In most countries, fresh produce sold at local markets is usually not analyzed for agricultural chemical residues as export products are, which raises concerns about the perceived safety levels of local food supplies in contrast with exported products. The aim of this study was to determine pesticide residue levels in fruits and vegetables sold at two of the biggest fresh produce markets in Africa. A total of 199 fruit and vegetable samples were collected between 2012 and 2014 and analyzed for 74 pesticides commonly used in the horticultural sector. Of the samples analyzed, 91% were compliant with set maximum residue levels (MRLs). The remaining samples either contained unregistered chemicals (8%) or exceeded set MRL values (1%). Products containing more than one pesticide residue constituted 4.02% of all samples tested. Imazalil and iprodione were found to be the most frequently detected pesticides (12 samples each). Boscalid, endosulfan, profenofos, and procymidone were associated with the most noncompliance, including exceeding MRL values or being unregistered for the specific crop. The establishment of a national pesticide monitoring program is essential for the country and would ensure that pesticides are used in accordance with good agricultural practices.
The increased worldwide concern over the safety status of the food supply has resulted in a growing demand for effective and transparent regulation and control. The major challenge in safe food provision is assurance that microbiological and chemical contamination does not compromise product integrity. Generally, plant pests and diseases are responsible for major crop losses, necessitating the use of pesticides to boost farm productivity (39). However, pesticides are xenobiotic, in general; hence, from a public health and environmental sustainability perspective, there is an urgent need to more effectively monitor and regulate their use (33). Consequently, most countries have set extensive pesticide monitoring programs to protect consumer health and improve management of agricultural resources while preventing economic losses (4). Such food surveillance programs generally focus on the proper use of pesticides in terms of authorization, registration, and disposal and ensure compliance with maximum residue limits (MRLs) (13, 41). The MRL levels are set for a specified crop representing scientific evidence that at the point of consumption, it will not be harmful to the consumer. The maximum allowable residues must thus be within legally permitted levels that are set on national and international standards based on Codex Alimentarius (26).
If pesticides are applied in accordance with basic good agricultural practices, by the time fresh produce reaches the markets or retail outlets, residue levels on crops should be well below legal limits (11, 19). To avoid exceeding MRLs, farmers need to comply with prescribed application rates, as indicated on pesticide labels, including withholding periods before crops are harvested (8, 37). However, a number of researches have reported pesticide residue levels exceeding those set in regulations (10, 16).
Residues on fresh produce resulting from the inappropriate use of pesticides are one of the most important food safety concerns in developing countries. South Africa (SA) has the highest usage of pesticide active ingredients in Southern Africa (15). Effective residue monitoring is important to indicate the potential risk of pesticide exposure on human health. All fresh produce intended for export is generally tested for pesticide residue levels to ensure compliance according to standards set by the destination country. Although this is done in SA for the export market, both locally produced and imported agricultural fresh produce sold at the national markets is mostly not analyzed for pesticide residues. The aim of this study was, therefore, to monitor pesticide residues in fruits and vegetables sold at two of the biggest fresh produce markets in Africa (Joburg and Tshwane Fresh Produce Markets, SA) to attain a general picture of the level of compliance of local food supplies.
MATERIALS AND METHODS
Site description and sample collection. The Joburg Fresh Produce Market is the largest in Africa and commands nearly 40% of the national market share in both volume (1,291,892 tons) and turnover ($397 million), as reported for the 2012 to 2013 financial year (6, 20, 34, 35). It also trades about twice the volume of fresh produce sold on the second biggest market, namely, the Tshwane Fresh Produce Market (641,404 tons [1 ton = 1,000 kg]; $182 million) for the 2012 to 2013 financial year (6, 20, 35).
A total of 199 fruit (apples, apricots, avocados, banana, berries, cactus pears, cherries, grapefruits, litchis, mangoes, melon-spanspek (cantaloupe), nectarines, oranges, papaya, peaches, pears, plums, pineapples, persimmons, strawberries, and table grapes [n = 87]) and vegetable (baby marrow [zucchini], beetroot, broccoli, brinjals [eggplant], butternut, cabbage, carrots, chillies, cucumbers, garlic, gem squash, green beans, green peppers, lettuce, mushrooms, potatoes, onions, spinach, sweet potatoes, and tomatoes [n = 112]) samples representing major and minor crops commonly consumed in SA were randomly collected from the Joburg and Tshwane Fresh Produce Markets from 2012 to 2014. Samples were screened for 74 different pesticides whose selection was based on the frequency of application and disease profiles prevalent in the country. The scope of pesticide residues are depicted in Table 1.
The sample size was at least 1 kg for small- and medium-sized fresh produce and 2 kg for large products, as based on the Department of Agriculture, Forestry and Fisheries standard operating procedure for MRL sampling, which is harmonized to the European Directive 2002/63/EC sampling procedure and Codex guidelines (14, 25, 57). Samples were collected in polythene bags, sealed, labeled with a sample identity code, and transported to the South African Bureau of Standards Chromatographic Services Laboratory in Pretoria, which is accredited under ISO/IEC 17025:2005 (accreditation no: T0270) (31). All samples were prepared on arrival and stored at −18°C.
Extraction. Samples were analyzed by using the South African Bureau of Standards in-house test method 029/2006 (58) modified with the quick, easy, cheap, effective, rugged, and safe mini-multiresidue method for the detection of pesticide residues in low-fat products (2). For each sample, 10 g of homogenate was transferred into an empty 50-ml centrifuge tube, followed by 10 ml of acetonitrile (Sigma-Aldrich, Johannesburg, SA) and 100 μl of the standard mixture with a concentration of 1 mg/kg used for spiking; this was vigorously shaken by hand for 1 min, as a Digitimer stopper was used to check the time (1, 2, 51). The homogenate was further mixed with 4 g of magnesium sulfate (MgSO4; Sigma-Aldrich), 1 g of sodium chloride (NaCl; Associated Chemical Enterprises [Pty] Ltd., Johannesburg, SA), 1 g of trisodium citrate dihydrate (C6H5Na3O·2H2O), and 0.5 g of disodium hydrogen citrate sesquihydrate (C6H6Na2O7·1.5H2O; Merck [Pty] Ltd., Modderfontein, SA) in the tube, vigorously shaken by hand for 1 min, and then centrifuged for 5 min at 3,000 × g to facilitate partitioning of the pesticides between the water phase of the sample and the acetonitrile phase. The pesticide standards were purchased from Merck (Pty) Ltd. The purity of the reference materials was between 90 and 100%.
A 1-ml aliquot of the acetonitrile layer was removed and added to the appropriate 2 ml of the dispersive solid-phase extraction tube, which contained 150 mg of MgSO4 for further water absorption and 25 mg of primary secondary amine (Chemetrix [Pty] Ltd., Johannesburg, SA) to remove acids (including fatty acids), sugars, and some pigments. Note that for samples with high amounts of chlorophyll or carotenoids, 2.5 mg of graphitized carbon was added. After the addition of the aliquot, the 2-ml of the dispersive solid-phase extraction tubes were all capped and shaken for 2 min and then centrifuged for 5 min at 3,000 × g. The aliquot was transferred into autosampler vials, and formic acid in acetonitrile was added to 10 μl of extract to adjust the pH to 5.0 to 5.5.
Pesticide residue analysis. The fruit and vegetable extracts in the autosampler vials were analyzed for pesticide residue by using gas chromatography (GC) and mass spectrometry techniques (2). Pesticide results were detected by GC–electron capture detectors and flame photometry detectors and confirmed by the GC–mass spectrophotometer detectors. Analysis of the pesticides was performed by using Agilent 6890 GC series on HP-5 (5% phenyl-methylpolysiloxane) capillary columns for organochlorines and Zebron 7HM-G016-17 series on ZB-Multiresidue-1 capillary columns for organophosphates. The HP-5 column (30 m length by 0.25-mm inner diameter) and coated with 0.25-μm film thickness for the stationary phase. The ZB-Multiresidue-1 column (30-m length by 0.32-mm inner diameter) was coated with 0.50-μm film thickness for the stationary phase. Helium was used as a carrier gas at a flow rate of 1.3 ml/min for organochlorines and a flow rate of 1.1 ml/min for organophosphates. The temperature program for all GCs was as follows: initial temperature 90°C for 1 min, raised at a rate of 25°C/min to 150°C, and finally raised at a rate of 5°C/min to 280°C, and then isothermal for 10 min.
Of the 199 samples analyzed over the 2-year period, 91% were found to be compliant with the national MRL (benchmarked to European Union standards) on crops, and the remaining samples had violations either as a result of unregistered chemicals (8%) or exceeding the set MRL (1%), as indicated in Table 2. Carrots were the crop found with the highest number of samples containing unregistered pesticides (three samples), followed by baby marrows, green peppers, spinach, and strawberries, with two samples each (Table 3). Fresh produce that had the lowest number of unregistered pesticides detected included beetroot, green beans, lettuce, mangoes, and pears, all with one sample each. Only two samples of tomatoes had pesticide concentration exceeding set MRLs (boscalid concentration of 0.05 mg/kg exceeded the MRL of 0.01 mg/kg). Pesticide residues were not detected in 135 (68%) of the samples, and 46 (23%) samples contained registered pesticides. Our findings showed boscalid to be the pesticide associated with the most noncompliance (five samples), followed by endosulfan (four samples), profenofos (three samples), and procymidone (two samples). Carboxamide (five samples) was the chemical group with the most noncompliance, followed by cyclodiene organochlorine and organophosphate (with four samples each), and dicarboximide and pyrethroid (two samples each). The chemical groups with the least noncompliance were arylpyrrole, demethylation inhibitors (DMI; imidazole) multisite (chloronitrile), phenylamide (acylalanine), and strobulin type (methoxyacrylate), with one sample each.
Imazalil and iprodione were the most detected pesticides with 12 samples each, followed by prochloraz (9 samples), azoxystrobin (7 samples), boscalid (6 samples), and endosulfan and chlorpyrifos (5 samples each). The least-detected pesticides were chlorfenapyr, chlorothalonil, cy-permethrin, cyprodinil, difenoconazole, fenpropathrin, cy-halothrin-lambda, and metalaxyl, with one sample each. The most detected chemical group was DMI (imidazole) with 20 samples: dicarboximide (16 samples), organophosphate (9 samples), strobulin type (methoxyacrylate; 7 samples), and carboxamide and cyclodione organochlorine (6 samples each). The least-detected chemical group was anilinopyrimidine, arylpyrrole, DMI (triazole), multisite (chloronitrile), and phenylamide (acrylalanine), with one sample each. Samples containing more than one pesticide that were detected constituted 4.02% of all samples tested.
To our knowledge, this is the first article that provides an extensive assessment of pesticide residue levels in various fruits and vegetables commonly sold in the domestic fresh produce markets in SA, considered the biggest of their kind in Africa. In general, 32.2% of the 199 samples were positive for at least one of the pesticides tested for in this study. Higher levels of pesticides were reported in similar studies in emerging markets or developing countries. Dogheim et al. (17) reported that of 1,579 samples from eight fresh produce markets in Egypt, 23.9% contained detectable pesticides. Lee et al. (42) indicated that 36.2% of 126 samples from seven fresh produce markets in Mauritius, and Knežević and Serdar (38) reported 25.8% of 240 samples in Croatia had detectable pesticides. Almost double these figures (43.5%) were reported in Ghana, based on 350 samples from six fresh produce markets (7).
In the present study, the majority of samples (23%) with detected pesticides were compliant with the national MRL levels, while 1% of the products exceeded these values. The numbers of products exceeding the national MRL standards in our study were found to be far fewer than those reported in other similar studies, i.e., Mauritius (2.3%), Egypt (2.59%), Croatia (7.5%), and Ghana (19%) (7, 17, 38, 42). All of these studies used MRLs based on the stringent European Union regulations, except for Mauritius, which were based on Food and Agriculture Organization and World Health Organization Codex Alimentarius Commission (3). Although the scope of the tested pesticides differed among the studies and the accreditation status of the laboratories could not readily be established from the publications, the findings provide baseline information on local fresh produce pesticide residue levels in mostly developing countries. Note that these studies were not done at the same time. The study in Croatia was conducted in June to December 2007 and in Ghana during August 2009 and June 2010, while the Egyptian and Mauritius studies were completed more than 10 years ago when MRLs were set less stringently. In recent years to meet health concerns, the established regulations regarding the MRLs in commodities have become more and more stringent (9, 28).
The relatively low proportion of samples exceeding set MRL values in this study (1%) indicates a high level of compliance that could be attributed to the strong export-oriented fresh fruit industry (18). A strong export-oriented sector will create a positive pull effect on the industry to generate fresh produce within a more compliant national context. More than 80% of the fresh produce sold in the local markets originates from commercial farms that are involved in exports (55). Of the products exceeding the MRL levels, two tomato samples were detected (boscalid concentration of 0.05 mg/kg, exceeding the MRL of 0.01 mg/kg). Note that vegetables are not commonly exported from SA, with fruit as the main export (59).
Irregularities were found in 9% of the samples tested, and this was mainly because of the presence of unregistered pesticides (8%). The use of unregistered pesticides in the current study was a matter of concern, which was almost double that reported in a study done in Italy, where Arienzo et al. (5) reported levels of 4.8% of unauthorized pesticides in local fresh produce. The authors ascribed the finding to incorrect agricultural practices. In contrast, Jardim and Caldas in Brazil (32) reported 13.2% of their samples contained nonauthorized active ingredients. The authors of the previously mentioned studies in Croatia, Mauritius, Egypt, and Ghana, did not report unregistered pesticides.
The use of unregistered pesticides in the current study was detected mostly in vegetables (baby marrow, beetroot, carrot, green beans, green peppers, lettuce, and spinach) and to a lesser extent on fruits (mangoes, pears, and strawberries). This disparity may be attributed to a large number of minor crops without adequate registered pesticide portfolios. Except for pears and baby marrow, the other crops are mainly produced for the local market. Hence, there is a need for local companies to register a broader spectrum of pesticides for minor crops. To accelerate the minor crop registration process, several initiatives have been launched in SA to support field trials for pesticide registration purpose (29). Despite these initiatives, minor crop registration is still taking excessive time due to a backlog in active ingredients awaiting registration (44).
Registered pesticide residues detected on crops where they are not permitted may also be a result of contamination from adjacent fields, soil, or water (12, 27). Deliberate use of nonauthorized pesticides has previously been attributed to low levels of education or awareness and a lack of access to information, limited or incorrect technical support, or failure to correctly read pesticide labels or understand their content (32). Moreover, decisions on the use of preferred pesticides are often also a result of previous experiences related to effectiveness, better costs, and product availability on the farm (53). In the case in which a banned, effective, and still available pesticide is used, it is often related to stocks still available in the system, either on farms or with dealers unscrupulously selling the product.
The detection of endosulfan on baby marrow, beetroot, table grapes, and spinach in this study, despite its ban, may reflect that available stocks were most likely not destroyed by the industry or producers (49). In addition, the detection of endosulfan is a serious violation of good agricultural practices, as it suggests continued use of the pesticide by some producers disregarding its ban as a crop pesticide (60). Cyclodiene organochlorine (endosulfan) is a highly toxic, environmentally persistent pesticide, whose production and use was approved for banning worldwide during the April 2011 Stockholm Convention (60). SA as a member of the convention was obliged to implement the ban on 30 April 2012 (56). The fact that this pesticide could still be detected on crops in 2013 and 2014 reflects, in all likelihood, that stocks are still present in the system (49).
The current study reported imazalil, iprodione, azoxystrobin, boscalid, endosulfan, and chlorpyrifos were the most-detected pesticide residues, and this is similar to the Petraitis et al. (52) study conducted in Lithuania. Imazalil, the most frequently detected pesticide, is an important postharvest fungicide for the citrus industry, which is used to control a wide range of fungi on fruits and vegetables (13). Imazalil is most commonly applied as a dip treatment in citrus packhouses (21). Research on application of postharvest fungicides in citrus packhouses has led to improved management of the fungicide bath, drench, and wax application systems to prevent exceeding set MRLs, while maintaining the required levels of disease control (23, 24, 36, 47, 48). Interestingly, Erasmus et al. (22) recently reported demonstrated practical resistance to imazalil in Penicillium spp. from several citrus packhouses. Hence, there is a need for effective industry-wide management of pesticides that will contribute to a more sustainable crop protectant arsenal and prevent buildup of resistance in the pathosystem. The residue levels of imazalil detected in mangoes in this study may be the result of some producers using the same packhouses for both citrus and mangoes in alternate seasons and not adequately sanitizing the packline (49).
Iprodione is the second most detected pesticide in this study and is a dicarboximide contact fungicide that inhibits the germination of spores, glycerol synthesis, and blocks the growth of mycelium. The pesticide is also used for the control of postharvest diseases, including Botrytis gray mold, brown rot (Monilinia fructicola), sclerotinia stem rot, and other fungal diseases in fruits and vegetables (61). Both imazalil and iprodione are applied at the postharvest stage, and sampling for residue levels usually takes place within 24 h posttreatment. Produce for sale can be moved into the local market from 2 to 7 days; thus, residue levels may still be present at the time of sale. Notwithstanding, postharvest application of fungicides must be carefully managed to avoid exceeding stringent MRL levels. Azoxystrobin is the third most detected pesticide and is a methoxyacrylate, systemic fungicide that is sprayed for preventive or curative purposes. This is followed by the fungicide boscalid and insecticide chlorpyrifos. Boscalid is used to control a range of plant pathogens, while chlorpyrifos is an organophosphate insecticide, mostly used against red scale, of particular importance on citrus crops (13, 50). The extensive use of these pesticides, as has been found in this study, indicates a high level of application and perhaps exceedance of dosages (43).
The detection of multiple pesticides in this study is a cause of concern, as previously reported from other studies (13). Multiple residues may be expected in some crops because the application strategy is usually to alternate pesticide classes to prevent buildup of pest resistance. However, the presence of multiple residues may also suggest that principles of good agricultural practice were not complied with (26).
Some of the pesticides detected in the current study, including chlorpyrifos, cypermethrin, chlorothalonil, endosulfan, iprodione, prochloraz, and procymidone, are recognized for their endocrine disrupting properties (40, 45, 46). In a recent study by Kugathas et al. (40), cypermethrin and imazalil were also shown to have endocrine disrupting properties. This is of particular importance because imazalil is one of the key postharvest pesticides for the citrus industry. Some of these chemicals, such as imazalil and prochloraz, also show extreme high potencies (50% inhibitory concentration [IC50]) in in vitro tests similar to those analgesics intended to block cyclooxygenase enzymes (40). These recent studies will most certainly raise new concerns within the regulatory sphere. The current main concern with endocrine disrupting pesticides is that it can, among others, lead to increased birth defects and sexual abnormalities and may increase the risk of cancers of the reproductive organs (30).
The detection of pesticides, including MRL violations, may unnecessarily lead to concerns among consumers because the general public often lacks appropriate information concerning the actual level of exposure. Most pesticide noncompliance in fresh produce in the local market is a result of mainly unregistered and prohibited agricultural chemicals. The relative high prevalence of unregistered pesticide residues in locally marketed fresh produce requires the setting of a pesticide control and monitoring program similar to the export inspection and testing services provided by the Perishable Product Export Control Board (54, 57). The improper use of pesticides is a major environmental and health risk factor (7). Therefore, there is a need to regulate retailed pesticide quality and to evaluate the potential impact of the observed residue levels on public health. The presence of pesticide residue levels in the informal market and on imported products must still be investigated.
A number of pesticides are being used on crops for which they are not registered, whereas others, such as endosulfan, are currently prohibited but are still being used on fruits and vegetables. Comparatively, exported products were more compliant with set pesticide regulations than locally sold produce. It is, therefore, necessary to set up more effective monitoring and control measures to regulate pesticide use on local food supplies. Regular monitoring of pesticide residues should be carried out to ensure compliance with national food safety standards, as is currently done for export products. Moreover, annual reports on pesticide-monitoring programs will ultimately help in reviewing the registration and use of pesticides. The study provides novel and important information on pesticide residues on fruits and vegetables sold on the major fresh produce markets in Africa.
The authors thank the Centre of Excellence in Food Security, a Department of Science and Technology–funded program managed by the National Research Foundation, South African Table Grape Industry, and the Department of Agriculture, Forestry and Fisheries, Directorate: Agricultural Inputs Control for the financial support and guidance. Thanks also to the Joburg and Tshwane Fresh Produce Markets for permission to collect fruit and vegetable samples for the study. We acknowledge Ms. Barbara English for her editorial input and Dr. Hennie Le Roux from Citrus Research International and Mr. Gerhard Nortje from Subtrop for critical comments on this article.