Research paper Uranium loss and aragoniteecalcite age discordance in a calcitized aragonite stalagmite Matthew S. Lachniet a, * , Juan Pablo Bernal b , Yemane Asmerom c , Victor Polyak c a Department of Geoscience, Mailstop 4022, University of Nevada Las Vegas, 4505 Maryland Parkway, Las Vegas, NV 89154, USA b Centro de Geociencias, UNAM, Campus Juriquilla, Blvd. Juriquilla 3001, Juriquilla, Querétaro 76230, Mexico c Department of Earth and Planetary Science, University of New Mexico, 200 Yale Blvd. NE, Albuquerque, NM 87131, USA article info Article history: Received 20 December 2011 Received in revised form 21 May 2012 Accepted 12 August 2012 Available online 21 August 2012 Keywords: Speleothem Aragonite Recrystallization Uranium series dating Laser ablation MC-ICPMS abstract We analyzed uranium-series concentrations and isotopic ratios in a mixed aragonite and calcite stalagmite from Juxtlahuaca Cave, from the Sierra Madre del Sur of Mexico. The U-series data for the aragonite layers return highly precise and stratigraphically correct ages over the past ca. 4300 years. In contrast, age determinations from calcite layers are too old by several hundred years relative to the precise aragonite ages, have analytical uncertainties an order of magnitude larger than aragonite ages, and yield ages that do not overlap the aragonite ages within analytical uncertainties. Based on geochemical and petrographic observations, we interpret the calcite layers to have formed from recrystallization of aragonite soon after primary aragonite deposition. Calcite occurs as discontinuous lenses on and off the growth axis, and laminae can be traced between aragonite and calcite layers, demonstrating that visible growth banding is not effaced in the recrystallization process. Paired aragonite-calcite U-series data from coeval stratigraphic layers demonstrate that uranium concentrations decrease by two orders of magnitude during calcitization, and result in decreased ( 234 U/ 238 U). Uranium loss during diagenesis mimics a need for an age correction using an initial 230 Th/ 232 Th ratio one to two orders of magnitude larger than the bulk Earth ratio of 4.4  2.2  10 6 . A need for apparent high initial 230 Th/ 232 Th ratios results from ingrowth of 230 Th during 234 U decay. Ó 2012 Elsevier B.V. All rights reserved. 1. Introduction 1.1. Calcitization of aragonite in speleothems The co-occurrence of aragonite and calcite in speleothems has been commonly observed (Frisia et al., 2002; Gonzalez and Lohmann, 1988; Ortega et al., 2005; Railsback et al., 1994; Woo and Choi, 2006). Mixed aragonite/calcite mineralogy has also been investigated under controlled laboratory conditions (de Choudens- Sánchez and González, 2009). In some stalagmites, the replace- ment of aragonite with calcite is clearly documented (Frisia et al., 2002; Spötl et al., 2002). An understanding of U-series systematics in aragonite speleothems is important because selected stalagmites have been shown to precipitate in oxygen isotopic equilibrium with drip waters (Li et al., 2011), may be well-calibrated to modern climate variability (Lachniet et al., 2012), and are increasingly being used to develop high-resolution and precisely dated paleoclimate records. However, because aragonite is less stable than calcite at typical cave temperatures, its recrystallization to calcite (calcitiza- tion) is possible if sufficient conditions are met. Some calcite spe- leothems recovered from caves that are currently precipitating aragonite may have been altered by aragonite calcitization. Evaluation for diagenetic calcite in speleothems is required prior to U-series analysis because such replacement invalidates the assumption of closed-system U-series behavior (Richards and Dorale, 2003). This is because uranium is highly soluble in precipi- tating fluids as the UO 2þ 2 uranyl ion complexes (Langmuir, 1978; Sandino and Bruno, 1992), whereas thorium has a low solubility (Langmuir and Herman, 1980; Ryan and Rai, 1987; Vandenborre et al., 2010). Uranium loss during calcitization of aragonite produces anomalously old ages (Hoffmann et al., 2009). If the open- system nature of the affected samples is not recognized from inde- pendent evidence for recrystallization (e.g., thin section micro- morphology), the ages could be easily misinterpreted. Correction of such ages to reflect the ‘true’ ages would require initial 230 Th/ 232 Th 1 ratios higher than those used for the unaltered samples. Typically, an * Corresponding author. Tel.: þ1 702 895 4388; fax: þ1 702 895 4064. E-mail addresses: matthew.lachniet@unlv.edu (M.S. Lachniet), jpbernal@ geociencias.unam.mx (J.P. Bernal), asmerom@unm.edu (Y. Asmerom), polyak@ unm.edu (V. Polyak). 1 Throughout the paper, activity ratios are given in parentheses, whereas mass (isotope) ratios are denoted without brackets. Contents lists available at SciVerse ScienceDirect Quaternary Geochronology journal homepage: www.elsevier.com/locate/quageo 1871-1014/$ e see front matter Ó 2012 Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.quageo.2012.08.003 Quaternary Geochronology 14 (2012) 26e37