3.1. SAMPLING

During rainfall events in September 2000, and July, September, and November 2001, supplemental grab samples of runoff were collected from S5 pasture and from W4, a similarly instrumented but unfertilized semi-native pasture similar in management history to soil sampling site 770 (Figure 1). The unfiltered water samples were collected in acid-washed, field-rinsed polyethylene bottles and shipped to U.S. Geological Survey laboratories in Reston, Virginia by overnight carrier.

To identify the isotopic characteristics of potential fertilizer and feed inputs, samples of locally applied superphosphate, ammonium sulfate, and P-bearing cattle feed supplements were obtained from commercial suppliers in Okeechobee, Florida.

3.2. SOIL EXTRACTS

Soil samples representing a range of depths and U concentrations in each of the three soil profiles were selected for U isotope analysis. Two grams (g) of dried soil was mixed with 100 milliliters (mL) of 0.1 molar (M) sodium bicarbonate solution (pH = 8.3) to extract loosely bound, readily-exchangeable U for isotopic analysis. The soil and solution mixtures were stirred for 24 h at room temperature followed by filtration through a 0.45-micrometer (µm) membrane under pressure from nitrogen gas.

3.3. ANALYSIS

3.3.1. Soils
Ash content of soil was determined by loss-on-ignition (LOI) at 550 °C. Total carbon (TC), total nitrogen (TN), and total sulfur (TS) concentrations were determined directly on aliquots of dried (60 °C) samples using a CNS Analyzer. Organic carbon (OC) content of these samples was determined with the CNS Analyzer after removing inorganic carbon (IC) using an acid-vapor method modified from that of Hedges and Stern (1984) and Yamamuro and Kayanne (1995). All soil samples were analyzed at least in duplicate. Analytical precision (1-sigma) expressed as a relative standard deviation (RSD) is ±2 percent for TC, ±4 percent for OC, ±3 percent for TN, and ±6 percent for TS. Inorganic carbon is reported as the calculated difference between TC and OC.

Total phosphorus (TP) concentrations were determined by a method described by Aspila et al. (1976), which was slightly modified for analyses of MAERC soils. Dried (60 °C) sediment samples (0.4-0.6 g) were baked at 550 °C for 2 h, then extracted in 1 M HCl for 16 h on a shaker to completely dissolve the phosphate. An aliquot of each extract was centrifuge filtered using a Durapore centrifuge filter (0.45 µm pore size), neutralized with a NaOH solution, and analyzed for phosphate using the standard phospho-molybdate method (Strickland and Parsons, 1973). Analytical precision (1-sigma) for the TP analysis is ±4 percent (RSD).

3.3.2. Water Samples
Runoff samples were filtered through a 0.45 µm membrane and analyzed for pH, specific conductance, dissolved ammonium, dissolved anions (phosphate, chloride, nitrate, fluoride, bromide, and sulfate), and dissolved uranium and its isotopic composition. Ammonium and phosphate were determined using standard colorimetric methods. Other anions were determined by ion chromatography. Estimated 1-sigma precisions are ±4 percent (RSD) for ion chromatography and ±5 percent (RSD) for colorimetry. Detection limits are 0.05 mg/L for anions and 0.5 µg/L for phosphate and ammonium.

3.3.3. Uranium Concentration and Isotopic Analysis
The concentration of U in dried samples of soil, fertilizer, and feed supplements was determined by a delayed neutron technique (Millard and Keaten, 1982). This nuclear activation technique has a detection limit of 0.75 µg U, which is equivalent to 0.09 parts per million (ppm) in a typical analyzed sample weight of 8 g. Precision (1-sigma) of U determinations in soil samples ranged from 5 to 13 percent (RSD) and was poorest for samples with lowest U content.

The U concentration and ²³⁴U/²³⁸U isotopic composition of soil extracts, surface waters, and acid-digested fertilizers and feed supplements were determined by isotope-dilution thermal ionization mass spectrometry (TIMS). Samples were combined with a mixed ²³³U-²³⁶U spike, evaporated to near dryness, and reacted with warm, ultrapure 7N HNO₃. Uranium was purified using anion exchange resin (AG 1-X8, 200-400 mesh, chloride-form) with ultrapure HNO₃ and HCl. Uranium was eluted with 0.5N HCl. This procedure was then repeated on a smaller resin column. Total procedural blank for U was ~5 picogram and has a negligible contribution. The purified U was then loaded on a single Re filament doped with colloidal graphite (Arden and Gale, 1974; Chen et al., 1986). Isotope measurements were performed using a VG-54E mass spectrometer equipped with an analogue Daly detector. A mixed ²³³U-²³⁶U spike was used to correct for mass discrimination. The ²³⁴U/²³⁸U activity ratio (AR) of the U isotopes was calculated from the measured mass ratio by normalizing to a secular equilibrium standard that was measured routinely with each batch of samples. Measured mass ratios of the two U isotopes are very precisely determined and have a 2-sigma analytical precision better than ?0.35 percent.

3.3.4. Sulfur Isotopic Composition
Sulfur in soil was oxidized to sulfate by fusion with Eschka's mixture (magnesium oxide and calcium carbonate) for 2 h at 800°C. The fused mixture was submerged in boiling deionized water for 30 min. The suspension was filtered and the recovered solution acidified to pH 4 with concentrated HCl. The acidified solution was heated to boiling and BaCl₂ added to precipitate the sulfate as barium sulfate. The precipitate was collected by filtering the solution through a 0.4 µm membrane. Dissolved sulfate in filtered runoff samples was similarly concentrated as barium sulfate. Barium sulfate precipitates were weighed into tin capsules, mixed with vanadium pentoxide, then combusted in an elemental analyzer to form SO₂ gas that was introduced into an in-line continuous-flow mass spectrometer. The sulfur isotopic composition is reported as a ³⁴S value, calculated as follows:

³⁴S (in per mil, ) = (R_sample/R_standard - 1) x 1000

where R is the atomic ³⁴S/³²S ratio and the standard is Vienna Canyon Diablo troilite (V-CDT). Accuracy and precision of the measurements was ±0.2 based on multiple determinations of S isotopic standards (NBS-127, IAEA-S-5, and IAEA-S-6).