Environmental Toxicology and Chemistry, Vol. 33, No. 5, pp. 1048–1058, 2014 # 2014 SETAC Printed in the USA
DEGRADATION RATE OF SODIUM FLUOROACETATE IN THREE NEW ZEALAND SOILS GRANT NORTHCOTT,y DWAYNE JENSEN,y LUCIA YING,y and PENNY FISHER*z yPlant & Food Research Ruakura, Hamilton, New Zealand zLandcare Research, Lincoln, New Zealand (Submitted 20 October 2013; Returned for Revision 20 November 2013; Accepted 23 January 2014) Abstract: The degradation rate of sodium fluoroacetate (SFA) was assessed in a laboratory microcosm study incorporating 3 New Zealand soil types under different temperature (5 8C, 10 8C, or 20 8C) and soil moisture (35% or 60% water holding capacity) conditions using guideline 307 from the Organisation for Economic Co-operation and Development. A combination of nonlabeled and radiolabeled 14CSFA was added to soil microcosms, with sampling and analysis protocols for soil, soil extracts, and evolved CO2 established using liquid scintillation counting and liquid chromatography–mass spectrometry. Degradation products of SFA and their rates of formation were similar in the 3 soil types. The major degradation pathway for SFA was through microbial degradation to the hydroxyl metabolite, hydroxyacetic acid, and microbial mineralization to CO2, which constituted the major transformation product. Temperature, rather than soil type or moisture content, was the dominant factor affecting the rate of degradation. Soil treatments incubated at 20 8C displayed a more rapid loss of 14C-SFA residues than lower temperature treatments. The transformation half-life (DT50) of SFA in the 3 soils increased with decreasing temperature, varying from 6 d to 8 d at 20 8C, 10 d to 21 d at 10 8C, and 22 d to 43 d at 5 8C. Environ Toxicol Chem 2014;33:1048–1058. # 2014 SETAC Keywords: 1080
Degradation time
Hydroxyacetic acid
Pesticide
Sodium fluoroacetate
Soil
natural soils. Temperature and moisture content, as regulators of microbial activity, were expected to affect the rate of SFA degradation in New Zealand soils [12]. In 2007, a regulatory reassessment of SFA [13] by the Environmental Risk Management Authority New Zealand (now Environmental Protection Agency, New Zealand) determined that existing data on the degradation pathways and rates of SFA in soil were limited in scope and applicability, and were not conducted in accordance with international test methods. The present study addressed a recommendation [13] to generate additional data using guideline 307 of the Organisation for Economic Co-operation and Development (OECD) [14] under aerobic test conditions. In addition to the test temperatures of 20 8C and 10 8C stipulated in the test guideline, an additional treatment of 5 8C was included to simulate the New Zealand winter conditions when SFA is most commonly applied. An additional test treatment simulating relatively dry soil conditions was also included. Test soils used needed to represent regions of New Zealand subject to aerial application of SFA bait, and encompass variability in soil organic matter and clay content.
INTRODUCTION
Sodium fluoroacetate (SFA; FCH2CO2Na) is registered as a pesticide in New Zealand and Australia, for delivery in bait to control pest mammals, and is commonly referred to as 1080. In New Zealand, aerial or bait station application of cereal pellet or chopped carrot bait containing up to 0.15% SFA by weight is used for broad-scale management of introduced brushtail possums (Trichosurus vulpecula), rodents (Rattus spp.) and, to a lesser extent, rabbits (Oryctolagus cuniculus). After application, bait remaining uneaten on the ground is subject to dissipation and degradation through exposure to rainfall [1]. Exposure to environmental moisture results in water-soluble SFA leaching from bait into litter and soil [2]. Thus, the degradation of SFA and the rate of formation of degradation products in soil are of interest for environmental risk assessment. Potential degradation pathways of SFA in soil include chemical hydrolysis in solution and biotic mechanisms of metabolism or mineralization. Degradation in solution begins with dissociation of the sodium ion followed by hydrolysis of fluoroacetate to produce fluoromethane and bicarbonate. Metabolism of SFA following ingestion by soil organisms occurs via conversion to fluorocitrate within mitochondria [3] or by root uptake of fluoroacetate in solution followed by metabolism by plants. Mineralization of SFA through microbial degradation produces fluoride ion and hydroxyacetic acid (HAA; also termed glycolate/glycolic acid). Previous studies have identified soil bacteria that could utilize fluoroacetate as a sole carbon source through enzymatic cleavage of the carbon– fluorine bond [4–6] as well as the enzymes responsible and their gene coding in specific bacteria [7]. Previous investigations [5,6,8–11] have identified a significant role of some bacteria and fungi in the degradation of SFA in
MATERIALS AND METHODS
Within the protocol described by OECD guideline 307 [14], we used a combination of nonlabeled and radiolabeled SFA to determine the rate of degradation and the nature and rates of formation and decline of transformation products. The guideline recommends that test soil moisture content should be 40% to 60% water-holding capacity (WHC). In the present study we used 60% WHC as 1 treatment variable and also included 35% WHC as another, to address the question of degradation of SFA under drier soil conditions. In addition to the test temperatures of 10 8C and 20 8C specified by OECD guideline 307 [14], an additional 5 8C treatment was incorporated into the experiment to better represent the range of temperatures prevalent in New Zealand, particularly as SFA baits are often applied during winter when soil temperatures are lowest. The assessment was
* Address correspondence to [email protected]. Published online 29 January 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2536 1048
Degradation of sodium fluoroacetate in soil
Environ Toxicol Chem 33, 2014
completed to Good Laboratory Practice standard at Plant & Food Research’s Food and Biological Chemistry laboratory between May 2009 and December 2010. More detailed technical description of the test system and analytical methods used in the assessment are provided by Northcott [15]. Test soil selection, sampling, and preparation
Three test soils were selected using spatial data to identify representative areas of New Zealand over which SFA bait has been applied aerially and is likely to be applied in future. The geographic information system (GIS) data were sourced from the VectorNet database (Animal Health Board, Wellington, New Zealand) for aerial bait applications for possum control during July 2007 to February 2008, and from the Department of Conservation (Research Development and Improvement Division, Christchurch, New Zealand) for aerial application for possum and rodent control over areas of conservation estate in the period 2006 to 2007. These GIS data were matched to areas of the dominant New Zealand Soil Classification Soil Orders (National Soil Database, Landcare Research) to produce a ranking of soil order by area exposed to aerial bait application. Three orders (brown soils, podzols, and pumice soils) represented the greatest combined area exposed to SFA baits after aerial application and the most diversity in terms of regional location and variation of physical, chemical, and biological characteristics. Collection sites (Table 1) were selected using the National Soils Database (Landcare Research) to identify existing reference sites and provide descriptive information.
1049
Soil from the 3 reference sites was sampled to a maximum depth of 20 cm. At each site, soil was taken from at least 5 locations approximately 10 m apart. Large stones, vegetation and leaf litter, visible earthworms, and the upper layer of grass and root material were removed by hand. The collected soil from each location was transferred into large plastic bags and loosely tied to allow exchange of air. The soil samples were transported chilled (3 cm were broken apart manually, and for each soil type the contents of all sample bags were mixed together by hand, with visible pieces of plant material and stone removed during mixing. The fully mixed composite soil for each sampling site and soil order were transferred into loosely tied, labeled plastic bags and stored at 5 8C. Approximately 10 kg of each composite soil sample was sieved to 2 mm, and soil dry weight and gravimetric water content (GWC) were determined by drying triplicate subsamples of sieved soil fractions to constant weight at 105 8C. Waterholding capacity was measured using the method of Hoper [16]. Maximum WHC was calculated as the corresponding GWC using the soil dry weight data. The GWC of sieved field-moist soils was either close to or greater than that required to achieve 60% or 35% WHC, necessitating partial air-drying of the soils before readjusting to the required experimental WHC, as described by Northcott [15]. Sterile soil for control treatments was prepared by autoclaving preweighed portions of soil 3 times. The soil was autoclaved at 121 8C (1 bar) for 30 min and maintained within the sealed
Table 1. Details of test soil sampling and descriptions
Order Podzol
Location and reference from database Orikaka Sandy Loam, West Coast, South Island GR L29 2425373 593 4737
Global positioning system coordinate and date of sample collection
Existing data
248250 35.500 E
ZOH, Humose orthic podzol
Coarse sand 13%
598340 71.400 N
Beech–podocarp forest, Profile description 0-B3
Medium sand 11%
a
2 January 2009
Brown Soil
Matiri, West Coast, South Island GR L30 2410730 5909418
248100 75.600 E 598090 39.200 N
BLA Acid allophanic brown Upland yellow-brown earth, lowland podocarp forest
8 January 2009
Pumice soil
Kaingaroa (sand), Taupo, North Island GR U18 2795100 6257100
278950 10.600 E
M/W Welded impeded pumice
628570 10.700 N 12 January 2009
Mānuka exotic forest, sphagnum
Physical and chemical characteristicsb
Fine sand 9% Silt 13% Clay 54% pH 4.6 Total C 12.6% Total N 0.50% MBC 1362 Coarse sand 1% Medium sand 6% Fine sand 30% Silt 39% Clay 24% pH 3.7 Total C 12.0% Total N 0.46% MBC 1790 Coarse sand 3% Medium sand 9% Fine sand 18% Silt 43% Clay 27% pH 5.4 Total C 7.68% Total N 0.33% MBC1014
a Horizon descriptions accessible on New Zealand soils database; search by series name or identity number: http://soils.landcareresearch.co.nz/contents/ SoilData_NSD_ReportsFlat.aspx?currentPage¼SoilData_NSD_ReportsFlat&menuItem¼SoilData b Particle Size Distribution testing of Fine Earth Fraction (New Zealand Classification), and measurements of pH, total carbon and nitrogen, and microbial biomass carbon (MBC; mg/kg) undertaken by Landcare Research, Hamilton, New Zealand.
1050
Environ Toxicol Chem 33, 2014
autoclave until it had cooled to 30 8C (5–6 h). The autoclave cycle was repeated, and the soils were then left to incubate at ambient temperature for 24 h to allow surviving bacterial spores to germinate. Following this period of incubation, the soils were autoclaved at 122 8C (1.35 bar) for 35 min. All subsequent manipulations of the sterilized soils were conducted under laminar-flow and sterile conditions. The 5 8C, 10 8C, and 20 8C treatments were housed in separate temperature-controlled rooms monitored by calibrated electronic monitors and minimum– maximum thermometers. Pre-equilibration of all test soils at 5 8C, 10 8C, and 20 8C was undertaken for a minimum of 4 wk, with any germinating seedlings removed by hand. Soil spiking procedure
An initial moisture adjustment was used to introduce SFA to soil as an aqueous spike using a modification of the method described by Brinch et al. [17], thereby avoiding the potential alteration of the sorption behavior of SFA in the test soils [18]. Soil within each flask was spiked with a mixture of nonlabeled and radiolabeled SFA (fluoroacetic acid, [1-14C] sodium salt) equivalent to 18 mg of SFA, which was the nominal quantity in a single 12-g cereal pellet bait containing 0.15% (w/w) SFA. The aqueous spike was evenly distributed over quarter portions of soil that were subsequently combined with the remaining volume of soil. Sterile soil treatments were spiked with the same procedure but using soil that had been autoclaved within the previous 2 d and a solution of SFA prepared in sterile MilliQ water (Millipore). Spiking of soil in the 3 temperature treatments was staggered to accommodate the sampling schedule. Immediately after spiking, subsamples (0.5–0.9 g) of soil from each treatment were weighed into individual paper combustion cones, with 6 replicates and 10 replicates taken from each spiked sterile and nonsterile treatment, respectively. These replicate subsamples were stored at –50 8C to 80 8C until analysis by sample oxidation to confirm the activity and homogeneity of the [14C]-SFA portion of the spikes. Flow-through incubation flask system
The test system was based on the design specified in OECD guideline 307 [14], and by the Soil Science Society of America [19]. Spiked soil treatments in 250-mL glass Schott bottles were incubated in flow-through, dark conditions under 3 temperature (5 8C, 10 8C, or 20 8C) and 2 soil moisture (35% or 60% WHC) treatments. Eighteen flasks were incubated at each temperature (n ¼ 6 for each soil type). Each group of 6 flasks comprised duplicate nonsterile and a sterile spiked soil for each of the 35% and 60% WHC treatments. Each flask contained an equivalent dry weight of 105 g of soil, but the total final weight of soil prepared for each varied depending on the adjusted GWC required to attain the treatment WHC. Flasks were sealed with a screw cap containing a Teflon plug and Viton O-ring insert, through which a controlled air flow (5 mL/min) was delivered and then directed through a trap containing a polyurethane foam (PUF) plug, and bubbled into glass scintillation vials containing alkali solution (5 mL of 1 M NaOH) to trap CO2/14CO2. Sample types and sampling intervals
The 10 8C treatments were spiked first, followed by the 20 8C treatments 4 d later and the 5 8C treatments 2 d later. The OECD guideline 307 [14] specifies that the study should normally not exceed 120 d, but we used a longer incubation period to ensure a complete characterization of the decline of SFA concentrations over time. Sampling and analyses included measurement of
G. Northcott et al.
evolved 14CO2 from alkali trap solutions; extraction of volatile PUF plug traps; analysis of soil subsamples for extractable residual radioactivity, parent SFA, and transformation products; and total residual radioactivity by total oxidation. Air flow to the incubation flasks was temporarily suspended during sampling procedures. Subsamples of soil (10 g) taken during the incubation period were stored in sealed capped vials at 20 8C to 50 8C until analysis. At soil sampling, each incubation flask was weighed, and any reduction resulting from moisture loss was compensated for by the addition of an appropriate volume of sterile water, which was mixed into the soil to maintain the required WHC. For the 5 8C and 20 8C treatments, incubation time was 135 d, and soil sampled at 0 d, 3 d, 4 d, 7 d, 10 d, 21 d, 22 d, 35 d, 64 d, and 135 d. The 10 8C treatments were incubated for 136 d, and soil was sampled at 0 d, 8 d, 14 d, 21 d, 28 d, 42 d, 63 d, and 136 d. Alkali trap solutions were sampled and changed at 1 d, 2 d, 3 d, 4 d, 5 d, 7 d, 10 d, 14 d, 21 d, 28 d, 35 d, 42 d, 56 d, 63 d, 70 d, 84 d, 98 d, 112 d, and 126 d after spiking, with some minor variations later in the incubation period (see Results section). The PUF plugs were sampled and changed on days 10, 21, 49, 77, and 112 of incubation, and stored in sealed glass vials at –20 8C until analysis. Sample extraction and analysis
Total oxidation of soil subsamples to determine the total residual activity of 14C was carried out using a PerkinElmer 307 oxidizer. Quality assurance measures included a control soil combustion blank, a control soil spiked with a solution of [14C]toluene (0.02 mL of 305 Bq/mL) to assess combustion efficiency, combustion blanks to assess the carryover of radioactivity between samples, and a comparative standard prepared from the [14C]-toluene Spec-Chec solution (PerkinElmer). Soil (0.25– 1 g) was weighed into paper combustion cones, 0.1 mL of Combustaid solution (PerkinElmer) was added, and the contents were capped with a combustion pad. Samples were combusted for 2.5 min. Evolved 14CO2 was trapped and eluted in 10 mL of Carbosorb E (PerkinElmer) and mixed with 10 ml of Permafluor E scintillation cocktail (PerkinElmer) in 23-mL vials; then 14C radioactivity was measured by liquid scintillation counting. Residues of SFA were extracted from soil subsamples (2.5 g), using a modification of the extraction method described by Wright et al. [20]. Quality assurance samples for each batch of samples included an extraction solution blank; a control soil blank; a control soil spiked with the internal standard chloroacetic acid, SFA, and HAA; and a comparative standard prepared by dispensing the same spike compound solutions into a 10-mL volumetric flask. Approximately 2.5 g of soil was spiked with the internal standard (chloroacetic acid) and extracted with 5 mL of magnesium carbonate–saturated aqueous solution by sonication for 20 min (Bandelin Sonorex Digital 10P), followed by shaking for 20 min at 300 rpm (IKA KS 501). The soil slurry was centrifuged (Hettich Rotanta 460R) to separate the soil and aqueous extract, which was decanted into a plastic centrifuge tube. The extraction of the soil was repeated, and the combined aqueous extract was filtered through a glass fiber filter (Labserve, 25-mm diameter). The concentration of SFA, HAA, and chloroacetic acid in the soil extracts was determined by liquid chromatography mass–spectrometry (LCMS), and total extractable [14]C residues by liquid scintillation counting. The LCMS analysis was performed on a ThermoFisher Scientific Surveyor liquid chromatography system coupled to an autosampler and heated column compartment. Separations of 5-mL injections of aqueous soil extracts were carried out using a
Degradation of sodium fluoroacetate in soil
Environ Toxicol Chem 33, 2014
Zorbax Extend-C18 column (150 mm 2.1 mm, 5 mm, Agilent) with an isocratic mobile phase consisting of 10% methanol–water with tributylamine (5 mM) and formic acid (5 mM) at a flow rate of 0.2 mL/min. The column temperature was maintained at 35 8C throughout. Mass spectrometric analysis was completed on a ThermoFisher Scientific LCQ Deca-ion trap mass spectrometer system using atmospheric pressure chemical ionization in the negative mode. Vaporizer temperature was 250 8C, sheath and auxiliary nitrogen gas flows were 35 U and 10 U, respectively, and the heated capillary was maintained at 200 8C. Mass spectral data were acquired between 4 min and 15 min using selected ion monitoring (SIM). The compound-specific SIM ions monitored during the analysis were 75 m/z for HAA, 77 m/z for SFA, and 93.5 m/z for chloroacetic acid. The alkali trapping solution was mixed with 15 mL of Ultima Gold XR scintillation cocktail (PerkinElmer), and the trapped and accumulated radioactivity was measured by liquid scintillation counting. Samples were counted for a period of 20 min or a threshold of 0.5% uncertainty, whichever was reached sooner. Counting data (counts/min) were quench-corrected to provide the corresponding activity in disintegrations/min (DPM). The activity of alkali trapping solutions and aqueous sample extracts was corrected against a quench curve constructed using certified 14C quench standards prepared in Ultima Gold liquid scintillation fluid (National Institute of Standards and Technology [NIST]).The activity of oxidized samples was corrected against a quench curve constructed from a dilution series of Carbosorb E in Permafluor E scintillation fluid spiked with a known amount of 14C radioactivity ([14C]-toluene). This in-house–prepared quench curve was validated against certified NIST 14C quench standards. The PUF plugs were transferred to empty 20-mL polypropylene solid-phase extraction tubes mounted in a vacuum manifold, and 5 mL of MilliQ water was added. Each plug was sequentially compressed and extended to adsorb and express water. This process was repeated 4 times to ensure even saturation; then the plug was fully compressed and vacuumapplied to collect the aqueous extracts in glass scintillation vials. The extraction process was repeated and combined with the first extract, and the extracted radioactivity was measured by liquid scintillation counting. Quality assurance samples consisting of a solution blank, control blank PUF plug, PUF plug spiked with 1.8 kBq of [14C]-1080, and [14C]-1080 spike comparative were prepared and extracted with each batch of extracted PUF plugs. Statistical analysis
Except for the 14CO2 mineralization profiles, all estimates were calculated on a soil dry weight basis. Statistical analysis was undertaken using Microsoft Excel 2003 and 2007. The decline of SFA concentration in soil with time was described according to first-order kinetics C ðtÞ ¼ C 0 expðktÞ
ð1Þ
where C0 is the initial concentration of SFA spiked in soil (mg/kg soil), C(t) is the concentration of SFA in soil (mg/kg soil) at time t, t is the time (d), and k is the degradation rate constant (d). The rate constant k was estimated by fitting Equation 1 to the experimental data with SigmaPlot for Windows Version 10.0 (Release 10.0.1.2.) using the Levenberg–Marquardt algorithm to determine parameter values. For nonlinear curve fitting, the initial parameter value for C0 was set to the theoretically applied amount of SFA initially spiked into the soil. SigmaPlot’s Automatic Initial Parameter Estimate Functions were used to
1051
estimate parameter k, and the initial concentration was optimized to obtain best fit to the data, based on the assumption that the initial concentration was subject to experimental error. Between 10 iterations and 27 iterations were performed to derive solutions. Parameter values together with their standard errors and the adjusted r2 values of the best-fit solution were obtained. Half-lives for the degradation of SFA were calculated from k using the standard equation t1/2 ¼ ln(0.5)/k, or t1/2 ¼ 0.6932/k. The 50%, 75%, and 90% disappearance times of SFA (DT50, DT75,and DT90 values) were estimated by substituting the SFA concentration at time zero (C0), defining C(t) as 0.5 C0 or 0.1 C0 , and entering the estimated k for each soil treatment into Equation 1 and solving for the time in days (t). Total recovered radioactivity obtained from each flow-through incubation flask was calculated as the sum of [14C]-radioactivity remaining in the soil within the flask at time x, the cumulative [14C]-radioactivity removed in each soil subsample at time x, and the cumulative radioactivity measured as [14C]-CO2 at time x. Total recovered [14C]-radioactivity for each flask was expressed as a percentage of the total [14C]-SFA radioactivity spiked into the soil and added to each flask at the initiation of the study and measured by total sample oxidation and liquid scintillation counting of [14C]-radioactivity. Quality assurance in soil spikes, temperature conditions, and analytical methods
Mean and standard deviation 14C-radioactivity of subsampled spiked soils (DPM [14C]/g) were used to calculate the percentage relative standard deviation (%RSD) as a measure of spike homogeneity. Mean ( 95% confidence interval [CI]), median, minimum, and maximum %RSD (n ¼ 36) were 4.8% 1.2%, 3.8%, 0.6%, and 13.8%, respectively, demonstrating a homogeneous distribution of SFA spikes within and between individual soil treatments. Higher %RSD values were obtained for spikes in the 60% WHC soil treatments and particularly the Matiri soil type, which maintained a higher level of clumping after shaking and mixing. Two outlier values obtained for spikes in Matiri soil at 60% WHC corresponded to minimum and maximum values of 26 500 DPM/g and 40 700 DPM/g soil. The 95% confidence limit for mean radioactivity across all spiked soil treatments (n ¼ 36) was 34 600 800 DPM/g soil, further demonstrating the reproducibility of the spiking procedure. The accuracy of the adopted spiking procedure was determined by comparing the activity of [14C]-SFA in spiked soil (by total sample oxidation) against that of comparative spike solutions prepared from the same [14C]SFA solutions used to spike the soil treatments, and reported as percentage spike recovery. The 95% confidence limit for the mean recovery of [14C]-SFA radioactivity and corresponding median were 97% 5% and 95%, respectively (n ¼ 34), excluding the 2 outliers identified. The acceptable temperature range specified by OECD guideline 307 [14] for 20 8C and 10 8C treatments is 2 8C, and a similar level of precision was applied to the additional 5 8C soil treatments. Temperatures in the 5 8C and 10 8C treatment facilities were maintained within the specified range for the duration of the study. In the 20 8C treatment facility, recorded temperatures spiked above the accepted range of 20 8C 2 8C on only 42 out of a total of 9302 individual temperature measurements when the facility was accessed during very warm outside conditions. The temperatures in excess of 22 8C within the facility occurred for a period of 10 h on the second to last day of the study, ranging from 22.01 8C to 22.34 8C; however, the controlled environment was maintained within the
1052
Environ Toxicol Chem 33, 2014
range specified by the OECD guideline for the duration of the transformation experiment. Mean radioactivity ( 95% CI) measured in blank control soils (n ¼ 23) by total soil sample oxidation was 62 DPM 6 DPM, equivalent to the background level of radioactivity. The 95% CI for combustion efficiency of the total sample oxidizer, as determined by the recovery of [14C]toluene spiked onto control soil and combusted, was 99 2% (n ¼ 23), and the 95% CI for the mean level of radioactivity carryover between consecutive combusted samples was 0.1 0.1% (n ¼ 23). The limits of detection for SFA, HAA, and chloroacetic acid in soil extracts analyzed by LCMS were 0.025 mg/mL, 0.1 mg/ mL, and 0.5 mg/mL, respectively, equating to limits of detection of 0.1 mg/kg, 0.4 mg/kg, and 2.0 mg/kg, respectively, on a soil dry-weight basis. The relatively higher limits of detection for chloroacetic acid was inconsequential because it was used as the internal standard at a nominal concentration of 60 mg/kg. Nominal concentrations of the SFA and HAA quality assurance recovery spikes added to control soil were equivalent to 37 mg/ kg and 50 mg/kg. Mean recoveries ( 95% CI) fell within the 70% to 110% range of acceptance specified by OECD guideline 307 [13], being 104 1% (n ¼ 459) for chloroacetic acid, 103 3% (n ¼ 22) for SFA, and 75 7% (n ¼ 22) for HAA. RESULTS
Mineralization of [14C]-SFA
At 20 8C the fastest rates of SFA mineralization occurred within the first 30 d in all soil types. At 133 d the mean cumulative percentages of original [14C]-SFA mineralized were 77% (Kaingaroa at 60% WHC), 70% (Kaingaroa at 35% WHC), 84% (Matiri at 60% WHC), 63% (Matiri at 35% WHC), 60% (Orikaka at 60% WHC), and 70% (Orikaka at 35% WHC). Mineralization occurred in 20 8C sterile controls at relatively slow rates, reaching cumulative percentage mineralized endpoints ranging from 21.8% to 44.4%. At 10 8C the fastest rates of mineralization occurred within the first 40 d, and by 143 d the mean cumulative percentages of original [14C]-SFA mineralized were 77% (Kaingaroa at 60% WHC), 64% (Kaingaroa at 35% WHC), 84% (Matiri at 60% WHC), 59% (Matiri at 35% WHC), 73% (Orikaka at 60% WHC), and 64% (Orikaka at 35% WHC). Some mineralization occurred in 10 8C sterile controls after the first 40 d to total cumulative percentage endpoints ranging from 3.2% to 21%. At 5 8C the fastest mineralization rates of SFA occurred over the first 70 d, and at 135 d the mean cumulative percentages of original [14C]-SFA mineralized were 66% (Kaingaroa at 60% WHC), 67% (Kaingaroa at 35% WHC), 75% (Matiri at 60% WHC), 74% (Matiri at 35% WHC), 83% (Orikaka at 60% WHC), and 80% (Orikaka at 35% WHC). Some mineralization occurred in 5 8C sterile controls after the first 40 d, reaching cumulative percentage endpoints ranging from 1% to 30%. As an example, Figure 1 shows the percentage of mineralized [14C]-SFA produced during the incubation of Orikaka soil at 35% WHC. [14C]-SFA residues in PUF plugs, soil, and soil extracts and radioactivity mass balance
Analysis of PUF plugs indicated that negligible volatilization of [14C]-SFA or [14C]-HAA occurred over the course of the study. The positioning of the 14C atom at carbon 1 within the SFA molecule corresponds to the carbon atom within the carboxylic acid functional group. Therefore the absence of 14C radioactivity in the PUF plugs demonstrates that volatilization of [14C]-SFA or [14C]-HAA did not occur, or was negligible. The
G. Northcott et al.
Figure 1. Percentage of mineralized sodium fluoroacetate ([14C]-SFA) produced during the incubation of Orikaka soil at 35% water-holding capacity (WHC). Maximum and minimum values are marked by smaller version of symbols corresponding to ^ ¼ 20 8C treatment, * ¼ 10 8C treatment, ~ ¼ 5 8C treatment, ^ ¼ 20 8C sterile treatment, ¼ 10 8C sterile treatment, and D ¼ 5 8C sterile treatment.
PUF plugs from 20 8C treatments sampled at 10 d of incubation (the period at which greatest mineralization of SFA occurred) showed radioactivity to a maximum of 1300 DPM, which represented
DEGRADATION RATE OF SODIUM FLUOROACETATE IN THREE NEW ZEALAND SOILS GRANT NORTHCOTT,y DWAYNE JENSEN,y LUCIA YING,y and PENNY FISHER*z yPlant & Food Research Ruakura, Hamilton, New Zealand zLandcare Research, Lincoln, New Zealand (Submitted 20 October 2013; Returned for Revision 20 November 2013; Accepted 23 January 2014) Abstract: The degradation rate of sodium fluoroacetate (SFA) was assessed in a laboratory microcosm study incorporating 3 New Zealand soil types under different temperature (5 8C, 10 8C, or 20 8C) and soil moisture (35% or 60% water holding capacity) conditions using guideline 307 from the Organisation for Economic Co-operation and Development. A combination of nonlabeled and radiolabeled 14CSFA was added to soil microcosms, with sampling and analysis protocols for soil, soil extracts, and evolved CO2 established using liquid scintillation counting and liquid chromatography–mass spectrometry. Degradation products of SFA and their rates of formation were similar in the 3 soil types. The major degradation pathway for SFA was through microbial degradation to the hydroxyl metabolite, hydroxyacetic acid, and microbial mineralization to CO2, which constituted the major transformation product. Temperature, rather than soil type or moisture content, was the dominant factor affecting the rate of degradation. Soil treatments incubated at 20 8C displayed a more rapid loss of 14C-SFA residues than lower temperature treatments. The transformation half-life (DT50) of SFA in the 3 soils increased with decreasing temperature, varying from 6 d to 8 d at 20 8C, 10 d to 21 d at 10 8C, and 22 d to 43 d at 5 8C. Environ Toxicol Chem 2014;33:1048–1058. # 2014 SETAC Keywords: 1080
Degradation time
Hydroxyacetic acid
Pesticide
Sodium fluoroacetate
Soil
natural soils. Temperature and moisture content, as regulators of microbial activity, were expected to affect the rate of SFA degradation in New Zealand soils [12]. In 2007, a regulatory reassessment of SFA [13] by the Environmental Risk Management Authority New Zealand (now Environmental Protection Agency, New Zealand) determined that existing data on the degradation pathways and rates of SFA in soil were limited in scope and applicability, and were not conducted in accordance with international test methods. The present study addressed a recommendation [13] to generate additional data using guideline 307 of the Organisation for Economic Co-operation and Development (OECD) [14] under aerobic test conditions. In addition to the test temperatures of 20 8C and 10 8C stipulated in the test guideline, an additional treatment of 5 8C was included to simulate the New Zealand winter conditions when SFA is most commonly applied. An additional test treatment simulating relatively dry soil conditions was also included. Test soils used needed to represent regions of New Zealand subject to aerial application of SFA bait, and encompass variability in soil organic matter and clay content.
INTRODUCTION
Sodium fluoroacetate (SFA; FCH2CO2Na) is registered as a pesticide in New Zealand and Australia, for delivery in bait to control pest mammals, and is commonly referred to as 1080. In New Zealand, aerial or bait station application of cereal pellet or chopped carrot bait containing up to 0.15% SFA by weight is used for broad-scale management of introduced brushtail possums (Trichosurus vulpecula), rodents (Rattus spp.) and, to a lesser extent, rabbits (Oryctolagus cuniculus). After application, bait remaining uneaten on the ground is subject to dissipation and degradation through exposure to rainfall [1]. Exposure to environmental moisture results in water-soluble SFA leaching from bait into litter and soil [2]. Thus, the degradation of SFA and the rate of formation of degradation products in soil are of interest for environmental risk assessment. Potential degradation pathways of SFA in soil include chemical hydrolysis in solution and biotic mechanisms of metabolism or mineralization. Degradation in solution begins with dissociation of the sodium ion followed by hydrolysis of fluoroacetate to produce fluoromethane and bicarbonate. Metabolism of SFA following ingestion by soil organisms occurs via conversion to fluorocitrate within mitochondria [3] or by root uptake of fluoroacetate in solution followed by metabolism by plants. Mineralization of SFA through microbial degradation produces fluoride ion and hydroxyacetic acid (HAA; also termed glycolate/glycolic acid). Previous studies have identified soil bacteria that could utilize fluoroacetate as a sole carbon source through enzymatic cleavage of the carbon– fluorine bond [4–6] as well as the enzymes responsible and their gene coding in specific bacteria [7]. Previous investigations [5,6,8–11] have identified a significant role of some bacteria and fungi in the degradation of SFA in
MATERIALS AND METHODS
Within the protocol described by OECD guideline 307 [14], we used a combination of nonlabeled and radiolabeled SFA to determine the rate of degradation and the nature and rates of formation and decline of transformation products. The guideline recommends that test soil moisture content should be 40% to 60% water-holding capacity (WHC). In the present study we used 60% WHC as 1 treatment variable and also included 35% WHC as another, to address the question of degradation of SFA under drier soil conditions. In addition to the test temperatures of 10 8C and 20 8C specified by OECD guideline 307 [14], an additional 5 8C treatment was incorporated into the experiment to better represent the range of temperatures prevalent in New Zealand, particularly as SFA baits are often applied during winter when soil temperatures are lowest. The assessment was
* Address correspondence to [email protected]. Published online 29 January 2014 in Wiley Online Library (wileyonlinelibrary.com). DOI: 10.1002/etc.2536 1048
Degradation of sodium fluoroacetate in soil
Environ Toxicol Chem 33, 2014
completed to Good Laboratory Practice standard at Plant & Food Research’s Food and Biological Chemistry laboratory between May 2009 and December 2010. More detailed technical description of the test system and analytical methods used in the assessment are provided by Northcott [15]. Test soil selection, sampling, and preparation
Three test soils were selected using spatial data to identify representative areas of New Zealand over which SFA bait has been applied aerially and is likely to be applied in future. The geographic information system (GIS) data were sourced from the VectorNet database (Animal Health Board, Wellington, New Zealand) for aerial bait applications for possum control during July 2007 to February 2008, and from the Department of Conservation (Research Development and Improvement Division, Christchurch, New Zealand) for aerial application for possum and rodent control over areas of conservation estate in the period 2006 to 2007. These GIS data were matched to areas of the dominant New Zealand Soil Classification Soil Orders (National Soil Database, Landcare Research) to produce a ranking of soil order by area exposed to aerial bait application. Three orders (brown soils, podzols, and pumice soils) represented the greatest combined area exposed to SFA baits after aerial application and the most diversity in terms of regional location and variation of physical, chemical, and biological characteristics. Collection sites (Table 1) were selected using the National Soils Database (Landcare Research) to identify existing reference sites and provide descriptive information.
1049
Soil from the 3 reference sites was sampled to a maximum depth of 20 cm. At each site, soil was taken from at least 5 locations approximately 10 m apart. Large stones, vegetation and leaf litter, visible earthworms, and the upper layer of grass and root material were removed by hand. The collected soil from each location was transferred into large plastic bags and loosely tied to allow exchange of air. The soil samples were transported chilled (3 cm were broken apart manually, and for each soil type the contents of all sample bags were mixed together by hand, with visible pieces of plant material and stone removed during mixing. The fully mixed composite soil for each sampling site and soil order were transferred into loosely tied, labeled plastic bags and stored at 5 8C. Approximately 10 kg of each composite soil sample was sieved to 2 mm, and soil dry weight and gravimetric water content (GWC) were determined by drying triplicate subsamples of sieved soil fractions to constant weight at 105 8C. Waterholding capacity was measured using the method of Hoper [16]. Maximum WHC was calculated as the corresponding GWC using the soil dry weight data. The GWC of sieved field-moist soils was either close to or greater than that required to achieve 60% or 35% WHC, necessitating partial air-drying of the soils before readjusting to the required experimental WHC, as described by Northcott [15]. Sterile soil for control treatments was prepared by autoclaving preweighed portions of soil 3 times. The soil was autoclaved at 121 8C (1 bar) for 30 min and maintained within the sealed
Table 1. Details of test soil sampling and descriptions
Order Podzol
Location and reference from database Orikaka Sandy Loam, West Coast, South Island GR L29 2425373 593 4737
Global positioning system coordinate and date of sample collection
Existing data
248250 35.500 E
ZOH, Humose orthic podzol
Coarse sand 13%
598340 71.400 N
Beech–podocarp forest, Profile description 0-B3
Medium sand 11%
a
2 January 2009
Brown Soil
Matiri, West Coast, South Island GR L30 2410730 5909418
248100 75.600 E 598090 39.200 N
BLA Acid allophanic brown Upland yellow-brown earth, lowland podocarp forest
8 January 2009
Pumice soil
Kaingaroa (sand), Taupo, North Island GR U18 2795100 6257100
278950 10.600 E
M/W Welded impeded pumice
628570 10.700 N 12 January 2009
Mānuka exotic forest, sphagnum
Physical and chemical characteristicsb
Fine sand 9% Silt 13% Clay 54% pH 4.6 Total C 12.6% Total N 0.50% MBC 1362 Coarse sand 1% Medium sand 6% Fine sand 30% Silt 39% Clay 24% pH 3.7 Total C 12.0% Total N 0.46% MBC 1790 Coarse sand 3% Medium sand 9% Fine sand 18% Silt 43% Clay 27% pH 5.4 Total C 7.68% Total N 0.33% MBC1014
a Horizon descriptions accessible on New Zealand soils database; search by series name or identity number: http://soils.landcareresearch.co.nz/contents/ SoilData_NSD_ReportsFlat.aspx?currentPage¼SoilData_NSD_ReportsFlat&menuItem¼SoilData b Particle Size Distribution testing of Fine Earth Fraction (New Zealand Classification), and measurements of pH, total carbon and nitrogen, and microbial biomass carbon (MBC; mg/kg) undertaken by Landcare Research, Hamilton, New Zealand.
1050
Environ Toxicol Chem 33, 2014
autoclave until it had cooled to 30 8C (5–6 h). The autoclave cycle was repeated, and the soils were then left to incubate at ambient temperature for 24 h to allow surviving bacterial spores to germinate. Following this period of incubation, the soils were autoclaved at 122 8C (1.35 bar) for 35 min. All subsequent manipulations of the sterilized soils were conducted under laminar-flow and sterile conditions. The 5 8C, 10 8C, and 20 8C treatments were housed in separate temperature-controlled rooms monitored by calibrated electronic monitors and minimum– maximum thermometers. Pre-equilibration of all test soils at 5 8C, 10 8C, and 20 8C was undertaken for a minimum of 4 wk, with any germinating seedlings removed by hand. Soil spiking procedure
An initial moisture adjustment was used to introduce SFA to soil as an aqueous spike using a modification of the method described by Brinch et al. [17], thereby avoiding the potential alteration of the sorption behavior of SFA in the test soils [18]. Soil within each flask was spiked with a mixture of nonlabeled and radiolabeled SFA (fluoroacetic acid, [1-14C] sodium salt) equivalent to 18 mg of SFA, which was the nominal quantity in a single 12-g cereal pellet bait containing 0.15% (w/w) SFA. The aqueous spike was evenly distributed over quarter portions of soil that were subsequently combined with the remaining volume of soil. Sterile soil treatments were spiked with the same procedure but using soil that had been autoclaved within the previous 2 d and a solution of SFA prepared in sterile MilliQ water (Millipore). Spiking of soil in the 3 temperature treatments was staggered to accommodate the sampling schedule. Immediately after spiking, subsamples (0.5–0.9 g) of soil from each treatment were weighed into individual paper combustion cones, with 6 replicates and 10 replicates taken from each spiked sterile and nonsterile treatment, respectively. These replicate subsamples were stored at –50 8C to 80 8C until analysis by sample oxidation to confirm the activity and homogeneity of the [14C]-SFA portion of the spikes. Flow-through incubation flask system
The test system was based on the design specified in OECD guideline 307 [14], and by the Soil Science Society of America [19]. Spiked soil treatments in 250-mL glass Schott bottles were incubated in flow-through, dark conditions under 3 temperature (5 8C, 10 8C, or 20 8C) and 2 soil moisture (35% or 60% WHC) treatments. Eighteen flasks were incubated at each temperature (n ¼ 6 for each soil type). Each group of 6 flasks comprised duplicate nonsterile and a sterile spiked soil for each of the 35% and 60% WHC treatments. Each flask contained an equivalent dry weight of 105 g of soil, but the total final weight of soil prepared for each varied depending on the adjusted GWC required to attain the treatment WHC. Flasks were sealed with a screw cap containing a Teflon plug and Viton O-ring insert, through which a controlled air flow (5 mL/min) was delivered and then directed through a trap containing a polyurethane foam (PUF) plug, and bubbled into glass scintillation vials containing alkali solution (5 mL of 1 M NaOH) to trap CO2/14CO2. Sample types and sampling intervals
The 10 8C treatments were spiked first, followed by the 20 8C treatments 4 d later and the 5 8C treatments 2 d later. The OECD guideline 307 [14] specifies that the study should normally not exceed 120 d, but we used a longer incubation period to ensure a complete characterization of the decline of SFA concentrations over time. Sampling and analyses included measurement of
G. Northcott et al.
evolved 14CO2 from alkali trap solutions; extraction of volatile PUF plug traps; analysis of soil subsamples for extractable residual radioactivity, parent SFA, and transformation products; and total residual radioactivity by total oxidation. Air flow to the incubation flasks was temporarily suspended during sampling procedures. Subsamples of soil (10 g) taken during the incubation period were stored in sealed capped vials at 20 8C to 50 8C until analysis. At soil sampling, each incubation flask was weighed, and any reduction resulting from moisture loss was compensated for by the addition of an appropriate volume of sterile water, which was mixed into the soil to maintain the required WHC. For the 5 8C and 20 8C treatments, incubation time was 135 d, and soil sampled at 0 d, 3 d, 4 d, 7 d, 10 d, 21 d, 22 d, 35 d, 64 d, and 135 d. The 10 8C treatments were incubated for 136 d, and soil was sampled at 0 d, 8 d, 14 d, 21 d, 28 d, 42 d, 63 d, and 136 d. Alkali trap solutions were sampled and changed at 1 d, 2 d, 3 d, 4 d, 5 d, 7 d, 10 d, 14 d, 21 d, 28 d, 35 d, 42 d, 56 d, 63 d, 70 d, 84 d, 98 d, 112 d, and 126 d after spiking, with some minor variations later in the incubation period (see Results section). The PUF plugs were sampled and changed on days 10, 21, 49, 77, and 112 of incubation, and stored in sealed glass vials at –20 8C until analysis. Sample extraction and analysis
Total oxidation of soil subsamples to determine the total residual activity of 14C was carried out using a PerkinElmer 307 oxidizer. Quality assurance measures included a control soil combustion blank, a control soil spiked with a solution of [14C]toluene (0.02 mL of 305 Bq/mL) to assess combustion efficiency, combustion blanks to assess the carryover of radioactivity between samples, and a comparative standard prepared from the [14C]-toluene Spec-Chec solution (PerkinElmer). Soil (0.25– 1 g) was weighed into paper combustion cones, 0.1 mL of Combustaid solution (PerkinElmer) was added, and the contents were capped with a combustion pad. Samples were combusted for 2.5 min. Evolved 14CO2 was trapped and eluted in 10 mL of Carbosorb E (PerkinElmer) and mixed with 10 ml of Permafluor E scintillation cocktail (PerkinElmer) in 23-mL vials; then 14C radioactivity was measured by liquid scintillation counting. Residues of SFA were extracted from soil subsamples (2.5 g), using a modification of the extraction method described by Wright et al. [20]. Quality assurance samples for each batch of samples included an extraction solution blank; a control soil blank; a control soil spiked with the internal standard chloroacetic acid, SFA, and HAA; and a comparative standard prepared by dispensing the same spike compound solutions into a 10-mL volumetric flask. Approximately 2.5 g of soil was spiked with the internal standard (chloroacetic acid) and extracted with 5 mL of magnesium carbonate–saturated aqueous solution by sonication for 20 min (Bandelin Sonorex Digital 10P), followed by shaking for 20 min at 300 rpm (IKA KS 501). The soil slurry was centrifuged (Hettich Rotanta 460R) to separate the soil and aqueous extract, which was decanted into a plastic centrifuge tube. The extraction of the soil was repeated, and the combined aqueous extract was filtered through a glass fiber filter (Labserve, 25-mm diameter). The concentration of SFA, HAA, and chloroacetic acid in the soil extracts was determined by liquid chromatography mass–spectrometry (LCMS), and total extractable [14]C residues by liquid scintillation counting. The LCMS analysis was performed on a ThermoFisher Scientific Surveyor liquid chromatography system coupled to an autosampler and heated column compartment. Separations of 5-mL injections of aqueous soil extracts were carried out using a
Degradation of sodium fluoroacetate in soil
Environ Toxicol Chem 33, 2014
Zorbax Extend-C18 column (150 mm 2.1 mm, 5 mm, Agilent) with an isocratic mobile phase consisting of 10% methanol–water with tributylamine (5 mM) and formic acid (5 mM) at a flow rate of 0.2 mL/min. The column temperature was maintained at 35 8C throughout. Mass spectrometric analysis was completed on a ThermoFisher Scientific LCQ Deca-ion trap mass spectrometer system using atmospheric pressure chemical ionization in the negative mode. Vaporizer temperature was 250 8C, sheath and auxiliary nitrogen gas flows were 35 U and 10 U, respectively, and the heated capillary was maintained at 200 8C. Mass spectral data were acquired between 4 min and 15 min using selected ion monitoring (SIM). The compound-specific SIM ions monitored during the analysis were 75 m/z for HAA, 77 m/z for SFA, and 93.5 m/z for chloroacetic acid. The alkali trapping solution was mixed with 15 mL of Ultima Gold XR scintillation cocktail (PerkinElmer), and the trapped and accumulated radioactivity was measured by liquid scintillation counting. Samples were counted for a period of 20 min or a threshold of 0.5% uncertainty, whichever was reached sooner. Counting data (counts/min) were quench-corrected to provide the corresponding activity in disintegrations/min (DPM). The activity of alkali trapping solutions and aqueous sample extracts was corrected against a quench curve constructed using certified 14C quench standards prepared in Ultima Gold liquid scintillation fluid (National Institute of Standards and Technology [NIST]).The activity of oxidized samples was corrected against a quench curve constructed from a dilution series of Carbosorb E in Permafluor E scintillation fluid spiked with a known amount of 14C radioactivity ([14C]-toluene). This in-house–prepared quench curve was validated against certified NIST 14C quench standards. The PUF plugs were transferred to empty 20-mL polypropylene solid-phase extraction tubes mounted in a vacuum manifold, and 5 mL of MilliQ water was added. Each plug was sequentially compressed and extended to adsorb and express water. This process was repeated 4 times to ensure even saturation; then the plug was fully compressed and vacuumapplied to collect the aqueous extracts in glass scintillation vials. The extraction process was repeated and combined with the first extract, and the extracted radioactivity was measured by liquid scintillation counting. Quality assurance samples consisting of a solution blank, control blank PUF plug, PUF plug spiked with 1.8 kBq of [14C]-1080, and [14C]-1080 spike comparative were prepared and extracted with each batch of extracted PUF plugs. Statistical analysis
Except for the 14CO2 mineralization profiles, all estimates were calculated on a soil dry weight basis. Statistical analysis was undertaken using Microsoft Excel 2003 and 2007. The decline of SFA concentration in soil with time was described according to first-order kinetics C ðtÞ ¼ C 0 expðktÞ
ð1Þ
where C0 is the initial concentration of SFA spiked in soil (mg/kg soil), C(t) is the concentration of SFA in soil (mg/kg soil) at time t, t is the time (d), and k is the degradation rate constant (d). The rate constant k was estimated by fitting Equation 1 to the experimental data with SigmaPlot for Windows Version 10.0 (Release 10.0.1.2.) using the Levenberg–Marquardt algorithm to determine parameter values. For nonlinear curve fitting, the initial parameter value for C0 was set to the theoretically applied amount of SFA initially spiked into the soil. SigmaPlot’s Automatic Initial Parameter Estimate Functions were used to
1051
estimate parameter k, and the initial concentration was optimized to obtain best fit to the data, based on the assumption that the initial concentration was subject to experimental error. Between 10 iterations and 27 iterations were performed to derive solutions. Parameter values together with their standard errors and the adjusted r2 values of the best-fit solution were obtained. Half-lives for the degradation of SFA were calculated from k using the standard equation t1/2 ¼ ln(0.5)/k, or t1/2 ¼ 0.6932/k. The 50%, 75%, and 90% disappearance times of SFA (DT50, DT75,and DT90 values) were estimated by substituting the SFA concentration at time zero (C0), defining C(t) as 0.5 C0 or 0.1 C0 , and entering the estimated k for each soil treatment into Equation 1 and solving for the time in days (t). Total recovered radioactivity obtained from each flow-through incubation flask was calculated as the sum of [14C]-radioactivity remaining in the soil within the flask at time x, the cumulative [14C]-radioactivity removed in each soil subsample at time x, and the cumulative radioactivity measured as [14C]-CO2 at time x. Total recovered [14C]-radioactivity for each flask was expressed as a percentage of the total [14C]-SFA radioactivity spiked into the soil and added to each flask at the initiation of the study and measured by total sample oxidation and liquid scintillation counting of [14C]-radioactivity. Quality assurance in soil spikes, temperature conditions, and analytical methods
Mean and standard deviation 14C-radioactivity of subsampled spiked soils (DPM [14C]/g) were used to calculate the percentage relative standard deviation (%RSD) as a measure of spike homogeneity. Mean ( 95% confidence interval [CI]), median, minimum, and maximum %RSD (n ¼ 36) were 4.8% 1.2%, 3.8%, 0.6%, and 13.8%, respectively, demonstrating a homogeneous distribution of SFA spikes within and between individual soil treatments. Higher %RSD values were obtained for spikes in the 60% WHC soil treatments and particularly the Matiri soil type, which maintained a higher level of clumping after shaking and mixing. Two outlier values obtained for spikes in Matiri soil at 60% WHC corresponded to minimum and maximum values of 26 500 DPM/g and 40 700 DPM/g soil. The 95% confidence limit for mean radioactivity across all spiked soil treatments (n ¼ 36) was 34 600 800 DPM/g soil, further demonstrating the reproducibility of the spiking procedure. The accuracy of the adopted spiking procedure was determined by comparing the activity of [14C]-SFA in spiked soil (by total sample oxidation) against that of comparative spike solutions prepared from the same [14C]SFA solutions used to spike the soil treatments, and reported as percentage spike recovery. The 95% confidence limit for the mean recovery of [14C]-SFA radioactivity and corresponding median were 97% 5% and 95%, respectively (n ¼ 34), excluding the 2 outliers identified. The acceptable temperature range specified by OECD guideline 307 [14] for 20 8C and 10 8C treatments is 2 8C, and a similar level of precision was applied to the additional 5 8C soil treatments. Temperatures in the 5 8C and 10 8C treatment facilities were maintained within the specified range for the duration of the study. In the 20 8C treatment facility, recorded temperatures spiked above the accepted range of 20 8C 2 8C on only 42 out of a total of 9302 individual temperature measurements when the facility was accessed during very warm outside conditions. The temperatures in excess of 22 8C within the facility occurred for a period of 10 h on the second to last day of the study, ranging from 22.01 8C to 22.34 8C; however, the controlled environment was maintained within the
1052
Environ Toxicol Chem 33, 2014
range specified by the OECD guideline for the duration of the transformation experiment. Mean radioactivity ( 95% CI) measured in blank control soils (n ¼ 23) by total soil sample oxidation was 62 DPM 6 DPM, equivalent to the background level of radioactivity. The 95% CI for combustion efficiency of the total sample oxidizer, as determined by the recovery of [14C]toluene spiked onto control soil and combusted, was 99 2% (n ¼ 23), and the 95% CI for the mean level of radioactivity carryover between consecutive combusted samples was 0.1 0.1% (n ¼ 23). The limits of detection for SFA, HAA, and chloroacetic acid in soil extracts analyzed by LCMS were 0.025 mg/mL, 0.1 mg/ mL, and 0.5 mg/mL, respectively, equating to limits of detection of 0.1 mg/kg, 0.4 mg/kg, and 2.0 mg/kg, respectively, on a soil dry-weight basis. The relatively higher limits of detection for chloroacetic acid was inconsequential because it was used as the internal standard at a nominal concentration of 60 mg/kg. Nominal concentrations of the SFA and HAA quality assurance recovery spikes added to control soil were equivalent to 37 mg/ kg and 50 mg/kg. Mean recoveries ( 95% CI) fell within the 70% to 110% range of acceptance specified by OECD guideline 307 [13], being 104 1% (n ¼ 459) for chloroacetic acid, 103 3% (n ¼ 22) for SFA, and 75 7% (n ¼ 22) for HAA. RESULTS
Mineralization of [14C]-SFA
At 20 8C the fastest rates of SFA mineralization occurred within the first 30 d in all soil types. At 133 d the mean cumulative percentages of original [14C]-SFA mineralized were 77% (Kaingaroa at 60% WHC), 70% (Kaingaroa at 35% WHC), 84% (Matiri at 60% WHC), 63% (Matiri at 35% WHC), 60% (Orikaka at 60% WHC), and 70% (Orikaka at 35% WHC). Mineralization occurred in 20 8C sterile controls at relatively slow rates, reaching cumulative percentage mineralized endpoints ranging from 21.8% to 44.4%. At 10 8C the fastest rates of mineralization occurred within the first 40 d, and by 143 d the mean cumulative percentages of original [14C]-SFA mineralized were 77% (Kaingaroa at 60% WHC), 64% (Kaingaroa at 35% WHC), 84% (Matiri at 60% WHC), 59% (Matiri at 35% WHC), 73% (Orikaka at 60% WHC), and 64% (Orikaka at 35% WHC). Some mineralization occurred in 10 8C sterile controls after the first 40 d to total cumulative percentage endpoints ranging from 3.2% to 21%. At 5 8C the fastest mineralization rates of SFA occurred over the first 70 d, and at 135 d the mean cumulative percentages of original [14C]-SFA mineralized were 66% (Kaingaroa at 60% WHC), 67% (Kaingaroa at 35% WHC), 75% (Matiri at 60% WHC), 74% (Matiri at 35% WHC), 83% (Orikaka at 60% WHC), and 80% (Orikaka at 35% WHC). Some mineralization occurred in 5 8C sterile controls after the first 40 d, reaching cumulative percentage endpoints ranging from 1% to 30%. As an example, Figure 1 shows the percentage of mineralized [14C]-SFA produced during the incubation of Orikaka soil at 35% WHC. [14C]-SFA residues in PUF plugs, soil, and soil extracts and radioactivity mass balance
Analysis of PUF plugs indicated that negligible volatilization of [14C]-SFA or [14C]-HAA occurred over the course of the study. The positioning of the 14C atom at carbon 1 within the SFA molecule corresponds to the carbon atom within the carboxylic acid functional group. Therefore the absence of 14C radioactivity in the PUF plugs demonstrates that volatilization of [14C]-SFA or [14C]-HAA did not occur, or was negligible. The
G. Northcott et al.
Figure 1. Percentage of mineralized sodium fluoroacetate ([14C]-SFA) produced during the incubation of Orikaka soil at 35% water-holding capacity (WHC). Maximum and minimum values are marked by smaller version of symbols corresponding to ^ ¼ 20 8C treatment, * ¼ 10 8C treatment, ~ ¼ 5 8C treatment, ^ ¼ 20 8C sterile treatment, ¼ 10 8C sterile treatment, and D ¼ 5 8C sterile treatment.
PUF plugs from 20 8C treatments sampled at 10 d of incubation (the period at which greatest mineralization of SFA occurred) showed radioactivity to a maximum of 1300 DPM, which represented
