COMPARING METEORIC 10BE, IN SITU 10BE AND NATIVE 9BE ACROSS THREE WATERSHEDS

By measuring ¹⁰Be produced in rock and soil (in situ ¹⁰Be - ¹⁰Be_is), one can obtain average denudation rates over millennia. However, ¹⁰Be_is analysis is methodologically limited to locations where quartz-bearing rocks crop out and to systems where one can isolate sand-size quartz grains. In other systems, meteoric ¹⁰Be (¹⁰Be_met) could provide a similar tool, but pedogenic processes, soil chemistry, and grain size control concentrations. Because ⁹Be weathered from crustal rocks has the same reactivity as ¹⁰Be_met, others have hypothesized that ¹⁰Be_met/⁹Be ratios of grain coatings normalize the ¹⁰Be_met inventory to remobilization and grain size dependency. To test this hypothesis, we determined the relationship between ¹⁰Be_met, ¹⁰Be_is and ⁹Be by extracting ⁹Be from three sets of fluvial sand samples previously measured for ¹⁰Be_met and ¹⁰Be_is.

We leached ⁹Be from 206 samples by reacting powdered sands in 6M HCl for 24 hours while reacting in a heated ultrasonic bath – a method that did not leach ⁹Be in mineral grains. This is significant because ⁹Be in grain coatings is subject to the same post-depositional influences as ¹⁰Be_met, while ⁹Be incorporated into mineral matrixes is uninfluenced by post-depositional conditions, similar to ¹⁰Be_is. The leachate was analyzed by ICP-OES.

Samples show different ¹⁰Be_met /⁹Be correlations with ¹⁰Be_is across sampling locations. Samples from the Mekong River watershed in China have a positive correlation between ¹⁰Be_is and ¹⁰Be_met (R²=0.46, n=119) and a stronger positive correlation between ¹⁰Be_is and ¹⁰Be_met/⁹Be (R²=0.69). Samples from the Barron River Watershed in Queensland Australia have no correlation between ¹⁰Be_is and ¹⁰Be_met (R²=0.05, n=16) but a positive correlation between ¹⁰Be_is and ¹⁰Be_met/⁹Be (R²=0.56). Samples from the Potomac River Watershed in east-central North America show no correlation between ¹⁰Be_is and ¹⁰Be_met (R²=0.06, n=71) or ¹⁰Be_is and ¹⁰Be_met/⁹Be (R²=0.00). Considering all data, correlations between ¹⁰Be_met/⁹Be ratios and ¹⁰Be_is (R²=0.70, n=206) are only slightly stronger than correlations between ¹⁰Be_is and ¹⁰Be_met (R²=0.66). These results suggest that while measuring ⁹Be can improve correlations between ¹⁰Be_met and ¹⁰Be_is, further study is needed to understand the variables that influence Be dynamics across a diversity of watersheds.