scieee AI-readable full text Open interactive document viewer

Southern Hemisphere Westerly Winds have modulated the formation of laminations in sediments in Lago Fagnano (Tierra del Fuego, Argentina) over the past 6.3 ka

Vizcaino, Alexis,Jimenez Espejo, Francisco Jose,Dunbar, Robert B.,Mucciarone, David,García-Alix, Antonio,Neugebauer, Ina,Ariztegui, Daniel

Abstract

The research has been supported by the Beatriu de Pinós postdoctoral programme at Agència de Gestió Acadèmica i Universitària at Generalitat de Catalunya, NSF grants ATM-0408668 and EAR-1103550 to R. Dunbar, and the Spanish Ministry of Science and Innovation (grant CTM2017-89711-C2-1-P), co-funded by the European Union through FEDER funds. A. García-Alix was also supported by Ramón y Cajal Fellowship RYC-2015-18966 of the Spanish Government (Ministerio de Economía y Competividad).

Full text

Southern Hemisphere Westerly Winds have modulated the formation of laminations in sediments in Lago Fagnano (Tierra del Fuego, Argentina) over the past 6.3 ka ALEXIS VIZCAINO , FRANCISCO J. JIMENEZ-ESPEJO , ROBERT B. DUNBAR , DAVID MUCCIARONE , ANTONIO GARC  IA-ALIX , INA NEUGEBAUER AND DANIEL ARIZTEGUI Vizcaino, A., Jimenez-Espejo, F. J., Dunbar, R. B., Mucciarone, D., Garc ıa-Alix, A., Neugebauer, I. & Ariztegui, D. 2023 (January): Southern Hemisphere Westerly Winds have modulated the formation of laminations in sediments in Lago Fagnano (Tierra del Fuego, Argentina) over the past 6.3 ka. Boreas, Vol. 52, pp. 124–138. https://doi.org/10. 1111/bor.12600. ISSN 0300-9483. Tierra del Fuego in Argentina is a unique location to examine past Holocene wind variability since it intersects the core of the Southern Hemisphere Westerly Winds (SHWW). The SHWWare the most powerful prevailing winds on Earth. Their variation plays a role in regulating atmospheric CO 2 levels and rainfall amounts and distribution, both today and in the past. We obtained a piston core (LF06-PC8) from Bah ıa Grande, a protected sub-basin at the southern margin of Lago Fagnano, the largest lake in Tierra del Fuego. This article focuses on the uppermost 185 cm of this core, corresponding to laminated sediment from the last ~6.3 ka. Laminations consist of millimetrescale paired dark and light layers. Previous studies and new geochemical analysis show that the dark and light layers are characterized by differing concentrations of Mn and Fe. We attribute the distribution of Mn and Fe to episodic hypolimnic oxic–anoxic variations. The age model suggests an approximately bidecadal timescale for the formation of each layer pair. We propose a new model of these redox changes with the SHWW variations. The most likely phenomenon to produce complete water-column mixing is thermobaric instability, which occurs in colder winters with low-intensity SHWW (El Ni~ no-like conditions). In contrast, windier winters are characterized by higher temperatures and reduced mixing in the water column, facilitating a decline in oxygen concentration. Laminations, and the inferred presence of periodic hypolimnion redox changes, are common features of the past ~6.3 ka. Geochemical proxy variability is compatible with an intensification of El Ni~ no/Southern Oscillation activity during the past ~2 ka. Alexis Vizcaino, Robert B. Dunbar and David Mucciarone, Earth System Science, Stanford University, 397 Panama Mall, Stanford, CA 94305, California, USA; Francisco J. Jimenez-Espejo (corresponding author: [email protected]), Instituto Andaluz de Ciencias de la Tierra, UGR-CSIC, Avda. de las Palmeras n°4, 18100, Armilla, Spain and Research Institute for Marine Resources Utilization (Biogeochemistry Program), Japan Agency for Marine-Earth Science and Technology, 2-15 Natsushimacho, Yokosuka, Kanagawa 237-0061, Japan; Antonio Garc ıa-Alix, Department Paleontolog ıa y Estratigraf ıa, Universidad de Granada, Avenida de la Fuente Nueva S/N, 18071, Granada, Spain; Ina Neugebauer, Department of Earth Sciences, University of Geneva, Rue des Maraichers 13, 1205, Geneva, Switzerland and Section Climate Dynamics and Landscape Evolution, GFZ German Research Centre for Geosciences, Wissenschaftpark “Albert Einstein”, Telegrafenberg, 14473, Potsdam, Germany; Daniel Ariztegui, Department of Earth Sciences, University of Geneva, Rue des Maraichers 13, 1205, Geneva, Switzerland; received 19th June 2021, accepted 16th July 2022. Lago Fagnano (LF) is the most southerly large lake outside of Antarctica, and it is a key location at which to understand the relationship between Antarctic and South American climate variations. Previous studies based on sediment cores in Lago Fagnano have allowed the reconstruction of Late Pleistocene and Holocene climate, environment, glacial fluctuations and tectonic activity (Waldmann et al.2008,2010,2011,2014;Moy et al.2011; Sanci et al.2021). Lago Fagnano’s location and orientation make this site an excellent place to examine the Holocene variations of the Southern Hemisphere Westerly Winds (SHWW), because of the weaker rainshadow effect owing to the lower elevation of the Andes running parallel to the westerlies (Fig. 1). The SHWW exert strong control over the amount and distribution of rainfall (Garreaud et al.2013; McCulloch et al.2020) and regional vegetation (Markgraf & Huber 2010), associatedwith Andean orographic effects (Garreaud 2007;Moyet al.2008; Lamy et al.2010). The SHWW, together with human activity (agriculture and cattle farming), have a significant influence on Patagonian life and water management. Globally, the SHWW are thought to play an important role in the regulation of atmospheric CO 2 levels through: (i) a direct effect related to higher gas transfer velocities governing the air–sea gas exchange at higher wind speeds (e.g. Ito et al.2010); and (ii) dynamic changes in ocean productivity and upwelling/downwelling that influence the CO 2 difference between ocean and atmosphere (Canadell et al.2007; Lamy et al.2010). Many studies have focused on understanding the SHWW regime and its past latitudinal movement (e.g. Saunders et al.2018; Xia et al.2018). There is evidence for a northward shift or expansion of the SHWWduring the Holocene (McCulloch et al.2000; Toggweiler et al.2006), yet we still have few details on how the winds changed and whether their strength varied by latitude. Previous insights are derived from glacial DOI 10.1111/bor.12600 ©2022 The Authors. Boreas published by John Wiley & Sons Ltd on behalf of The Boreas Collegium. This is an open access article under the terms of the Creative Commons Attribution-NonCommercial License, which permits use, distribution and reproduction in any medium, provided the original work is properly cited and is not used for commercial purposes. geomorphology (Mercer 1976; Rabassa & Clapperton 1990; Coronato et al.2009), pollen (Musotto et al.2017; Navatini et al.2019), charcoal analyses (Huber et al.2004; Mansilla et al.2016), dendrochronology (Aravena et al.2002; Masiokas & Villalba 2004) and multiproxy sediment records (Rogers & van Loon 1982; Ariztegui et al.2007;Moyet al.2008, 2011; Borromei et al.2010; Lamy et al.2010; Waldmann et al.2010; Kilian & Lamy 2012; van Daele et al.2016; Quade & Kaplan 2017; Zolitschka et al.2019). These palaeoclimate reconstructions reveal apparently contradictory data and regional syntheses remain controversial (Kilian & Lamy 2012). The high-resolution millimetric-scale geochemical study presented by Neugebauer et al.(2022) demonstrates that the lamination observed at Lago Fagnano is associated with Fe and Mn enrichments that can be explained as lacustrine bottom-water redox variations. In our study, we aim a step further, to provide a comprehensive conceptual model relating the lake redox changes to the SHWW, their evolution and frequencies at a high-resolution scale in this area. Therefore, our goal is toinfer thepossiblerelationbetweenclimateandlaminae formation in Lago Fagnano (Fig. 1). Study area Tierra del Fuego is an archipelago at the southernmost tip of South America between Chile and Argentina, separated from the mainland by the Strait of Magellan. The archipelago consists of Isla Grande de Tierra del Fuego (also called Tierra del Fuego) and many islands and islets. This archipelago is influenced by volcanic and seismic activity related to the South American and Scotia–Antarctic plate boundaries (Klepeis 1994). One of its most prominent features is Lago Fagnano (known inChileasLagoCami), locatedat26 m a.s.l. Itis100 km long and 5–15 km wide, trending east–west and traversing Isla Grande de Tierra del Fuego at ~54°S/68°W. It is the largest freshwater lake on the island, and is the world’s most southerly large lake outside of Antarctica (Herdendorf 1982) (Fig. 1). Lago Fagnano was formed Fig. 1. A. MapofsouthernSouthAmerica,indicatinginredIslaGrandedeTierradel Fuegoand,panelBinblack.B.Terrainmapindicatingwitha red dot the position of the piston core LF06-PC8 from Bahia Grande. Google Maps (accessed 17 October 2020). C. Satellite image of Bahia Grande and surroundings superimposedby the bathymetry. The reddot indicates the location of coreLF06-PC8. Image taken from Google Earth(accessed 10 February 2022, with images acquired on 4 February 2015 by CNES/Airbus Maxar Technologies). BOREAS Formation of laminations in sediments in Lago Fagnano, Argentina over the past 6.3 ka 125 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License in the depression created in an asymmetric pull-apart basin within the principal displacement zone of the Magallanes–Fagnano fault system, which is part of the diffuse left-lateral plate boundary (Onorato et al.2021). The strike-slip rate along this boundary is approximately 6.61.3 mm a 1 (Smalley et al.2003). The most severe earthquake measured in Tierra del Fuego (M 7.8) occurred along the eastern end of Lago Fagnano in 1949 (Pedrera et al.2014). Lago Fagnano is divided into two main sub-basins (230 mintheeast and180 mdeepinthewest;Waldmann et al.2008). The outflow connecting Lago Fagnano to the Pacific Ocean is Rio Azopardo, via the Whiteside Channel and the Strait of Magellan. The lake water residence time is about 20 years, based on a lake volume of 43 km 3 . Lago Fagnano is an oligotrophic lake (Mariazzi et al.1987; Ritcher et al.2010). Its main source of water is precipitation. However, several small cirque glaciers between 900 and 1300 m a.s.l. supply meltwater, with more significance towards the western part of the lake. The altitude of the present equilibrium line at ca. 1100 m a.s.l. suggests that most of these glaciers are in rapid recession (Coronato et al.2008). Indeed, Bah ıa Grande also has a small catchment area that supplies sediment directly into it. The present-day climate is semi-arid and cold (annual average temperature 5.8 °C, precipitation 550 mm; Moy et al.2011). The principal feature of the Fuegian climate is the westerly wind system. Lago Fagnano’s location and orientation make this site an excellent location to examine the Holocene variation of the SHWW. The lake runsparallelto thewind fieldandtheupwindcordillerais relatively low in elevation (reaching 2469 m a.s.l. (relative to ~3000 m a.s.l. of the Patagonian Ice Fields) and above 4000 m a.s.l. further north), resulting in a weaker rainshadow effect that supports forests of Nothofagus pumilio and Nothofagus antarctica (Coronato et al.2009). The westerlies are important throughout the year, but as a general rule the SHWW intensity is stronger in summer (December to February) than in winter (June to August), when the core of the SHHW is displaced northwards (Lamy et al.2010). Precipitation follows wind speed only during the summer months (Garreaud et al.2013). The strength of the westerly winds and surface air temperature (SAT) are positively correlated in winter and inversely correlated in summer (Garreaud et al.2013). At a larger scale, in Patagonia El Ni~ no events are associatedwith weaker SHWW, whereas during La Ni~ na stronger SHWW occur (Schneider & Gies 2004). Additionally, the SHWW at the latitude of Tierra del Fuego are influenced by the El Ni~ no/Southern Oscillation (ENSO) (Rees et al.2015). Garreaud et al.(2013) correlate Tierra del Fuego summer rainfall anomalies with the Antarctic Oscillation and winter anomalies to the strength of the SHWW. Garreaud et al.(2013) also observe a positive relation between the SHWW and the SAT in winter. In summer the SHWW and SAT are inversely related. Material and methods In thiswork we focus on the uppermost 185 cm of piston core LF06-PC8 (54°35026.1600S, 68°29021.8700W, water depth 69 m; Fig. 2). The core was acquired in 2006 using a Kullenberg-type coring system deployed from a 12 m research vessel, the R/V Neecho (Stanford University, USA), from Bah ıa Grande in Lago Fagnano (Figs 1, 2). Physical properties and magnetic susceptibility Immediately after core splitting and surface cleaning, we used the GEOTEK Multi-Sensor Core Logger at ETH Zurich (Switzerland) to take continuous digital photographs of core LF06-PC8, and analysed for magnetic susceptibility (MS) and density at 1-cm spatial resolution. Magnetic susceptibility was additionally reacquired at a higher resolution (0.5-cm intervals) with a point sensor, on the wet sediment cores after core splitting using the GEOTEK Multi-Sensor Core Logger at the USGS Coastal Marine Geology facility in Menlo Park, CA (USA). Unfortunately, a small section of this core (between 236 and 253 cm) was lost in transfer to the photography laboratory and therefore could not be photographed, creating a noticeable gap in the full core photograph (Fig. 2C). Organic geochemical analyses Thestablecarbon andnitrogenisotopiccompositionand C and N concentrations of bulk organic matter were derived from 5 ml (wet) samples collected uniformly throughout the core at 1 cm intervals. Brodie et al.(2011) report analytical bias in the d 13 Candd 15 N values of bulk organic matter from various acid treatments. As a result, core LF06-PC8 was tested to determine whether the removal of inorganic carbon was necessary (Mucciarone 2021). No discernible difference was found in the carbon concentration between 1MHCl treated and untreated samples, thus untreated samples were freeze-dried and weighed into tin capsules and analysed on a Carlo Erba NA1500 Series 2 elemental analyser, coupled to a Finnigan Delta Plus isotope ratio mass spectrometer via a Finnigan ConFlo II open split interface, at the Stanford University Stable Isotope Biogeochemistry Laboratory (USA). The results are presented in standard delta notation, with d 13 C reported relative to the VPDB carbonate standard and d 15 N relative to air. We used L-glutamic acid USGS-40 (NIST RM8573) as our standard for calibration, and the precision reached on 102 USGS-40 measurements is 0.1&for d 15 N, 0.08&for d 13 C, 0.22% for N and 0.75% for C weight percent concentrations. 126 Alexis Vizcaino et al. BOREAS 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Inorganic geochemical analysis Major elementswere measured using the Avaatech X-ray fluorescence (XRF) core scanner at the Texas A&M IODP (Integrated Ocean Drilling Program) facility (USA) at 2-mm resolution in the laminated sediment intervals and at 5-mm resolution on the debrite intervals. The target elements (Al, Si, S, K, Ca, Ti, Mn, Fe, Ba, Ni, Cu, Br, Rb, Sr and Zr) were measured as counts per second (cps) using X-rays at both 10 and 30 kV. For this study only the elements Al, Ca, Fe, Ti and Mn are represented and interpreted. Data were collected for 40 and 20 s for 2and 5-mm intervals, respectively. To corroborate the various factors influencing the geochemical composition of the sediment, we conducted principal component analysis (PCA) using the software package PAST version 4.40 (Hammer et al.2001; Fig. 3). The normalized element counts were standardized by subtracting the mean and dividing by the standard deviation (Davis 1986), following Bahr et al.(2014). For the PCA we used only those elements with the highest intensities, that is, Al, Si, K, Ca, Ti, Mn, Fe, S and Ba. The chemical composition of the glass particles of glass (tephra deposit) wasdetermined using a JEOL 8900 electron microprobe with a 15 kV, 10 nA beam, defocused (10 lm) to avoid mobilizing the alkali metals. Chronology Dating lacustrine sediments in a cold environment presents a significant challenge owing to the scarcity of carbon suitable for radiocarbon analysis. The chronology for core LF06-PC8 is based on four radiocarbon analyses of wood fragments and bryophytes (Table 1) carried out at the Center for Accelerator Mass Spectrometry at Lawrence Livermore National Laboratory (USA). Calibrated ages were obtained by means of CALIB 8.2 software (Stuiver et al.2021) and the SHcal20 (Hogg et al.2020) calibration curve at the 95% confidence interval. Since dates at depths 228 and 253.5 cm were obtained from turbidite/debrite deposits, they were not used in the age model. An additional age was provided by a tephra layer, identified as Hudson H1 (Stern et al.2016). Therefore, we decided to show all of Fig. 2. A. High-resolution seismic profile of Bah ıa Grande indicating the sediment core and a schematic acoustic facies interpretation. B. False colour satellite image of the western part of Lago Fagnano and Bah ıa Grande, with a red line indicating the seismic profile (A). C. Image of core LF06-PC8,lithologyandvisuallydescribedgrainsize (vfs=veryfindsand;fs=finesand;ms=mediumsand;cs=coarsesand).D. Detailedimageof the upper part of core LF06-PC8 with characteristic lamination. BOREAS Formation of laminations in sediments in Lago Fagnano, Argentina over the past 6.3 ka 127 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License the data in depth along with the obtained 14 C calibrated ages, and to perform a simple linear age model for the uppermost hemipelagite. Optical microscopy We undertook microscopic observation of the sediment samples using a petrographic microscope at US Geological Survey, Menlo Park. This method was especially relevant to understanding the biogenic composition of the lighter sediments lying on top of the turbidites. Qualitative observational data helped us to correlate the LF06-PC8 tephra to the H1 event. The andesitic glass exhibits a distinct green/brown tint when observed through a microscope. Results Sediment stratigraphy Six sediment facies characterize core LF06-PC8: hemipelagites, debrites, turbidites, diatom layers, glacioFig. 3. A. Principal component analysis using PAST version 4.40 software (Hammer et al.2001) of XRF data of only the laminated intervals, avoiding turbidites, and the elements Al, Si, K, Ca, Ti, Mn, Fe, S and Ba. B. Table of principal components (PCs), and their eigenvalues and variance. C. Bar chart indicating the values corresponding to PC1, responsible for 53.4% of the variance of XRF data. D. Bar chart indicating the values corresponding to PC2, responsible for 17.6% of the variance of XRF data. E. Correlation matrix with significant positive correlation (r >0.60 and p-value <0.01) between Al, Si, K, Ca, and Ti (green), and significant negative correlation between Fe and the other elements (red) for the laminated sediment intervals. Table 1. AMS radiocarbon ages from core LF06-PC8’s sedimentary record. Calibrated ages were obtained using Calib 8.2 (Stuiver et al. 2021) and the SHCal20 calibration curve (Hogg et al. 2020) at 95% confidence intervals. The age of the Hudson H1 tephra represents the mean 14 Cage determined at various lakes (Stern et al. 2016). AMS laboratory reference Core depth (cm) Sample material Radiocarbon age (a BP1r) Median age (cal. a BP) 2rrange (cal. a BP) 154 730 2.5 Wood 19030 174 0–283 N92831 136.0 Wood 413530 4617 4447–4817 N92832 228.0 Wood 610545 6925 6753–7156 N92833 253.5 Bryophytes 696540 7757 7671–7917 Hudson 1 8434 8379–8537 128 Alexis Vizcaino et al. BOREAS 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License lacustrinesediments,andtephra(Fig. 2).Theuppermost 185 cm of the succession is dominated by hemipelagic sediments. Below 185 cm, the sediment succession is dominatedbyveryfinetocoarsesandturbidites(between 5and 55-cm thick) and debrites (ranging from a few centimetres to almost 1.5-m thick). Turbidites and debrites generally underlie a grey layer with a high diatom content (microscopic observation; Fig. 2). A tephra deposit is observed at 397.5 cm top and 401.4 cm bottom. The bottommost section corresponds to a succession of millimetric very fine sand layers related to glaciofluvial processes. Focusing our attention on the core’s uppermost 185 cm, we found hemipelagites and turbidities. HemipelagitefaciesinLagoFagnanoarecharacterizedbyolive and dark olive laminations that exhibit a homogeneous clayeygrain size.Thereafter,weconsideredalaminatobe a pair of olive and dark olive clay sediment layers. The uppermost185 cmrepresentthebest-preservedrecordof hemipelagic sediments in the core, and they contain 275 laminae (visual count). The mean thickness of each lamina is ~0.65 cm, and they range from several millimetres to several centimetres (Figs 2, 4, 5). Hemipelagic sediments are characterized by MS values between 2 and 10 SI and densities between 1.2 and 1.35 g cm 3 . There are four exceptions at 46, 66, 108 and 112 cm depth, with densities of 1.1, 0.6, 0.6 and 0.9 g cm 3 , respectively. Three thin (~2 cm) turbidite deposits interrupt the hemipelagic sedimentation at the core top, at 42 and 127 cm. At 185 cm, a massive (182-cm-thick) multievent debrite divides the sediment core’s upper interval, dominated by hemipelagic sediments, from its lower one, dominated by mass transports deposits (Fig. 2). Turbidites are olive-grey clayey to fine sand layers characterized by a fining sequence upwards, sharp bases and density values around 1.5 g cm 3 and MS values vary from 5 to 12 SI. Age model We analysed four radiocarbon samples to obtain a chronological framework (Table 1). However, only the uppermost two samples, which correspond to pieces of wood deposited within the hemipelagic interval at 2.5 and 136 cm depth (~170 and ~4620 cal. a BP, respectively) were used for the age model, since the lowest two samples were obtained from debrites (Table 1). Finally, at ~400 cm we found a deposit corresponding to the Hudson H1 tephra, whose age and composition are very well constrained in the region (Figs 2and 6; the mean age of the various lakes is 8434 cal. a BP; Stern et al.2016). Because of the complexity of the stratigraphy and the lack of datable sediments, we could develop only a basic agemodelforthecore’suppermost185 cm.Basedonour frail or poorly constrained age model, the average rate of hemipelagic sedimentation in the Bahia Grande basin was~30 cm ka 1 ,whiletherateintheadjacentdeepbasin of Lago Fagnano was 18 cm ka 1 (Moy et al.2011). Geochemical proxies Isotopic composition of organic matter. – In core LF06PC8, hemipelagic sediments show C/N ratios between 8 Depth (cm) Depth (cm) Ln Mn/AlLn Fe/AlLn Ti/Al A B 2 1 0 3 2 1 5 6 4 3 5 4 118 128 138 148 158 168 178 118 128 138 148 158 168 178 Fig. 4. Scanned image of core LF06-PC8 showing laminations and the greyish diatom layers. A. Ln Mn/Al ratio plotted on the image to illustrate the connections between laminations and geochemistry on a logarithmic scale (Weltje et al.2015). B. Ln Ti/Al and Ln Fe/Al ratios, showing that several enrichments are not overlapped. Colour bars indicate Ln Mn/Al ratio (grey bars) and Ln Fe/Al ratio (orange bars) paired enrichment, interpreted as preserved palaeo-redox fronts. BOREAS Formation of laminations in sediments in Lago Fagnano, Argentina over the past 6.3 ka 129 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License and 10 (Fig. 7). Light layers have average values of C/N, d 13 C, and d 15 N of 10.11.7, 25.40.6&,and 2.90.5&, respectively, whereas dark layers have mean values of 9.70.6, 25.20.3&, and 30.4&, respectively. Therefore, we did not observe significant differences between the light and dark layers. Turbidites exhibit higher C/N peaks, and the C/N ratios associated with these C/N peaks range from 12 to 20 (Fig. 7). The d 13 C and d 15 N values (26.0 to 27.5&and 1.3 to 2.3&, respectively) decrease relative to the rest of the sediment facies. This is especially evident in the turbidites at ~44 and ~125 cm (Fig. 7). Major elements. – The relative abundance of major elements can be used to infer sedimentary processes as well as diagenetic phenomena. We differentiated two groups of major elements. First, Al, Ti and Ca exhibit correlation coefficients close to 1 (Fig. 3), suggesting a relationship between Ca and the detrital fraction, probably caused by low carbonate productivity as observed in other lacustrine systems (Mesa-Fern andez et al.2018). The second group, Mn and Fe, relates to redox processes occurring within the sediments rather than with detrital input, in both the lacustrine and marine records (e.g. Davison et al.1982; Naeher et al.2013; JimenezFig. 5. Progressive cross-sections (south to north) of Bah ıa Grande in Lago Fagnano, indicating the stages of lamination formation and detailed schemesofchemicalreactionsatthewater–sedimentinterface. A, A’.Initial stage with mixed andventilatedlakewateroccurring duringwinter (low temperatures and low SHWW), and initial conditions of C/N, Ti and Mn/Al. B, B0. Laminae production with dysoxic/anoxic hypolimnion when warmconditionsprevail(probablyassociatedwithintense SHWWorhigherinsolation) anddecreaseinC/NandTi and certain increasesin Mn and Fe. C, C0. Third stage corresponds to a normal winter, with mixed lake water and an absence of hypolimnion, and C/N decreases to similar levels to stage A’. High enrichment in Mn at the boundary between dark and light lamina (see article for details). C″. Eventual persistence of C0stage conditions. During this stage oxygenatedwaters originate a redox front. The oxidant front penetrates downwards (dashed orange line), forming the lower Mn/Al peak. D, D0. Similar conditions to (B) stage. E, E0. Mass transport event with higher C/N ratio, as upland plants are contained in the sediment, and higher Ti owing to more terrigenous influence. F, F0. Diatom blooms associated with the increase in nutrients promoted by a mass transport event; the diatom layer has a high Si content. 130 Alexis Vizcaino et al. BOREAS 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Espejo et al.2020). Additionally, we conducted a PCA on the XRF data for only the laminated intervals to obtain a more detailed insight into the main factors governing the elemental composition. The first two principal components (PCs) explain 70.9% of the variance (Fig. 3B). PC1 explains 53.4% of the total variance, showing high scores for Al, Si, K, Ca and Ti, negative for Mn and Ba, and strongly negative for Fe (Fig. 3C). This evidence indicates that the Fe record at Lago Fagnano is not associated with siliciclastic elements but with other mineral phases. Non-detrital Fe-enriched phases are common redox-sensitive minerals, typically pyrite and iron oxides, and their presence is confirmed by previous millimetre-scale studies in this core (Neugebauer et al.2022). PC2 (17.5% of total variance) shows a high score for Ca, S and Ba, slightly positive for Mn and Fe, andnegativefor Al,Siand K(Fig. 3D). PC2points again to the lackof covariance between Fe and Mn and detrital elements such as Al and K. To correct for dilution effects, Mn and Fe concentrationswerenormalized to Al. This normalization assumes that the Al in sediments is contributed by aluminosilicates (Calvert 1990). Most Mn/Al enrichments are coincident with Fe/Al enrichments (Fig. 4) and commonly found in the transition from dark to light layers or within the thin dark layers between light laminae (Fig. 4). Tephra analyses. – Core LF06-PC8 contains a tephra at ~400 cm, and in the studied core this level is characterized by the presence of andesitic volcanic glass enriched by FeO and TiO 2 (Fig. 6). Selected proxies and interpretation Provenanceand deliverymechanismswereinferredusing bulk organic matter d 13 C and d 15 N values, as well as elemental data derived from XRF scanning. The C/N ratio is commonly usedto differentiate betweenalgal and terrestrial sources of the organic matter in lacustrine sediments (Meyers & Teranes 2001). Terrestrial organic matter derived from vascular land plants typically has C/ N atomic ratios greater than 20, whereas lacustrine algae typically have C/N values lower than 10 (Meyers 2003). Because lacustrine organic matter is a mixture of terrestrial and aquatic organic matter, C/N ratios may be used to identify the relative contributions of these two endmembers to the sediment (Meyers & Teranes 2001). Moreover, grain size may exert control on the C/N ratios (Ghazoui et al.2019), and d 13 C and d 15 N can be considered as indicators of productivity (Meyers & Teranes 2001). Nevertheless, it must be noted that the sampling intervals inherent to each of these three methods are different: while d 13 Candd 15 N values (at 1-cm intervals) are useful to identify centennial variation, XRF core-scan data (at 2-mm intervals) potentially provide decadal-scale resolution. Discussion This palaeoenvironmental study of Lago Fagnano focused on the uppermost 185 cm of core LF06-PC8, corresponding mostly to hemipelagic sediments. Sedimentary processes Tephra and mass transport deposits. – The tephra layer analysed has been geochemically correlated to the major Early Holocene eruption H1 of the Hudson volcano (Table 2, Fig. 6). Glass from the Hudson event is distinct from other large eruptions of the Andean Volcanic Zone as its andesitic composition (rich in FeO and TiO 2 ) differs from that of the rhyolitic glass associated with other tephra (Fig. 6) (Stern 2008). Glass from the Fig. 6. Elemental composition of glass from major late Quaternary volcanic eruptions of the Andean Volcanic Zone (Naranjo & Stern 1998; Stern 2008). Plot showing the weight percents of: (A) SiO 2 vs. K 2 O and (B) TiO 2 and FeO. In red, we show the analysed glass in this study. BOREAS Formation of laminations in sediments in Lago Fagnano, Argentina over the past 6.3 ka 131 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License Hudson 1 tephra is also notable for its relatively high K 2 O content (Stern 2008). The Late-Holocene Aguilera tephra is the only regional major Holocene event to contain higher K 2 O values. Turbidites and debrites are sedimentary deposits resulting from rapid and ephemeral mass transport processes (Stow & Zeinab 2020). In the Bah ıa Grande sediment core’s uppermost 185 cm, the only mass transport deposits observed correspond to turbidites exhibiting higher C/N peaks, indicative of terrestrial organic matter (C/N values 12.42.5), compared with more‘algal’hemipelagites (9.91.1). All turbidites show high C/N ratios and significantly lower d 13 C and d 15 N, influenced byorganic matter from littoral areas, from the catchment basin or associated with grain size variations described in turbidite layers (Fig. 7). Diatom-rich layers occur above each of the turbidites, probably related to an increase in nutrient availability in the lake water following mass transport events (Mackay et al.1998). The age of this uppermost turbidite is compatible with the historic 1949 Fagnano earthquake. Situated along an active plate boundary with regular and strong seismic events, earthquakes areaprobabletrigger mechanism for mass transport episodes (Costa et al.2006; Waldmann et al.2008). The synchroneity found between this turbidite and the seismic record opens the door to more palaeoseismic studies in the region. Hemipelagic sediments: formation of laminations. – The hemipelagic interval is characterized by conspicuous millimetric lamination of olive and dark olive clays. In coreLF06-PC8weinferredanalgal originforthe organic matter in the hemipelagic facies, based on its C/N ratio (Fig. 7), but we did not observe this distinction between darker and lighter levels. However, since the isotopic composition between light and dark layers is highly homogeneous, we used XRF data to characterize the hemipelagic deposits and to describe the laminations. Al, Ti, and Ca variability exhibit high correlation coefficients between each other (Fig. 3), but the Ti/Al ratio shows a weak relationship to the dark/light laminations (Fig. 4). In contrast, Mn and Fe are enriched in the dark laminae (Fig. 4). Most Mn/Al enrichments are coincident with Fe/Al enrichments (Fig. 4) and commonly occur at the transition between dark to light layers or in thin dark layers between light laminae (Fig. 4; Neugebauer et al.2022). Also, conspicuous differences are discernible with depth regarding the content of Mn and Fe at the transitions between a dark and a light layer. SuchMn–Fe oxides typicallyreflect pastlakehypolimnic redox conditions (Davison et al.1982; Eusterhues et al.2005; Rush 2010; Kasper et al.2013;Dr € ager et al.2019). ReactiveMnis deliveredtothe lakebyrainfall andalso by diffusion from coastal sediments. In addition, dissolved Mn +2 is present in aquatic systems in low concentrations. However, changes in redox conditions led to a change in dissolved manganese (Mn 2+ ) and iron (Fe 2+ ) content within the water column and/or pore waters, with consequent precipitation of Fe and Mn oxides (Fig. 5C0; e.g. Schaller & Wehrli 1997; Pakhomova et al.2007). Similar models have been proposed for other lakes (e.g. Schaller & Wehrli 1997; Haberzettl et al.2006). In thick dark laminae the Mn/Al enrichments often exhibit double peaks (Fig. 5A). This pattern can be explained by a redox front associated with betterventilated deep water entering the basin after sediments havebeen deposited underlesswell-ventilatedconditions (Fig. 5C; Rutten et al.1999; Mangini et al.2001). The distribution of Mn and Fe within the laminated sediment succession can be explained by episodic anoxic/ oxichypolimnion changes(Fig. 5), asdescribed forother lakes worldwide (Dean 1997;Dr € ager et al.2019). Low levels of primary production in Lago Fagnano and/or poor organic matter preservation (Mucciarone 2021) suggest that these variations cannot be explained by biological processes alone. Rather, they are best explained by periodic oxygenation of the hypolimnion responding to energetic water column mixing and associated deep lake ventilation (Neugebauer et al.2022). The same link between Mn variations and water oxygenation has been described for another Patagonian lake (Jouve et al.2013) as well as for lacustrine (Schaller et al.1997; Eusterhues et al.2005; Rush 2010; Naeher et al.2013;Dr € ager et al.2019) and marine environments (e.g. L€ owemark et al.2008;De Lange et al.2008; Jimenez-Espejo et al.2020). This differential variation in Mn and Fe preservation within a lacustrine basin has been described in other regions and linked to variations in local conditions and sedimentation rates (e.g. Granina et al.2004). Taken together, the geochemical data indicate that laminations in the upper sediment column of Lago Fagnano result from the presence of a variable redox boundary within the water column. In lacustrine systems such redox boundaries are typically associated with incomplete water column mixing and consequent hypolimnetic anoxia/dysoxia (e.g. Naeher et al.2013; Dr€ ager et al.2019). In lakes such as Lago Fagnano, water column mixing is caused by a numberof processes, such as wind-generated wave forcing and cooling. The sediments forming under these well-oxygenated conditions are light in colour. In contrast, dark laminae are indicative of poor ventilation. These laminations are probably controlled by winter climatic conditions, yet warmer and calmer periods during summers might also generate water column stratification and increase biological production, a model similar to other locations (e.g. Davies et al.2004; Naeher et al.2013; Makri et al.2021). We note that the presence of persistently stratified conditions has not been described in the few water column studies conducted at Lago Fagnano (Mariazzi et al.1987). To test our interpretation of laminae origin related to winter climatic conditions, a 132 Alexis Vizcaino et al. BOREAS 15023885, 2023, 1, Downloaded from https://onlinelibrary.wiley.com/doi/10.1111/bor.12600 by Csic Organización Central Om (Oficialia Mayor) (Urici), Wiley Online Library on [20/02/2024]. See the Terms and Conditions (https://onlinelibrary.wiley.com/terms-and-conditions) on Wiley Online Library for rules of use; OA articles are governed by the applicable Creative Commons License