Variation in chemical composition of Harpagophytum species as function of age and locality

English: Dried tubers, growing on secondary roots of Harpagophytum procumbens subsp. procumbens are widely used as an analgesic and anti-inflammatory medicine against arthritis and other chronic pain conditions. Tubers are harvested in large quantities from South Africa, Botswana and Namibia. The dried material is exported to European countries. This industry makes a big contribution to poverty alleviation in rural areas. The large quantities exported (about 600 tons per annum), has raised concerns about the sustainability of wild harvesting. Apart from threatening the plant with extinction, overexploitation of the resource will also threaten the economic survival of rural communities. This thesis aims to study the variation in chemical composition between samples from different regions, species, harvesting regimes and age of tubers. Only tubers from secondary roots are removed and a large percentage of harvested plants survive. This thesis is part of a bigger study of which the aim is to determine the optimum wild harvesting regime resulting in the best yield of the active ingredient in a sustainable harvesting industry. Plant material (tubers) was collected from December 2007 to February 2008 in the North West Province, Limpopo Province, Namibia, Caprivi in Zimbabwe. In the North West Province, plant materials were collected in 6 different areas and/or farms: Cassel, Ganyesa, Moswana, Molopo Nature Reserve, Terra firma and Lafras. Samples from Cassel represents tubers from the same plants with different inter harvest periods (one to five years). The freshly collected tubers were sliced, sun dried and analysed with HPLC-UV for six different analytes (harpagide, harpagoside, 8-p-coumaroylharpagide, verbascoside, isoverbascoside and 6-acetylacteoside). These analytes can be divided into two structurally related groups. An analytical method was developed to quantify the six analytes routinely. The method is based on water and methanol as eluent and a reverse phase analytical column. A stepwise isocratic procedure ( 3% MeOH for 1 min, 50% MeOH for 20 min, column cleanup with 95% MeOH for 5 min and regeneration with 3% MeOH for 5 min) was found to be the best for our purposes. The method was validated for specificity, accuracy, precision, robustness and linearity. Internal standards were used for calibration. The data from more than one thousand analyzed samples were statistically processed using StatSoft, inc. (2008), STATISTICA, version 8.0. to answer the following questions: 1. Are there meaningful differences in chemical composition between populations from the five different inter harvest periods (1 to 5 years) at Cassel in the North West Province (Are there meaningful differences in chemical composition of tubers that are one, two, three, four or five years old?). 2. Are there meaningful differences in chemical composition between the six populations from the North West Province (Cassel, Ganayesa, Moswana, Terra Firma, Lafras and Molopo Nature Reserve)? 3 Are there meaningful differences in the chemical composition of populations from North West Province and Namibia? 4. Are there meaningful differences in the chemical composition of populations from North West Province/ Namibia (H. procumbens subsp. procumbens, from the Northern Province (H. zeyheri subsp. zeyheri) and from Zimbabwe (Victoria Falls) (H. zeyheri subsp. sublobatum)? Factor analysis of the six variables (harpagide, harpagoside, 8-p-coumaroylharpagide, verbascoside, isoverbascoside and 6-acetylacteoside) yielded two, three or four factors depending on the level of degree of variance that we required. Multivariate and univariate analysis (MANOVA and ANOVA) of normal and transformed data indicated highly significant differences. Levine’s test was used to test homogeneity of variances. In correlation analysis all six variables were used because the factors did match clusters based on the chemical structure of the six variables. In all four experiments significant variations were observed and described. Cluster analysis, using scatter plots of three factors (factor 1: verbascoside, isoverbascoside and 6-acetylacteoside, factor 2: harpagide and factor 3: 8-p-coumaroylharpagide) identified three distinct populations. These three populations are from three different geographical regions and correspond with the corresponding taxonomic classification of the different populations. (H. zeyheri subsp. zeyheri from the Limpopo Province, H. zeyheri subsp. sublobatum from Zimbabwe (Victoria Falls) and Captivi and H. procumbens subsp. procumbens from the North West province and Namibia.