal volcanics, following the termination of arc-related magmatism in Pennsylvanian.

Continued oblique subduction of oceanic crust inPermian generated incipient S-type melts within thethickened crust of the central Peru, while progressivestrike-slip duplexing resulted in formation oftranstenisonal basins filled with the Permian rift-related magmas further south during Triassic. Weexclude the possibility of extending the Arequipaterrane north of its present isotopic borders duringthis time and consider it either non-existent, orreserve its removal from the Peruvian segment ofthe Gondwanan margin before Carboniferous.

REFERENCES

Haeberlin, Y., 2002. Ph.D. thesis, Department of Mineralogy,University of Geneva, Terre et Environement, v. 36, 182 p.

Jacay, J., Sempere, T., Carlier, G., and Carlotto, V., 1999. 4thISAG Conference Extended Abstracts, Göttingen,Germany, p. 358- 362.

Kontak, D.J., Clark, A.J., Farrar, E., and Strong, D.F., 1985. In:Pitcher, W.S., Atherton, M.P., Cobbing, J., and Beckinsale,R.D., (Eds.), Magmatism at a plate edge: the PeruvianAndes. London, Blackie & Son, p. 36-44.

Macfarlane, A.W., Tosdal, R.M., Vidal, C.E., and Paredes, J.,1999. In: Skinner, B.J., ed., Geology and ore deposits ofthe Central Andes: Economic Geology Special PublicationSeries, v. 7, p. 267-279.

Mégard, F., 1978. Travaux et Documents de l’ORSTOM, Paris,v. 86, 310 p.

Mpodozis, C., and Kay, S. M., 1992. Geological Society ofAmerica Bulletin, v. 104, p. 999-1014.

Mukasa, S. B., and Tilton, G. R., 1985. In: Pitcher, W.S.,Atherton, M.P., Cobbing, E.J., and Beckinsale, R.D., (Eds.),Magmatism at a plate edge; the Peruvian Andes: London,Blackie & Son, p. 203-207

Polliand, M., Schaltegger, U., Frank, M., and Fontbote, L. 2005.International Journal of Earth Sciences (Geol. Rund.), v.94, p. 231-242.

Sempere, T., Carlier, G., Soler, P., Fornari, M., Calotto, V.,Jacay, J., Arispe, O., Neraudeau, D., Rosas, S., andJimenez, N., 2002. Tectonophysics, v. 345, p. 153-181.

Soler, P., 1991. Thèse de doctorat d’Etat, Université Pierre-et -Marie-Curie (Paris VI), 950 p.

Vidal, C.E., Paredes, J., Macfarlane, A.W., and Tosdal, R.M.,1995. Sociedad Geológica del Perú, Lima, volumen jubilarA., p. 351- 377.

Until recently, models for the formation of adakiticmagmas have focused on partial melting of oceaniclithosphere in regions where young, hot slabs aresubducted. Nevertheless, adakitic-like rocks also oc-cur in continental arcs related to subduction of colderoceanic lithosphere, where they have been ex-plained in terms of remelting of basaltic material un-derplated at the base of thickened orogenic crust oras the result of modification of the sub-arc mantlerelated to tectonic erosion of the forearc crust. Ne-vado de Longaví volcano (NLV; 36°12’S – 71°10’ W)located just to the south of the region that has beenstrongly affected by Tertiary crustal shortening andthickening and associated eastward arc migration, isthe only occurrence of Quaternary magmas with anunequivocal adakitic signature in the accessible partof the Andean Southern Volcanic Zone (SVZ: 33-41°S). In this contribution we propose a fractional crys-tallization model to explain the occurrence of ada-kites at NLV related to highly hydrous mafic melts.

NLV is a mainly andesitic late Quaternary edificewhose magmatic suite progressively evolves fromearly basalts and basaltic andesites towards dacitesby a trend characterized by a low increase rate ofK2O and other incompatible trace elements com-pared to the rest of SVZ volcanoes, and decreasingconcentrations of Y and HREE, contrary to the ob-served trend for the other volcanoes in the arc.These features are most extreme in the Holocenedacitic products (63-65 wt% SiO2) which are char-acterized by high Sr and low incompatible elementcontents (especially K, Y and HREE), a mineral as-semblage with amphibole as the main mafic phase,an unusually high fO2 (NNO+1.7) and elevated watercontents of 5-6 wt % H2O (as inferred from experi-mental results on closely comparable Pinatubo 1991dacite; Scaillet and Evans 1999). On the other hand,mafic magmas preserved as enclaves on thesedacites (53-56 wt% SiO2; MgO <6 wt%) are in many

respects unlike other SVZ mafic magmas. They haverelatively low contents of many incompatible ele-ments, notably Th, U, Zr, Nb, Hf and REE, in combi-nation with high B (19-25 ppm), Be, Cs, and Li con-tents and high Ba/Th, Ba/Zr, Pb/Th ratios. Thesefeatures are consistent with these mafic magmasbeing derived from high degrees of melting of themantle source as a consequence of being fluxed byanomalously high amounts of slab-derived fluids.This highly wet character of enclaves is put in evi-dence by an amphibole rich (30 vol %) mineralogyand a notably oxidized (NNO+2) character.

The low incompatible element contents of NLVdacites are inconsistent with assimilation of uppercrustal rocks as well as with fractionation of typicalanhydrous pyroxene+plagioclase dominated assem-blages that are proposed for the rest of the arc. Nev-ertheless, the similarities between adakitic dacitesand mafic quenched enclaves (low incompatibleelement contents, oxidized, water-rich), suggest apossible cogenetic relation between them. In order toevaluate the feasibility of fractional crystallization toproduce the NLV adakitic melts from the wet maficmelts, we have developed a major elements mass-balance model combined with Rayleigh fractionalcrystallization that considers 50% fractionation of anassemblage composed of: 0.5 Hbl + 0.37 Plag +0.07 Opx + 0.03 Aug + 0.03 Mgt + 0.007 Ap + 0.02Gt. This model successfully reproduce the major andtrace element characteristics of NLV dacites and issupported by abundant amphibole-bearing cumu-lates whose mineralogy agrees with the assemblageconsidered (except for garnet). This combination offractionating phases also explains the observedY+HREE depletions in NLV andesites and dacitesrelative to mafic magmas, as well as minimal en-richments in elements that are incompatible relativeto anhydrous silicates.

Nevado de Longaví volcano (Chilean Andes, 36.2 ºS):Adakitic magmas by fractional crystallization

from hydrous mafic melts

1Rodríguez, C., 1Sellés, D., 1Dungan, M., 2Langmuir, C. & 3Leeman, W.

1 Université de Genève, Section des Sciences de la Terre, Département de Minéralogie. 13 Rue des Maraîchers. 1205Geneva, SWITZERLAND. Carolina.Rodriguez@terre.unige.ch

2 Department of Earth Science MS-126, Rice University, 6100 Main St., Houston, TX, 77005, U. S. A.3 Department of Earth and Planetary Sciences, Harvard University. 20 Oxford Street, Cambridge, MA 02138. U. S. A.

The proposed model requires high water contentsin the mafic melts in order to stabilize early amphi-bole instead of anhydrous mafic phases and to re-duce the stability field of plagioclase (e.g. Grove etal. 2003). Highly wet melts also permit the crystalli-zation of garnet at crustal pressures similar to thoseexpected in the lower crust under NLV (35-40 km)(Müntener et al. 2001; Ulmer et al. 2003). The pro-jection of the Mocha fracture zone (Eocene-ageNazca plate) under NLV is our favored candidate toexplain the occurrence of the unusually wet maficmelts at NLV (Sellés et al. 2004). This feature of theslab probably contains serpentinized bodies that de-hydrate during subduction and release largeamounts of fluid to the mantle, resulting in high de-grees of melting and the generation of water-richmafic melts. This also explains the extremely localoccurrence of the NLV adakites in the geodynamiccontext of SVZ.

We propose that adakites in cold subductionzones can alternatively be formed by fractionalcrystallization of amphibole-rich assemblages fromhydrous mafic melts. NLV constitutes a case of studyin which lack of evidence for upper crustal assimila-tion, hornblende-bearing cumulates throughout thevolcano, increasing modal abundances of horn-blende toward Holocene magmas, and the unusuallyincompatible element-depleted character of themelts appear to be ultimately related to an excep-tionally high fluid-flux from the subducted MochaFracture Zone which projects beneath NLV, andwhich is inferred to have generated water-rich but in-compatible element-poor basalts through flux-melting.

REFERENCES

Grove, T., Elkins-Tanton, L., Parman, S., Chatterjee, N.,Müntener, O. & Gaetani, G. (2003): Fractional crystalliza-tion and mantle-melting controls on calc-alkaline differen-tiation trends. Contributions to Mineralogy and Petrology,Vol. 145, No. 5, p.515-533.

Müntener, O., Kelemen, P. & Grove, T. (2001): The role of H2Oduring crystallization of primitive arc magmas under up-permost mantle conditions and genesis of igneous pyrox-enites: an experimental study. Contributions to Mineralogyand Petrology, Vol.141, p.643-658.

Scaillet, B. & Evans, B. (1999): The June 15, 1991 eruption ofMount Pinatubo. I. Phase equilibria and pre-eruption of P-T-fO2-H2O conditions of the dacite magma. Journal of Pe-trology, Vol.40, p. 381-411.

Sellés, D., Rodríguez, C., Dungan, M., Naranjo, J. &Gardeweg, M. (2004) : Geochemistry of Nevado de Lon-gaví volcano (36.2°S): a compositionally atypical arc vol-cano in the Southern Volcanic Zone of the Andes. RevistaGeológica de Chile, Vol.31, No.2, p.293-315.

Ulmer, P., Müntener, O. & Alonso-Pérez, R. (2003): Potencialrole of garnet fractionation in H2O-undersaturated andesiteliquids at high pressure: an experimental study and a com-parison with the Kohistan arc. Geophysical Research Ab-stract, Vol.5, 08308.

THE PROJECT

In order to understand the interhemispheric telecon-nections and global climatic change, the lack of pa-leoclimate data from the Southern Hemisphere hasto be overcome. South-Central Chile is located at thewindward side of the Andes and at the northern limitof the Southern Westerlies influence, with dry sum-mers and humid winters. This area is, therefore, ex-pected to be sensitive to climatic changes (Bertrandet al. 2005).

Our project aims at producing a (sub)decadal,multi-proxy paleoclimate reconstruction for the lastca. 1000 years based on sediments from high-elevation pro-glacial lakes in the Central Andes ofChile (33°S). The study of this period of time canprovide information that may answer the question towhat extent the 20th century is unusual in the light ofthe recent past.

Fig. 1: Overview of the study site cored lakes

THE RESEARCH PLAN

During the field campaign in March 2004 short sedi-ment cores (38-48 cm) from several potential studylakes at altitudes between 2500 and 3400 m a.s.l.were retrieved. Preliminary sediment analysis showin some cores structures at the (sub-) cm scale. Onthe basis of these results, we select the 3-4 mostpromising lakes for further investigations including asecond field campaign in October 2005.

In a next step these lakes, their catchments andadditional sediment cores will be analysed using se-lected geochemical and sedimentological standardmethods, dating methods and digital image proc-essing.

PRELIMINARY RESULTS

Lake catchment analysis. Data from SRTM (Shut-tle Radar Topography Mission) were obtained in or-der to compute digital elevation models for the lakecatchments. Orthofotographies are calculated basedon topographical maps (1:50’000) and aerial views.This information will serve as a basis for detailedgeomorphological maps and description of surfaceprocesses within the lake catchments.

Geochemical and sedimentological analysis. Ourdata set contains the following parameters for allsediment cores: magnetic susceptibility, density (_-ray absorption), C, N measurement and biogenicsilica measurements, grain size, thin section andsmear-slide analysis.

Dating. Radiogenic Pb and Cs were measured andshow that dating methods work in this environment.Additional 14C Dating was carried out.

Lake sediments and Medieval Climate Anomaly (MCA) and Little IceAge Type Events (LIATES) in the South-Central Andes of Chile

*Salvetti, Ch, *von Gunten, L. & **Grosjean, M.

* Institute of Geography, University of Bern, Hallerstr. 12, CH-3012 Bern

** NCCR Climate, Erlachstr. 9a, CH-3012 Bern

salvetti@giub.unibe.ch; lucien.vongunten@giub.unibe.ch; grosjean@giub.unibe.ch

Table 1. Preliminary results from geochemical andsedimentological analysis

Laguna El

Ocho

Laguna del

Encañado

Laguna Ne-

gra

Laguna del

Inca

C/N ratio 6.10 11.01 6.64 6.24

biSi [%] 12.71 1.53 2.38 2.48

Ø grain size

[_m]

6.11 11.01 13.72 31.25

RESEARCH QUESTIONS

1. Which climate parameters and processes arerecorded today in the lake sediments?

2. Can this information be used for downcore ex-trapolation for the last ca. 1000 years and cli-mate reconstruction?

3. How do the sediment records compare withother regional climate reconstructions in adja-cent areas (Luckmann & Villalba 2001; Villalbaet al. 2003) and inter-hemispheric, circum-Pacific or global phenomena for the recentpast?

For selected windows of time during the last 1000years (Medieval Climate Anomaly, Late MaunderMinimum and Dalton Minimum) our data set will becompared with data derived from GCM ensembleruns carried out within the framework of NCCR Cli-mate (e.g. Raible et al., 2005).

REFERENCES

Bertrand, S. et al. (2005): Temporal evolution of sediment sup-ply in Lago Puyehue (Southern Chile) during the last 600 yrand its climatic significance. Quaternary Research, 64, 163-175.

Luckmann, B., Villalba, R. (2001): Assessing the Synchroneityof Glacier Fluctuations in the Western Cordillera of theAmericas during the last Millennium. In: Markgraf, V. (Ed)Interhemispheric Climate Linkages, Academic Press, SanDiego.

Raible, C.C, Stocker, T.F., Yoshimori, M., Renold, M., Beyerle,C., Casty, C., Luterbacher, J. (2005): Northern hemispherictrends of pressure indices and atmospheric circulation pat-terns on observations, reconstructions and coupled GCMsimulations. Journal of Climatology, in press.

Villalba, R. et al. (2003): Large-scale temperature changesacross the southern Andes: 20th century variations in thecontext of the past 400 years. Climatic Change, 59, 177-232.

The Oligocene magmatism of Central Chile consti-tutes a tholeiitic suite generated during a period ofcrustal thinning and mantle upwelling. MORB-likeREE patterns together with low 87Sr/86Sri ratios andhigh e-Nd values are consistent with relatively drymantle-derived magmas that did not assimilate im-portant amounts of crustal material and that evolveddominantly by low-pressure fractional crystallization.In the area of Santiago (33-34°S), the volcanic se-quence is cross-cut by coeval to slightly youngersubvolcanic intrusive bodies, most of which are iso-topically and chemically similar to the effusive rocksand can be considered the plutonic roots of the Oli-gocene arc. However, a subset of slightly youngerintrusive bodies (20-15 Ma; Manquehue-type stocks)shows a distinctive adakitic chemical signature thatis not observed in previous or subsequent volcanics.The front of the volcanic activity shifted eastwardsimmediately after or simultaneously with intrusion ofthese adakitic stocks, coinciding with the beginningof a period of increased convergence rate and pro-gressive crustal thickening and uplift. Compared tothe Oligocene lavas, the Miocene arc magmas arewetter and they incorporated higher proportions ofcrustal material, and the residual mineral assem-blage suggests higher pressures of fractionation withthe involvement of some amphibole but no garnet. Atthe latitude of Santiago, no other Neogene magmaticunit is known to show a residual garnet signature,and it is only in Quaternary times that the volcanicarc, located above a ~60 km thick crust, indicatesagain garnet involvement coupled to crustal con-tamination. In the El Teniente Area (34°S), however,intrusive units related to porphyry copper mineraliza-tion show garnet signature during the latest Mioceneto Pliocene (6-4 Ma; Kay et al. 2004).

The adakitic Manquehue-type stocks in the areaof Santiago are low- to medium-K andesites todacites (56-72% SiO2). They are commonly porphy-

ritic in texture, with plagioclase and amphibole as themain phenocrysts. They have a distinctive chemistryrelative to other magmas of the area in having lowincompatible element contents (K, Rb, Zr, etc.), highSr and positive Eu anomalies, and high La/Yb andCe/Y ratios. Compared to the modern SVZ, Man-quehue-type stocks have very modest increases inincompatible elements relative to silica, similar to thetrend depicted by Nevado de Longaví volcano low-Rb magmas (Sellés et al. 2004). Both Manquehueand Longaví differ from Quaternary volcanics of thenorthern Southern Volcanic Zone (SVZ) in that highLa/Yb ratios are not accompanied by substantial en-richments of incompatible elements, which arguesagainst crustal sources for the garnet signature. In-terestingly, the less evolved compositions arebroadly similar to mafic magmas from the SVZ, sug-gesting a common mantle source but different evolu-tionary paths. The absence of significant crustalcontamination is further demonstrated by Sr and Ndisotopic ratios. e-Nd values of Manquehue-typestocks are within the range of Oligocene lavas (+6 to+5), although initial 87Sr/86Sr ratios are slightly higher(0.7038-0.7042). Sr and Nd isotopic compositionsare moreover uncorrelated with increasing La/Yb(Figure 1).

Although garnet is often inferred to make part ofresidual assemblages of continental arc magmas, itis rarely actually observed because it tends to frac-tionate from the host magma or is resorbed at lowpressure conditions. One of the Manquehue-typestocks preserves euhedral to subhedral garnetcrystals up to 2 mm in diameter. The host rock is al-most aphanitic in texture, the only phenocrysts being~1% garnet and ~5% plagioclase (An40-35). Despitethis particular mineralogy, this body is chemicallysimilar to the amphibole-bearing stocks. Garnetcrystals are unzoned, almandine-rich (Al73-Py11-Sp6-Gr10), and contain abundant randomly oriented inclu-

Garnet phenocrysts in Early Miocene intrusivesin Central Chile.

Evidence for a crystal fractionation origin of adakite-like magmas

Sellés, D., Dungan, M., *Gana, P. & Rodríguez, C.

Department of Mineralogy, Earth Sciences Section, University of Geneva, Switzerland

* National Geology and Mining Survey, SERNAGEOMIN, Chile

sions, mainly needle-like apatite, iron-rich orthopy-roxene, Fe-Ti oxides and minor amphibole. Garnetsof similar composition have been interpreted else-where as a primary phase crystallizing from hydrousmantle-derived calc-alkaline melts. Recent experi-ments have proven that garnet can crystallize atmoderately high pressures (≥0.8 GPa) from hydrousand oxidized andesitic magmas (e.g. Müntener et al.2004).

Figure 1. Sr and Nd isotopic composition of Manquehue-typestocks compared SVZ magmas (top) and Cenozoic magmaticunits at 34°S (bottom, modified from Kay et al. 2004).

Previous interpretations on the petrogenesis of thesestocks have assumed that garnet was a restiticphase, either from a subducted slab source or fromcrustal slivers dragged down by subduction erosion(Kay et al. 2004). Newly acquired isotopic data andpetrologic observations suggest that garnet was aprimary phase, crystallizing from intermediate hy-drous mantle melts evolved at lower to middlecrustal pressures. The hydrous character of Man-quehue-type magmas is suggested by the abun-dance of amphibole and paucity of pyroxene phe-nocrysts. Also, high Al2O3, Sr and positive Euanomalies indicate that plagioclase did not thor-oughly fractionate, which is also expected to happenin hydrous melts. Moreover, enrichments of fluid-mobile trace elements over less mobile ones sug-gest important participation of slab-derived fluids. Inthe Quaternary Nevado de Longaví volcano (Selléset al. 2004), similar characteristics are interpreted tobe consequence of the subduction of an oceanicfracture zone that can host serpentinized bodies,potentially efficient water carriers to the subarc man-tle. Fluids released upon serpentine breakdownshould have the isotopic composition of sea water,which could explain the relatively elevated Sr iso-topic ratios. Elevated fluid flux to the mantle gener-ates high-degree hydrous melts that evolve mainlyby fractionation of amphibole, keeping incompatibleelement contents low.

The recognition of two independent cases of wa-ter-rich, mantle-derived adakitic magmas in the An-dean context opens new perspectives in the inter-pretation of high-La/Yb arc magmas.

REFERENCES

Kay, S.M., Godoy, E., Kurtz, A. (2004): Episodic arc migration,crustal thickening, subduction erosion, and magmatism inthe south-central Andes. GSA Bulletin, v.116, no. 11/12.DOI: 10.1130/B25431.1.

Müntener, O., Perez-Alonso, R. & Ulmer, P. (2004): Phase re-lations of garnet, amphibole and plagioclase in H2O under-saturated andesite liquids at high pressure and implicationsfor the genesis of lower arc crust. Geophysical ResearchAbstracts, v. 6: 05183.

Sellés, D., Rodríguez, A.C., Dungan, M.A., Naranjo, J.A.,Gardeweg, M. (2004): Geochemistry of Nevado de Longavívolcano (36.2°S): a compositionally atypical arc volcano inthe Southern Volcanic Zone of the Andes. RevistaGeológica de Chile, v. 31, No. 2, p. 293-315.

The ‘Lluta collapse’ is a prominent geomorphic fea-ture in the landscape of the western escarpment ofthe Andes of northern Chile that resulted fromprobably one of the oldest recognizable landslide (>2.5 Ma) in a continental setting, and from subse-quent modification of the landslide scar by backwarderosion. The combination of geomorphic and geo-logical information from the ‘Lluta collapse’ andsedimentological observations from the landslidedeposits imply that a total of 25 km3 of mass wasdisplaced by landsliding. Subsequent modification ofthe landslide scar occurred by backward erosion, re-sulting in the establishment of a dendritic drainagenetwork and the removal of an additional ca. 24 km3

of material. It appears that this mass was producedby mass wasting in the headwaters, and exported byhigh-concentrated debris flows in channels.

The source area of the ‘Lluta collapse’ is bor-dered by an amphitheater-shaped scarp. This scarp-line – corresponding to the first-order geometriclength-scale in the landscape – is made up of coa-lescing units of lower-ordered length-scales that alsodisplay concave geometries, and that have resultedfrom hillslope mass wasting. It appears, therefore,that the geomorphic evolution of the ‘Lluta collapse’and the establishment of a dendritic geometry in theheadwaters have been governed to large extents bymass wasting processes of different length-scales. Incontrast, high-concentrated flows have controlled theexport of mass, which, in turn, has allowed the base-level to lower to sufficiently low magnitudes to initiatefurther landslides. Hence, the data suggest thatwhereas the geometrical development of the ‘Llutacollapse’ has been controlled by gravitational masswasting, the rates of the development of this geo-morphic unit have been limited by the export rates ofmass and hence by the transport capacity of theflows.

Controls of gravitational mass wastingon the geomorphic evolution of headwaters:

The ‘Lluta collapse’, northern Chile

*Strasser, M., & **Schlunegger, F.

* Geological Institute, ETH Zurich, Switzerland, strasser@erdw.ethz.ch

** Institute of Geological Sciences, University of Bern, Switzerland, fritz.schlunegger@geo.unibe.ch

The interandean Cauca depression corresponds toa deep sedimentary basin (Fig. 1) filled byapproximately 6000 meters of sediments ranging inage from Early Tertiary to recent. This researchattempts to establish the mechanisms, which mayhave led to the opening of this basin by studyingthe fluvio-lacustrine Zarzal Formation (Neuwerth etal., 2005). This formation is associated with thelatest distensive tectonic phase and related with thevolcaniclastic mass flows derived from the CentralCordillera, which form the volcaniclastic fans ofQuindío, Pereira and Cartago (Fig. 1C; Gorin et al.this symposium; Guarin et al., 2005).

The Cauca Depression has been exposed to aregional oblique compressive tectonic regime,which has generated different pull-apart basinsalong the Cauca-Romeral fault system. The latterdefines the boundary between continental andoceanic basements. From west to east, the studiedarea comprises the foothills of the WesternCordillera, the Cauca River valley, the foldedTertiary sediments of the Serranía de Sta Barbaraand the La Vieja River valley (Fig. 1B and C). TheSerranía de Sta Barbara forms a natural barrierbetween the western (Valle del Cauca) and eastern(Quindío) parts of the basin, which coalesce in theCartago Fan to the north.

Figure 1. A) Location of study area. B) Digital ElevationModel (DEM) of the studied area and its surroundings.Location of profiles AA' and BB' shown in Figure 2. C)Simplified geological map of studied area.

In the eastern part ofthe basin and in theCartago Fan (Fig. 1C),the fluvio-lacustrinesediments of the ZarzalFormation have a clearvolcanic origin and arei n t e r b e d d e d w i t hvolcanic mass flows. Inthe western part of thebasin, the Zarzal

Formation is sourced from the Tertiary sedimentsof the Serranía de Sta Barbara and from theWestern Cordillera (Fig. 1C). Preliminarypalynological investigations have yielded an age ofless than 1 m.y. Sediments of the Zarzal Fm exhibitnumerous soft-sediment deformations interpretedas seismites (Neuwerth et al., 2005), which provethe important seismic activity of this area.

Depositional model for the Quaternary Zarzal Formation (Colombia) and itsstratigraphic relationship with the volcaniclastic mass flows

derived from the Central Cordillera

*Suter, F., *Neuwerth, R., **Guzman, C. & *Gorin, G.

* Department of Geology-Paleontology, University of Geneva, SwitzerlandFiore.Suter@terre.unige.ch, Ralph.Neuwerth@terre.unige.ch, Georges.Gorin@terre.unige.ch

** Department of Geological Sciences, University of Caldas, Manizales, Colombiacguzmanl@yahoo.com

Figure 2. Relation betweenfield observations and theproposed depositional model(see figure 1B for location ofprofiles). Gravitational flowsare represented in grey andblack, fluvio-lacustrine sedi-ments in white. This modeldoes not take into accountthe tectonic activity. More-over, the frequency of gravi-tational f lows underesti-mated.

Field data and palynological results allow theproposal of a depositional model (Fig. 2). Thefolding of the Serranía de Sta Barbara, subdividedthe basin into two parts. While the western partwas infilled by the fluvial sediments of a paleo-Cauca River, volcanic mass flows derived from theCentral Cordillera accumulated in the eastern part.Each of these mass flow unit behaved like atopographical barrier temporarily damming thebasin and creating a lake. Following the erosion ofthe damming mass flow, the lake dried out andturned into a floodplain. This cycle was repeatedeach time a new significant mass flow dammed thevalley. The eastern basin continued to infill until thevolcanic mass flows could flow over the lower-relief, northern part of the Serranía de Sta Barbaraand spread into the Cauca Valley to form theCartago Fan (Fig. 1C). Subsidence rates in theCauca Valley seem to have been higher than in theeastern part.

These preliminary results establish thesynchroneity of the deposition of the Zarzal Fmwith that of the volcaniclastic fans. The Zarzal Fmbears the imprint of an intense tectonic activitypartly related with the deposition of the fans(Guarin et al., 2005 and this symposium). Thesejoint studies will help to refine the dynamicinterpretation of this basin.

This research is supported by the Swiss NationalScience Foundation (grant no. 21-67080.01). Partof the field work was supported by the SwissAcademy of Sciences.

REFERENCE

Gorin, G., Guarin, F., Neuwerth, R., Suter, F., Espinosa, A. &Guzman, C. (2005): Contribution of Quaternary sedimentsto the understanding of the tectonic history in CentralColombia: the volcaniclastic fans in Quindío-Risaraldaand the Zarzal Formation in the Cauca Valley. 3rd SwissGeosc. Meeting Zurich 2005.

Guarin, F., Gorin, G.& Espinosa, A. (2005): A Pleistocenestacked succession of volcanic mass flows in CentralColombia: the Quindío-Risaralda Fan. Acta Vulcanologica(in press).

Guarin, F., Gorin, G.& Espinosa, A. (2005): The interandeanQuindío-Risaralda basin in Central Colombia and itsPleistocene infill by stacked volcaniclastic mass flowsderived from the Central Cordillera. 3rd Swiss Geosc.Meeting, Zurich 2005.

Neuwerth, R., Suter, F., Guzman, C. & Gorin, G. (2005): Soft-sediment deformations in a tectonically active area : thePleistocene Zarzal Formation in the Cauca Valley(Western Colombia). Sedimentary Geology (in press).

The Ecuadorian Andean mountains comprise the to-pographically distinct Eastern Cordillera and West-ern Cordillera. These topographic ridges are sepa-rated by the Interandean Depression (IAD), which isan elongate, tectonic structure that has been activesince the Late Miocene, resulting in an extensive to-pographic depression, and the formation of isolatedintermontane basins. The IAD straddles the dis-membered Late Cretaceous suture between al-lochthonous oceanic rocks exposed in the WesternCordillera and continental crust in the Eastern Cor-dillera. Within Ecuador, the origin and composition ofthe basement of the IAD is mainly unknown becauseof the extremely restricted extent of basement inlierswithin the post-Oligocene volcanic cover.

This contribution aims to determine the tectonicprovenance and the age of the basement of the IADand to reconstruct the post-Late Miocene tectonichistory of the IAD structure, based on field observa-tions and radiometric ages of stratified volcanoclasticrocks.

BACKGROUND INFORMATION

The crystalline Cretaceous basement of the WesternCordillera partly comprises Late Cretaceous oceanicplateau basalts (Pallatanga Terrane), which are intectonic contact with a Late Cretaceous, tholeiitic is-land arc sequence (Rio Cala Arc). These units ac-creted against the continental margin during the LateCretaceous (Hughes and Pilatasig, 2002; Jaillard etal. 2004; Spikings et al., 2005). The suture is partlyrepresented by the Calacali-Pallatanga Fault, whichdefines the western border of the IAD. Metamor-phosed continental rocks comprise the Eastern Cor-dillera, which is juxtaposed against the IAD via thePeltetec Fault. Undated slivers of mafic rocks crop-out in anastomosed zones along the Peltetec Fault.

We performed geochemical and isotopic analysesof various ultramafic and mafic rocks that comprisethe basement of the IAD within the depression andalong its bounding faults (Pallatanga, Rio Cala andPeltetec Units). 40Ar/39Ar analyses of stratified vol-canic tuffs exposed in intermontane basins in theIAD are in progress.

RESULTS - Mafic basement

Basement rocks exposed in the northern IAD aregeochemically similar to the Rio Cala arc lavas, andare characterised by high LILE/HFSE ratios, nega-tive primitive mantle normalised Nb-Ta anomaliesand LREE enrichment relative to HREE’s ((La/Yb)n~6.) The IAD basement in the south is geochemicallysimilar to the Pallatanga Unit, and is characterisedby enriched LILE’s and HFSE’s relative to primitivemantle and flat REE patterns ((La/Yb)n ~0.9). Maficrocks of the Peltetec Unit yield both plateau-like((La/Yb)n 1.3 – 1.8) and subduction related((La/Yb)n >2.4) characteristics.

These preliminary data reveal broad geochemicalsimilarities between the basement of the IAD andallochthonous, Late Cretaceous mafic rocks of theWestern Cordillera, implying that the plateau andsubduction derived basement units of the WesternCordillera extend beneath the IAD. Furthermore, thesame Late Cretaceous basement units may alsocrop-out along the western flank of the Eastern Cor-dillera, where they either partly or completely com-prise the Peltetec Unit. Consequently, pending im-minent geochronological (40Ar/39Ar) analyses of theundated Peltetec Unit, there is no evidence for theexistence of a previously proposed (Litherland et al.,1994) Early Cretaceous terrane beneath the IAD.The Peltetec fault may also represent part of theLate Cretaceous, ocean-continent suture in Ecuador.

Nature and origin of the Interandean Depressionin Ecuador

*Villagomez, D., *Spikings, R., **Winkler W. & *Gorin G.

* Section des Sciences de la Terre, Université de Genève, Rue des Maraîchers 13, CH-1211 Genève 4, SwitzerlandDiego.Villagomez@terre.unige.ch

** Geologisches Institut, ETH-Zentrum, Zurich CH-8092, Switzerland,winkler@erdw.ethz.ch

RESULTS - Post-Late Miocene history of the IAD

Radiometric analyses of basal volcanic depositsshow that the intermontane basins within the IADyoung from ~ 6 Ma in the north to ~ 3 Ma in thesouth (Winkler et al. 2005). Prior to 5.5 Ma, therewas no IAD, implying that one cordillera existed inEcuador (Spikings and Crowhurst 2004, Spikings etal. 2005). The basement of the IAD probably sharesthe same origin as that exposed in the Western Cor-dillera, although it has since been segmented byfault activity. The complicated and dense fault arraywithin the IAD and its basement has resulted in asegmented morphology and has controlled the sitesof volcanic emplacement.

CONCLUSIONS

- The basement of the IAD is probably an extensionof allochthonous, Late Cretaceous accreted oceanicrocks exposed in the WC.

- The undated Peltetec Unit shows broad geochemi-cal similarities with Late Cretaceous accreted maficrocks that are currently exposed in the WC, andhence may also comprise the same Late Cretaceousbasement units.

- Previously identified Late Cretaceous suture zones(e.g. the Calacali-Pallatanga fault zone) may be Mio-cene or younger structures and the Late Cretaceousocean-continent suture is probably located further tothe east, partly comprising the Peltetec Fault.

- The IAD began to open at ca. 6-5 Ma in northernEcuador and propagated southward (Winkler et al.

2005). Syn-sedimentary deformation prevailed dur-ing most of the life-span of the IAD (Spikings andCrowhurst 2004).

ACKNOWLEDGEMENTS

DV was supported by a Swiss Federal GovernmentGrant (2004-2005).

REFERENCES

Jaillard E., Ordoñez M., Suárez J., Toro J., Iza D. Lugo W.(2004): Stratigraphy of the late Cretaceous–Paleogene de-posits of the cordillera occidental of central ecuador: geo-dynamic implications. Journal of South American EarthSciences 17: 49-58.

Spikings, R., Crowhurst, P. (2004): (U-Th)/He thermochrono-metric constraints on the Late Miocene – Pliocene tectonicdevelopment of the northern Cordillera Real and the Inter-andean Depression, Ecuador Journal of South AmericanEarth Sciences 17: 239 – 251.

Spikings, R.A., Winkler, W., Hughes, R.A., Handler, R. (2005):Thermochronology of the Cordillera Occidental and theAmotape Complex, Ecuador: unravelling the accretionaryand post-accretionary history of the Northern Andes. Tec-tonophysics 399: 195 – 220.

Litherland M., Aspden J., Jemielita R. (1994): The metamorphicbelts of Ecuador. British Geological Survey, Quito, Over-seas Memoir 11: 147 pp.

Winkler W., Villagomez D., Spikings R., Abegglen P., Tobler S.,Eguez A. (2005): The Chota basin and its significance forthe inception and tectonic setting of the Inter-Andean De-pression in Ecuador. Journal of South American Earth Sci-ences 19: 5-19.

In the Rio Tinguiririca valley in the Main Cordillera ofthe Andes of central Chile, 35° south, parts of astratigraphic section ranging from the late Jurassic tothe Quaternary are exposed. Fission track analysiswas carried out on samples from all the stratigraphicunits exposed in the area in order to gain informationon the low-grade metamorphic history of the CentralAndes and to test older models for the metamorphicand tectonic evolution of the area.

The sequence exposed in the Rio Tinguiriricavalley is characterised by several distinctive featuresnot found in other localities in the Central Andes(Charrier et al., 1996); the deposits of the middleCretaceous (Aptian to Albian) Colimapu Formationand of the middle to late Micoene Farellones Forma-tion are completely missing in the study area. In-stead, a volcanic tuff layer, the White Tuff, and a unitconsisting of fan deposits and alluvial plane depos-its, the Brownish-Red Clastic Unit, unconformablyoverlie the Late Jurassic deposits of the Baños delFlaco Formation.

The fission track data give some indications ofthe style and timing of metamorphic events in thestudy area, enable more accurate constraint of theage of the Brownish-Red Clastic Unit and allowsome statements on the tectonic evolution of thestudy area from the Late Jurassic to present.

Burial metamorphism has been proposed byvarious authors as the main mechanism to producelarge suites of rocks altered at low grades in theCentral Andes. The results of this study indicate that,on the contrary, hydrothermal alteration connected tomagmatic and/or volcanic activity was the maincause of alteration of the rocks and that burialmetamorphism played at most a very minor role.Pulses of hydrothermal activity appear to have oc-curred from Cretaceous to almost recent times andled to alteration of the rocks at slightly varying

metamorphic conditions at different times in differentparts of the study area.

A new model is proposed for the tectonic evolu-tion of the study area. Fission track analysis of de-trital zircons from the Brownish-Red Clastic Unitshows that the unit must have been deposited duringthe latest Cretaceous (Maastrichtian) and that it iscertainly younger than the White Tuff.

Thermal modeling shows that considerable ex-humation of the lower part of the Rio Damas Forma-tion occurred during the Late Cretaceous to EarlyTertiary. This exhumation is thought to be connectedto tilting and erosion of the Mesozoic units in thearea prior to the formation of an extensional basin inthe Late Eocene. Data from the Eocene to MioceneCoya Machali Formation imply that sedimentationwithin the Tertiary basin continued somewhat longerthan hitherto supposed.

REFERENCE

Charrier, R., Wyss, A.R., Flynn, J.J., Swisher III, C.C., Norell,M.A., Zapatta, F., McKenna, M.C., and Novacek, M.J.(1996): New evidence for Late Mesozoic-Early Cenozoicevolution of the Chilean Andes in the Upper TinguiriricaValley (35[deg]S), Central Chile: Journal of South AmericanEarth Sciences 9: 393-422.

Constraints from fission track analysis on the evolution of the RioTinguiririca valley area in the Main Cordillera of the Andes,

Central Chile

*Waite, K., **Fügenschuh, B. & ***Schmidt, S.

* Institute of Mineralogy and Petrology, University of Basel, Bernoullistrasse 30, 4056 Basel, Switzerland,

k.waite@unibas.ch

** Institute of Geology and Paleontology, University of Innsbruck, Austria

*** Department of Mineralogy, University of Geneva, Switzerland

The Ecuadorian Andean Cordillera represents adouble-vergent orogenic belt, which generally com-prises accreted oceanic plateau and arc series in theCordillera Occidental (CO), and metamorphic conti-nental basement and sedimentary series in the Cor-dillera Real (CR) (e.g. Spikings et al. 2001, 2005).The Andean Amazon Basin (AAB) of Ecuador devel-oped to the east of the evolving Andean chain and tothe west of the Amazon craton (Guyana Shield) fromthe mid-Cretaceous to the Recent. At least since theMaastrichtian (Tena Fm.), the basin shows the typi-cal characteristics of a variable, very shallow marineto continental facies retro-arc foreland basin with re-spect to the growing orogen. Detrital provenanceanalyses (heavy minerals), and fission track (FT)thermochronology of detrital zircons (source rockexhumation ages) are used for monitoring the longlasting history of the supplying Andean chain andadjacent regions. The investigated samples are de-rived from the northern and southern Subandeanzone (including the Napo and Tena uplift areas, re-spectively) and proximal parts of the Oriente Basin.

In the mid- to Late Cretaceous sediments (Hollinand Napo fms.), zircon-tourmaline-rutile dominatedheavy mineral spectra (ZTR association) imply thatthe basin was supplied from non- and very low-grademetamorphic granitic source rocks (referred to asshallow continental crust provenance). Since theMaastrichtian (Tena Fm.) and during the Paleogene,an increasing amount of detrital medium-grademetamorphic grains (epidote-group, chloritoid andgarnet) is observed. This is culminating with the fur-ther occurrence of high-grade metamorphic grains(kyanite, sillimanite) during the Neogene. This trendshows the continuous exhumation of deeper crustallevels in the supplying CR. The additional appear-ance of mafic minerals (augite, hypersthene, diop-side, chromian spinel, olivine) infers that oceanicbasement rocks in the CO started to contributed to

the clastic supply since the Late Oligocene (post-dating a major accretionary event in the forearc).From other circumstantial arguments, it can be sug-gested that since the Late Miocene an orographicsituation similar to today has existed.

The exhumation age of the detrital zircons wasmeasured in the same sandstones as used for theheavy mineral analysis in the northern SubandeanZone (e.g. Ruiz et al. 2004). Various populations ofdetrital zircon FT ages were discriminated by statisti-cal methods in the sandstone samples. The calcu-lated lagtime (cooling/closure age minus depositionalage), in assuming negligible time of transport of thegrains into the basin, describes the time necessaryfor exhuming the source rocks from depth (ca260oC) to the surface with subsequent re-sedimentation. Typical lagtimes are in the rangebetween 0 My to 400 My (Fig. 1). The absence oflagtimes shorter than depositional ages proves thatno post-depositional heating at the zircon FT an-nealing temperature has occurred.

The zircon FT age populations describe distinctpopulation paths in time (D1-n in Fig. 1); the young-est population path (D1) is believed to portrayclosely the tectonic activity in the sediment sourceterranes. This is corroborated by the combinedheavy mineral analysis. In considering two extremecases: (1) decreasing lagtime, going even to zero isassociated with increasing supply from metamorphicsource rocks, and (2) increase of lagtime correlateswith change to a new source blocks, which have ex-perienced earlier cooling. However, a steady lagtimeat frequent changes of sources is observed duringthe last 10-15 Ma of the orogenesis (Fig. 1).

The orogenic growth of the Andean cordillera inEcuador is summarized as follows: The Early Creta-ceous Peltetec compressive tectonic event has givenrise of deep erosion of the earlier Misahualli volcanic

The Andean Cordillera of Ecuador: timing and mode of orogenicgrowth as revealed from sediments in the Amazon Basin

(heavy minerals and detrital zircon fission-tracks)

Winkler, W., Seward, D., Ruiz, G.M.H. & Martin-Gombojav, N.

Department of Earth Sciences, Geological Institute, ETH Zurich, Switzerland

arc, and the creation of a primordial CR. The subse-quently forming Amazon Basin (Hollin and Napofms.) was supplied from the very slowly exhumingAmazon craton (D1), from Paleozoic-Early Mesozoicmagmatic/volcanic and sedimentary rocks (D2), andfrom the moderately exhuming CR (D3) (Fig. 1). De-creasing lagtimes to zero during upper Napo andTena time (Santonian-Maastrichtian) correlate withstarting exhumation of medium-grade metamorphicsource rocks in the rapidly exhuming CR, and withthe vanishing of the Amazon cratonic source to theeast, i.e. the CR became the main source of detritalmaterial in proximal parts of the AAB. The corre-sponding decrease of lagtimes in path D2 corrobo-rates that these supplying rocks also were situated inthe rapidly exhuming CR. From Eocen until Oligo-cene, the exhumation of very high-grade meta-morphics, the frequent changes of sourcerocks/drainage, respectively, and the radical de-crease of lagtimes in path D2 infers a major phase oforogenic growth in the CR. Since ca 15-20 Ma (Mio-cene) the constant lagtime (30-40 My) associatedwith frequent change of source rocks representspreferential erosion of different blocks, that werebeing exhumed, but not sufficiently for the reset zir-cons to reach the surface. The relatively long lag-time, also seen in the modern river sediment, is pos-sibly due to a regional exhumation event in the LateEocene-Early Oligocene. Hence, the younger historyof the Andean chain is rather characterized by mod-erate, but generalised uplift.

ACKNOWLEDGEMENTS

This work was supported by the Swiss ScienceFoundation Grants No. 21-050844.97 and 20-056794-99.

REFERENCES

Spikings, R.A., Winkler, W, Seward, D. & Handler, R. (2001):Along strike variations in the thermal and tectonic responseof the continental Ecuadorian Andes to the collision withheterogeneous oceanic crust. Earth and Planetary ScienceLetters 186: 57-73.

Ruiz, G. M. H., Seward, D. & Winkler W. (2004): Detrital ther-mochronology – a new perspective on hinterland tectonics,an example from the Andean Amazon Basin, Ecuador. Ba-sin Research 16: 413-430.

Spikings, R.A., Winkler, W., Hughes, R.A., Handler, R. (2005):Thermochronology of the Cordillera Occidental and theAmotape Complex, Ecuador: unravelling the accretionaryand post-accretionary history of the Northern Andes. Tec-tonophysics: 399, 195-220.

Figure 1. Detrital zircon fission track age populations correlatedwith heavy mineral variations in the proximal Andean AmazonBasin, including the interpretation of the active source areasand dynamics. 1/1 represents the stratigraphic correlation line,D1-D3 are population paths of the detrital zircon FT ages intime. Modified from Ruiz et al. (2004).

Salina de Ambargasta is a seasonal playa systemlocated in subtropical South America (29°S). To-gether with other saline environments, Ambargastaoccupies a topographically closed depression knownas Cuenca Saliniana (Álvarez et al., 1990) in thebroken foreland of the Sierras Pampeanas of Argen-tina. This type of broken foreland basin is character-ized by successive basement uplifts produced bythick-skinned deformation throughout the Tertiaryperiod (Jordan & Allmendinger, 1986 and Allmend-inger et al., 1997). Regional and local tectonic set-tings play a main role generating orographic rainfalland, thus, contributing to the prevalence of a trulysemiarid climate in the region (~500 mm/year). Pre-sent average low precipitation rates generate amudflat-ephemeral lake complex. The sedimentaryrecord of Ambargasta, therefore, is monitoring recentenvironmental changes providing a unique archive ofthe dominant climatic conditions that have fluctuatedthroughout the Quaternary, in the eastern Andes ofcentral Argentina.

Combining geomorphology, instrumental data andsatellite images analyses allowed the characteriza-tion of the different modern environments and its dy-namics. At present, this multicomponent systemshows major seasonal changes in the dynamics ofthe subenvironments that are ruled by the regionalhydrology and climate. Dry mudflats (DM) occupythe highest western portions of the playa whereasthe eastern low areas include ephemeral and inter-mittent lakes, ringed by clastic and saline mudflats(CM and SM). These lakes fill with brine during theearly austral summer (December to March) and be-gin shrinking by evaporation by late summer wherethe subenvironment switch to CM and SM surfacesuntil the next rainy season.

The integrated study of the modern system pro-vided an analogue to investigate older sediments.Thus, the saline mudflat (SM) and the clastic/dry

mudflat boundary (CM/DM) were cored and studiedusing a quantitative multiproxy approach includingpetrophysical properties, microstratigrapy and bothorganic and inorganic geochemistry. Ongoinginvestigations include mineralogy analyses, stableisotope geochemistry and dating. Density variationscan be clearly identified in both cored sites, howeverstrong fluctuations mostly characterize the CM/DMenvironment. Water content correlates well withdensity probably caused by a variable content ofevaporites. Sedimentary cores in both SM andCM/DM areas show very high magnetic susceptibility(MS) values with sharp fluctuations and a decreasingtrend throughout depth. The sedimentary sequenceconsists mainly of brownish red and red massiveclays, alternating with gray silty clay beds and yellowblack mineral-bearing sands. Some levels areevaporite-rich (probably gypsum), either sand orclay-sized sediments. The high values in MS couldbe related to the reddish colour of the sedimentsindicating the presence of abundant iron-minerals.Organic matter and carbonates contents in SM andCM/DM environments are quite small although arelative enrichment can be observed in the CM/DMcore.

These preliminary data indicate substantialchanges in the hydrological budget that are shownby a conspicuous response of the sedimentary fa-cies throughout the Late Quaternary record, whichrange from more clastic-dominated mudflats toephemeral lake sediments. The results of ongoingmineralogical and geochemical analyses of the Am-bargasta sedimentary record integrated within a wellconstrained chronological framework will allow us tounravel the environmental history of this systemduring the Late Quaternary. The further combinationof this record with results from similar studiessteaming from this region of South America (e.g.Laguna Mar Chiquita, Piovano et al., 2002) will help

Recent hydrological changes in subtropical Argentina,east of the Andes:

the sedimentary record of Salina de Ambargasta

*Zanor, G., *Piovano, E. & **Ariztegui, D.

* CIGES, Universidad Nacional de Córdoba, Argentinagzanor@efn.uncor.edu

** Institute Forel and Department of Geology and Paleontology, University of Geneva, Switzerland

to clarify the still controversial role of tropical regionsduring intervals of global reorganization in the cli-mate system.

REFERENCES

Allmendinger, R.W., Jordan, T.E., Kay, S.M. & Isacks, B.L.(1997): The evolution of the Altiplano-Puna plateau of theCentral Andes. Annual Rev. Earth Planet. Science, 25 (1):39-74.

Álvarez, L.A., Fernández Seveso, F., Pérez, M.A. & Bolatti, N.D. (1989): Interpretación del subsuelo en los bolsones deSierras Pampeanas en base a la información Geofísicadisponible y Geología de superficie. Inédito YPF, BuenosAires.

Jordan, T.E. & Allmendinger, R.W. (1986): The Sierras Pam-peanas of Argentina: a modern analogue of Rocky Moun-tain foreland deformation. American Journal of Science286: 737-764.

Piovano, E., Ariztegui, D. & Damatto Moreiras, S. (2002): Re-cent environmental changes in Laguna Mar Chiquita (Cen-tral Argentina): A sedimentary model for a highly variablesaline lake. Sedimentology 49: 1371-1384