STRATEGIES FOR SUCCESSFUL U-TH DATING OF PALEOLAKE CARBONATES: AN EXAMPLE FROM THE BOLIVIAN ALTIPLANO

We report over 90 U-Th dates from lacustrine carbonates deposited around four paleolakes on the Bolivian Altiplano, that range in age from 120 to 10 ka. Petrographic and chemical data for carbonate tufas and the silicate detritus contained within them allow (1) assessment of possible diagenetic effects (2) a successful strategy for selection of carbonate samples with low initial Th contents, and (3) assessment of the uncertainty due to initial Th. This strategy allows us to produce precise and accurate U-Th dates from lacustrine carbonates

Carbonate textures are evaluated for potential diagenetic effects, and such alteration is minimal. Petrographic thin sections of carbonates from the Bolivian Altiplano appear generally free from secondary carbonate or visible alteration and pore spaces are free from secondary calcite. The most conservative argument for a closed system with respect to U and Th and the validity of these dates is their reproducibility at numerous locations and with diverse carbonate textures.

The principal consideration in dating lacustrine carbonates is the composition and quantity of initial Th incorporated into the carbonate. Lacustrine deposits present special challenges to U-Th dating because they almost always require a correction for initial Th. Additional uncertainty is introduced because this initial Th may have two sources: 1) a siliciclastic fraction with a chemical composition that approximates that of the upper continental crust, and 2) a component with “unsupported” or “hydrogenous” Th with an elevated ²³⁰Th/²³²Th ratio. Isochron plots, measured ²³⁰Th/²³²Th ratios, and X-ray diffraction and trace element chemistry of silicic residues all favor regional soil and dust as the sole source of initial Th in carbonates from the Bolivian Altiplano. This situation simplifies single-sample dating of Altiplano carbonates. In contrast, several global paleolakes where the surrounding terrain is dominated by carbonate bedrock have initial ²³⁰Th/²³²Th activity ratios above 2. Such samples present a much more challenging situation for single-sample dating compared to Bolivia. Paleolake systems hosted in non-carbonate terrains and with relatively low U concentrations may be good candidates for future U-Th dating efforts, as carbonates from such lakes may be free of initial unsupported Th.