Elemental composition of Cyrtanthus obliquus and Lippia javanica used in South African herbal tonic, Imbiza

2.1 Sample collection

C. obliquus bulbs were purchased from traders at eight different market sites in KwaZulu-Natal. The chosen market sites were A1-Durban, A2-Umzinto, A3-Eshowe, A4-Verulam, A5-Stanger, A6-Tongaat, A7-KwaNkalokazi and A8-Ixopo (Fig. A.1). L. javanica leaves were collected from ten different sites in KwaZulu-Natal. The chosen sites were B1-Amandawe, B2-Eshowe, B3-Ntumeni, B4-Mangeti, B5-Maphumulo, B6-Stanger, B7-Ndwedwe, B8-Ixopo, B9-KwaNkalokazi and B10-Highflats (Fig. A.2). Plant and soil samples were collected at the beginning of April 2012; the average temperature of the five day sampling period was 27 °C with no rain or wind but sunshine during this time. All plant samples were washed with double distilled water. Bulb samples were cut into smaller pieces with a stainless steel knife. All samples (both plant and soil) were then dried in an oven at 50 °C, overnight. Dried samples were crushed into a fine powder using a food processor and stored in labelled polyethylene bags until analysed.

2.2 Reagents and chemicals

All chemicals used were supplied by Merck and Sigma Chemical Companies and were of analytical-reagent grade. Double distilled water was used throughout the experiments. Working standards were made up with double distilled water and 10 mL of 70% HNO₃ to match the sample matrix. To minimize the risk of contamination all glassware and other equipment were cleaned with 6.0 M HNO₃ and rinsed off with double distilled water.

2.3 Extraction of water soluble metals

In accordance with the conventional method provided by the traditional healers, C. obliquus bulbs (0.375 g) were placed into a 200 mL beaker containing 50 mL of deionised water and brought to a boil (at medium heat) for 10 min on a hotplate. The resulting solution was filtered by gravity into a 50 mL volumetric flask and made up to the mark with distilled water. The solution was then transferred into polyethylene bottles. L. javanica leaves (0.200 g) were prepared in a similar manner and the resultant solution was stored in polyethylene bottles. All extractions were done in quintuplicate (n = 5).

2.4 Elemental analysis

The bulb and leaf samples were digested prior to analysis using the microwave-assisted closed vessel technique. Digestions were performed using the CEM MARS (CEM Corporation, USA) microwave reaction system with patented Xpress technology. Five replicates of each sample were digested to improve accuracy and precision. Plant samples (0.5 g) were accurately weighed into 50 mL liners to which 10 mL of 70% HNO₃ was added. For digestion, the power was set at 100% at 1600 W and the temperature was ramped to 180 °C (ramp time 15 min) where it was held for 15 min. Digested samples were removed from the liners, transferred into 50 mL volumetric flasks, diluted to the mark with double distilled water and stored in polyethylene bottles for elemental analysis.

All plant and soil samples were analysed for the following elements As, Ca, Cd, Co, Cr, Cu, Fe, Mg, Mn, Ni, Pb, Se and Zn. Elemental analysis was carried out by Inductively Coupled Plasma-Optical Emission Spectrometry (ICP-OES) using the Perkin Elmer Optima 5300 DV. The analysis of a sample was done in triplicate. The accuracy of the elemental determination was measured by the use of the certified reference material (CRM), lyophilized brown bread (BCR 191), from the Community Bureau of Reference of the Commission of the European Communities. All samples were analysed in quintuplicate (n = 5).

3.1 Elemental composition of C. obliquus bulbs

The accuracy of the method for elemental analysis was measured by comparing results obtained with certified results (Table B.1). Values for Cu, Fe, Mn and Zn are certified and those for Ca and Mg are indicative so no standard deviations were provided. Measured values compared well with certified results.

Fig. A.3 shows the total and the water extractable concentrations of the elements detected in C. obliquus bulbs. High concentrations of Ca and Mg were found in the bulbs. Both these elements are said to be amongst the most abundant in plants (Broadley et al., 2008; Sedaghathoor et al., 2009). Calcium in C. obliquus bulbs ranged from 3022 to 4617 μg g⁻¹, which is comparable to typical Ca concentrations in plants (5000 μg g⁻¹) (Epstein, 1994). Magnesium concentrations in the bulbs ranged from 506 to 1316 μg g⁻¹. Bulbs from site A1 had the highest concentration of Ca and those from site A6 had the highest total concentration of Mg. The exact location from which the bulbs were obtained in the Eastern Cape is not known, as these were purchased from traders who obtained them from suppliers. However, the variance in concentrations in the bulbs from diverse market sites suggests that the bulbs were obtained from sites with differing soil properties.

The minor elements that were investigated were Cr, Cu, Fe, Mn, Pb, Se and Zn. Of the minor elements studied, Fe concentrations were high in the bulbs. Iron in plants is reportedly stored by a specialized protein, ferritin which helps plants cope with high levels of the element (Connolly and Guerinot, 2002; Briat et al., 2010). Sites A3, A4 and A8 had high concentrations of Fe in the bulbs, while sites A5 and A6 had lower concentrations (Fig. A.3). The distribution of Cr, Cu, Se and Pb (Table B.4) in the bulbs was somewhat similar at all sites except for sites A5, A6 and A1, where no Cr or Pb was detected. The concentrations of Mn and Zn varied at the different sites. Manganese concentrations were highest at site A1 and lowest at site A5. The variation in Mn concentration was not as significant as that in Zn where Zn concentrations at the different sites ranged from 8.2 to 42.2 μg g⁻¹. Site A3 had the highest concentration of Zn. The concentrations of the elements in C. obliquus bulbs were, typically, in the decreasing order of Ca > Mg > Fe > Zn > Mn > Cu = Se > Pb > Cr.

Concentrations of elements Ca, Fe, Mg, Mn and Zn in C. obliquus bulbs were compared to concentrations of these elements in the water extracts which more closely represents concentrations in the herbal tonic, Imbiza (Fig. A.3). The extraction percentages of the elements were also determined to ascertain what fraction of the elements from the bulbs was extracted into solution. If present in the water extracts, Cr, Cu, Pb and Se were below the instrument detection limits. Calcium concentrations extracted from the bulbs at all sites were relatively low. The extraction percentages ranged from 9.2% to 31.4% (Table B.2). The extraction percentage for Mg ranged from 17.7% to 77.6% with relatively high extraction percentages at most sites. Although the extraction percentages for Mg were high, the concentration of Mg in the solution was still lower than Ca. The concentrations of Ca and Mg extracted are within acceptable limits for human consumption. Extraction percentages of Fe at all sites ranged from 35.8% to 81.4% with more Fe being extracted from bulbs obtained at sites A4 and A8. The percentage of Mn extracted was between 11.0% and 49.3%. The percentage of Zn extracted was relatively high compared to the other elements with extraction percentages ranging from 30.9% to 91.5%. For most sites, extracted Zn more closely resembled bulb concentrations (Fig. A.3). The concentrations of extracted Zn are within acceptable limits for human consumption.

The extraction percentages give an indication of what fraction of the elements from the bulbs is transferred to the solution used to make the herbal tonic, Imbiza. For the detected elements, the concentrations of elements in the water extracts were, generally, in the decreasing order of Ca > Mg > Fe > Zn > Mn.

3.2 Elemental composition of L. javanica leaves

Total concentrations of selected elements in L. javanica leaves and water extracts from ten different sites in KZN are shown in Fig. A.4. The plant was located in the forest edges, where it is known to grow abundantly. The climate was humid and subtropical. The soil was generally sandy or loamy in texture. The total concentration of Ca in the leaves ranged from 2856 to 9225 μg g⁻¹ and Mg ranged from 1598 to 5619 μg g⁻¹. The ratio of Ca to Mg in the leaves at most sites was 1:3. According to Sedaghathoor et al. (2009), Ca and Mg are amongst the most abundant elements in tea plants.

Concentrations of the minor elements were dependant on site location. Iron in the leaves at Site B7 was relatively high compared to the other sites. Lead in the leaves from different sites was within a small range of variation and between 0.38 and 1.18 μg g⁻¹. Selenium was detected in leaves at seven sites with concentrations ranging from 0.86 to 4.96 μg g⁻¹ (Table B.5). Zinc in the leaves ranged from 15.6 to 27.3 μg g⁻¹. For the detected elements, the concentrations of elements in L. javanica leaves were typically in the decreasing order of Ca > Mg > Fe > Zn > Mn > Cu > Se > Cr > Pb > Co > Cd.

The elements extracted into water from L. javanica leaves were Ca, Cr, Cu, Fe, Mg, Pb and Zn. If present in solution, Cd, Co, Mn, and Se were below the instrument’s detection limits. Calcium concentrations in L. javanica leaves were high, but the amount extracted was considerably lower. The extraction percentage was between 16.2% and 23.4% (Table B.3), which was similar to that obtained for C. obliquus bulbs. Even though the extraction percentages are similar, the concentrations of the element in the plants differ significantly with higher Ca concentrations in L. javanica leaves than C. obliquus bulbs. The extraction percentage for Mg was between 31.4% and 37.3% in the leaves which was also similar to that obtained for C. obliquus bulbs. Again, the concentrations of the element in the plants differ significantly with higher Mg concentrations in L. javanica leaves than C. obliquus bulbs.

The extraction percentage for Cr ranged from 71.8% to 93.9%, which is relatively high. This indicates that more than 70% of Cr from the leaves will be extracted into solution and consumed. However, Cr is considered a safe nutrient as there are no observed adverse effects even with intakes 300 times the acceptable limit and vomiting is likely to expel Cr before any toxicity damage can occur (Dourson, 1994). Copper concentrations in the leaves were relatively low and for the very low concentrations Cu was not extracted by water (Fig. A.4). When Cu was extracted, the percentage ranged from 21.5% to 48.8%.

The extraction percentage for Fe from L. javanica leaves was between 23.1% and 28.7% which was lower than that obtained for C. obliquus bulbs (44.6% to 61.5%). The extraction percentage estimated for Pb was 37.0% to 68.8%; a very wide range was obtained probably due to the extremely low concentrations in the leaves. For most sites, Zn in leaves was more than double the extracted amount and the extracted percentage for Zn ranged from 29.1% to 89.1%.

For the detected elements, the concentrations of elements in the water extracts were, generally, in the decreasing order of Ca > Mg > Fe > Zn > Cu > Cr > Pb.