Environ. Sci. Technol. 2004, 38, 6147-6154
Uptake of Weathered DDT in Vascular Plants: Potential for Phytoremediation ALISSA I. LUNNEY, BARBARA A. ZEEB, AND KENNETH J. REIMER* Environmental Sciences Group, Royal Military College of Canada, P.O. Box 17000, Station Forces, Kingston, Ontario K7K 7B4, Canada
Since the discovery of its insecticidal properties and its subsequent widespread use, DDT [2,2-bis(chlorophenyl)1,1,1-trichloroethane] has accumulated in the environment, having a wide range of adverse effects on nontarget species. Due to their hydrophobicity, DDT and other persistent organic pollutants are difficult to remove from contaminated soils, and increasingly so through time as weathering occurs. Phytoremediation is an emerging plant-based technology that may be used to cost-effectively remove or neutralize contaminants in the environment. For some phytoremediation strategies, it must first be possible to translocate hydrophobic chemicals across the root and through the shoot via an aqueous transpiration stream. The objective of this study was to compare the ability of five plant varieties (zucchini, tall fescue, alfalfa, rye grass, and pumpkin) to mobilize and translocate DDT. Plants were grown in the greenhouse in soil contaminated with DDT and its metabolites, DDD and DDE (ΣDDT refers to all of DDT, DDD, and DDE) at two concentrations (high ∼3700 ng/g, and low ∼150 ng/g). All trays were covered with laboratory Parafilm to limit volatilization. Cucurbita pepo species (pumpkin and zucchini) achieved the highest translocation and bioaccumulation factors, and also extracted the highest absolute amounts of ΣDDT from both the high and low ΣDDT soils. In the high ΣDDT soil treatment, pumpkin accumulated 1519 ng of ΣDDT in the roots and 57 536 ng of ΣDDT in the shoots, and zucchini accumulated 2043 ng of ΣDDT in the roots and 35 277 ng of ΣDDT in the shoots. With the exception of alfalfa and pumpkin, principal component analysis detected no preferential translocation or transformation of ΣDDT compounds within the plant. The success of the Cucurbita pepo species in this study to extract and translocate such hydrophobic molecules may be related to their high transpiration volume, large above-ground biomass, and composition of root exudates. This suggests potential for their application in phytoremediation.
Introduction Persistent organic pollutants (POPs) are a class of hydrophobic organic contaminants that are characterized by long half-lives in the environment and a potential for bioaccumulation through the food chain (1). DDT [2,2-bis(chlo* Corresponding author e-mail:
[email protected]. 10.1021/es030705b CCC: $27.50 Published on Web 10/08/2004
2004 American Chemical Society
rophenyl)-1,1,1-trichloroethane] has accumulated in the environment since World War II in conjunction with its use as an insecticide against forest and agricultural pests, and against insect vectors of typhus and malaria (2). It has two isomeric forms (2,4-DDT and 4,4′-DDT), depending on the relative positioning of chlorine atoms on the two phenyl rings in its structure. In the environment, DDT can be degraded microbiologically or abiotically to DDD [1,1-dichloro-2,2bis(chlorophenyl)ethane] or DDE [1,1-dichloro-2,2-bis(chlorophenyl)ethylene] (3), and like DDT, these compounds each have a 2,4- and a 4,4′- isomeric form. Under anaerobic conditions, DDT can be converted sequentially to DDE and then DDD, whereas in an aerobic environment, it can be converted directly to DDD (4). Technical grade DDT (TG-DDT) is the active ingredient in commercial pesticide formulations, where the primary ingredients are 4,4′-DDT (77.1%) and 2,4-DDT (14.9%); DDD and DDE compounds are present in the mixture as byproducts of the manufacturing process (4,4′-DDE, 4.0%; 2,4-DDE, 0.1%; 4,4′-DDD, 0.3%; 2,4-DDD, 0.1%) (5). When DDT is measured in the environment, it is common to report the sum of all of its isomers and metabolites as total DDT (ΣDDT). Within a decade of its use, it became evident that DDT accumulates in the fatty tissues of mammalian and avian species, causing a number of adverse effects on their reproduction, growth, and immunocompetence (6, 7). As a result, DDT has since been banned in Canada, the United States, and other nations. Under the Stockholm Convention of 2001, DDT is now restricted for use as an anti-malaria agent in approximately nineteen countries where no other safe, affordable, and effective alternatives are available (8). Given the number of sites previously and currently being contaminated with DDT, this compound is a worthwhile candidate for phytoremediation, an emerging lower-cost alternative to conventional remediation strategies (9). By definition, phytoremediation is the use of vegetation for the treatment of contaminated soils, sediments, groundwater, or surface water (9, 10). Depending on the nature of the contaminant, plant, and soil, phytoremediation may be achieved in several different ways. First, plants may indirectly degrade organic contaminants in the biologically active rhizosphere, a process termed rhizodegradation (9). Some organic contaminants may be taken up with soil water, enter and traverse the plant root, and then be translocated (i.e., phytoextracted) to the aboveground portion of the plant where they may be metabolized to intermediates, or mineralized to carbon dioxide and water by catabolic enzymes such as dehalogenases in the case of organochlorines (9). This process is termed phytodegradation. Alternatively, if a contaminant cannot be metabolized, it may become associated with lignin and be stored in the plant. This is termed lignification or phytoaccumulation (9). Finally, by a process known as phytovolatilization, contaminant molecules may be transported to the leaves and released during transpiration (9). Hydrophobic chemicals have octanol-water partition coefficient (log Kow) values greater than 3.5; for DDT, this value is around 6 (6). Sicbaldi et al. (11) have found that in soybean plants, hydrophobic compounds with midrange log Kow values between 2 and 3 are translocated most efficiently within the plant, and translocation efficiency decreases for compounds with log Kow on either side of that optimum. That is, highly hydrophobic pesticides easily permeate plasma membranes, but do not partition well into the xylem sap because of their affinity for lipidic sites in the cell, and vice versa for more hydrophilic compounds (11). VOL. 38, NO. 22, 2004 / ENVIRONMENTAL SCIENCE & TECHNOLOGY
9
6147
The highly hydrophobic nature of DDT makes mobilization and transport of this compound through a weathered soil matrix and into plant systems, at least in theory, very problematic. In weathered soil that has been exposed to natural wetting and drying cycles, water molecules from precipitation entering the soil matrix cause hydrophobic contaminants to become more strongly sorbed to organic matter also present in the soil environment (12). This renders contaminants less bioavailable to uptake by plants. Nonetheless, there have been some interesting greenhouse and field-scale studies of weathered DDE uptake in legume and grass species, as well as plants from the genera Cucurbita and Cucumis (13-16). Similarly promising results obtained by the Environmental Sciences Group (ESG) at the Royal Military College of Canada (RMC) for phytoremediation of PCBs also influenced the present study (17). The work described here extends upon those results with a greenhouse-scale study of phytoremediation of weathered DDT-contaminated soil. Specifically, this study examined the potential of each of five plant species for phytoremediation of weathered DDT-contaminated soil at two concentrations. Plants involved in this study are rye grass (Lolium multiflorum), tall fescue (Festuca arundinacea Schreb.), alfalfa (Medicago sativa), zucchini (Cucurbita pepo L. cv. Senator), and pumpkin (Cucurbita pepo cv. Howden). These species were chosen based on promising results in other phytoremediation studies and general physiological characteristics that might contribute to their success as phytoremediators (14, 18-21).
Experimental Section (1) Greenhouse. The zucchini (Cucurbita pepo L. cv. Senator hybrid) seeds used in this study were purchased from The Cooperative Company (Kingston, ON), and pumpkin (Cucurbita pepo cv. Howden) seeds were purchased from the Ontario Seed Company (Waterloo, ON). The tall fescue (Festuca arundinacea cv. Common koknee), rye grass (Lolium multiflorum cv. Common forage), and alfalfa (Medicago sativa cv. Common variety No. 1) seeds were purchased from Bishop Seeds (Belleville, ON). Zucchini, tall fescue, rye grass, and alfalfa plants were grown in the RMC greenhouse in July through September 2002. Pumpkin plants were grown in January through March 2003. For both the summer and winter trials, the light regime was kept constant using growth lights, and the temperature in the greenhouse was maintained at 23 ( 2 °C. All five species were planted separately in bottomperforated 28 × 1 × 6 cm seed trays in a 6 cm soil depth and were grown under the following conditions in singular replicate: (i) uncovered, uncontaminated potting soil; (ii) covered, uncontaminated potting soil; (iii) covered low concentration ΣDDT soil (∼150 ng/g); (iv) covered high concentration ΣDDT soil (∼3700 ng/g). Contaminated soil samples were collected by the Environmental Sciences Group (ESG) at RMC from Kittigazuit, NWT, a former Long-Range Aid to Navigation (LORAN) site in the western Canadian Arctic that received DDT applications from 1947 to 1950 (22). Interestingly, due to low ambient temperatures, the DDT mixture has degraded very little since its application; soils collected from this area have largely maintained the ratio of isomers and byproducts found in the original TG-DDT. Soil samples collected from different areas on the site were sieved at 0.4 mm and homogenized for the low and high ΣDDT soil treatments. Control soil was an ordinary potting soil mix. After seedling emergence, covered treatments had a layer of laboratory Parafilm stretched over the seed tray to limit volatilization and subsequent loss of ΣDDT to the atmosphere. For zucchini and pumpkin, small holes were carefully made in the Parafilm to allow the seedlings to emerge. For alfalfa, rye grass, and tall fescue, individual 6148
9
ENVIRONMENTAL SCIENCE & TECHNOLOGY / VOL. 38, NO. 22, 2004
sprouts pushed themselves through the Parafilm with little or no assistance. Seed trays were bottom-watered as needed and plant heights were measured on a regular basis. Plants were harvested when plots of height data showed a decline in growth rate, which occurred at approximately 50 days after planting. For the Cucurbita pepo plants, this was generally coincident with the flowering stage. The sex of each flower was not specifically noted, however most flowers fell off the plants and did not lead to fruit formation. Zucchini, tall fescue, rye grass, and alfalfa were each harvested by removing an entire plant from the soil, and subsequently separating plant roots and shoots. Pumpkin plants were similarly harvested by separating roots and the lower stem, defined as the above-ground portion of the plant below the oldest leaf; the upper stem above the oldest leaf; the leaves, and the flowers. All plant samples were washed thoroughly under running water to remove any soil particles present. The fresh weights and lengths of all plant samples were recorded. Bulk low and high ΣDDT soils were sampled prior to planting to determine initial contaminant concentrations. For zucchini, alfalfa, and rye grass, composite soil samples were subsequently obtained at harvest from each seed tray to account for any variability in ΣDDT concentration in the soil. For pumpkin, a greater effort was made to ensure that soil sampled from the tray had been affected by the roots; composite soil samples were obtained from the near root zone, which was defined as the soil that fell off the root after harvest, as per White (13). (2) Laboratory. All soils and plants were analyzed at the Analytical Services Unit (ASU) at Queen’s University in Kingston, ON. (a) Soils. For quality assurance/quality control, all analytical runs included a blank, control, and sample duplicate. Blanks were prepared with approximately 40 g of anhydrous sodium sulfate and 20 g of Ottawa sand, then spiked with 100 µL of a surrogate, decachlorobiphenyl (DCBP). Controls were prepared as above, but were additionally spiked with 50 µL of an organochlorine pesticide mix prepared at ASU (Supelco product 49151). Soil samples from the greenhouse were prepared in the same way as the blanks with anhydrous sodium sulfate, Ottawa sand, and DCBP. For each soil sample, 10 g (( 0.5 g) of wet soil was weighed into thimbles. A subsample of each sample was dried for at least 24 h at 100 °C and then weighed for moisture determination. Samples were extracted in a Soxhlet apparatus for 4-6 h with 250 mL of methylene chloride (Fisher Scientific, D151-4). The extract was concentrated by rotoevaporation to 1 mL, and the solvent was exchanged to hexanes (Fisher Scientific H303-4) using a Bu ¨ chi Rotavapor R-114. The extract was applied to a Florisil column and diluted with hexanes to 10 mL. A fraction of the eluant was transferred to a vial for analysis by gas chromatography/electron capture detection (GC/ECD) to determine the composition and quantity of ΣDDT in soils. Concentrations of ΣDDT in soils were calculated from a DCBP standard. DCBP concentrations of 10, 50, 100, and 200 ppb were used to develop an initial calibration curve, and a 100 ppb standard was used for each batch of samples once the calibration was established. All samples were corrected for surrogate recovery. GC/ECD analysis used an Agilent 6890 gas chromatograph and electron capture detector. Results obtained for 4,4′-DDT, 2,4-DDT, 4,4′-DDE, 2,4-DDE, 4,4′-DDD, and 2,4-DDD were expressed as nanograms of pesticide per gram dry weight of soil (ng/g). The analytical detection limit for this method was 10 ng/g for all of the DDT, DDE, and DDD compounds. All blanks were below sample detection limits (
