PII S0016-7037(00)00541-X Apparent gibbsite growth ages for regolith in the Georgia Piedmont PAUL A. SCHROEDER, 1, *NATHAN D. MELEAR, 1 PAUL BIERMAN, 2 MICHAELE KASHGARIAN, 3 and MARC W. CAFFEE 3 1 Department of Geology, 210 Field St., University of Georgia, Athens, Georgia 30602-2501, USA 2 Department of Geology, Perkins Hall, University of Vermont, Burlington, Vermont 05405-0122, USA 3 Lawrence Livermore National Laboratory, 7000 East Ave., Livermore, California 94550-9234, USA (Received December 29, 1999; accepted in revised form September 14, 2000) Abstract—–Carbon bound in gibbsite, collected from a residual weathering profile developed on a Paleozoic granite in the Georgia Piedmont, was examined for its 14 C content and found to be geologically young. The study site, located at the Panola Mountain Research Watershed, has developed a granite–saprolite–soil regolith in which 14 C-gibbsite model ages deep within the profile (C-horizon) average about 8000 years. Near the surface (A- and B-horizon) 14 C-gibbsite model ages range from 2100 to 4200 years. Quartz has acquired 26 Al and 10 Be inventories suggesting a near-surface residence time of at least 90,000 years. This age disparity supports the notion that secondary minerals undergo significant recrystallization as weathering fronts prop- agate into the landscape. Combining the results of 14 C, 26 Al, and 10 Be analyses offers the potential to assess differential rates of chemical weathering and continental denudation to understand better the links between rates of silicate rock weathering, climate, and soil residence times. Copyright © 2001 Elsevier Science Ltd 1. INTRODUCTION Global rates of chemical weathering are intimately linked to the denudation of silicate terrains in humid-temperate to trop- ical climates (Berner and Berner, 1997). This link provides an important feedback to the abundance of atmospheric CO 2 (i.e., climate) as bicarbonate production returns carbon to the oceans (Berner, 1994). One valuable approach to constraining land surface denudation rates (mass loss per unit time) is the use of in situ produced cosmogenic nuclides such as 10 Be and 26 Al (e.g., Bierman, 1994; Granger et al., 1996; Nishiizumi et al., 1991). For example, measuring the abundance of these isotopes in granites and granitic saprolite allows for an assessment of the length of time over which a weathering profile has been ex- posed to the surface (Bierman et al., 1995; Bierman, 1993). Near-surface mineral residence times, rates of mineral dissolu- tion, and colluvial export rates are key parameters that are needed to understand the connection between climate change and landscape change. One facet of the complex process of landscape evolution and denudation that has been difficult to measure is the timescale over which secondary minerals form. Knowledge about the duration and timing of secondary mineral formation offers insights into the pathways of mineral genesis and the “meta- bolic” properties of weathering profiles (e.g., long-term soil respiration rates and estimates of mass loss). This article intro- duces a new approach for estimating the relative age of sec- ondary mineral authigenesis with the intent of better under- standing models for the breakdown of silicate regolith in humid-temperate to tropical climates. 2. MATERIALS AND METHODS A soil–saprolite–rock sequence was collected from the Panola Mountain Research Watershed (PMRW) operated by United States Geological Survey (USGS) in the Georgia Piedmont of the southeast- ern United States (Huntington et al., 1993; Stonestrom et al., 1998) (Fig. 1). The bedrock is dominated by the Panola Granite, which has a very homogeneous, medium-grain, biotite– oligoclase– quartz–micro- cline composition with no discernible foliation (Melear, 1998). Sam- ples were collected at a summit site in the PMRW where the weath- ering profile is believed to be wholly residual (i.e., not colluvial). The same sample suite was recently studied for its mineralogical, chemical, and stable carbon isotopic properties (Schroeder and Melear, 1999; Melear, 1998). Samples analyzed for 10 Be and 26 Al were taken from a soil pit at depths ranging from 2 to 10 cm below the surface. Samples were chemically and physically separated at the University of Vermont to produce 40 g of pure quartz by using a technique modified from Kohl and Nishiizumi (1992). Subsequent treatment involved HF dissolution of the quartz and the preparation of BeO and Al 2 O 3 targets for AMS analysis (Bierman and Gillespie, 1997). All targets were analyzed by Accelerator Mass Spectrometry (AMS) at the Lawrence Livermore National Laboratory (Davies et al., 1990). Procedures for extraction of CO 2 from the soil minerals, calcite and goethite, for paleoclimate studies have been developed and applied by Cerling (1984) and Yapp and Poths (1991). More recently, it was recognized that gibbsite sequesters carbon as it crystallizes in the soil–saprolite. This carbon is released as CO 2 as the gibbsite thermally breaks down at 230°C to 240°C under vacuum in the laboratory. We used a modified version of the method developed by Yapp and Poths (1991) and Schroeder and Melear (1999). Five samples were selected for 14 C measurements. Pretreatments included particle size separation (2 m retained), repeated digestions in 30% H 2 O 2 and low-temperature (200°C) O 2 combustion to remove organic matter. Stable carbon isotopic analyses were conducted at the University of Georgia Stable Isotope Facility. CO 2 produced from the breakdown of gibbsite was cryogenically trapped into break-seal tubes and sent to the LLNL AMS facility, where graphite targets were prepared and analyzed. Line blanks were run to ascertain the contri- butions of 14 CO 2 from other phases such as kaolin group minerals and from the extraction line itself. The 8720-year age of the kaolin blank has limited meaning because negligible gas was extracted (0.28 mol CO 2 g -1 ) compared with most samples. The low CO 2 yield is consis- tent with previous work by Schroeder and Melear (1999) who showed kaolin group minerals produced extremely little CO 2 during the 230°C treatment. The apparent 14 C age of 45,710 years for the marble indi- cates that the extraction line is effectively devoid of external sources of 14 C. Sample backgrounds have been subtracted on the basis of the measurements of the 14 C-free marble. Backgrounds were scaled by relative sample size. *Author to whom all correspondence should be addressed (schroe@gly.uga.edu). Pergamon Geochimica et Cosmochimica Acta, Vol. 65, No. 3, pp. 381–386, 2001 Copyright © 2001 Elsevier Science Ltd Printed in the USA. All rights reserved 0016-7037/01 $20.00 + .00 381