Research Article Received: 2 June 2009; Revised: 5 March 2010; Accepted: 25 March 2010 Published online in Wiley Online Library: 24 June 2010 (wileyonlinelibrary.com) DOI 10.1002/pca.1231 Characterisation of Two South American Food and Medicinal Plants by Chemometric Methods based on their Multielemental Composition Miguel A. Cantarelli,a,b Roberto G. Pellerano,b Luis A. Del Vitto,c Eduardo J. Marchevskyb,d and José M. Camiñaa,d* Introduction – The chemometric characterisation of two plants frequently used as food and medicinal species, Achyrocline satureioides and Achyrocline venosa (Asteraceae: Gnaphalieae), was carried out based on their mineral composition. Both species, known by the common name of ‘marcelas’, are very similar in their morphological features but they have different medicinal and food properties. Objective – To develop multivariate models for the classification of A. satureiodes and A. venosa based on their mineral content. Methodology – The analytic determinations were made by means of inductively coupled plasma optical emission spectrometry from aerial parts of the plants. An internal standard was used to evaluate the accuracy in the sample treatment and the recovery of toxic elements was studied. The multivariate methods used include principal components analysis, cluster analysis and linear discriminant analysis. Results – Classification for both A. satureioides and A. venosa was successful in all cases using only four variables: aluminium, iron, magnesium and sulphur content. The concentrations of the following elements were determined: Al, As, Ba, Ca, Cd, Co, Cr, Cu, Fe, K, La, Mg, Mn, Mo, Na, Ni, P, Pb, S, Sr, Ti, V, Y and Zn. Conclusions – This method is useful to identify both species in raw material in order to detect eventual errors of selection. Copyright © 2010 John Wiley & Sons, Ltd. Keywords: Achyrocline satureioides; Achyrocline venosa; chemometric methods; toxic and trace elements Introduction 550 Over recent years there has been a significant increase in the use of herbal medicine. Medical doctors are also prescribing herbal teas and herbal extracts as a supplementary type of treatment for everyday problems caused by modern civilisation. The use of medicinal plants in both crude and prepared forms has greatly increased (Eisenberg et al., 1998; Yeh et al., 2002), and although herbal remedies are often perceived as being natural and therefore safe, they are not free from adverse effects (Ernst, 2002 a,b). Considering the complexity of these drugs and their inherent biological variation, it is necessary to evaluate their safety, efficacy and quality (WHO, 1991). Achyrocline satureioides (Lam.) DC. (Asteraceae : Gnaphalieae), known under the common name of ‘Marcela’, is a South American plant used as traditional medicine and also in the beverage industry. It is used in infusions because of its digestive and antispasmodic properties, and its extract is the base of some bitter drinks that are widely consumed in the region known as ‘amargos’. Many studies have been carried out regarding the chemical composition and properties of A. satureioides (Ruffa et al., 2002; Kadarian et al., 2002; Gugliucci and Menini, 2002; De Souza et al., 2002, 2004; Arredondo et al., 2004; Cosentino et al., 2008), including its relation to weight loss (Dickel et al., 2007). Because of the extensive use of this plant, the acute toxicity of its extract in rats was determined (Rivera et al., 2004). The properties of its essential oils such as trypanocydal (Rojas de Arias et al., 1995) and mosquito repellent (Gillij et al., 2008) have been proved. In view of such evidence, A. satureioides is an important food and medicinal resource in Argentina, and hence a priority plant for conservation studies (Martinez et al., 2006). In contrast, A. venosa is not known regarding its nutritional and medicinal properties, although it can be mistaken for A. satureioides due to the similar morphological features, and wrongly used in the food and beverage industry. Frequently, these two species are * Correspondence to: J. M. Camiña, Facultad de Ciencias Exactas y Naturales, Universidad Nacional de La Pampa, Av. Uruguay 151 (6300) Santa Rosa, La Pampa, Argentina. E-mail: jcaminia@yahoo.com a Facultad de Ciencias Exactas y Naturales, Universidad Nacional de La Pampa, Av. Uruguay 151 (6300) Santa Rosa, La Pampa, Argentina b Instituto de Química de San Luis (INQUISAL), Facultad de Química Bioquímica y Farmacia, Universidad Nacional de San Luis, Chacabuco y Pedernera (5700) San Luis. Argentina c Departamento de Farmacia y Herbario, Projects 8702 and 22/Q616, Universidad Nacional de San Luis, Av. Ejército de los Andes 950 (5700) San Luis, Argentina d CONICET, Consejo Nacional de Investigaciones Científicas y Técnicas, Argentina. Av. Rivadavia 1917 (C1033AAJ) Ciudad de Buenos Aires, Argentina Copyright © 2010 John Wiley & Sons, Ltd. Phytochem. Anal. 2010, 21, 550–555 Characterization of Two South American Plants harvested together for medicinal and industrial purposes, because only an experienced botanist can distinguish them. For this reason, it is important to find a method to identify both plants in order to avoid errors or adulterations that can affect human health. Furthermore, there are no reports concerning the major, minor, trace and toxic elements in A. satureioides and A. venosa, although such knowledge is significant considering the wide use of these plants in South America. The presence of heavy metals in plants may cause serious health problems including renal failure, chronic toxicity and liver damage (Shaw et al., 1997; Andrew et al., 2003). According to the World Health Organisation (WHO, 1991), the concentration of lead, cadmium, chromium and other heavy metals must be controlled in medicinal plants in order to ensure their safety. Thus, the analysis of toxic and trace elements in food and medicinal plants has been targeted by several research groups worldwide (Abou-Arab and Abou Donia, 2000; Lemberkovics et al., 2002; Marchisio et al., 2005; Maiga et al., 2005; Nookabkaew et al., 2006; Gómez et al., 2007; Kara, 2009). For the purposes of classification, the most commonly used multivariate calibration methods (Massart et al., 1997; Mongay Fernandez, 2005) are: (1) Principal components analysis (PCA)—the original data are reduced to few new variables, known as principal components (PCs). This approach permits the detection of several characteristics that cannot be seen in the original data, which can help to identify the differences within the original data and to obtain groups or families. (2) Cluster analysis (CA) – based on the ability to establish relationships between data using different criteria of similitude. The use of amalgamation and distance criteria permit the construction of a tree or dendogram that shows the possible groups and their differences. (3) Linear discriminant analysis (LDA)—used to confirm the classification made through PCA or CA, which are unsupervised methods, in which the origin of grouping cannot be confirmed and that grouping can proceed from an unknown source. LDA is a supervised method that uses a model (discriminant equations) that can classify mathematically different groups within unknown samples (Mongay Fernandez, 2005). This paper discusses the multielemental composition of A. satureioides and A. venosa, including Al, As, Ba, Ca, Cd, Co, Cr, Cu, Fe, K, La, Mg, Mn, Mo, Na, Ni, P, Pb, S, Sr, Ti, V, Y and Zn, which were analysed by inductively coupled plasma optical emission spectrometry (ICP-OES), and the classification of both plants by means of the multivariate methods PCA, CA and LDA. Experimental Samples Phytochem. Anal. 2010, 21, 550–555 Ultrapure nitric acid was acquired from Sigma (St. Louis, MO, USA). All standard solutions were prepared using spectroscopic grade Merck (Darmstadt, Germany) reagents. Ultrapure water (18.2 MW cm) was obtained using a Barnstead Easy-Pure RF compact water system (Dubuque, IA, USA). Sample preparation and extraction The aerial parts of the plants were dried in a oven at 75°C during approximately 24 h until constant weight. Afterwards the material was ground using a Wiley 3379 series grinder and passed through a sieve (0.50 mm diameter). The powder (2.0 g) was transferred to a porcelain crucible and 5 mL of indium solution (500 mg/L) was added as internal standard to evaluate the degree of recovery. The mixture was covered with a cap and reduced to ashes in a furnace at 500°C for 6 h. The ashes were dissolved with 5.0 mL of concentrated nitric acid until the evolution of gases had stopped. Then the solution was transferred to a 50 mL volumetric flask and completed to the mark with deonised water (Abou-Arab and Abou Donia, 2000; Maiga et al., 2005). Analytical procedure The concentrations of the 24 elements and internal standard were determined by direct nebulisation using a Varian ICP-OES model ICP-OES Vista Pro, with a Czerny-Turner monocromator, holographic diffraction grid and a VistaChip charge coupled device (CCD) array detector. The wavelengths (nm) used for measurement of each element are indicated in Table 1. Calibration curves for each element were constructed in triplicate using five different concentrations. The regression coefficient (r2) values of such curves ranged from 0.995 to 0.999. Software PCA was carried out using the Unscrambler 6.11 software package (Trondheim, Norway). Cluster analysis was carried out using the InfoStat software package (Córdoba, Argentina). For CA, the amalgamation criterion was complete linkage and the selected distance was the Euclidean one. For LDA, the InfoStat software was also employed. Recovery assay In order to demonstrate the validity of the analytical procedure, a recovery study was carried out. A synthetic solution containing four toxic elements (Cr, Pb, Cd and As) was prepared and added to the dried samples Table 1. Wavelengths used for ICP-OES analysis Element Al As Ba Ca Cd Co Cr Cu Fe In K La Mg a Wavelengtha Element Wavelengtha 308.215 193.696 455.403 317.933 226.502 228.616 267.716 324.754 259.940 303.963 766.491 398.850 279.079 Mn Mo Na Ni P Pb S Sr Ti V Y Zn 257.610 202.032 588.995 231.604 213.618 220.353 181.972 181.972 337.280 292.402 371.031 213.856 Expressed in nm. Copyright © 2010 John Wiley & Sons, Ltd. 551 Nineteen samples of each plant population (A. satureioides and A. venosa) were harvested from the mountains near the city of San Luis, Argentina, at 850 m above sea level, in the summers of 2007 and 2008. The species were authenticated by comparison with herbarium samples deposited at the Universidad Nacional de San Luis (UNSL) and identified as follows: Achyrocline satureioides (Lam.) DC (voucher specimen: L.A. Del Vitto no. 8603) and Achyrocline venosa Rusby (voucher specimen: L.A. Del Vitto & al. no. 6765). Reagents View this article online at wileyonlinelibrary.com M. A. Cantarelli et al. (2.0 g) of Achyrocline venosa (Table 2). The extraction and analyses of the concentrations of the four elements were determined according to the protocol described earlier and the recovery percentage was determined. Results and Discussion Validation and recovery The method of standard addition is considered as a validation method (ICH, 1994; Prichard et al., 1996; Gómez et al., 2007). The addition of indium as internal standard serves for evaluating the recovery percentage as well as the losses in the ashing step (Kara, 2009). In the present work, the addition of indium was performed prior to the mineralisation step and the mean recovery value was 98.2 ⫾ 3.1% (n = 38), similar to the results reported by Abou-Arab and Abou Donia (2000). The recovery percentages obtained in the four-metal-recovery assay were considered satisfactory ranging from 98.7 to 102.0 % (Table 2). Multielement composition analysis Both plants contained high concentrations of K, Ca, Mg, P and S. The concentrations of K, Ca and Mg were similar to those reported previously for Hungarian herbs (Lemberkovics et al., 2002; 16276, 6281 and 1472 mg/kg, respectively), demonstrating that these elements are the most abundant in medicinal plants. In contrast, the concentrations of P and S in A. satureioides and A. venosa were smaller compared with those of Hungarian herbs (Lemberkovics et al., 2002; mean values: 2730 and 2085 mg/kg, respectively). The toxic elements Pb, Cr, Cd and As were also present in A. satureioides and A. venosa respectively, although the concentrations were below the permissible limits in foods, which are 10, 120, 0.3 and 4 mg/kg, respectively (WHO, 1991). The concentrations of Pb, Cr and Cd were lower in comparison with those of medicinal plants from Egypt (Abou-Arab and Abou Donia, 2000; mean values: 3.87, 16.01 and 1.45 mg/kg, respectively), but similar to those of Hungarian tea (Lemberkovics et al., 2002; <1.26 mg/kg for Pb and Cd; 3.47 mg/kg for Cr). All the other elements were present at low concentrations. Table 4 shows the acceptable limits for Zn, Ni, Cu and Fe (Maiga Monitoring of toxic metals in crop and medicinal plants is particularly important in developing countries where such plants are essential for the diet and health of the population (Maiga et al., 2005). Table 3 shows the concentrations of the 24 elements analysed in A. satureioides and A. venosa. Table 4. Permissible limits for medicinal herbs for Zn, Ni, Cu and Fe; and concentration ranges of Al, Ba, and Sr found in A. satuterioides and A. venosa Element Cr Pb Cd As Addeda Founda,b Recoveredc 4.52 ⫾ 0.82 2.06 ⫾ 0.13 0.21 ⫾ 0.02 0.58 ⫾ 0.21 2.0 1.5 0.3 1.0 6.56 3.54 0.51 1.57 102.0 98.7 100.0 99.0 60a 10a 40a 45b Zn Ni Cu Fe Table 2. Validation method for the recovery of toxic elements Initial valuea Limit Range ASa 52–1136 56–105 39–77 Al Ba Sr AVa 223–403 26–57 26–37 a Concentration in mg/kg. Mean value (n = 7). c [(found-initial)/added] ¥ 100. a b b Expressed in mg/kg. Expressed in mg per day. Table 3. Concentrations of major, minor, trace and toxic elements in Achyrocline venosa and Achyrocline satureioides Plants Element concentration (mg/kg dry weight )a Ba Ca Al As A. venosa A. satureioides 324 ⫾ 42 606 ⫾ 169 0.58 ⫾ 0.14 0.96 ⫾ 0.32 45 ⫾ 8 66 ⫾ 10 A. venosa A. satureioides Cr 4.52 ⫾ 0.55 3.33 ⫾ 1.38 Cu 17 ⫾ 2 12 ⫾ 1 A. venosa A. satureioides Mn 108 ⫾ 6 152 ⫾ 16 A. venosa A. satureioides S 1032 ⫾ 121 732 ⫾ 40 a Cd Co 6044 ⫾ 364 7448 ⫾ 51 0.21 ⫾ 0.02 0.16 ⫾ 0.02 0.70 ⫾ 0.16 0.36 ⫾ 0.08 Fe 482 ⫾ 14 639 ⫾ 22 K 10838 ⫾ 607 11737 ⫾ 812 La 0.25 ⫾ 0.08 1.75 ⫾ 0.47 Mg 1699 ⫾ 238 1438 ⫾ 280 Mo 0.99 ⫾ 0.13 1.07 ⫾ 0.15 Na 148 ⫾ 16 166 ⫾ 23 Ni 3.12 ⫾ 0.23 2.12 ⫾ 0.65 P 1672 ⫾ 161 1228 ⫾ 177 Pb 2.06 ⫾ 0.09 1.96 ⫾ 0.28 Sr 31 ⫾ 3 44 ⫾ 8 Ti 22 ⫾ 2 25 ⫾ 2 V 1.07 ⫾ 0.16 1.43 ⫾ 0.40 Y 0.17 ⫾ 0.05 0.41 ⫾ 0.11 Zn 39 ⫾ 2 28 ⫾ 5 552 Average ⫾ standard deviation (n = 19). View this article online at wileyonlinelibrary.com Copyright © 2010 John Wiley & Sons, Ltd. Phytochem. Anal. 2010, 21, 550–555 Characterization of Two South American Plants et al., 2005) and the concentration ranges for the nonessential elements Al, Ba and Sr detected in A. satureioides and A. venosa. Such concentrations were similar to those reported previously for other species (Nookabkaew et al., 2006; mean values: 1124, 59.65 and 60.29 mg/kg, respectively). In A. venosa, the concentration of Co was two-fold higher than that of herbal teas from Thailand (Nookabkaew et al., 2006; mean value: 0.260 mg/kg) and Turkey (Kara, 2009; mean value: 0.23 mg/kg), whereas in A. satureioides the concentration of Co was comparable to those detected in the same teas. The concentration of V was similar to that found in herbs from Thailand (Nookabkaew et al., 2006; mean value: 0.84 mg/kg) and Argentina (Gómez et al., 2007; mean value: 1.29 mg/kg). The concentration of La in A. satureioides was fivefold higher than in medicinal tea leaves from Turkey (Kara, 2009; mean value: 0.37 mg/kg), whereas in A. venosa the concentration 80 of La was similar to that found in the same tea. The concentrations of Ti and Mo were higher than those reported for other herbal medicines (Lemberkovics et al., 2002; mean values: 2.02 and <0.05 mg/kg, respectively). Principal component analysis The PCA model was obtained using three PCs, which can explain 99% of the original information on the variables. Through this model, it was possible to obtain a good classification. Figure 1 shows a 3D score plot, in which A. venosa is shown on the right (samples 1–19) and A. satureioides is shown on the left (samples 20–38). The final model was obtained using as variables the elements Al, Fe, Mg and S present in the 38 samples (19 for each species), since their influence is higher than that of the other elements. Figure 2 shows a 3D loading plot, which explains the influence of original variables on the model. S had an important 10 4 15 12 60 7 2535 2326 20 29 33 21 36 31 22 34 32 -20 0,4 8 16 3 17 2 186 9 13 1 5 11 19 30 -40 Al PC 3 0 Mg 0,6 14 -400 -300 -200 -100 0 100 200 300 PC 1 400 0,2 Fe 0,0 -0,2 80 40 0 -40 -80 -0,4 PC 2 PC 3 20 -60 0,8 24 28 38 3727 40 -0,6 -0,4 S -0,2 0,0 PC Figure 1. Three-dimensional score plot for 38 samples of Achyrocline genus: samples 1–19 belong to A. venosa (left group); samples 20–38 belong A. satureioides (right group). 0,8 0,6 0,4 0,2 1 0,2 0,4 PC 0,0 0,6 0,8 -0,2 2 Figure 2. Three-dimensional loadings plot to evaluate the influence of variables on the PC for Achyrocline genus classification. 1 6 5 2 3 8 4 18 9 13 10 14 7 11 16 17 12 15 19 20 21 26 23 24 27 38 28 29 22 33 30 32 34 25 31 35 36 37 0 100 200 300 400 500 600 700 Distance Dendogram plot of 38 samples of Acyrocline genus. Samples 1–19 A. venosa; samples 20–38 A. satureioides. Phytochem. Anal. 2010, 21, 550–555 Copyright © 2010 John Wiley & Sons, Ltd. View this article online at wileyonlinelibrary.com 553 Figure 3. M. A. Cantarelli et al. Table 5. Results of classification obtained by linear discriminant analysis A. venosa A. satureioides Total Error (%) 16 0 16 0 16 16 16 16 32 0 0 0 3 0 3 0 3 3 3 3 6 0 0 0 Training AV AS Total Prediction AV AS Total influence on PC1, while the influence of Mg, Al and Fe was not so great. Al, Mg and S greatly influenced PC2, whereas Fe had little influence. Regarding PC3, Mg was the most important variable, followed by Al and Fe; S seemed to present little influence. Cluster analysis The dendogram (Fig. 3) shows the classification obtained for A. venosa (samples 1–19) and A. satureioides (samples 20–38) using the variables Al, Fe, Mg and S. Both plants were successfully classified. Linear discriminant analysis Once again, Al, Fe, Mg and S were used as variables to confirm the results obtained with PCA and CA. The classification of Achyrocline plants according to LDA is shown in Table 5. Sixteen samples of each species were used for the training step, while for the prediction step only three samples of each species were used. All samples were predicted in the correct form without error for both cases in the training and prediction steps. The canonical discriminant function obtained using the variables mentioned was D = -20.06 + 0.01 Al + 0.02 Fe—0.02 Mg—0.01 S (Mongay Fernandez, 2005). The present work represents a valuable contribution for the improvement of knowledge regarding the elemental composition of A. satureioides and A. venosa. In addition, it was demonstrated that the two species, as well as their by-products, can be classified by three different methods according to their use as food and medicine. Hence the methods reported can be applied in the quality control of these plants and their derivatives, to detect possible adulterations or to avoid mistakes in the selection of raw materials for the beverage and phytopharmaceutical industries. Acknowledgements The authors are grateful to Professor Roberto Olsina (Universidad Nacional de San Luis) for providing The Unscrambler 6.11 software. We would also like to thank Facultad de Agronomía, Universidad Nacional de La Pampa for providing the InfoStat software. We are grateful for research grants from Consejo Nacional de Investigaciones Científicas y Técnicas, Universidad Nacional de La Pampa and Universidad Nacional de San Luis that supported this research. References 554 Abou-Arab AAK, Abou Donia MA. 2000. Heavy metals in Egyptian spices and medicinal plants and the effect of processing on their levels. J Agric Food Chem 48: 2300–2304. View this article online at wileyonlinelibrary.com Andrew AS, Warren AJ, Barchowsky A, Temple KA, Klei L, Soucy NV, O’Hara KA, Hamilton JW. 2003. Genomic and proteomic profiling of responses to toxic metals in human lung cells. Environ Health Persp 111: 825–838. Arredondo MF, Blasina F, Echeverry C, Morquio A, Ferreira M, Abin-Carriquiry JA, Lafon L, Dajas F. 2004. Cytoprotection by Achyrocline satureioides (Lam) D.C. and some of its main flavonoids against oxidative stress. J Ethnopharmacol 91: 13–20. Cosentino M, Bombelli R, Carcano E, Luini A, Marino F, Crema F, Dajas F, Lecchini S. 2008. Immunomodulatory properties of Achyrocline satureioides (Lam.) D.C. infusion: A study on human leukocytes. J Ethnopharmacol 116: 501–507. De Souza KCB,Schapoval EES, Bassani VL. 2002. LC determination of flavonoids: separation of quercetin, luteolin and 3-O-methylquercetin in Achyrocline satureioides preparations. J Pharmaceut Biomed 28: 771– 777. De Souza KCB, Andrades ME, Da Silva EG, Bonatto F, Heydrich J, Dal-Pizzola F, Schapoval EES, Bassani VL, Moreira JCF, Polydoroa M. 2004. Antioxidant, pro-oxidant and cytotoxic effects of Achyrocline satureioides extracts. Life Sci 74: 2815–2826. Dickel ML, Kuze Rates SM, Rejane Ritter M. 2007. Plants popularly used for loosing weight purposes in Porto Alegre, South Brazil. J Ethnopharmacol 109: 60–71. Eisenberg DM, Davis RB, Ettner SL, Appel S, Wilkey S, Rompay MV, Kessler RC. 1998. Trends in alternative medicine use in the United States, 1990–1997: results of a follow-up national survey. J Am Med Assoc 280: 1569–1575. Ernst E. 2002a. Toxic heavy metals and undeclared drugs in Asian herbal medicines. Trends Pharmacol Sci 23: 136–139. Ernst E. 2002b. Heavy metals in traditional Indian remedies. Eur J Clin Pharmacol 57: 891–896. Gillij YG, Gleiser RM, Zygadlo JA. 2008. Mosquito repellent activity of essential oils of aromatic plants growing in Argentina. Bioresource Technol 99: 2507–2515. Gómez MR, Cerutti S, Sombra LL, Silva MF, Martinez LD. 2007. Determination of heavy metals for the quality control in argentinian herbal medicines by ETAAS and ICP-OES. Food Chem Toxicol 45: 1060–1064. Gugliucci A, Menini T. 2002. Three different pathways for human LDL oxidation are inhibited in vitro by water extracts of the medicinal herb Achyrocline satureoides. Life Sci 71: 693–705. ICH. Harmonised Tripartite Guideline. Text on Validation of Analytical Procedures. International Conference on Harmonization: Geneva, 1994. Kadarian C, Broussalis AM, Miño J, Lopez P, Gorzalczany S, Ferraro G, Acevedo C. 2002. Hepatoprotective activity of achyrocline satureioides (LAM) D.C. Pharmacol Res 45: 57–61. Kara D. 2009. Evaluation of trace metal concentrations in some herbs and herbal teas by principal component analysis. Food Chem 114: 347– 354. Lemberkovics E, Czinner E, Szentmihályi K, Balázs A, Szoke E. 2002. Comparative evaluation of Helichrysi flos herbal extracts as dietary sources of plant polyphenols, and macro- and microelements. Food Chem 78: 119–127. Maiga A, Diallo D, Bye R, Paulsen BS. 2005. Determination of some toxic and essential metal ions in medicinal and edible plants from Mali. J Agric Food Chem 53: 2316–2321. Marchisio PF, Sales AM, Cerutti S, Marchevsky EJ, Martínez LD. 2005. On-line preconcentration/determination of lead in Ilex paraguariensis samples (mate tea) using polyurethane foam as filter with USN-ICPOES. J Hazard Mater 124: 113–118. Martinez GJ, Planchuelo AM, Fuentes E, Ojeda M. 2006. A numeric index to establish conservation priorities for medicinal plants in the Paravachasca Valley, Córdoba, Argentina. Biodivers Conserv 15: 2457–2475. Massart DL, Vandeginste BGM, Buydens LMC, De Jong S, Lewi PJ, SmeyersVerbeke J. 1997. Handbook of Chemometrics and Qualimetrics, Vol. A. Elsevier: Amsterdam; 366–368. Mongay Fernandez, C. 2005. Quimiometría. Publicaciones Universidad de Valencia: Valencia; 247–280. Nookabkaew S, Rangkadilok N, Satayavivad J. 2006. Determination of trace elements in herbal tea products and their infusions consumed in Thailand. J Agric Food Chem 54: 6939–6944. Prichard E, Mackay GM, Points J. 1996. Trace Analysis: A Structures Approach to Obtaining Reliable Results. The Royal Society of Chemistry: London; 38. Rivera F, Gervaz E, Sere C, Dajas F. 2004. Toxicological studies of the aqueous extract from Achyrocline satureioides (LAM) D.C. (Marcela). J Ethnopharmacol 95: 359–362. Copyright © 2010 John Wiley & Sons, Ltd. Phytochem. Anal. 2010, 21, 550–555 Characterization of Two South American Plants Rojas de Arias A, Ferro E, Inchausti A, Ascurra M, Acosta N, Rodriguez E, Fournet A. 1995. Mutagenicity, insecticidal and trypanocidal activity of some Paraguayan Asteraceae. J. Ethnopharmacol 45: 35–41. Ruffa MJ, Ferraro G, Wagner ML, Calcagno ML, Campos RH, Cavallaro L. 2002. Cytotoxic effect of Argentine medicinal plant extracts on human hepatocellular carcinoma cell line. J. Ethnopharmacol 79: 335– 339. Shaw D, Leon C, Kolev S, Murray V. 1997. Traditional remedies and food supplements. A 5-year toxicological study (1991–1995). Drug Safety 17: 342–356. WHO. 1991. Guidelines for the Assessment of Herbal Medicines. World Health Organisation: Munich; 28.6. Yeh GY, Eisenberg DM, Davis RB, Phillips RS. 2002. Use of complementary and alternative medicine among persons with diabetes mellitus: results of a national survey. Am J Pulic Health 92: 1648–1652. 555 Phytochem. Anal. 2010, 21, 550–555 Copyright © 2010 John Wiley & Sons, Ltd. View this article online at wileyonlinelibrary.com