(234U)/(238U) DISEQUILIBRIUM AND TRACE ELEMENTS IN GROUNDWATER THROUGH TIME: A SPELEOTHEM RECORD FROM SOUTHERN ILLINOIS
There are few methods for examining water-rock interaction over geologic time scales. One paleohydrologic record comes from speleothems. Here we present combined U-series chronology and geochemistry of a 65 cm long stalagmite from Fogelpole Cave in Southern Illinois. The host cave lies within the St. Louis limestone and is overlain by Illinoisan glacial sediments. We characterized growth fabric using petrography and CL and determined 238U-234U-230Th and 235U-231Pa ages by Thermal Ionization Mass Spectrometry. Ages for the two independent systems are concordant for 7 samples, range from ~80,000 years to ~30,000 years and progressively decrease up the sample. The growth rate of the stalagmite decreases significantly through time from ~230mm/yr to ~3 mm/yr. Despite changing growth rates, U concentrations are continuously high (40-60 ppm) while Th concentrations are low (~ppb). d 18O ( 4.5~ 3 per mil) and d 13C ( 4.5 ~ 2.5 per mil) do not appreciably vary, suggesting fairly stable vegetation over this time. This contrasts with data from nearby Crevice Cave (Dorale et al., 1998) that indicate significant shifts between forest and prairie vegetation at this time.
The initial activity ratio for 234U/238U (d 234U0) varies from -210 to -70 and correlates well with the growth rate of the stalagmite and Mg concentrations. Both d 234U0 and Mg likely reflect changes in water-rock interaction in the overlying sediments. However, 87Sr/86Sr, another indicator of water-rock interaction, shows no variation over the length of the stalagmite.
An examination of the global database shows that speleothem growth rates often correlate withd 234U0. Most studies find d 234U0 > 0 that increases as growth rate decreases. In contrast, our stalagmite shows d 234U0 < 0 that decreases with decreasing growth rate. A self-consistent explanation for these observations is that changes in d 234U0 reflect mixing of U from two endmember components: 1) mineral interiors having d 234U0 close to zero and 2) mineral surfaces having d 234U0 far from equilibrium. During wetter conditions greater amounts of mineral dissolution occurs so that d 234U0 is closer to zero compared to drier conditions where surfaces dominate. If growth rate reflects paleo-precipitation, variations in d 234U0 through time may provide another important record of paleoclimate change.