Carbonate weathering, phosphate fertilizer, and hydrologic controls on dissolved uranium in rivers in the US Corn Belt: Disentangling seasonal geogenic- and fertilizer-derived sources.

Soil and bedrock weathering and phosphate (P) fertilizers may both contribute to the uranium (U) load of rivers in agricultural regions, but controls over their relative influence are not well known. This study investigates the U sources to rivers in Ohio, United States, part of the Eastern Corn Belt in the Mississippi River watershed. We present a regional picture of seasonal U sources to rivers based on four analyses: 1) a spatial analysis of legacy soil and water data, 2) new measurements of U and carbonate weathering products from rivers at 50 locations across the state collected seasonally over two years, 3) a weekly time series with additional 234U/238U (n = 5) and 87Sr/86Sr (n = 5) measurements from an agricultural river, and 4) a mass-balance approach to U addition to the landscape based on reported P fertilizer use. Uranium concentrations in surface waters collected statewide ranged 0.1-21 nM (n = 132), with significantly higher concentrations in the glaciated agricultural portion of the state (mean = 7.3 nM; n = 105) than the non-glaciated portion (mean = 2.0 nM; n = 24). Concentrations in the glaciated region were highest during the spring and summer and decreased during baseflow. In the time-series, concentrations were ~7 nM during baseflow and ~14 nM during intermediate seasonal discharge conditions, indicating a second more surficial endmember source of U in addition to bedrock weathering that is well correlated with other carbonate weathering products. Systematic increases in 87Sr/86Sr and decreases in 234U/238U with increasing discharge confirm a changing source of carbonate and U weathering and a third surficial endmember during high discharge events. Our mass balance approach and geochemical analysis suggest that elevated U concentrations are the result of carbonate weathering deep in the soil column during elevated seasonal flow. Further work on U dynamics in agricultural rivers is required to understand mechanism controlling seasonal changes in U concentrations and 234U/238U in downstream rivers and U flux.

Quantification of Pb pollution sources in complex urban environments through a multi-source isotope mixing model based on Pb isotopes in lichens and road sediment.

Despite a growing focus on anthropogenic toxic metal pollution in urban environments, few studies have addressed the problem of quantification when more than two pollution sources are likely present, particularly within complex urban settings in the United States (U.S.). In this study, we utilize the MixSIAR package in R for source apportionment based on Pb isotopic signatures in lichen and road sediment in two urban-industrial centers in SW Ohio (OH). We show that ranges of pollutant contributions are more useful than only visualizing mean or raw values of source apportionment, because this avoids overinterpretation of data when certain sources have a large range of uncertainty. We point out both the dominance of industrial pollution as well as the legacy of leaded gasoline pollution in typical mid-sized U.S. cities, which is evident in both road sediment and lichens. Leaded gasoline contribution to Pb in Middletown, OH lichens mostly vary between ~10 and 25%, while in Hamilton, OH the contribution to lichens and road sediment tends to be relatively negligible except for two road sediment samples and one lichen sample, where median contributions are ~20-30%. Industrial combustion pollution source contributions vary between ~25 and 75% in Hamilton, and ~50-100% in Middletown, OH. Furthermore, comparing pollution sources in lichens to modern particulate matter can provide a record of how pollutant sources change over time, such as our traffic lichen (Sample Li-9) plotting closer to leaded gasoline on a bivariate mixing diagram than modern traffic particulate matter, or our coke plant lichen containing slightly less Pb contribution from industrial combustion sources relative to modern coke plant particulate matter. Lastly, when applicable, multi-source mixing models should be complimented in future studies with additional isotopic source tracers such as Cu, Zn, Nd, and Os to further elucidate unique sources of metal pollutants in addition to Pb.

Osmium isotope geochemistry of steel plant emissions using tree bark biomonitoring.

We report for the first time the Os isotopic composition of tree bark samples from a steel town. Osmium concentrations and 187Os/188Os isotopic ratios of ashed bark samples range from 1.40 to 24 ppt and 0.70 to 1.54, respectively, with the lowest 187Os/188Os recorded in samples close to the steel plant. Compositional variations in the bark samples can be explained by mixing between at least two sources with different Os isotopic signatures: a radiogenic source consistent with crust-derived materials and a relatively less radiogenic source consistent with mantle-derived chromite. The exact origin of the radiogenic Os component cannot be constrained, as background signatures and crustal materials used in the steel industry (e.g., coal and iron ore) likely have overlapping radiogenic signatures. Cr shows a similar distribution pattern to Os, indicating that both metals have a common origin, which provides further evidence that the Os budget in the bark samples is controlled primarily by the chromite used in the steel manufacturing. This study shows that Os isotopes are an effective tool for tracing steel production-related emissions.

Crystals reveal magma convection and melt transport in dyke-fed eruptions.

The processes and ranges of intensive variables that control magma transport and dyke propagation through the crust are poorly understood. Here we show that textural and compositional data of olivine crystals (Mg/Fe, Ni and P) from the tephra of the first months of Paricutin volcano monogenetic eruption (Mexico, 1943-1952) record fast growth and large temperature and oxygen fugacity gradients. We interpret that these gradients are due to convective magma transport in a propagating dyke to the Earth's surface in less than a few days. The shortest time we have obtained is 0.1 day, and more than 50% of the calculated timescales are < 2 days for the earliest erupted tephra, which implies magma ascent rates of about 0.1 and 1 m s-1. The olivine zoning patterns change with the eruptive stratigraphy, and record a transition towards a more steady magma flow before the transition from explosive to effusive dynamics. Our results can inform numerical and experimental analogue models of dyke propagation, and thus facilitate a better understanding of the seismicity and other precursors of dyke-fed eruptions.

Uranium mobility across annual growth rings in three deciduous tree species.

Black walnut (Juglans nigra), slippery elm (Ulmus rubra), and white ash (Fraxinus americana) trees were evaluated as potential archives of past uranium (U) contamination. Like other metals, U mobility in annual growth rings of trees is dependent on the tree species. Uranium concentrations and isotopic compositions (masses 234, 235, 236, and 238) were analyzed by thermal ionization mass spectrometry to test the efficacy of using tree rings to retroactively monitor U pollution from the FFMPC, a U purification facility operating from 1951 to 1989. This study found non-natural U (depleted U and detectable 236U) in growth rings of all three tree species that pre-dated the start of operations at FFMPC and compositional trends that did not correspond with known contamination events. Therefore, the annual growth rings of these tree species cannot be used to reliably monitor the chronology of U contamination.

Uranium isotopes in tree bark as a spatial tracer of environmental contamination near former uranium processing facilities in southwest Ohio.

Inappropriate handling of radioactive waste at nuclear facilities can introduce non-natural uranium (U) into the environment via the air or groundwater, leading to anthropogenic increases in U concentrations. Uranium isotopic analyses of natural materials (e.g. soil, plants or water) provide a means to distinguish between natural and anthropogenic U in areas near sources of radionuclides to the environment. This study examines the utility of two different tree bark transects for resolving the areal extent of U atmospheric contamination using several locations in southwest Ohio that historically processed U. This study is the first to utilize tree bark sampling transects to assess environmental contamination emanating from a nuclear facility. The former Fernald Feed Materials Production Center (FFMPC; Ross, Ohio) produced U metal from natural U ores and recycled nuclear materials from 1951 to 1989. Alba Craft Laboratory (Oxford, Ohio) machined several hundred tons of natural U metal from the FFMPC between 1952 and 1957. The Herring-Hall-Marvin Safe Company (HHM; Hamilton, Ohio) intermittently fabricated slugs rolled from natural U metal stock for use in nuclear reactors from 1943 to 1951. We have measured U concentrations and isotope signatures in tree bark sampled along an ∼35 km SSE-NNW transect from the former FFMPC to the vicinity of the former Alba Craft laboratories (transect #1) and an ∼20 km SW- NE (prevailing local wind direction) transect from the FFMPC to the vicinity of the former HHM (transect #2), with a focus on old trees with thick, persistent bark that could potentially record a time-integrated signature of environmental releases of U related to anthropogenic activity. Our results demonstrate the presence of anthropogenic U contamination in tree bark from the entire study area in both transects, with U concentrations within 1 km of the FFMPC up to ∼400 times local background levels of 0.066 ppm. Tree bark samples from the Alba Craft and HHM transects exhibit increasing U concentrations within ∼5 and ∼10 km, respectively of the FFMPC. The 236U/238U isotopic ratios in tree bark from both transects increase progressively towards the FFMPC with values as high as 2.00 × 10-4 at the FFMPC. Tree bark sampled within 1 km of the FFMPC exhibits clear evidence for both enriched and depleted uranium with 235U/238U values from 0.00461 to 0.00736, with 234U/238U activity ratio ranging from 0.53 to 0.96, and 236U/238U from 6.05 × 10-5 to 1.05 × 10-4. Tree bark from transect #1 between 1 and 30 km from the FFMPC exhibits depleted and natural 235U/238U values ranging from 0.00552 to 0.00726 [234U/238U activity ratio: 0.69-1.04; 236U/238U: 2.49 × 10-6 - 2.00 × 10-4]. Tree bark from transect #2 sampled between 1 and ∼20 km away from the FFMPC exhibits evidence of enriched and depleted U in the environment with 235U/238U ranging from 0.00635 to 0.00738 [234U/238U activity ratio: 0.83-0.98; 236U/238U: 1.43 × 10-5 - 2.00 × 10-4]. Results from scanning electron microscopy with energy dispersive spectrometry provides evidence for U-rich particles as the source of contamination found in tree bark growing within 1-3 km of the former FFMPC. Such observations are consistent with the previously observed 14 µm U-rich particle identified in tree bark sampled within 1 km of the FFMPC (Conte et al., 2015). Overall, this study shows the usefulness of a tree bark sample transect to assess the areal extent of atmospheric contaminant U stemming from nuclear facilities.