Evaluation of Macro- and Microelement Levels in Black Tea in View of Its Geographical Origin

The aim of this study was to evaluate the elemental composition of black tea samples and their infusions in view of their geographical origin. In total, 14 elements were analyzed, 13 (Ca, K, Mg, Na, Mn, Fe, Zn, Cu, Cr, Ni, Co, Cd, and Pb) by flame atomic absorption spectrometry, and P by UV-Vis spectrometry, after mineralization of samples. It was found that K was the most abundant macroelement in the analyzed samples, whereas among microelements, the highest concentration was found for Mn. Based on the obtained data, the percentage of elements leached into the infusions as well as the daily elemental intake from tea were calculated. The daily intake from tea was compared to the recommended daily allowances (RDAs), and the highest percentages of the RDAs were found for Mn (15 %) and Co (10 %). To study the relations between elemental composition and country of origin of samples, factor analysis and cluster analysis were applied. These multivariate techniques proved to be efficient tools able to differentiate samples according to their provenance as well as plantation within the common regions.


Introduction
According to the Food and Agriculture Organization of the United Nations Statistics Division (FAOSTAT), total tea production was estimated at 5.3 million tons in 2011, which makes it the most commonly consumed beverage around the world. Tea production in Europe reaches 513,000 tons per year [1]. The main world tea producers are China and India (1.4 and 0.92 million tons, respectively) [2], whereas in Europe, the Russian Federation and the UK (184,000 tons and 129,000 tons, respectively) are the leading producers [1]. The countries with the highest rate of tea consumption include Paraguay and Afghanistan, 7.93 and 4.55 kg/year/person, respectively [2]. According to FAOSTAT data, Malta has the largest tea consumption per capita in Europe amounting to 2.24 kg/year [1], followed by the UK and Ireland, whereas Poland is in seventh place with 0.92 kg/year [1]. According to Polish literature, the average Polish resident consumes 2-3 cups of tea per day, and 20 % of the population consumes 4-5 cups per day [3]. Many studies have reported positive effects of tea consumption on human health, such as cancer prevention, diabetes management by reducing glucose and cholesterol levels in blood, and improved immune defense [4]. Most of the tea world consumption is black tea (80 %), whereas the remaining 20 % belong to green, oolong, red, and yellow types of this beverage. The price of tea products as well as consumer interests are usually connected with the certified geographical origin [5]. Therefore, it is important to have tools that are able to ensure good quality of products, which is usually associated with their geographical provenance. There are several studies of authentication of tea origin, using multivariate chemometric techniques, such as principal component analysis (PCA), factor analysis (FA), cluster analysis (CA), or linear discriminant analysis (LDA) [5]. These techniques have been applied to elemental and organic composition data, obtained using various analytical techniques including HPLC [6][7][8], GC-MS [9][10][11], 1 HNMR [12], FT-NIRS [13], ICP-MS [5,[14][15][16], and FAAS [17][18][19][20]. Our previous studies have showed that FA and CA are efficient tools of green, fruit, and Pu-erh tea diversification [19,20]. These two multivariate techniques applied to green tea data enabled differentiation of samples according to their provenance [19]. Moreover, it was also possible to diversify Pu-erh tea according to its type of confection, whereas fruit teas were differentiated according to their type [20]. Such information is of great importance to consumers, who expect that price is an equivalent of good quality products.
Therefore, in this study, we aimed to verify the geographical origin of commercialized black tea samples applying factor analysis (FA) and cluster analysis (CA) to their elemental composition. These techniques were found useful to determine the country of origin of the tea as well as its provenance within a single country. In addition, the percentage of elements leached into the tea infusions was determined, and the daily elemental intake from tea was calculated and compared to the recommended daily allowances (RDAs). For Cd and Pb, the intake from tea was compared to the provisional tolerable weekly intake (PTWI).

Samples
The analyzed tea samples (loose form and tea bags) were purchased from markets and tea shops (original tea) in Poland, and analyzed for their content of 14 elements: K, Na, Ca, Mg, Mn, Zn, Cu, Fe, P, Co, Ni, Cr, Cd, and Pb. In total, 118 types of black tea from different producers were tested, i.e., 708 analytical samples of black tea leaves and their infusions were prepared. Of the 118 tea types analyzed, 43 were purchased in tea shops (12 Chinese, 17 Indian, 10 Ceylon, and 4 Kenyan teas), and 75 in markets (loose form and tea bags) ( Table 1).

Preparation of Samples and Elemental Analysis
The bulk teas were homogenized and representative samples were mineralized in an electric furnace and then analyzed by flame atomic absorption spectrometry (FAAS) according to the previously published procedure by Brzezicha-Cirocka et al. [19,20].

Method Validation
The limit of detection (LOD) and limit of quantification (LOQ) for all the analyzed elements were calculated based on the independently prepared blank samples measurements. According to the method described by Konieczka and Namieśnik [21], LODs were set to blank means +3SD, where blank mean is a result of all blank samples measurements and SD is their standard deviation; whereas LOQ was calculated by multiplying LOD by a factor of three. Data for the validation procedure are given in Table 2.
The reliability of the method was determined using certified reference materials, i.e., Oriental Basma Tobacco Leaves (INCT-OBTL-5) and Polish Virginia Tobacco Leaves (INCT-PVTL-6). They were prepared according to the same procedure as the analytical samples. Recoveries of the studied elements ranged between 87 and 113 % (RSDs between 0.02-10.3 %) of the certified values for all the elements (Table 3).

Statistical Analysis
The Shapiro-Wilk test showed that the data were not normally distributed; therefore, nonparametric tests were applied [22]. Moreover, data standardization was adopted, and the correlation analysis was performed using Spearman rank analysis. Kruskal-Wallis test, factor analysis, and cluster analysis were conducted in order to obtain statistically significant information about the quality and origin of samples.

Macroelements
The highest contents of Ca and Na were found in marketed teas (264 and 89.3 mg/100 g, respectively), whereas Indian and Kenyan teas were characterized by the lowest Na level (30.7 and 21.8 mg/100 g, respectively). Products from China, India, Ceylon, and Kenya had similar amounts of Ca (153-168 mg/ 100 g). Comparable Ca results (215 mg/100 g) were obtained by Dambiec et al. [23], whereas similar Na levels (88 mg/100 g) were reported by Soomro et al. [24] and Yemane et al. [25]. Among all macroelements, the highest levels were found for K (2349-2981 mg/100 g). Considerable variation was found in the content of Mg, as the highest levels amounted to 822 mg/100 g in Indian teas, and the lowest to 518 mg/100 g in marketed teas ( Table 4). The latter value is higher than the one obtained by Gerbresadik and Chandravanshi [26] (354 mg/100 g). Indian, Ceylon, and Kenyan products contained similar amounts of P (305-359 mg/100 g), and Chinese tea had the highest P concentration (408 mg/100 g). Malik et al. [27] published comparable results for P (366 mg/100 g). Dambiec et al. [23] estimated a percentage of Na leaching (45.3 %) similar to ours (40.0 %), while Szymczycha-Madeja et al. [17] classified this macroelement as highly extractable (>55 %), which was confirmed in the case of teas from Kenya (62.0 %). The percentages of Ca leaching were the highest in Chinese teas (17.2 %) and the    Many factors influence the contents of trace metals in tea leaves. According to Milani et al. [33], variations in mineral composition of tea leaves can be explained by the age of leaves used in the production (old or young), soil composition, rainfall amount, and growing conditions in general. Unfortunately, commercial teas often have unknown geographical origin, as they are a mixture of leaves from different locations [33]. Black tea is produced through leaf fermentation in contrast to green tea, which results in higher levels of certain trace elements [34]. Regulations have been established for many vegetable products in many countries, but the European Union has no specified regulations about the acceptable metal content in tea.

Toxic Metals
In general, the analyzed samples contained more Pb than Cd. The levels of Cd varied little, i.e., 0.003-0.005 mg/100 g. Similar results for Cd were obtained by Milani et al. [33] (0.001-0.002 mg/100 g) and significantly lower ones by Barone  The contents of heavy metals in tea leaves may be a result of contamination that can be caused by many factors such as use of various manufacturing and agronomic processes and of fertilizers [29,35,36]. Among the main sources of Pb in the environment are leaded fuel, waste incineration, and industry [37]. Lead pollution is correlated with urbanization and population density. Moreover, higher levels of this heavy metal in tea samples can be attributed to contamination during the process of tea production and its packing [37]. Cadmium found in tea leaves might be a result of phosphate and zinc fertilizers usage [38]. It is estimated that Cd from phosphate fertilizers constitutes >50 % of the total input to agricultural land not heavily polluted or heavily industrialized [39,40]. Sarma et al. [41] reported that heavy metal contaminations of tea leaves might be explained by the position of the tea cultivation area. In that study, an Assam plantation was situated in the vicinity of the oldest crude oil exploration station, and accidental spillage during drilling and transportation could be the main sources of tea field contamination. Moreover, it has been established that heavy metals enter into plant bodies in acidic soil, in which tea usually grows [42].

Correlation
Nonparametric Spearman's rank test was performed at three levels of significance (p < 0.05, p < 0.01, p < 0.001), and both positive and negative correlations were found between the analyzed elements. The most significant positive correlations (p < 0.001) were found for P-Zn, Mn-Co, and Mn-Cr for all the countries. There were also recorded important interelement correlations among black tea samples from Indian and Ceylon plantations. Strong interdependences (p < 0.01) were also found between Ca-Mn, P-Zn, Mn-Co, Zn-Ni, and Cu-Co.

Kruskal-Wallis Test
Through the Kruskal-Wallis test, it was possible to determine statistically significant differences in the analyzed database. Relationships were found between the geographical provenance of tea and concentrations of elements including the following: Na ( Table 5. Kruskal-Wallis test and Dunn's test (Table 6) were also performed for tea samples from Asia, i.e., India (plantation Assam and Darjeeling) and the island of Ceylon. Relationships were found between provenance of tea and concentrations of sev-

Factor Analysis
The results of the factor analysis (FA), conducted on raw data sets of black tea obtained from tea shops, are shown in Fig. 1a and b. Factor analysis performed with all the analyzed metals did not give clear output, and thus, we decided to narrow the data set to the elements Na, Mn, Ni, Cu, Fe, P, and Zn. Factor analysis was applied to data of elements which were found significant in the Kruskal-Wallis test. The final choice of descriptors was done by a series of factor analyses, which were performed in order to verify the clarity of the outcome. As a result, two factors were obtained, i.e., F1 (35.7 % of the total variance) and F2 (23.5 % of the total variance). Both factors cumulatively explain 59.2 % of the total variance, whereas the eigenvalues for F1 and F2 are 2.14 and 1.41, respectively. As can be seen in Fig. 1a, samples from all the analyzed regions can be distinguished, i.e., Ceylon, China, India, and Kenya. The scatter plot of loadings was drawn for F1-F2 in order to identify elements responsible for the grouping of objects (Fig. 1b). Higher values of F1 correspond to Indian and Chinese samples, which were described by Ni, Cu, Fe, P, and Zn. Phosphorus, Fe, and Zn were responsible for differentiation of Chinese tea, whereas the highest amounts of Ni and Cu were noted in Indian samples. Such high levels of these metals in samples of these two origins could possibly be explained by product contamination during its manufacture or area pollution.
Simultaneously, high amounts of P in Chinese samples may be related to the increasing use of fertilizers, which was confirmed by research conducted by Mupenzi et al. [30]. Lower values of F1 characterize Kenyan and Ceylon tea samples which are described by Mn and Na (Fig. 1a, b). Kenyan tea was differentiated by Mn as significant differences were found in Mn concentrations between these two groups of tea. Sodium was found to be a descriptor of Ceylon teas, which may be associated with geographical location of the plantation, as Ceylon is an island surrounded by the waters of the Indian Ocean. Higher values of F2 described Ceylon, Chinese, and partly Indian tea samples, which can be identified by Na, Fe, P, and Zn. Lower values of F2 were associated with Indian and Kenyan samples, which were differentiated by Mn, Ni, and Cu (Fig. 1a, b).
Factor analysis was also performed for Indian plantations, i.e., Assam and Darjeeling, and Ceylon samples in order to differentiate samples within plantations (Fig. 2a, b). It was  . The factor analysis clearly differentiated between tea samples originating from different plantations. Ceylon samples were described by higher F1 values and Na, which might be associated with its significant content in soils of this country (Fig. 2a, b). Lower values of F1 corresponded to Assam and Darjeeling samples, which were described by P, Fe, Zn, Ni, Cr, Cu, and Mn. Tea samples from Darjeeling were differentiated by P, Fe, and Zn, whereas Assam tea samples were significantly correlated with Cr, Cu, and Mn. Factor 2 differentiated samples from India (Assam and Darjeeling) as its higher values corresponded to Darjeeling plantations and lower values to Assam ones. Although Ni is characterized with positive F2 loading, there were no significant variations in its content in Assam and Darjeeling samples.

Cluster Analysis
The cluster analysis (CA) was based on Ward's method with the usage of the Euclidean distance. Application of CA made it possible to differentiate samples according to their origin, i.e., Ceylon and Indian plantations (Fig. 3). Assam plantations were discriminated by Cr, Cu, and Mn, and Darjeeling ones by Fe, Zn, and P. Ceylon tea samples corresponded to Na, which was already confirmed to be a discriminative element for this region. However, only few samples were assigned to the improper cluster, which might be due to the similarity between samples, especially in view of their contamination with heavy metals. What is more interesting, it was noted that Darjeeling samples are more similar to Ceylon samples than to the other Indian plantation (Assam) (Fig. 3).
Thus, it can be concluded that CA similarly to FA is able to distinguish tea samples in view of their geographical origin, which might be helpful when authenticity assessment and fraud detection is necessary.

Recommended Dietary Intake
The daily intake of bioelements from tea was evaluated in view of the latest available Polish [43] and American recommended dietary intakes (RDA) [44]. Consumption of one cup of black tea (200 mL) results in intakes of Ca, K, Na, Mg, and P in the range of 0.02-1.30 % of their respective RDAs. Thus, black tea is not a rich source of these macroelements, despite the fact that their concentrations in tea leaves are the highest among all the analyzed elements ( Table 7). Realization of the recommended daily intakes for microelements such as Fe, Zn, Cu, and Ni is less than 1.0 %. For Cr, intake of 200 mL of tea infusion supplies 2.8 and 4.0 % of the RDA for men and women, respectively. Black tea is a significant source of Mn and Co, one cup daily provides 15 % of the RDA for Mn and 10 % for Co. However, Mn bioavailability amounts up to 40 % [45]. Therefore, only 6 % of Mn present in one cup of black tea will be absorbed. Moreover, according to the World Health Organization [46] there is no quantitative information available to indicate toxic levels of manganese in the human diet. High Co content in tea leaves may possibly be explained by the use of Co-containing fertilizers. There is evidence that Co in higher plants promotes the formation of chlorophyll and plant growth [47]. According to Kabata-Pendias and Szteke [47], approximately 50 % of Co will be absorbed in the gastrointestinal tract. Furthermore, Co absorption can be increased among individuals who are Fe deficient. There were also assessed levels of heavy metals such as Pb (Table 7) and Cd in the infusions, but the latter was under the limit of detection of the method applied.

Conclusions
Although specified regulations concerning tea quality are not established in the European Union, it is important to control it as this beverage is one of the most commonly consumed. Therefore, we determined the elemental composition of various black tea samples originating from China, India, Ceylon, and Kenya. We conclude that there is no significant health risk associated with consumption of the analyzed tea samples, but  [43] c In the form of vitamin B 12 that tea can constitute a valuable source of manganese in the human diet. Based on the obtained elemental data, the percentage of leaching as well as daily intake realization were assessed. The highest level of RDAs' realization was noted for Mn (15 %) and Co (10 %). Verification of the interdependences between elemental composition and country of origin of samples was done by multivariate techniques such as factor analysis and cluster analysis. They allowed on differentiation of teas according to the country of origin, i.e., China, India, Ceylon, and Kenya. Moreover, they were found helpful in diversification of teas originating from various plantations within a single country. Thus, they proved to be good tools able to differentiate samples in view of their provenance as well as plantation within the common region.
Compliance with Ethical Standards The manuscript does not contain clinical studies or patient data.

Conflict of Interest
The authors declare that they have no conflict of interest.
Funding Open Access This article is distributed under the terms of the Creative Comm ons Attribution 4.0 International License (http:// creativecommons.org/licenses/by/4.0/), which permits unrestricted use, distribution, and reproduction in any medium, provided you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons license, and indicate if changes were made.