Environ. Sci. Technol. 2006, 40, 4788-4793
Trichloroethylene Uptake by Apple and Peach Trees and Transfer to Fruit BRANDON K. CHARD, WILLIAM J. DOUCETTE,* JULIE K. CHARD, BRUCE BUGBEE, AND KYLE GORDER† Utah State University, Utah Water Research Laboratory, 8200 Old Main Hill, Logan, Utah 84322-8200
A greenhouse study was conducted to quantify 14Ctrichloroethylene (TCE) uptake and transfer into the edible fruit of apple and peach trees. Trees were subsurface irrigated with solutions of 14C [TCE] that bracketed groundwater concentrations (5 and 500 µg/L) found in residential areas surrounding Hill Air Force Base, UT, where trace amounts of TCE had been found in several fruits during a preliminary field survey. Nondosed control trees were grown within the canopy of the dosed trees and in a separate greenhouse. Tissue samples were analyzed for 14C and TCE using combustion/liquid scintillation counting (LSC) and headspace/gas chromatography/mass spectrometry (HS/ GC/MS). Tissue was also extracted and analyzed by GC/ MS for dichloroacetic acid (DCAA), trichloroacetic acid (TCAA), and trichloroethanol (TCEt), three specific TCE metabolites that have been previously identified in laboratory and field studies. No 14C was detected in the nonexposed control trees. Exposed trees contained levels of 14C that were proportional to the exposure concentration. 14C concentrations were greatest in leaves followed by branches and fruits. At the end of the study, TCE was detected only in roots implying that the 14C in the leaves, branches, and fruit was associated with unidentified nonvolatile TCE transformation products and/or is nonextractable. However, TCAA and DCAA were positively identified only in leaves collected during the first year from an apple tree exposed to the high dose treatment. Additional data for other chemicals and fruit trees are needed to better understand the potential transfer of organic compounds to edible fruit.
Introduction Trichloroethylene (TCE) has been widely used as a cleaning and degreasing agent at industrial and military sites around the United States, including Hill Air Force Base (AFB) in northern Utah. Past use and disposal practices resulted in TCE contamination of shallow aquifers underlying Hill AFB and surrounding communities. During natural attenuation and phytoremediation assessments at the base, TCE was found in mature trees growing above shallow contaminated groundwater at concentrations roughly proportional to the TCE groundwater concentrations in the vicinity (1-3). The * Corresponding author phone: (435) 797-3178; fax: (435) 7973663; e-mail:
[email protected]. † Present address: Environmental Management Division, Hill Air Force Base, UT. 4788
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 15, 2006
detection of TCE in on-base trees, together with the migration of shallow groundwater plumes into residential areas containing numerous fruit trees and gardens, prompted concerns that TCE could be taken up and transferred into edible fruit. A preliminary field survey conducted during the fall of 2001 identified TCE in several fruits collected from trees growing around Hill AFB using a headspace/gas chromatography method previously developed for tree cores. TCE concentrations ranged from the method detection limit (0.1 µg/kg) to 18 µg/kg fresh weight of fruit (4). A screening-level human health risk calculation considering the potential carcinogenic effects associated with ingestion of TCE in fruit suggested that concentrations above 15 µg/kg (fresh weight) may raise regulatory concerns. This screening-level calculation was made with the following exposure assumptions: exposure duration ) 30 years (24 as an adult and 6 as a child), body weight ) 70 kg (adult) and 15 kg (child), ingestion rate for fruit ) 500 g/day, exposure frequency ) 350 days/ year, and an averaging time of 70 years. The California EPA oral slope factor of 0.013 (mg/kg day)-1 was used, and the target risk was 1 × 10-6. While many laboratory and field studies have documented the uptake of TCE into trees such as hybrid poplars (e.g., refs 5-8), the potential transfer into fruits has received much less attention. Trapp et al. (9) developed a model specifically for fruit trees that predicted the transfer of TCE into apples based on the physical-chemical properties of TCE and physiological characteristics of apple trees, but no experimental data were available for model validation. Previous laboratory studies using 14C [TCE] reported only limited transfer of 14C into tomatoes and carrots and did not identify TCE above method detection limits (10, 11). In an attempt to better understand the preliminary field survey results and the potential transfer of 14C [TCE] into edible fruit, a two-year greenhouse study using dwarf apple and peach trees was conducted. The goal of this study was to determine the extent of 14C and TCE uptake by fruit trees as a function of two exposure concentrations and to examine the potential transfer of 14C and TCE into edible fruit.
Materials and Methods Experimental Design. Seven-year old dwarf apple (Malus domestica Borkh cv. ‘Golden Delicious’) and five-year old peach (Prunus persica Batsch cv. ‘Redhaven’) trees were obtained from the Utah Agricultural Experiment Station at Utah State University. The trees were previously used in fruit production experiments and had a history of reliable fruit production. To avoid transplant shock, the small caliper trees (1.9 cm diameter and 1.2 m tall) were kept in their original containers (apple: 5 gal (18.9 L) and peach: 7 gal (26.5 L)) that contained a heterogeneous mix of soil, sand, and organic matter. The soil mixture was supplemented with a slowrelease fertilizer (Osmocote, Scotts Co., Columbus, OH) prior to the start of the study. A soil-water sampler (Irrometer Co., Riverside, CA), tensiometer (Soil Moisture Equipment Co., Goleta, CA), and subsurface irrigation system were installed in each container (Figure 1). The irrigation system consisted of a drip emitter (Rainbird Co., Glendale, CA) fitted with a manual control valve. The irrigation lines were connected to 5 gal (19 L) carboys that contained water or water containing known concentrations of TCE. The plants were manually watered when the soil moisture content fell below field capacity (less than -30 kPa) as indicated by the tensiometer. Watering was performed to minimize the generation of any leachate, yet maintain adequate soil moisture. However, as a precaution, 10.1021/es060156k CCC: $33.50
2006 American Chemical Society Published on Web 06/28/2006
FIGURE 1. Schematic of TCE exposure system for individual trees. a secondary container filled with a layer of activated carbon was used to capture any leachate that reached the container bottom. The surface of each container was also covered with a layer of activated carbon (2.5 cm) separated from the soil with a layer of sand to minimize volatilization of TCE into the greenhouse. Twelve trees (six of each species) were exposed in triplicate to a nominal concentration of either 5 or 500 µg/L 14C [TCE] (specific activity: 12.5 mCi/mmol), while six (three of each species) served as nondosed controls. One control tree of each species was grown within the canopy of the treatment trees to distinguish between translocated and atmospherically deposited 14C. Two additional control trees of each species were kept in a separate greenhouse under the same conditions. The greenhouses were temperature-controlled at 24/ 18 °C day/night and equipped with high pressure sodium lamps on a 16 h photoperiod. Light levels at the top of the canopy ranged from 200 to 700 µmol m-2 s-1, depending on the amount of natural sunlight. The supplemental lighting, used to maintain a minimum photosynthetic photon flux of 200 µmol m-2 s-1, was turned off when sufficient natural light levels were attained. After the first fruit harvest, the trees were watered and moved to a walk-in cooler (5 °C) for about 4 months (mid-November to mid-March). Irrigation and Soil Water Sampling. Prior to dosing, replicate background samples of the irrigation water and soil water were collected and analyzed by liquid scintillation counting (LSC) for 14C and by headspace/gas chromatography/mass spectrometry (HS/GC/MS) for TCE. For 14C analysis, 2 mL samples were collected directly into 7 mL scintillation vials containing 5 mL of scintillation cocktail. For TCE analysis, 10 mL samples were collected in 20 mL headspace vials and immediately sealed. These samples were collected to verify that no TCE was initially present and to determine background 14C levels in the soil or water. After the initial dosing, triplicate samples of irrigation water were collected daily for the first week, then weekly (before and after refilling the carboys) for the duration of the study. Soil water was sampled at regular intervals early in the study but was discontinued after several weeks when the ceramic sampling cups became clogged and reliable samples could not be collected. Plant Tissue Sampling. Plant tissue samples (leaves, branches, and fruit) were collected in triplicate and analyzed for 14C and TCE when the fruits were ripe. Fruit and leaves were collected from the tree by gloved hand, and samples of branches were cut from the tree with stainless steel pruning shears. For mature fruit, samples of peel and flesh were separately collected and analyzed. For TCE analysis, samples (2-5 g fresh weight) were placed directly into a 20 mL
headspace vial containing 10 mL of matrix modifying solution (MMS) [saturated sodium chloride solution adjusted to pH 2 using a phosphoric acid] and analyzed within 14 days by HS/GC/MS (4). For 14C analysis, samples were stored in 500 mL wide-mouthed glass jars at 4 °C (e14 days), cut into small pieces, mixed, and analyzed by combustion/LSC (∼1 g fresh weight). Extra tissue not used for 14C or TCE analysis was frozen and later used for metabolite analysis. Whole Tree Sampling: 14C and TCE Distribution. At the end of the second growing season, four trees (one high dose and one control each of the peach and apple) were sacrificed for more rigorous determination of 14C and TCE distribution. Samples of leaves, branches, trunk (every 15 cm from the top of the tree), and roots from each of the four trees were analyzed for 14C and TCE as previously described. Samples of the growth media (soil/organic matter mixture), the activated carbon and sand used to cover the growth media, and activated carbon contained in the secondary container were also collected and analyzed using the same general approach as described for the plant tissue analysis. 14C Analysis by Liquid Scintillation Counting. The amount of 14C in the aqueous samples was determined directly by liquid scintillation counting (LSC) (Beckman LS 1701, Beckman Instruments, Inc., Fullerton, CA), after adding 2 mL of sample to an appropriate volume of Ready Gel scintillation cocktail. To determine 14C concentrations in the plant tissues, triplicate ∼1 g (fresh weight) samples of tissue were combusted at 900 °C (R. J. Harvey Model OX-600 biological oxidizer, Hillsdale, NJ), and the evolved 14CO2 was collected in a solution of 50% Ready Gel/40% methanol/10% monoethanolamine and directly analyzed by LSC. Method detection limits (MDLs) were calculated from the results of the seven replicate lowest concentration spikes and were determined to be 0.85 µg/L for high dose aqueous samples, 0.01 µg/L for low dose aqueous samples, 7.7 µg/kg fresh weight for high dose vegetation samples, and 0.11 µg/kg fresh weight for low dose vegetation samples. The differences in detection limits for the high and low dose treatments are due to differences in the ratio of labeled to unlabeled TCE used for each treatment. TCE Analysis by HS/GC/MS. Headspace samples (2 mL) were introduced into a Hewlett-Packard 6890 GC/5973 MS (running EnviroQuant, Chemstation G1701AA version D.03.00 data acquisition and analysis software) using a Tekmar 7000HT Headspace Analyzer/Autosampler. The autosampler platen/sample temperature was set to 50 °C, the sample equilibrium time was 10 min, and the transfer line and sample loop temperatures were 180 °C. Chromatographic conditions were as follows: DB-624, 30 m × 0.25 mm, 1.4 µm film thickness column (J&W Scientific, Folsom, CA); helium carrier gas at 0.7 mL/min (3.52 psi); temperature program 35 °C for 2 min to 170 °C at 30 °C/min, then 170-230 °C at 70 °C/min with a 1 min hold at the final temperature (total run time ) 8.36 min); split/splitless inlet vent flow 10.4 mL/min; and split ratio 15:1. The concentrations of TCE in the plant tissue samples were determined indirectly from the concentration of TCE in the headspace. External standards (minimum of five different concentrations), made by spiking known amounts of a commercial TCE standard (Supelco, Bellefonte, PA) into MMS, were used to define the relationship between the headspace and the MMS concentrations. The standards were made directly in headspace vials just prior to calibration. The GC/MS was operated in Selected Ion Monitoring (SIM) mode. Headspace methods are based on the equilibrium distribution of the compound of interest between the three phases within the vial (headspace, MMS, and tissue). Spike recovery experiments indicated that about 25-45% of the mass of TCE within the headspace vial at equilibrium was VOL. 40, NO. 15, 2006 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
4789
sorbed to the plant tissue (i.e., fruit and trunk core) depending on sample size and tissue type. The plant tissue concentrations reported in this paper were not corrected for spike recoveries. Sample-specific MDLs depend on the tissue type and sample size, but an average MDL for all tissue types was approximately 0.1 µg/kg fresh weight. TCE Metabolite Analysis. Tissue samples were also analyzed for three specific TCE metabolites previously identified in the laboratory (e.g., refs 5 and 12) and field (e.g., refs 13-15) studies: dichloroacetic acid (DCAA), trichloroacetic acid (TCAA), and trichloroethanol (TCEt) following the procedure described by Orchard et al. (5). Fresh leaf, branch, and root samples were homogenized using a commercial coffee grinder. Known weights of homogenized tissue were placed into 60 mL centrifuge tubes with 10 mL of a 0.25 N sodium hydroxide solution, shaken for 15 min with a reciprocating shaker, and then centrifuged for 10 min at 9000 rpm to separate the solution from the tissue. The supernatant was transferred into another centrifuge tube, and the extraction procedure was repeated two additional times. The combined aqueous extracts were acidified to a pH of less than 1 with 50% sulfuric acid and subsequently extracted three additional times with 7 mL of methyl t-butyl ether (MTBE) by shaking for 15 min followed by centrifugation for 5 min at 3000 rpm. The supernatants from the triplicate MTBE extractions were combined and brought to exactly 25 mL in a volumetric flask. The extracts were dried over anhydrous sodium sulfate, derivatized with diazomethane-saturated ether (2 mL aliquot), and analyzed by GC/MS using the same conditions as previously described for TCE. With these conditions, the retention times were 11.59 min for TCEt, 11.15 min for TCAA, and 9.60 min for DCAA.
Results and Discussion Trichloroethylene Concentrations in the Irrigation Water. Average TCE concentrations (( standard deviation) in the irrigation water, calculated from 14C measurements, were approximately 4.4 ( 1.2 µg/L (n ) 810) for the low dose treatment and 450 ( 124 µg/L (n ) 780) for the high dose treatment during year one. In year two, concentrations were approximately 5.3 ( 1.1 µg/ L (n ) 190) for the low dose treatment and 671 ( 122 µg/L (n ) 210) for the high dose treatment. The slightly higher concentrations in the second year were likely due to improvements in sampling and the water delivery system that decreased volatilization losses of TCE from the reservoirs. The ratio of unlabeled TCE to 14Clabeled was 1:1 for the low dose and 66:1 for the high dose for both years. The concentrations of TCE determined by LSC were within 10-20% of those determined by HS/GC/ MS. Transpiration/Fruit Production. The average rate of water use for the peach trees (0.6 L/day in year one and 0.7 L/day in year two) was nearly twice that of the apple trees (0.3 L/day year one and 0.4 L/day year two) because of their greater leaf area. Total water use per tree for both years ranged from 104.9 to 183.0 L for the peaches and from 39.3 to 85.5 L for the apples. Water use by both species was independent of exposure concentration, and no indications of phytotoxicity were observed. The apple trees averaged 14 fruit per tree the first year (average mass at harvest, 66.7 g/fruit), and the peach trees averaged 3 fruit per tree (average mass at harvest, 112.6 g/fruit) with two peach trees (one high dose and one low dose) not producing any fruit. In the second year, the peaches produced more fruit than the apple trees. The apple trees averaged 2 fruit per tree (average mass at harvest, 84.2 g/fruit), with three trees (one high dose and two low dose) not producing any fruit, while the peaches produced an average of 16 fruit per tree (average mass at harvest, 101.9 g/fruit). 4790
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 40, NO. 15, 2006
TABLE 1. Summary of 14C Concentrations Obtained at Mature Fruit Harvest (Fruit Harvested with 5 Days of Each Other) Expressed as µg of TCE equiv/L of Water or kg of Wet Plant Tissue (95% Confidence Interval for Triplicate Samples from Triplicate Trees) apple high exposure
apple low exposure
peach high exposure
peach low exposure
456 (19.8) 198 (55.7) 310 (53.6) 10.60 (0.46) 3.27 (0.60)
4.64 (0.22) 1.61 (0.17) 3.64 (2.24) 0.13 (0.04) 0.09 (0.04)
606 (30.9) 259 (89.5) 562 (214.9) 63.6 (5.27) 44.2 (4.98)
7.62 (1.27) 3.18 (0.66) 8.44 (2.4) 0.73 (0.10) 0.55 (0.04)
Year 1 (2002) irrigation water leaves branches fruit peel fruit flesh
446 (21.6) 294 (87.1) 189 (30.1) 8.17 (0.81) 9.23 (1.23)
irrigation water leaves branches fruit peel fruit flesh
626 (38.4) 1210 (492) 483 (197) 66.6 (8.02) 23.1 (2.42)
4.32 (0.20) 2.51 (0.71) 2.20 (0.52) 0.28 (0.02) 0.12 (0.01)
Year 2 (2003) 5.42 (0.32) nsa ns ns ns
a ns ) not sampled. Vegetation samples were not collected for low exposure apple trees in 2003 as they did not produce fruit.
The biennial cycle of fruit production observed in this study is common for healthy, productive fruit trees where the fruit is not thinned (16). 14C concentration in Aboveground Tissues. All aboveground tissue samples (leaf, stem, and fruit) collected from the dosed apple and peach trees during both years were found to contain 14C (Table 1). No 14C was detected in any samples collected from the control trees (nondosed), indicating that the 14C was transferred from root to shoot via the transpiration stream rather than by volatilization from the soil surface and deposition onto the various aboveground tissues. The 14C concentrations (µg of TCE equiv/kg of dry plant tissue) in the high dose plant tissues were about 100 times higher than the corresponding (branches, leaves, and fruit) low dose trees, roughly proportional to the differences in the irrigation water concentrations. The relative plant tissue concentration differences as a function of exposure concentration has been observed in previous studies with hybrid poplar trees (e.g., refs 5 and 17). The concentrations of 14C in the dosed apple trees were generally greatest in the leaves, followed by the branches and then the fruit. There was a significant difference (P < 0.05 ANOVA: Single Factor) in 2003 but not in 2002 for leaves and branches. The trend of 14C concentrations being higher in leaves than the branches follows the trend observed for TCE metabolites in previous laboratory (5) and field (15) studies. In the dosed peach trees, leaf concentrations were significantly greater than branch concentrations in 2003 but not 2002. In both years and for both species, branch and leaf concentrations of 14C were significantly higher than fruit concentrations (Table 1). Overall, 14C concentrations in the fruit were typically a factor of 10 lower than in the branches or leaves for both the apples and the peaches (Table 1). In a separate ancillary experiment, 14C-labeled TCE was spiked directly into the center of several mature apples purchased from a local grocer using a microliter syringe. After spiking, the syringe hole was immediately sealed with a silicone caulk. The 14C was found to move rapidly from the center of the apple to the peel and eventually disappeared approximately 2 weeks after the single spike (data not shown). This suggests that volatilization is a potential loss mechanism for TCE introduced into the fruit. Water loss through transpiration is also well-documented for many types of fruit (18). Generally, all tissue types contained higher levels of 14C (P < 0.05 ANOVA: Single Factor) in the second year of
TABLE 2. Final Average Concentrations (95% Confidence Interval) of TCE and 14C (TCE equiv) for the Two Destructively Sampled Treesa TCEb (µg/kg)
14Cb
(µg/kg)
distribution fresh moisture of 14C within c tree (%)d mass (g) content (%)
Apple
fruite leavese branches trunk roots
