Horizontal and Vertical Transport of Uranium in an Arid Weapon-Tested Ecosystem

Abstract

Armor-penetrating projectiles and fragments of depleted uranium (DU) have been deposited in soils at weapon-tested sites. Soil samples from these military facilities were analyzed by inductively coupled plasma-optical emission spectroscopy and X-ray diffraction to determine U concentrations and transport across an arid ecosystem. Under arid conditions, both vertical transport driven by evaporation (upward) and leaching (downward) and horizontal transport of U driven by surface runoff in the summer were observed. Upward vertical transport was simulated and confirmed under laboratory-controlled conditions, to be leading to the surface due to capillary action via evaporation during alternating wetting and drying conditions. In the field, the 92.8% of U from DU penetrators and fragments remained in the top 5 cm of soil and decreased to background concentrations in less than 20 cm. In locations prone to high amounts of water runoff, U concentrations were reduced significantly after 20 m from the source due to high surface runoff. Uranium was also transported throughout the ecosystem via plant uptake and wild animal consumption between trophic levels, but with limited accumulation in edible portions in plants and animals.

Graphical Abstract

1. INTRODUCTION

Uranium is a naturally occurring trace element, but due to increase contamination in particular regions, there is a concern to public and environmental health.^1,2 These concerns are not limited to the radioactive nature of U but also in response to U being a chemically toxic heavy metal. In soils all over the world, U is naturally found at levels averaging 2 mg/kg U.^3–5

Depleted uranium (DU) is the byproduct of the U enrichment process. Due to its pyrophoric properties and high density, DU has uses in military applications. Its high density allows for DU to be used as an effective material for armor-piercing projectiles known as penetrators and for armor-plating of tanks. When fired at high velocity, the penetrators can pierce armored vehicles and aerosolize and produce metal fragments, which can instantaneously combust,⁶ creating a cloud of DU particulates small enough for inhalation.^7–10 Several hundred tons of DU have been used in military conflicts over the past forty years, and unknown amounts of DU are expected to remain buried or exposed in battlefields and testing sites.^2,9,11

In the environment, U migrates physically through runoff and leaching as well as ecologically through plant and animal uptake. Mobile, non-metallic forms of U such as schoepite, metaschoepite, or soluble uranyl (UO₂ ²⁺) and U⁴⁺ species in colloidal forms can be found in soil systems.¹² Leaching through soils is linked to precipitation events and has been shown to vary greatly.¹³ Downward transport of heavy metals due to percolation of surface water typically occurs with gravity. However, with repetitive wetting and evaporation cycles, U transports toward the surface through capillary action.¹⁴ However, the detailed mechanisms of upward U transport in arid environments are not fully understood. The mechanism of transport can be affected by properties such as U speciation, soil pH, redox, conductivity, texture, and soil components such as carbonates, metal oxides, and organic matter.^12,13,15,16 However, transport may also be affected by hydraulic variations and soil topography. These properties and soil components vary based on the environment and specific locations. The migration and transport of U are a concern in areas of conflict using DU, mining areas, or areas of agriculture that are contaminated with U from fertilizer application.^16,17 The source of U as well as the heterogeneity of environmental properties affirms the need for individual assessments being conducted in differing contaminated locations.

Heavy metal contaminants in soil are not static; plants can accumulate and animals can further continue the cycling of U in an ecosystem.^18,19 Where plants are tolerant to environmental stresses, i.e., arid regions, they are better able to accumulate toxic metals. The availability of U for uptake is influenced by its affinity for soil components, forms of U, and other physicochemical properties.^20–25 Accumulation is known to vary based on environmental factors as well as specific species. In organisms, U can accumulate in different organs.^26,27 The pathway among soil → plants → animals is a simple model of U transfer to herbivores that can be extended if the animal is prey to other organisms.

Few studies have spatially explored the processes of U transport in contaminated environments due to surface runoff, downward leaching, upward vertical transport, or the uptake of U through plants and animals in the field. This study is a multifaceted spatial analysis combined with laboratory simulations, providing essential information on the transport of U from penetrators left behind in DU tested sites in an arid ecosystem. Our data suggests that U from DU munition-testing sites transports naturally through vertical migration (downward due to leaching and upward due to capillary action and evaporation of soluble U), horizontal migration, and ecologically by plant absorption and animal uptake.

2. METHODS

2.1. Study Area and Sampling.

The location of this study is Yuma Proving Ground (YPG) in Arizona. YPG is a testing site for an array of projectiles that has been test firing DU since 1982. Two sampling areas were investigated at YPG: an area of concentrated DU impacts and a long-term DU weathering site. The highly impacted area is Gun Position 20 (GP 20), where DU was test-fired approximately 1 km downrange from the firing location. The weathering site is where DU penetrators were buried in 2003 to analyze DU corrosion.^28,29

YPG is located in the Sonoran Desert near the ArizonaCalifornia border. It is arid with a low annual precipitation of 3.09 inches and average temperatures reaching as high as 107°F during summer and as low as 46°F during winter.³⁰ Rainfall is sporadic, and high intensity flash floods cause erosion and significant runoff.³¹ The valley at YPG is a stony alluvial fan being described as a gravelly sandy loam. The sandy components allow for good drainage in the soil with water passing through rapidly.³²

At GP 20, sampling varied based on observed DU and terrain (Figure 1). The field was considerably flat, sloping gently to a large-channel ditch with vegetation that collects runoff and could be distinguished by impact craters that were lighter in color from the original scorched ground. In the field, penetrators showing early signs of corrosion were located. To analyze leaching, soil samples below the rods were collected at various depths. Soil samples were also collected at multiple distances from penetrators to analyze horizontal transport.

Figure 1.

The landscape and the sampling area at Gun Position 20 (GP 20) at Yuma Proving Ground in Arizona. The sampling area included the flat field with impact areas from ballistics (shown as disturbed terrain) and large drainage ditch with a high abundance of vegetation. Open field soil samples, red dots; soil samples beneath exposed penetrators, blue dots; soil samples along drainage ditch, yellow dots.

The field was sampled at a 12 cm depth, leading toward the ditch. Five samples were collected at each distance of 30, 15, and 5 m away from the ditch, each being 10 m apart. Ditch soil was collected at 10 m intervals at a top 5 cm depth over a span of 80 m. After taking four samples, an embankment dam was present to reduce particulate runoff. At the weathering site, soils from two penetrator rods buried at 30 cm were sampled.^28,29 Soil was collected at 5 cm intervals to the penetrators and then below.

To analyze exposure, transport, and uptake through trophic levels, plants, insects, and available artifacts of native fauna were collected. Artifacts included small mammal scat (n = 4), burrow scat (n = 6), and ant beds (n = 3). In the same area, plant shoot (n = 10) and roots (n = 4) along with ant communities (n = 3) were also collected.

2.2. Chemical Analysis.

All samples were processed in duplicate and checked against reagent blanks for QA/QC. Due to background soils containing significant amounts of U, the real soil would not be ideal for determining the detection limits as defined as blank.³³ Prior to analysis, soil samples were passed through a 1 mm sieve. All containers and glassware used were acid-washed and rinsed with deionized water.

2.2.1. Soil Characterization.

Polypropylene digestion tubes were used during soil characterization experiments. The pH and electrical conductivity (E.C.) were measured in 1:1 soil–water saturated paste with a calibrated Oakton p110 series pH meter and Fisher brand Traceable conductivity meter.^34,35 Organic matter (OM) was determined with the wet combustion (H₂SO₄-K₂Cr₂O₇) method, and carbonate was determined by the gravimetric method for loss of carbon dioxide.³⁴ The texture and particle size was determined with the hydrometer method using an ASTM 152H Soil Hydrometer standardized at 68 °F.^36,37 Active iron and manganese oxides were analyzed with the citrate-bicarbonate-dithionite method and then analyzed by inductively coupled plasma-optical emission spectrometry (ICP-OES), Perkin Elmer OPTIMA 3300 DV.³⁴

2.2.2. Characterization of U.

Oxidized U was characterized using X-ray diffraction (XRD) spectrometry using a MiniFlex 600. XRD analysis was performed with an accelerated voltage of 45 kV and a current of 15 mA. Data was recorded in the 10–80° range with a step of 0.02° and a speed of 5°/min. Results were processed with JADE software to determine mineralization.

2.2.3. Total U in Soil.

The total concentration of U was determined with 4 M HNO₃ extraction for all soil samples to analyze U transport.³⁸ Extraction was carried out by adding 25 mL of 4 M HNO₃ with ~1 g of soil in a 50 mL polypropylene digestion tube and heating in an 80 °C water bath for 16 h.^38,39 Samples were allowed to settle, decanted, and then filtered through a 0.45 μm filter, diluted, and analyzed by ICP-OES.

2.2.3.1. Upward Transport Simulation.

Column experiments were prepared to simulate the upward transport of U contaminants. We considered the mimicking field conditions and feasibility of lab-scale simulation in designing the dimensions of the columns.^40–42 The column diameter was desired to be 40 times the grain-size diameter of soil for a chemical experiment, and the ratio of column length to column diameter was ≥4.^40–42 Thus, the column had a diameter of 4 cm with a length of 15 cm with five series of drilled holes from the bottom to the top with an interspace of 2.5 cm between holes.¹⁴ Holes were temporally sealed with silicone gel. Uranium sources used were uranyl, UO₂, UO₃, and schoepite that were ground to a powder. A thin layer of filter paper was placed at the bottom of the column, and a small layer of clean Yuma soil was placed above the filter paper of the bottom. U sources were then carefully put above the layer of clean soils at the end of the columns. Afterward, clean Yuma soil was added and packed to the columns to the top of 15 cm. The bottom of each column was placed in a liquid matrix of various aqueous solutions. Aqueous solutions contained different salts (1 mmol/L CaHCO₃ solution, 1 mmol/L MgHCO₃, 1 mmol/L MgHCO₃-CaHCO₃ solution, or 1 mmol/L MgCl₂ solution). These salts were selected to mimic the presence of salts in arid soils as capillarity solutions for the upward transport of U. Total U concentrations in soils were measured with XRF after various periods of time.^14,43 Wet soil was then placed in an oven at 45 °C, simulating the Arizona summer temperature. After a week of drying, wetting the soil was repeated. At the end of the experiment, soil and subsamples were analyzed with XRD to determine the U form and with X-ray fluorescence (XRF) or ICP-OES for U concentrations in soils.¹⁴

2.2.4. Total U in Biomass.

The total concentration of U in biomass was used to analyze the exposure and uptake at different trophic levels by the total digestion of living organisms or animal artifacts. This procedure was performed on a Hot Block 200 by digesting ~1 g of biomass with concentrated HNO₃ followed by 30% H₂O₂, allowing to reflux at 95 ± 5 °C as described by EPA Method 3050B. Samples were then filtered through a 0.45 μm filter, diluted, and analyzed by ICP-OES.

2.3. Data Analysis.

Uranium concentrations in soil and biomass samples were analyzed, and Pearson’s correlation coefficient (R) was used to determine relationships with physicochemical factors. Analysis of variance single factor (ANOVA) was also used to determine significant difference between locations and strata. A p < 0.05 or lower was used to indicate significant difference.

3. RESULTS AND DISCUSSION

3.1. Site Characterization and Properties.

Soil characterization for both ditch and field samples at GP 20 and the weathering site are presented in Table 1. Soils from all sites were predominantly sand with higher percentages of silt and clay in the field and lower percentages in the ditch. Soil pH was alkaline across the entire sampling area, ranging from 7.75 to 9.56. All soil properties with exception of pH and sand were observed to have higher percentages in the field. The higher percentage of carbonate in the field was likely in desert soils due to a prominent calcic horizon that was typical in arid soils.⁴⁴ However, the lower carbonate concentration and texture variations in the ditch were presumed to be from erosion or precipitation events that may allow for leaching.^45,46 The weathering site was expected to vary due to soil disturbances that occurred during the placement of the DU penetrators in 2003.

Table 1.

Characterization and Physicochemical Properties of Soils at the GP 20 Field, GP 20 Ditch, and Weathering Site in Yuma Proving Ground^a

location		pHc	E.C. (mS)b,c	CaCO3 (%)b	OM (%)c	Fe oxide (%)b,c	Mn oxide (%)b	sand (%)	silt (%)	clay (%)
GP 20 field	mean	8.43	3.96	11.77	2.42	0.53	0.014	51.8	32.5	15.7
	std dev	0.44	2.66	2.67	0.61	0.06	0.001
	CV%	5.2	67.1	22.7	25.4	12.2	10.3
	max	9.56	12.11	20.95	4.68	0.74	0.017
	min	7.75	0.73	6.34	1.47	0.45	0.011
GP 20 ditch	mean	8.63	0.24	5.36	2.19	0.43	0.011	81.9	12.1	6.0
	std dev	0.24	0.12	1.21	0.48	0.03	0.002
	CV%	2.8	50.2	22.6	22.0	6.2	14.8
	max	8.93	0.44	7.71	3.10	0.48	0.014
	min	8.16	0.11	3.31	1.44	0.38	0.008
weathering site	mean	8.78	0.84	8.88	1.88	0.37	0.013	67.2	23.9	8.9
	std dev	0.48	1.24	2.59	0.44	0.03	0.001
	CV%	5.5	146.7	29.1	23.3	6.8	5.2
	max	9.28	4.49	19.41	3.08	0.46	0.015
	min	7.79	0.16	6.51	0.93	0.31	0.012

^aElectric conductivity (E.C.); organic matter (OM); coefficient of variation (CV); and standard deviation (std dev).

^bStatistical difference between samples in the field and ditch.

^cStatistical difference between samples in GP 20 and the weathering site.

3.2. Mineral Analysis of U in Soil.

Uranium mineralogy is important in determining forms of U produced from DU. Schoepite was determined to be the major mineral family at each location, being the dominant corrosion product (Figure 2). This is the typical product in well-oxidized, sandy soils⁴⁷ and is in agreement with a previous study of the area.^48,49 The XRD patterns resemble that of uranium(VI) oxide hydrate (UO₃·2H₂O) and uranium trioxide (UO₃). Schoepite is the oxidized form of U that could solubilize as uranyl in surface waters. Complexation may occur with carbonate or form precipitates with metal oxides in well oxidized and alkaline soils.^5,7 This soluble form allows for the leaching, upward transport, and runoff of U.

Figure 2.

X-ray powder diffraction (XRD) analysis of soil samples from the field and ditch at GP 20 and from penetrators that were previously fired and unfired from the DU weathering site.

3.3. Downward Transport of U in Soil.

At GP 20, penetrators were found exposed on the surface in the field area and were corroded with black and yellow oxidation products, indicating UO₂ and schoepite, respectively.⁵⁰ Analyzing depth profiles below the penetrators, which were on the top ground surface, each layer decreased in U concentration (Figure 3). Within the first depth of 0–2 cm below, the corroded penetrator U concentration was 5027 mg/kg. In the depths of 2–5 and 5–10 cm below the penetrator, U concentrations decreased by 79.0 and 91%, respectively. The top 5 cm beneath the exposed DU averaged 3416 mg/kg U. Having a higher concentration of U in the upper profile may indicate more U in particulate form. This may explain the high standard deviation in Figure 3 beneath the DU penetrator where the top 0–2 cm.^50,51 This U trend was correlated with pH (R = 0.72, p < 0.05). This relationship of pH and other heavy metals was widely observed, demonstrating sorption sites on metal oxides in alkaline soils.^51–53 In a similar soil study with loamy sand and alkaline soils, surface samples contained high concentrations of U, which was later determined to be associated with silica clays, amorphous silica, or coated U particulates.⁵¹ The results in our depth profile was in agreement with Johnson et al.⁵¹ in that a major portion of U was retained in the top 2 cm of soil. These results were similar to another field studies, where soils 0–20 cm beneath DU after the Kosovo conflict ranged from 2100 to 7600 mg/kg U.⁵⁴

Figure 3.

Uranium distribution at GP 20 for soil profiles directly beneath an oxidized DU rod. Samples were collected at 0–2, 2–5, and 5–10 cm intervals. Error bars represent standard deviations.

On the other hand, a strong leaching of U was observed under the deep buried penetrator at the weathering site (Figure 4). Below the unfired penetrator, the U concentrations at 35–40, 40–45, and 45–50 cm were 391, 787, and 1022 mg/kg, respectively, indicating that downward leaching was also an important transport mechanism in arid soils and continued to occur after 16 years. The higher concentrations at lower strata may indicate either greater retention or a lower rate of mobility. Based on column experiments, U transport is mainly dependent on schoepite dissolution in the form of uranyl and that particulate transport of U was minimal.^50,51

Figure 4.

Uranium distribution at the DU weathering site for soil profiles directly over and below both fired and unfired DU penetrators. Samples were collected at 5 cm intervals. Error bars represent standard deviations.

3.4. Horizontal Transport and Runoff of U along the Field and Ditch.

Our samples exhibited a decrease of magnitude in U concentration in soils horizontally from a corroded penetrator at 20 cm with 357 mg/kg. At 5 cm away from the DU, the U decreased by 84.2% to 481 mg/kg (Figure 5). This indicates that the majority of the U remained within the first 5 cm of the exposed DU. The U concentration at 20 cm away from the penetrator may be due to the four decades of U use at this site, resulting in high background-contamination levels of U. This was in agreement with a previous field study with average U at 279 mg/kg in the open field at GP 20.⁴⁸ It was reported that one year after the Kosovo conflict, at 30 cm from a penetrator, the decrease in U concentration was three magnitudes lower.⁵⁴

Figure 5.

Uranium distribution at GP 20 at various distances from the oxidized DU rod. Bars represent standard deviation.

We observed the horizontal transport of U in the open field, a 1000 m² area. During intense rain events in the summer, runoff flows to the large-channel ditch. The mean U concentration of this field was 383 (SD ± 50) mg/kg U, but at 30, 15, and 5 m from the ditch, U concentrations increased from 308 to 341 and 395 mg/kg U, respectively. Only in soil at 5 m, U concentrations were determined to be significant (Figure 6). Due to intense flash floods, one may assume that as runoff height consolidates, more scouring would remove U in its soluble or particulate form as described by previous reports.^55,56 However, runoff simulation by Ward and Stevens⁵⁷ observed that a finer grain material was transported by runoff early in hydraulic events, that these U particulates quickly settled and became stable, and also that raindrop impacting the soil was required to disturb the soil for transport. The U concentration in this area was most correlated with Mn oxides (R = 0.55, p < 0.05) being a possible particulate form. This is consistent with Kazery et al.⁴⁸ who observed U with easily reducible oxides, i.e., Mn oxides, where runoff occurs. The high U concentration in soil at 5 m from the ditch may be explained by the horizontal transport of U particulates settling during surface runoff, as well as the decreased friction at the water–soil interface.⁵⁸ This is in agreement with DU fragments not being found with suspended sediments in runoff experiments.⁵⁷

Figure 6.

Uranium concentrations from unimpacted soils at GP 20 at various distances leading to the ditch. Bars represent standard deviations.

Figure 7 presents the U concentrations along a wide-channel ditch for runoff at GP 20 where samples were collected at 10 m intervals. After the fourth sample, there was an embankment dam to deter the flow of particulates. The samples before and after the dam were significantly different with mean values of 685 (SD ±362) and 339 (SD ±34) mg/kg U, respectively. The first soil sample was taken in close proximity of a large source of yellow schoepite and was determined to have 1215 mg/kg U. This source could be washed away or solubilized over time as reported in a three-year study implementing soil columns, demonstrating that with a weekly application of rain water, a maximum of 10 mg/L U was released from oxidized DU sources in effluents.⁵⁹ After 20 m (Sample 3), U concentrations dropped to 384 mg/kg, which was still significantly higher than the soils after the dam, but demonstrates a decreasing trend. Along a large channel, similar to the ditch sampling area, it was originally hypothesized that U would be transported by a higher energy flow, causing increased friction across the soil surface, being more efficient at distributing DU particles over a larger area.⁵⁷ However, this decrease in U concentration indicates that U transport was minimal. Again, our findings were consistent with the reports from Wards and Stevens⁵⁷ assessing runoff experiments of hydraulic events across a sloping terrain. They determined that DU fragments were not found with suspended sediment, inferring that transport occurred along the soil or within the soil bed. Also, any transport was minimal, showing that only 7% of DU was transported after extreme hydraulic runoff, just as our samples exhibited minimal transport along the ditch. The ditch soils after the dam may be in a contaminated equilibrium with U due to years of DU use, which was similar to the field samples at GP 20 and consistent with a previous report.⁴⁸ One may infer that this is due to soluble U species at a soil capacity consistently distributed in these soils during equilibrium. Carbonate and texture conditions were different in the ditch where less carbonate, more sand, and approximately one-third of the clay content were observed (Table 1). U in the ditch was observed to be more correlated with Fe oxides (R = 0.46, p < 0.05) and negatively correlated with Mn oxides (R = −0.48, p < 0.05). This negative relationship was also observed in other studies^48,60 which was not observed in the GP 20 field, which is due to U being passed to Fe oxides as Mn oxides may be reduced during inundation.^52,61,62

Figure 7.

Uranium concentrations within the ditch at GP 20. Samples were collected at 10 m intervals. After Sample 4, a physical barrier was present.

3.5. Upward Transport of U Observed in the Field.

The U concentrations in the soil profile above the unfired penetrator were observed in Figure 4. The U concentrations above a textile barrier, placed over the DU penetrators were statistically similar, averaging 305 (SD, ±9.1) mg/kg U above both DU penetrators. The consistent concentration at each layer above the penetrators indicates that U had reached equilibrium with the soil. With ample time, metals could reallocate in a consistent distribution in the quasi-equilibrium.²¹ These concentrations above the penetrators were due to the upward transport of soluble U, as observed in the simulations below, where no precipitate was observed. This was shown by column experiments with Yuma soils that during repeated wet and dry cycles, soluble U and schoepites, a corroded product of DU penetrators, were transported toward the surface by capillary action driven by evaporation of water. After four months of a two week alternating cycle of wetting and drying, uranyl was observed to migrate up to 7.5 cm.¹⁴

3.6. Laboratory Simulations of Upward Transport of U.

This simulation was carried out to confirm that U accumulates in the surface layers of arid soils due to the upward transport of soluble U driven by water evaporation in arid ecosystems. Various forms of U were analyzed including soluble U as uranyl, UO₂, UO₃, and schoepite (Figure 8). The first laboratory simulation (n = 5) that compared the transport of uranyl, UO₂, and UO₃ over a 10 month period using water demonstrated that upward transport occurred as the soluble uranyl form (Figure 8A). The other U forms did not demonstrate upward migration. This appears due to the reduced UO₂ being insoluble, while UO₃ is less soluble than the uranyl form.^16,63,64

Figure 8.

Depth profiles in centimeters (cm) of U upward transport above the U source in laboratory simulations using Yuma soil. U concentrations (mg/kg) were detected in the soil column of various U sources (A: uranyl, UO₂, and UO₃ and B: schoepite) at different depths above the U source at the bottom of the column due to upward transport.

Ollila and Ahonen⁶⁴ demonstrated that the U solubility is slightly higher for a bicarbonate solution when compared to deionized water, which is initially observed at four months in our simulation. Aerobic and alkaline conditions tend to favor U-carbonate complexes that limit adsorptions to mineral surfaces.¹⁶ During the four month observation, the Ca-based matrix raised soluble U to 4.5 cm above the original U source and the Mg-based matrix raised to 7.5 cm, which were not significantly different. A supplemental study was conducted using schoepite, a direct corroded product of the DU penetrators. After 100 days of wet–dry cycles, the U migration from the schoepite source in both water and salt matrixes, i.e. MgCl₂, reached to 8 cm above the original depth. (Figure 8B).

3.7. Ecological Uptake.

3.7.1. Plant Uptake and Distribution.

Contamination of U in any ecosystem has considerable consequences to both environment health and the public through the food chain by possible transfer of U. Creosote-plant samples, consisting of stem/leaves and roots, were collected to observe U accumulation. The contamination in the present vegetation is due to plant uptake of soluble U through the roots from the contaminated soil rather than surface deposition. This can be observed by the compartmentalization within the plants. The roots accumulated significantly higher U concentrations (91.0 mg/kg) compared to the stems and leaves compartments (11.6 mg/kg) (Table 2). The ranges of these two compartments varied greatly, with some root samples a magnitude higher at 176 mg/kg U. This is to be expected as described by other reports, that while trace elements do translocate within plant tissues, U tends to accumulate in the root systems with reduced concentrations of U, migrating to leaves and fruit.^22,23,65 This reduced migration can be seen as a beneficial mechanism to reduce the transport of U to higher trophic levels.

Table 2.

Statistical Summaries of U Concentrations (U mg/ kg) Observed in the Stems/Leaves (n = 20), Roots (n = 8), and Animal Artifacts: Ant Beds (n = 6), Ants (n = 6), Rabbit Scat (n = 8), and Burro Scat (n = 12) Sampled from the Ditch at GP 20^a

	stems/leaves	roots	ant beds	ants	rabbit scat	burro scat
mean	11.6	91.04	754.9	215.0	56.89	61.91
std dev	3.756	54.50	149.9	46.16	6.001	7.021
CV%	32.4	59.9	19.9	21.5	10.6	11.3
max	18.13	175.8	913.6	423.7	90.13	125.5
min	3.656	30.32	537.8	131.1	39.32	42.04

^aCoefficient of variation (CV); standard deviations (std dev).

Total values are a simplistic method for quantifying contamination; however, transfer of U to other trophic levels can be investigated by determination of a bioconcentration factor (BCF).^66,67 The BCF is the ratio of U concentration that is in biomass compared to the exposure concentration.^38,68–70 The mean U concentration was determined from an 80 m range along the ditch and was determined to be 493 mg/kg. After analyzing the ditch vegetation, the BCF was determined to be 0.02. When determining transfer factors, such as BCF, concentrations are from the edible portions, i.e., stems/leaves.⁷⁰ This BCF indicates that 2% of the U in the environment is available to be passed along the food chain. However, the U used to determine the BCF does not consider animal exposure from inhalation or contaminated U deposited on the vegetation.^31,65

Uranium uptake in the Creosote bushes from both the field and the ditch areas at GP 20 was also examined. Uranium concentrations near the plants were observed by XRF and recorded. The use of XRF has been observed in laboratory studies to be an acceptable and rapid screening method in contaminated areas.⁴³ The plants were collected, digested, and analyzed by ICP-OES. The U concentrations in the Creosote bushes were determined to be 3.54 and 6.06 mg/kg in the field and ditch, respectively. The U concentrations in the other bushes ranged from 0.48 to 13.92 mg/kg in the field and 0.03 to 33.06 mg/kg in the ditch and were determined to be statistically similar. These values agreed with the stems and leaves compartment from Table 1. When each plant sample was compared to the U exposure from the surrounding soil, the mean BCF was determined to be 0.03 in the field and 0.08 in the ditch. This may lead one to presume that more contamination led to more uptake, but these values in both areas were determined to be statistically similar, which was likely be due to the difference in soil properties.

3.7.2. Animal Uptake.

Within organisms, U absorption varies depending on age, sex, nutritional demand, exposure time, and species variability.^17,71,72 Mode of U uptake also varies, where ingestion was observed in large organisms such as mammals to be the main exposure pathway,³¹ while invertebrates may adsorb U directly on surfaces.¹⁷ Ants (Formicidae) and ant beds were relatively scarce at our sample sites and only observed along the ditch at GP 20. Being a contaminated area provided opportunity to analyze U uptake and exposure. The ants were digested and determined to have a mean of 215 mg/kg U and ranged between 131 and 424 mg/kg (Table 1). These values fell within the lower spectrum of a 1996 study at YPG that reported local insects to range from 130 to 180,000 mg/kg U.³¹ The present study showed that U concentrations along the ditch varied greatly, with a mean of 493 mg/kg, while the ant beds were determined to range from 538 to 914 mg/kg U with a mean of 755 mg/kg U (Table 1). Based on U means, there was an accumulation of U in the local ant beds, which was also observed by Hill et al.⁷³ from a desert with naturally high concentrations of U.

It was demonstrated that arthropods may accumulate U through ingestion but also retain U by adsorption. This is due to chitinous exoskeletons and glucosamine having a high affinity for uranyl ions.¹⁷ The BCF for ants was determined to be 0.28, which was a magnitude higher than the determined ant concentration factor of U activity in a semi-natural ecosystem of 0.02 from Dragovic et al.⁷⁴ study.

With each trophic level, concentrations of U are expected to vary as transport occurs. This is demonstrated as animals have a low absorbance along the digestive tract with the majority of U being retained in the liver, kidneys, or intestines and less U in the muscle to be consumed by human.^65,70,72 Uptake of U in herbivores was determined by collecting fecal artifacts (scat). Multiple samples of rabbit (Leporidae) and burro (Equidae) scat were collected at GP 20. The concentration of U determined from the scat was considered to be exposure, represented as daily dietary uptake of U as determined and reported by Harrison.⁷⁵ The present study indicates that the rabbit dietary uptake was 56.9 mg/kg U and was not statistically different from the burro dietary uptake of 61.9 mg/kg U (Table 1). The U in these organic wastes may be underestimated since some of U may have been leached with the time on the soil surface. According the Ebinger et al., ³¹ the local Leporidae, Black-tailed jack rabbit, and Desert cottontail were reported to contain up to 11 and 300 mg/kg U in the muscle, with up to 19 and 120 mg/kg U in the livers and with up to 66 and 390 mg/kg U in the kidneys, respectively. This demonstrates that U may be transported further along in these tissues. By using the accumulation in the muscle and our dietary intake as exposure, we can roughly calculate the BCF in the lagomorphs to be 0.19 in the jackrabbit and 5.27 hyperaccumulating in the cottontail.

4. CONCLUSIONS

Non-metallic DU was predominately found in the sandy soils as schoepite. Uranium-laden soils allow for migration and transport by leaching, runoff, upward transport, and throughout the food chain. From exposed DU, the downward leaching and wash from runoff was limited. Uranium remained in the top 5 cm of soil and in a 5 cm radius, possibly due to soil texture. Uranium concentrations in the field were most correlated with the soil fractions Mn oxides and Fe oxides and may further be affected by hydraulic processes that occur from different water levels and transport particulates less in deeper profiles of water. From buried DU, U concentrations were found to be consistent in the depths above the DU source averaging 305 mg/kg. The upward migration was the result of years of wet–dry cycles, promoting capillary action. Laboratory simulations demonstrated the ability of U to transport upward through soil as soluble U. When compared to water, carbonate promoted upward transport of U but was shown to be hindered by calcium.

Uptake of U in vegetation was shown to dominate in the root systems, having reduced amounts of U in the edible portions of the plants. The GP 20 vegetation demonstrated accumulation of 3.6 to 18.1 mg/kg U for the next trophic level to ingest. This can vary by location as noted when observing the Creosote bushes having BCF values of 0.03 in the field and 0.08 in the ditch. Animal uptake was also observed to vary with other factors such as species. It was determined that ants have a BCF of 0.28 containing 215 mg/kg U based on exposure. In vertebrates, U uptake can be limited by ingestion, intestinal absorption, and translocation to specific organs. This limitation of transport is beneficial and should be determined if the organism is hunted or used for human consumption due to muscle being the main tissue eaten.