14-7 Detrital Contributions to Lake Junín (Peru) During Glacial Periods Revealed by Clumped Isotopes: Application of a Two-Point Δ₄₇ Mixing Model to Mixed-Source Carbonate and Implications for Isotope-Based Paleoclimate Reconstructions

Session: Lake Sedimentary Records of Past Climate and Environment

Presenting Author:

Sarah Katz

Authors:

Katz, Sarah A.¹, Levin, Naomi E.², Passey, Benjamin H.³, Rodbell, Donald T.⁴, Abbott, Mark B.⁵, Wostbrock, Jordan A.G.⁶, Smith, Brianna⁷, Rao, Anita⁸

(1) Department of Earth and Planetary Sciences, Yale University, New Haven, , (2) Department of Earth and Planetary Sciences, University of Michigan, Ann Arbor, , (3) Department of Earth and Environmental Sciences, University of Michigan, Ann Arbor, , (4) Geoscience Department, Union College, Schenectady, , (5) Department of Geology and Environmental Science, University of Pittsburgh, Pittsburgh, , (6) Department of Earth and Planetary Sciences, Yale University, New Haven, , (7) Department of Earth and Environmental Sciences, University of Michigan, Ann Arbor, , (8) Department of Earth and Environmental Sciences, University of Michigan, Ann Arbor, ,

Abstract:

Lacustrine carbonate oxygen isotope records are a powerful tool for reconstructing terrestrial hydroclimate. However, this approach often assumes lake carbonates are entirely authigenic, which is not always valid, especially if the lake catchment contains sources of detrital carbonate (e.g., carbonate bedrock). Although detrital carbonates in lake sediment records hinder isotope-based paleoclimate interpretations, it can be difficult to physically separate or screen for these contaminants using mineralogical or sedimentological means. Therefore, detrital carbonate is unaccounted for in many studies. Here, we demonstrate an approach to screen for detrital carbonate using carbonate clumped isotopes (Δ₄₇) and a two-endmember mixing model which accounts for non-linear Δ₄₇ mixing. We validate the model using a synthetic dataset and show that the proportion of bedrock within a bulk sample can be accurately estimated from bulk Δ₄₇ values. Further, in some cases (e.g., when bedrock fraction is low and bedrock δ¹⁸O is precisely known), it may be possible to estimate the δ¹⁸O value of authigenic carbonate from the bulk δ¹⁸O value.

 

We then apply this approach to cores from Lake Junín, Peru. At Lake Junín, the clastic abundance within the bulk sediment is highest during glacial periods and minimal during interglacial periods. This implies that carbonates within the glacial intervals may be partially derived from limestone bedrock in the catchment. Local bedrock has an average Δ₄₇ value of 0.484 ‰ (ICDES90), corresponding to a formation temperature of ~71°C, while modern authigenic carbonate from Lake Junín and other nearby lakes have average Δ₄₇ and temperature values of 0.638 ‰ and ~11°C, respectively. We use these values to represent model endmember conditions for Lake Junín. For Lake Junín glacial sediments (e.g., Last Glacial, MIS 14, MIS 16), some samples have cooler Δ₄₇-derived temperatures than modern lake carbonates, which we interpret as a primary climate signal. However, other glacial samples reflect warmer Δ₄₇-derived temperatures (~20-50°C), which contradicts our expectations of local temperatures at the time. We interpret these anomalously warm samples as a mixture of authigenic and detrital carbonate, with the latter comprising up to ~60% of the total carbonate in the sample. Beyond application of this approach at Lake Junín, this framework can be readily adapted to evaluate detrital contaminants and improve the accuracy of paleoclimate reconstructions in other modern and ancient lake systems.

© Copyright 2025 The Geological Society of America (GSA), all rights reserved.