Study setting
The research was conducted in Monze district in the southern province of Zambia. Monze town is situated 200 km south of Lusaka (Zambia’s capital city) and 300 km north of Livingstone. Monze district covers an area of 6687 km2. It lies at 1128.72 m above sea level, 16 degrees 18 min south and 26 degrees 28 min east. Monze district has a projected 2014 population of 206,693 (CSO 2010).
Economically, the district depends on farming, mainly subsistence crop farming (including fruits and vegetable growing) and animal husbandry. Pesticides are largely used to control pests and diseases to improve their market value. Vegetable growers are concentrated along the Magoye River that passes through the district on the east side of Monze town; mainly around Hachaanga, St Mary’s and Chikuni areas.
Study design
Mixed method (convergent parallel) design was used in this study. The quantitative approach used the cross section analytical study design while the narrative study design was used for the qualitative approach. Samples were procured from conveniently sampled fruit and vegetable farmers around Hachaanga and St. Mary’s areas of Monze. Prior to sample collection, farmers with fruits and vegetables at maturity stage and had sprayed them with dichlorvos were identified. Key Informant Interviews (KIIs) were conducted with farmers and public authorities to understand pesticide use in Hachaanga and St. Mary’s areas of Monze.
Sampling
A total of 30 samples of cabbage, rape, tomatoes and oranges were procured from nine farmers as per FAO guidelines (Alba 2004); 2 kg of each vegetable and 1 kg of oranges per sample were collected 1 day after the application of the pesticide (observed withdraw period). Samples were prepared for residual analysis according to Codex Alimentarius Commission guidelines (FAO/WHO 2014). All the 30 samples were analyzed for residues of dichlorvos using gas chromatography-mass spectrometry (GC-MS).
Sample preparation and analysis
About 1 kg of each vegetable and orange sample was thoroughly washed and chopped using a laboratory knife. This was blended using a waring laboratory blender with a variable speed to form a composite sample. A total of 50 g was then taken from the composite sample. After milling, a 20 g sample was placed into a 250 ml beaker, 5.0 g sodium chloride was added followed by adding dichloromethane, extracted by ultrasonic for 30 min. After that 10 g anhydrous sodium sulfate was added and stayed for 2 min. The extraction was transferred to a column packed with 4.0 g anhydrous sodium sulfate and rinsed twice (total of 5 ml) with dichloromethane, all the eluents were collected and evaporated to near dryness under nitrogen stream at 45 °C, the residue was re-dissolved with about 1 ml Hexane.
A florisil SPE column was conditioned with 6.0 ml mixture of hexane-acetone (4:1) and 5.0 ml hexane, the concentrated extraction was loaded on the top of the cartridge, and followed by eluting with 6.0 ml mixture of hexane-acetone (4:1), the eluent was collected, and evaporated to near dryness under nitrogen stream at 45 °C. The dried extract was then re-dissolved in acetonitrile solvent and final extract of 1 ml injected in the gas-mass spectrometry chromatography. The gas chromatography-mass spectrometry from Shimadzu was used for measurement of dichlorvos residues. External standard calibration was used to prepare the calibration curve, using concentrations of 5, 6, 7, 8, 9 and 10 μl/ml. The curve type was linear, original not forced; weighted method: None equation was:
$$ \begin{array}{l}\mathrm{Y}=2146.629\mathrm{x}+1151.629\hfill \\ {}\mathrm{R}\hat{\mkern6mu} 2=0.9844324\hfill \\ {}\mathrm{R}=0.9921856\hfill \end{array} $$
The amount of the pesticide in each sample was calculated based on the slope of the standard curve.
Data analysis and management
Pesticide residue values were reported as measured in mg/kg and compared to the codex Alimentarius MRL and the Zambia Food & Drug Act standards. Those that were found to be above MRLs were regarded to have violated the standards while those at or below the MRLs did not.
Stata version 13IC (StataCorp, College Station, Texas, USA) was used to derive median and interquartile ranges of pesticide residue levels in fruits and vegetable types.
Estimated average daily intake calculation
The estimated average daily intakes (EADI) were found by multiplying the median residual pesticide concentration (mg/kg) by the consumption rate (kg/day). To evaluate the safety of consumers, exposure estimates were assessed and compared to the acceptable daily intake (ADI) established by FAO/WHO (FAO/WHO 2004). Health risk indices were computed using data obtained from Zambia Food consumption and Micronutrient status survey report (Halimatou et al. 2014) as 21.9 kg/year for vegetables and 54.75 kg/year for fruits. This was based on assumptions that; everyone consumes vegetables on daily basis as Zambian diet is mainly vegetarian. One hundred percent of the vegetables are consumed after 24 h (observed withdrawal period) of pesticide application. The estimated hazard indices were calculated by dividing the EADI (mg/kg/day) by the corresponding value of WHO/FAO acceptable daily intake (Table 2).
Qualitative study
Population and data collection
Key informants were purposively sampled based on their knowledge on pesticide use, handling and regulation. Identification of farmers was done through agriculture extension officers manning the two study areas. Farmers with fruits and vegetables at maturity stage and had sprayed them with dichlorvos were eligible for KIIs and their produce for residue testing. Other key informants included agriculture extension officers, block extension, senior crop officer, public health department officer (from MOH) and ZEMA officer. Apart from one policy maker who was based in Lusaka, all the key informants were Monze residents. All KII were purposively selected as they were viewed to have information on the study subject. Only key informants who consented were interviewed.
Ten farmers and six officers were identified bringing the sample size to 16. All 16 eligible respondents were approached to participate in the study, but two (one farmer and one health officer from council) refused to consent, and this reduced the sample size to 14 as the two who declined to participate could not be replaced. Key informant interviews (KIIs) were conducted using semi-structured interviews using two different structured guides, one for farmers and another for policy makers or regulators. KIIs were audio recorded with the permission of the participants. Field notes were also taken during KIIs which lasted between 20 and 40 min. Secondary data reviewed included the Zambia Daily mail article by ZEMA which was posted on 15th December, 2015.
Key informant interviews with individual farmers were conducted in the local language (Tonga) and translated to English by the principle investigator who is a native Tonga speaker. Interviews with agriculture, public health and ZEMA officers were done in English. All KIIs were then transcribed and formatted in Microsoft word before importation into Nvivo10 software for the analysis process. Guiding questions focused on farmers’ practices on pesticide use and handling, knowledge on pesticide use and handling, knowledge on health effects of pesticide residues and issues of training, pesticide regulation and monitoring.
Key Informant Interviews were analyzed using thematic analysis according to the six-stage process of thematic analysis starting with familiarization, generation of initial codes, searching for themes, reviewing themes, defining and naming themes and finishing with the final report (Lawson 2014). Data was transcribed as part of the familiarization stage. Transcribed data was transferred to Nvivo version 10 for arrangement, coding and merging into themes.
Transcripts were carefully and thoroughly read and re-read after which initial coding and categorization of themes was done by the principle investigator. Responses that were related through content and context were categorized as themes until no new themes emerged. Codes were categorized according to similar contents and then developed into broader themes.