Thallium isotope cycling between waters, particles, and sediments across a redox gradient

Abstract

Thallium (Tl) isotopes are an emerging tool for tracking the history of molecular oxygen in seawater. Use of Tl isotopes in this manner requires a thorough understanding of the modern Tl isotope cycle, especially within the anoxic settings that typified Earth’s past oceans. To this end, we generated Tl concentration and isotope data for waters, particles, and sediments collected during three spring-summer-fall field seasons in Siders Pond, a salt-stratified and sulfide-rich (‘euxinic’) pond on Cape Cod (Massachusetts, USA). Over short timeframes (i.e., days to weeks), we observed a large degree of variability in the abundance (Tldiss = 9 pM–42 pM) and isotopic composition (ε205Tldiss = –3.9 ± 0.4; 2SD to –0.1 ± 0.7; 2SD) of dissolved Tl in pond surface waters. Thallium drawdown was common and favored removal of the lighter-mass Tl isotope. We surmise that biological Tl uptake plays a role in this phenomenon, but more work is warranted to confirm this hypothesis. Manganese oxides persisted in pond surface waters on many sampling days but had no obvious effect on Tldiss, only a slight effect on particulate ε205Tl values on the day of most prolific Mn oxide accumulation (driving ε205Tlpart as high as +0.9 ± 0.8; 2SD). In euxinic waters below the oxycline, rapid drawdown of dissolved Tl was observed on all sampling days together with a complimentary increase in particulate Tl (up to Tlpart = 21 pM). Strong associations between Tl and sulfide, and between Tl and other chalcophile metals (Mo and Cd), suggest that Tl is probably removed from euxinic waters in association with particulate sulfides, and seemingly with no resolvable isotopic fractionation effect despite non-quantitative Tl removal. The sediment data track the longer-term Tl cycle in the pond (i.e., across years) and reveal limited isotopic variability. Thallium isotope compositions leached from surface sediments match those predicted for contemporaneous waters, or nearly so, at all depths in the pond (above, within, and below the oxycline). Apparently, only a small fraction of the short-term Tl isotope variability found in the oxic surface waters of Siders Pond gets transferred to anoxic waters, and essentially none is transferred to sediments. Our results reaffirm the notion that Tl isotopes are uniquely capable of tracking long-term sedimentary Mn oxide burial – not mere Mn oxide formation. Our results also verify the ability of sediments formed under reducing conditions to capture the Tl isotope composition of contemporaneous waters. More broadly, the results of our investigation bolster our knowledge of the modern Tl isotope cycle and thereby permit more confident inferences of this cycle in Earth’s past.