1. Introduction
The remediation and management of soils, waters and sediments contaminated with persistent organic pesticides (POPs) has become a global environmental priority, mainly because POPs are hazardous to life. POPs, originally manufactured for pest and disease control, crop production and other industrial processes, have a high degree of persistency in the environment, potential for long-range transportability, the ability to bioaccumulate in the lipid components of living systems and a high toxicity, even at very low concentrations [
1,
2]. Considering these problems associated with the use of POPs, the latter were banned worldwide [
3], and research into effective clean-up methods evolved over time. Most of the developed world has intensive clean-up programs for POP-contaminated sites by developing suitable and effective methodologies [
4]; however, the research on such methodologies, especially on phytoremediation in SSA countries, is still rather scanty.
All clean-up methods currently in use can be grouped into three main categories. Category one is a collection of methods based on the principle of containment in which the contamination is sealed in a protective barrier to limit its release, and it may include practices such as land filling [
5] and solidification and stabilization practices [
6]. Category two involves a collection of technologies designed to destroy the POPs either through (a) non-combustion methods, such as dehalogenation, or (b) combustion methods like incineration or thermal desorption, which break down POPs to simple compounds such as CO
2, methane (CH
4) and water (H
2O) [
7,
8,
9]. Category three, on the other hand, includes a collection of technologies that involve the extraction of the contaminant from the matrix (soil, sediment or water) through either (i) concentration or (ii) the liberation/stripping of the contaminant, paving way to the treatment of the liberated contaminant through convenient methods of category one or two above. Examples of methods in category three include ex situ soil washing, ex situ solvent extraction, in situ soil flushing, soil vapor extraction, ex situ bioremediation and in situ bioremediation [
10,
11].
The deployment of any one or a combination of the methods mentioned above for the remediation of contaminated sites can be influenced by considerations for technical, economical and operational feasibility. As a rule of thumb, high technology demanding, expensive, time-sensitive and highly efficient technologies such as incineration can be feasible in the developed world, while time consuming, low-input yet efficient technologies such as bio- and phytoremediation are both suitable and feasible in most of the developing world.
The use of higher plants as detoxifiers, filters or traps of POPs and other pollutants is a well-proven approach, but this approach, also known as phytoremediation, is slow and difficult due to the inherently low bioavailability of POPs to plants [
12,
13]. A handful of research efforts have reported the successful testing of various higher plants such as
Cucurbita pepo and
Cichorium intybus as potential tools for the phytoremediation of POPs, including DDT- and metabolite-contaminated sites [
14,
15].
The success of phytoremediation depends mainly on the properly selected plant species [
16], which show fast growth, large biomass production in a short time, a developed root system, a higher tolerance to pollutants and the ability to accumulate toxins in above-ground parts, and resistance to diseases, pests and weather conditions. The test plants used in this study meet all or most of these properties. Furthermore, the uptake and accumulation of pollutants in crop plants vary with species or cultivars, the typology of pollutants as well as the level of contamination [
17].
In some cases, edible crop species can exhibit all or most of the above properties of a good phytoaccumulator plant. Such plants can be used as agents of phytoremediation efforts if the produced phytobiomass is protected from animal and human consumption to avoid their movement up the food chain [
18]. This study was designed, therefore, to test the potential of calabash—a locally available non-edible plant—along with simsim, pumpkin, sweet potato and finger millet to phytoaccumulate DDT from contaminated soils. The results would inform both the health risks associated with growing the edible species on the contaminated soils as well as their potential for use in the phytoremediation of contaminated soils.
2. Materials and Methods
2.1. Description of Study Sites
The Plant Protection Office (PPO) at the Tengeru site is in Arusha, Tanzania at a longitude and latitude of 3.39092 and 36.799200. Records show that the PPO Tengeru site was once used as a locust control center for Tanzania and thus had tons of supplies of pesticides, including malathion and lindane, which are stored there [
19]. In 2000, pesticides packed in drums and stored in open air started rusting and were subsequently buried in soil at the PPO Tengeru backyard (
Scheme 1).
NHC Morogoro study site, on the other hand, is located at a latitude and longitude of 6.083333 and 37.666667 in the heart of the Morogoro Municipality (
Scheme 1). The site was originally used as a center for the formulation and repackaging of DDT and endosulfan, among other pesticides, for onward distribution to end users across the country [
3]. Following an international ban on the production and use of these chemicals due to perceived side effects to humans and the environment, activities were stopped, and the site was eventually abandoned in 1997 [
3].
2.2. Soil Sampling and Analysis
Soil sampling at the NHC Morogoro and PPO Tengeru sites was carried out for two main purposes, namely for soil fertility and physicochemical analysis, on one hand, and to estimate the level of contamination by POPs, on the other hand. Soil samples for physicochemical and soil fertility analysis were taken using an auger to a depth of 30 cm from the soil surface. The hand auger used for surface soil sampling was rinsed with distilled water after each sampling. Samples were transferred to the pre-cleaned amber bottles. Each soil sample was 500 g, enough for specific analysis and a control sample. Samples were transported to the laboratory at Sokoine University of Agriculture for analysis. Soil fertility analysis parameters included macronutrients N, P, K, S, Ca and Mg and micronutrients Mn, Fe, Cu and Zn as well as CEC, EC and OC. All of these nutrients were determined following the procedure described in [
20]. The physical and chemical characteristics of the soils at the study sites are as shown in
Table 1.
Separately, surface soil samples meant for POP testing were put in pre-cleaned amber bottles and immediately placed in cool boxes with ice packs to be transported at the end of the sampling day to the laboratory for analysis. Special drilling equipment hired from the water solutions drilling company (WSDC) were used for sample collection from up to seven meters below the soil surface at the PPO Tengeru site. At the NHC Morogoro site, samples up to 2 m below the soil surface were taken using open ditch profile openings. The soil samples from both sites were placed in cleaned amber bottles and transported to two separate laboratories, namely the Tanzania Plant Health and Pesticides Authority (TPHPA) and Tshwane University of Technology’s environmental and water chemistry laboratory in South Africa. At both laboratories, soil samples were analyzed for DDT, lindane, fenthion, diazinon and permethrin.
Soil samples (2.5 g) were weighed in pre-cleaned cellulose thimbles and were thereafter transferred into cleaned Soxhlet apparatus for extraction. The samples were extracted for 16 h using a mixture of n-hexane/acetone (2:1, v/v). Upon completion, the extracts were allowed to cool down to room temperature. The extracts were carefully transferred into pre-cleaned round-bottom flasks and were rotary-evaporated to approximately 2 mL. Two laboratory reference soil samples were spiked with known amounts of the targeted compounds, and they were similarly prepared following the procedures named above. Spiking the reference samples serves as a quality assurance measure to assess the efficiency of the extraction method by estimating the analytical recoveries of the targeted compounds.
The soil extracts were purified using deactivated silica gel packed into glass columns. Prior to the clean-up procedure, the packed columns were conditioned using 25 mL of n-hexane to remove trapped air and possible interfering contaminants. The concentrated extract was quantitatively transferred into the glass column and was eluted under gravity with 40 mL of n-hexane/acetone (2:1, v/v). The eluate was rotary-evaporated to approximately 2 mL and transferred into an amber vial, where it was further concentrated with high-purity nitrogen gas until it reached incipient dryness. The extracts were re-constituted with 1 mL of Toluene, followed by the addition of 50 µL of 500 pg µL−1 of DDT-d8 that was employed as an internal standard.
Quantitative estimation of all of the targeted compounds was performed using an Ultra-trace 2010 Shimadzu GC equipped with QP 2010 Ultra mass spectrometer (Shimadzu, Kyoto, Japan) operated in EI mode. The chromatographic separation of these compounds was achieved using DB-5 MS (15 m, 0.25 mm i.d., 0.10 µm film thickness) capillary column. The optimal conditions employed for the GC-EI-MS instrument are shown in
Table 1. To enhance the sensitivity of the instrument and to overcome the inherent problems of interfering co-extractants, the mass spectrometer (MS) acquisition was carried out in selected ion monitoring (SIM) mode. In this case, a target ion and two reference ions were selected for each targeted compound as well as internal standard (DDT-d8) for their identification and quantification.
2.3. Pot and Field Experiments
Parallel field and screenhouse potted soil experiments were performed to assess the efficacy of phytoaccumulator test plants for DDT from the soil. The potted soil screenhouse experiment was laid out with a split-plot arrangement in a completely randomized design (CRD) with three replications. The main plots comprised two levels of DDT, i.e., a low level of DDT concentration (total DDT = 417 mg kg
−1) and a high level of DDT concentration (total DDT = 2308 mg kg
−1). The two levels used in this study were chosen randomly to reflect actual concentration scenarios in contaminated sites available in Tanzania [
19]. The test crops, namely pumpkin (
Cucurbita pepo), calabash (
Lagenaria siceraria), sweet potato (
Ipomoea batatus), simsim (
Sesamum indicum) and finger millet (
Eleusine coracana), were considered as subplots. Plants were regularly watered (every other day) each time the soil moisture was raised to about field capacity. About 250 mL of irrigation water per pot was enough to bring the moisture status to field capacity. Three weeks after planting, pre-calculated amounts of all deficient essential plant nutrients were added to recommended quantities to support the growth and development of test plants by mixing them with irrigation water.
The field experiments were laid out in a randomized complete block design (RCBD) with three replications. The sloping terrain was used as a blocking factor, and, like in the potted soil experiment, the test crops were pumpkin (
Cucurbita pepo), calabash (
Lagenaria siceraria), sweet potato (
Ipomoea batatus), simsim (
Sesamum indicum) and finger millet (
Eleusine coracana). Plants were established through direct seeding on 2 × 4 m plots. All deficient nutrient elements (
Table 1) were corrected to recommended levels using mineral fertilizers.
2.4. Plant Sampling and Analysis
Plant sampling in the pot and in the field experiments was carried out at flowering, roughly 60 days after planting. Plants sampled included shoots (above-ground biomass) and roots (below-ground biomass). For the potted soil experiment, shoots were harvested by cutting at about 1 cm above the soil, and then soil in the pots was poured on to a manila sheet, and the roots were collected/removed by hand and washed with tap water and then rinsed in distilled water. The samples (shoots and roots) were kept in separate amber bottles, clearly labeled and transported to TPHPA laboratory for analysis of DDT, its metabolites and other pesticides. A duplicate set of samples was sent to the Tshwane University of Technology’s environmental and water chemistry laboratory in South Africa for analysis. Prior to analysis in the laboratory, fresh weights of roots and shoots were measured. After weighing, shoots and roots were dried at about 50 °C for one week before being finely ground for chemical analysis.
The root and shoot parts of all test plants were prepared for the determination of residual levels of the targeted persistent organic pollutants. Approximately 2.5 g of the plant samples was weighed into pre-cleaned amber bottles. The samples were soaked overnight with 50 mL of n-hexane/acetone (1:1,
v/
v), followed by ultrasonic-assisted extraction for 30 min. The set-up was allowed to cool down to room temperature, and the extract was carefully transferred into a clean round-bottomed flask. The extraction was repeated using the same volume of extraction solvent and time. The extracts were combined and subjected to rotary evaporation as was previously indicated for soil samples. Two laboratory reference plant samples (lettuce) were spiked with known amounts of the targeted compounds, and they were similarly prepared following the aforementioned procedures. The spiked reference samples were employed for the estimation of the analytical recoveries of the target compounds. The plant extracts were purified using deactivated silica gel packed into glass columns as explained for soil samples, and quantitative estimation followed similar procedure as that described for soil samples. Optimized conditions for the GC-EI-MS employed for the analysis of target compounds are summarized in
Table 2.
2.5. Disposal of Polluted Soils and Plants
All polluted materials (remaining soil samples and harvested plants) generated from this study were disposed through a government-certified agent following standard procedure provided by government chemist laboratories.
2.6. Statistical Analysis
The data on concentrations of DDT, its metabolites and other pesticides were subjected to Analysis of Variance (ANOVA) using the GenStat (14th edition) statistical software. Mean separation was carried out according to Duncan’s New Multiple Range Test (NMRT) at p = 0.05 significance level.
4. Conclusions
The success of any phytoremediation project is highly dependent on the selection of effective accumulator plants. We have shown in the current study that sweet potato and calabash, as well as pumpkin, can be deployed to help decontaminate NHC Morogoro and PPO Tengeru from DDT pollution. When an edible crop species such as sweet potato or pumpkin is used, the resulting phytobiomass should be protected against consumption or any other form of exposure to animals and human beings, as this may pose a health risks. Plants like calabash—which is neither used as human food nor consumed by livestock in most African countries, including Tanzania—should be given priority in future projects to decontaminate the NHC Morogoro and PPO Tengeru sites in Tanzania. Due to the huge amounts of DDT and its metabolites in the contaminated soil, especially at the NHC Morogoro site, remediation would require longer periods of time with repeated cycles of planting, removal and the proper disposal of bioaccumulator plants. Where possible, the intercropping of various competent bioaccumulator plants should be tested in future research.