U N I V E R S I D A D D E C O N C E P C I Ó N DEPARTAMENTO DE CIENCIAS DE LA TIERRA

CONGRESO GEOLÓGICO CHILENO 10° 2003

HOLOCENE SEDIMENTATION IN ICALMA AND PUYEHUE (SOUTHERN ): INSTANTANEOUS VS CONTINUOUS RECORDS

BERTRAND, S.1,*, CHARLET, F.2, RENSON, V.1, . BOES, X.1, VARGAS-RAMIREZ, L.3, ARNAUD, F.4, LIGNIER, V.5, CHAPRON, E.6, BECK, S.7, MAGAND, O. 8, FAGEL,N.1, & DE BATIST, M.2

1 U.R. Argiles et Paléoclimats, Department of Geology, University of Liège, Liège, Belgium 2 RCMG, University of Ghent, Belgium 3University of Liege, Paleobotany, Paleopalynology, Micropleontology 4Université des Sciences et Techniques de Lille, Lille, France 5 ENS, Lyon, France 6 Geological Institute, ETH, Zürich, Switzerland 7 LGCA, Université de Savoie, Le-Bourget-du-Lac, France 8 LGGE, CNRS Grenoble, France * Corresponding author: [email protected]

INTRODUCTION El Niño is one of the most important present-day climatic phenomena affecting the Earth’s climate. But a better understanding of ENSO variability through time requires high-resolution studies of past records in suitable areas. In order to document Holocene climate variability at middle latitudes in Southern Chile and to test the possible influence of El Nino phenomena, two lakes directly submitted to the influence of the Westerlies in the District area, have been selected: Icalma and Puyehue lakes. Based on high-resolution seismic reflection investigations (Charlet et al., this volume), two long coring sites where selected in each lake and within the main distal clastic environments (interflow and underflow deposits) in order to provide high- resolution reconstructions and to document past flooding activities. Although the active geodynamic setting of this part of Chile might complicate the normal layering of lacustrine sedimentary infills, a detailed understanding of sedimentary features and sedimentary processes has the potential to provide reconstructions of climate changes, paleoseismicity and former volcanic activity.

GENERAL SETTING OF SELECTED LAKES Icalma is a small (11.65 km2) but deep (135 m) oligotrophic lake from the IX region (Parra et al, 1993), located in the Cordillera de Los at an elevation of 1140 m near the source of the Bio-Bio River (Fig. 1). This narrow and over deepened basin is mainly of glacial origin. The glacial valley contain several rock bars delimitating sub-basins in the lake, the main one being closely associated to the development of a fault affecting the basement (Mardones et al, 1993). The catchment area covers 147 km² and is surrounded by several active volcanoes (i.e., Lonquimay and Sollipulli) and especially the Llaima, one of the most active of America.

Todas las contribuciones fueron proporcionados directamente por los autores y su contenido es de su exclusiva responsabilidad.

Puyehue is a typical oligotrophic and moraine-dammed lake from the Lake District (X region). This large lake (164 km2) is located at the foothill of the Cordillera de Los Andes at an elevation of 185 m (Fig. 1) and is characterized by a rather limited water depth with respect to the other lakes of the area (maximum depth of 123 m). Several islands resulting from glacial erosion but also the development of inter-stadial moraines in the middle of the lake resulted in the subdivision of several sub-basins. The main tributary of the lake (Rio Golgol) forms a large delta in the deepest sub-basin of the lake along is eastern side. The watershed of (1267 km²) is delimited by several active volcanoes (i.e., Puyehue–Cordon de Caulle, Antillanca and Osorno), but the vegetation cover is here more dense than around Lake Icalma. The watersheds of these two lakes are covered by a several meters thick soft sedimentary cover composed of post-glacial volcanic ashes deposited by several volcanic eruptions during the Holocene (Bertrand & Fagel, submitted). Actual andosoils (i.e., Trumaos) are developed on these ashes. In such a setting, depending on the type and density of the vegetation cover, heavy rains will erode and rework Trumaos from the catchment area to the lake basin during major tributary floods.

CORING OPERATIONS Coring operations involved an Uwitec platform (Fig. 2) and a 3 m-long piston coring system, as well as a classical short gravity coring device. With this system, composite long cores are resulting from a succession of several 3 m-long sections cored at different depths and at two nearby locations, as show in Figure 3. On each coring site, five short gravity cores where also taken, in order to provide enough material for 210Pb and 137Cs dating and to test different climate proxies. To reconstitute the composite lithology of a long coring site, sediment cores back to Europe where measured using a GEOTEK multi sensor track providing the bulk density, the P- wave velocity and magnetic susceptibility of the sediments every two centimeters. These sediment cores where then splitted into halves, described and photographed. Composite lithology reconstructions were then based on the macroscopic reconnaissance of key horizons and on the petrophysical properties measured with the GEOTEK.

Fig. 1 - Location map of Icalma and Puyehue lakes in the North and the Centre of the Chilean Lake District. These two basins are submitted to heavy precipitations directly related to the Westerlies.

Fig. 2 - UWITEC platform: coring on Puyehue Lake, February 2002.

Fig. 3 - Composite logs of the four cores obtained on Icalma and Puyehue lakes. Overlapping 3 meters sections have been correlated with the help of macroscopic descriptions and high resolution magnetic suceptibility measurements. Grey shaded areas are sections which have been selected for sampling.

In each lake, two long coring sites (Fig. 4) where selected within the distal clastic depocenters associated to the main tributary of the lake (interflow and underflow flood deposits) in order to provide high resolution reconstructions and to document large floods periodicities (e.g., Chapron et al., 2002). Coring locations presented in Figure 4 where selected on the base of the morphology of the basin and of high-resolution seismic reflection profiling (Charlet et al., this volume).

The two cores collected in Icalma Lake are ICA-I (interflow deposits, 77 m water depth) and ICA-II (underflow deposits, 135 m water depth) and their composite logs reach a length of ~8 m (Fig. 3 and 5). Due to bad climatic conditions, the location of ICA-I had to be adapted, and resulted next to sediment slides visible on seismic profiles. These gravity reworking phenomena seems to be mainly related to the development of sub-aerial canyons along the steep northern flank of the catchment and to the generation of debris flows during heavy rains storms (Don Callucello-Carillo, Oral. Com.).

Cores collected in Puyehue Lake are PU-I (underflow deposits, 122 m water depth) and PU-II (interflow deposits, 48 m water depth). The compaction of decimeter-thick sandy layers with high water content during coring operations in site PU-I and the presence of gas in the sediments, produced important coring disturbances (fine-grained sediments liquefaction) and this coring site could not be properly recovered. A more than 11 m long composite core has been obtained on the interflow site of Puyehue Lake (PU-II, Fig. 3 and 5) while the 3 m long sections of the core collected in the underflow site of this lake did not show good overlaps (PU-I).

Fig. 4 - Long cores locations on Icalma and Puyehue lakes. Bathymetric map of Puyehue Lake according to Campos et al. (1989). ICAI = interflow location, ICAII = underflow location. PUI = underflow location, PUII = interflow location. DATING The preliminary chronologies of these cores are based on 210Pb and 137Cs profiles for the last millennium on short gravity cores and some AMS 14C dates on the long cores. 210Pb and 137Cs profiles where measured at the Laboratoire de Glaciologie et de Géophysique de l’Environnement (LGGE, France), and revealed rather low sedimentation rates for the last century: 17 mg/cm2.yr at ICA-II site (corresponding to approximately 0.7 mm/yr); 10 mg/cm2.yr at ICA-I site (~0.5 mm/yr) and 29-25 mg/cm2.yr at PU-II site (~1.08 mm/yr). Radiocarbon datings on wood samples (frequently observed in event deposits) or on bulk sediment are in progress at the Poznan Radiocarbon Laboratory (Poland).

HOLOCENE SEDIMENTATION

ICALMA LAKE Sediment cores from Icalma Lake consist in sandy volcanic layers intercalated in silty-clay laminated sediments (Fig. 5). Laminations are composed of diatoms, volcanic minerals and allophane, a typical secondary product coming from the weathering of volcanic ashes very widespread in the watershed (Bertrand & Fagel, submitted). Moreover, centimetric pumices have been described in both cores. Based on these new cores it appears that the volcanic activity of the area has been more intense in the recent past than it is today. The thicknesses of the laminations in the sediments are also changing down cores, indicating that sedimentation rates and the sedimentary processes might have been quite different during the Holocene.

The total thickness of the volcanic sandy layers in ICA-II core (underflow location) reaches 190 cm while the half of the core coming from interflow location (ICA-I) contains volcanic sands (Fig. 5). These sandy sediments have several possible origins: - in situ air-fall , - volcanic ashes eroded from the catchment during heavy rains and transported to the deep basin (ICA-II) through underflows during major floods of Rio Icalma, - volcanic ashes eroded from the catchment during heavy rains and directly transported to the coring sites by hill slopes (especially at ICA-I through steep canyons draining the northern slopes of the catchment nearby the depocenter of interflow deposits), - sub-aqueous gravity reworking of sediments affecting the steep slopes of the sub- basins (i.e. slides, turbidites, homogenites) triggered by (Sturm et al, 1995; Beck et al, 1996; Chapron et al, 1999) and/or strong volcanic eruptions.

Whereas in situ air-fall tephras should be found in both coring sites and can be followed by diatom blooms, volcanic ashes eroded from the catchment and transported to the coring sites after heavy rains might not be synchronous and recorded basin-wide. Particles transported through major floods of the tributary or directly transported into the lake by hill slopes processes will also have different grain-size properties. Turbidites and homogenites can as well be characterized on the base of grain-size properties and sedimentary structures.

Fig. 5 – Schematical logs of composite cores from Icalma and Puyehue lakes. Because sediment disturbances, PU-I core has not been represented.

Five thick sandy events only recorded in ICA-I between 1.5m and 4 m (Fig. 5) are graded coarse volcanic sands much probably originating from the canyons along the hill slopes. They could document a period characterized by more frequent heavy rains.

Two pumices layers have been described on both cores. Mineralogical and geochemical analyses carried out on these two pumice layers on ICA-II core have shown that they are identical. Moreover, their glasses have the same chemical composition than glasses of the Alpehue pumice outcropping in the watershed and dated at 2900 yr BP (i.e., 3045 cal. BP) (Naranjo et al., 1993; De Vleeschouwer et al., in prep.). The deepest pumice layer found on both coring sites seems to originate from a direct lacustrine depositional pattern during the eruption. But the shallowest pumice layer is included in thick deposits of volcanic sands on both cores and, seems to result from reworking processes. In core ICA-II this upper pumice layer is incorporated around 3 m into the sandy base of the 2 m thick homogenite (Fig. 5). This thick sedimentary event, visible on seismic profiles, is only present in the deepest part of the basin, and could reflect a large sediment slide triggered by an . In ICA-I, the upper pumice layer is included in one of the five coarse volcanic sandy layers and could therefore be reworked from the catchment after heavy rains. In the drainage basin of Lake Icalma, a second pumice layer coming from another volcano has been described near 10.000 cal. BP. (Naranjo & Moreno, 1991; De Vleeschouwer et. al., in prep.). Based on mineralogical and geochemical analyses, this pumice layer has not been found at the bottom of our cores, making them younger than 10.000 cal. BP.

PUYEHUE LAKE Because PU-I record is disturbed by gas bubbles and coring operations, this core has not been studied in details. The core collected on interflow location (PU-II) reaches more than 11 m (Fig. 5). From a sedimentological point of view, this core presents a very finely laminated clayey sediment rich in diatoms and contain 79 thin layers (Fig. 5). Grain-size measurements on these centimetric sandy layers will help to distinguish air-fall tephras from possibles turbidites reworking volcanic sands. Around 1 m in the core, two thicker sandy layers are characterized by load-structures, indicating sediment liquefactions. These horizons might result from massive air- fall tephras and/or earthquake shaking. Mineralogical analyses have shown a quite stable terrigenous sediment supply. Bulk mineralogy is dominated by amorphous minerals (mainly volcanic glasses and non-crystalline clays) and plagioclases. The combination of 210Pb and 137Cs profiles together with lamination counting on thin sections in the short gravity core from site PU- II revealed that laminations made of diatom layers and clastic layers are varves (i.e. an annual signal). This conclusion is strongly supported by the correlation of a sandy volcanic layer with the historical eruption of Osorno volcano in AD1835, and by the reconnaissance on thin section of a mixed layer (Marc & Agnon, 1995) associated to the 1960 earthquake (Chapron et al., in prep.). This mixed layer is a 3 cm thick horizon consisting of in situ deformed and liquefied laminae. It is recovered by a thin green clayey level that might to be related to the pumice fall during the eruption of the Cordon de Caulle that was linked to the .

Because this core has the potential to cover more than the Holocene, several sedimentological proxies are in course of study: high-resolution magnetic susceptibility and grain-size analyses (at a resolution of 0.5 cm, i.e. ~8 yr.), bulk and clay mineralogy, organic matter content, biogenic silica measurements and geochemical analyses with a 10 cm resolution (i.e. ~160 yr.).

AGE-DEPTH MODELS Given the geodynamic setting of these lakes, age-depth models on sediment cores should first of all be based on the reconnaissance of continuous sedimentation versus sedimentary events of climatic (flood deposits, hill slope erosion) or geodynamic (tephra, seismites) origin. Robust chronologies on these cores should be based on 210Pb, 137Cs and 14C datings, together with palynological studies and lamination counting on thin sections, in order to test 14C chronologies and to avoid reworked wood samples and possible contamination of dated organic matter by CO2 coming from tephras (e.g., Zolitscka & Negedank, 1996). Finally, age-depth model should be supported by the correlation of sedimentary events with historical data (earthquakes, volcanic eruptions) and or with regional well dated prehistorical eruptions.

PRELIMINARY CONCLUSIONS 1) The main core from Puyehue Lake (PU-II) contains a continuous sedimentary record. This core is studied by a high-resolution multi-proxy approach in order to test possible climatic impact of ENSO in Southern Chile. The establishment of the chronology in this core will provide a detailed record of past volcanic eruptions and maybe strong paleo-earthquakes in the southern part of the Lake District.

2) The two cores from Icalma Lake mainly contain informations about the regional geodynamic and/or volcanic activity, but can also provide some paleoclimatic reconstructions based on the evolution of flood deposits.

ACKNOWLEDGMENTS This research is supported by the Belgian SSTC project EV/12/10B “A continuous Holocene record of ENSO variability in Southern Chile”. We would like to thank Maria Mardones, Roberto Urrutia (U. de Concepcion) and Mario Pino (U. Austral de Chile, Valdivia) for their very helpful logistic participation during our 2001-2002 fieldwork mission in Chile. We are grateful to Waldo San Martin and Alejandro Peña (Centro EULA, Concepcion) for their hard work during the coring part of the mission. We acknowledge the CONAF and especially the wardens of for their field knowledge. Finally, we would like to thanks the Chilean population encountered during our fieldwork for their hospitality, sympathy and availability.

REFERENCES Beck, C., Manalt, F., Chapron E, Van Rensbergen, De Batist, M. 1996. Enhanced seismicity in the Early post-glacial period: evidence from the post-Würm sediments of Lake Annecy, Northwestern Alps. Journal of Geodynamics, 22, ½, p. 155-171. Bertrand, S. & Fagel, N. (submitted) – New evidences for volcanic origin of Trumaos parental material from the Lake District (Chile, 40°S). Submitted to Revista Geologica de Chile, March 2003. Bentley, M.J. 1997. Relative and radiocarbon chronology of two former glaciers in the Chilean Lake District. J. Quatern. Sci. 12, 25-33. Campos, H., Steffen, W., Agüero, G., Para, O., Zúñinga, L. 1989. Estudios limnologicos en el Lago Puyehue (Chile): Morfometria, factores fisicos y quimicos, plankton y productividad primaria. Medio Ambiente 10,36-53. Chapron, E., Beck, C., Pourchet, M., Deconinck, J.F. 1999. 1822 earthquake-induced homogenite in Lake Le Bourget (NW Alps). Terra Nova, 11, 2/3, p. 86-92. Chapron E, Desmet M, De Putter T., Loutre M.F., Beck C., Deconinck J.F. 2002. Climatic variability in the northwestern Alps, France, as evidenced by 600 years of terrigenous sedimentation in Lake Le Bourget. The Holocene, 12, 2, p. 177-185. Charlet, F., Marchand, C., Volland S., Pino, M., Muller J., Chapron, E. & De Batist, M. (this volume). Reflection- seismic study of six lakes in South-Central Chile (37°S-42°S): Lagos Laja, Lleulleu, Icalma, , Puyehue & Todos Los Santos. Marco, S. & Agnon, A. 1995. Prehistoric earthquke deformations near Masada, Dead Sea graben. Geology, 23, 8, p. 695-698. Mardones M., Ugarte E., Rondanelli M., Rodriguez A., Barrientos C. 1993. Planificacion ecologica en el sector Icalama-Liucura (IX Region): proposicion de un metodo. Monografias Cientificas, vol. 6, F. Faranda & O. Parra (Eds), EULA, 91 p. Naranjo, J.A., Moreno, H. 1991. Actividad explosiva postglacial en el volcán Llaima, Andes del Sur (38°45'S). Revista Geológica de Chile 18, 69-80. Naranjo, J.A., Moreno, H., Emparan, C. & Murphy, M. 1993. Volcanismo explosivo reciente en la caldera del volcàn Sollipuli, Andes del Sur (39°S). Revista Geológica de Chile, 20, 167-191. Para O., Campos H., Steffen W., Aguero G., Basualto S., Aviles D., Vighi M., 1993. Estudios limnologicos de los lagos Icalama y Gualletue: lagos de origen del Rio BioBio (Chile Central). Monografias Cientificas, Vol 12, “Evaluacion de la calidad del agua y ecologia del sistema limnetico y efluvial del rio Biobio, EULA, F. Faranda & O. Para (Eds), p. 161-188 Sturm, M., Siegenthaler, C. & Pickrill, R.A., 1995. Turbidites and ‘homogenites’. A conceptual model of flood and slide deposits. 5ème Congrès Français de Sédimentologie ASF, Paris, Publication ASF 22, p. 170. Zolitschka, B., Negendank, J.F.W. 1996. Sedimentology, dating and paleoclimatic interpretation of a 76.3 ka record from Lago Grande di Monticchio, Southern Italy. Quatern. Sci. Rev. 15, 101-112.