Materials and methods 1. Analytical reagents

Nitric acid (55%) used for digestion and acidification was purchased from Merck (Johannesburg, South Africa). All reagent stock solutions: 1000 mg L⁻¹ stock solution of the metals salts of zinc sulphate, potassium dichromate and lead acetate 3-hydrate analytical reagent grade bought from Merck (Johannesburg, South Africa) were used to prepare standard solutions. Calibration curves were constructed for each metal ion to be tested and the correlation coefficients were determined.

49 2.2. Materials

A portable Multi-probe Boeco pH meter purchased from Rochelle (Johannesburg, South Africa) was used to take the pH, electrical conductivity (EC), temperature and dissolved oxygen (DO) of the water samples. Milli-Q for de-ionised water was purchased from Sigma Aldrich (Johannesburg, South Africa) was used for dilutions of all solutions. All glass and sampling water bottles used were first soaked in dilute HNO3, thoroughly washed with liquid soap and then rinsed with de-ionised water and acetone. Thereafter, all glassware was dried in the oven at 100°C for 24 h, while the plastic were left to dry at room temperature and later used for water sample collection. Sieves from Rochelle (Johannesburg, South Africa) of different sizes were used to sort the dried plant powder samples into different size ranges. A refrigerator was used to store the samples at 4°C.

2.3. Instrument

A Perkin Elmer Pinaacle 900T Atomic Absorption spectrometer bought from Perkin Elmer (Johannesburg, South Africa) was used to analyse all the samples in the laboratory. Milli-Q water was used to blank the instrument before analysis. The auto-sampler cups were washed and rinsed using de-ionized water prior to use. Argon gas flow rate of (0.25 L min⁻¹) was used to protect and purge the graphite tubes during the furnace program procedures, and the data acquisitions were carried out using Syngistix for AA Spectra software. The absorbance of each metal was measured at appropriate wavelength using default instrument parameters for the GF- AAS.An amount of 20 µL of the working solutions was drawn into the graphite tube by the auto-sampler of the instrument and the absorbance readings were measured using a hollow cathode lamp at the appropriate wavelength for each metal: Zn at 216.86 nm (slit width 0.70 nm, lamp current 15 mA); Cr at 357.87 nm (slit width 0.70 nm, lamp current 25 mA) and Pb at 217.0 (slit width 0.70 nm, lamp current 13 mA). The heating programme for each metal was as follows: Zn (drying 110°C, ashing 700°C, atomisation 1800°C); Cr (drying 110°C, ashing 1500°C, atomisation 2300°C) and Pb (drying 110°C, ashing 850°C, atomisation 1600°C). All analyses were done in triplicate.

2.4. Sampling area

The Mvudi River in Thohoyandou (23.013 S, 30.2828 E) is a tributary of the Luvuvhu River, located in the north-eastern part of South Africa (Figure 1). The river is south of the University of Venda. It falls in the Lowveld of Limpopo Province, which forms part of the greater Limpopo River Basin with an elevation of 546 m above sea level (Edokpayi et al. 2015;

50

Mzezewa et al. 2010). It has a semi-arid climate, which is classified as humid subtropical with a subtropical dry forest biozone. Daily temperature in the catchment varies between 20 – 40°C (wet season) and 12 – 22°C (dry season) (Odiyo et al. 2012). The catchment average annual rainfall is about 800 mm, but it often varies between 340 mm and 2000 mm (DWAF 2003).

The river is mainly used for domestic, recreational and agricultural purposes. Several land use activities in the river catchment that could constitute possible sources of pollution include agriculture, human settlements, schools, hospitals, solid waste disposal sites and effluents from wastewater treatment plants. Mvudi River is a major source of water to Nandoni dam (Edokpayi et al. 2016). The dam supplies water to the urban areas of Louis Trichardt, Thohoyandou and rural communities. However, on its course to Luvuvhu River it passes through a wetland around Maungani village. This is where wastewater from a cement brickmaking factory runs off into the river passing through the wetland. Fig. 1 shows a map of the cement the brick making factory, wetland and Mvudi River.

Fig. 1. Sampling site of water and plants samples encircled in the vicinity of a cement brickmaking company in Thohoyandou, South Africa.

2.5. Sample collection

The propylene containers were used for water sample collection. Before use, the propylene containers (500 mL) were first rinsed with the water samples and thereafter, immersed about

Maungani

Mvudi River

University of Venda

Muledane Wetland

W1b

Brickyard W2b

W3b

WW2 W1a

51

15 cm below the surface of the river. Water samples were collected at the river just before, within and after the wetland area, making a total of 5 samples and labelled according to the sampling site. The water samples were collected during summer season (wet and high flow).

After collection, pH electrical conductivity, temperature and dissolve oxygen were determined in the field using portable measuring and data recording equipment. The samples for metal concentrations were acidified in the field with 5% HNO3, to avoid microbial activity.

Three different plant species were selected and collected for the analysis of heavy metals of interest absorbed by the roots into their stem and leaves. Fig. 2 shows the pictures of the plants that were collected along the Mvudi River. The plants were later identified and characterised for their ability to absorb heavy metals at University of Venda (Botany department). The plants were collected, separated into leaves, stem and roots and dried at room temperature. The dried plants were then ground using a mortar and pestle and sieved through a size 2 mm sieve.

Fig. 2. Pictures of the sampled plants (a) Xanthium strumarium, (b) Phragmites pustralis and (c) Bidens pilosa.

2.6. Sample preparation

All water samples collected were preserved by being stored at 4°C. The samples were also centrifuged at 200 rpm for 10 min and filtered before analysis. The plant samples were washed with de-ionised water to remove any particulate deposited from the atmosphere and dried under room temperature, partitioned into three parts, namely root, stem, and leaves. These samples were weighed to find the initial weight and further dried until a constant weight for each part was obtained. The samples were cut into small pieces and grounded using porcelain mortar and pestle. A 0.2 g portion of each sample was transferred into a 250 mL beaker with 25 mL HNO3

and heated to 80 – 100°C in fume hood. After heating for 20 min, the content of the flask was left to cool and later 10 mL of each sample solution was added to a centrifuge vial.

52

A 40 mL of de-ionised water was added to each solution in the centrifuge vial to bring the final volume to 50 mL. The centrifuge flasks were marked before determination of metals by GF-AAS.