Introduction

Taros (Colocasia esculenta) are tubers belonging to the Araceae family. Even though the exact origin of taros is unclear, several authors claim their origin to be Asia, specifically India or the Malay Peninsula (Englberger et al. 2008; Perdomo 2000). According to Onwueme (1999), taros were introduced in Africa approximately 2000 years ago. In recent years, taro cultivation has spread to many countries (mainly in tropical and subtropical regions) and is now one of the world’s staple food crops (Lebot et al. 2009; Lewu et al. 2010). The estimated annual world production in 2011 was 9.53 million tons, grown on a cultivated area of 1.24 million ha (FAOSTAT 2011). Taro cultivation is firmly established in the Canary Islands because of its high local consumption, as well as for traditional and economic reasons. According to official statistics, taro production in the Canary Islands was 524.9 tons in 2011, occupying 18.1 ha, and out of that total, 211.7 tons was produced on the island of Tenerife, in 7.3 ha (EAC 2011).

Food consumption is one of the most important pathways for human exposure to metals (Essumang et al. 2007; Turconi et al. 2009; Yan et al. 2007). Tubers have high ability to accumulate metals from the environment through the roots and foliar surfaces (Luis et al. 2011). Several metals like copper (Cu), zinc (Zn), and iron (Fe) are essential micronutrients for living organisms because they are part of the chemical makeup of enzymes. Other metals, such as cadmium (Cd) and lead (Pb), constitute non-essential, toxic elements of particular concern as environmental food contaminants for their toxicity at low concentrations (Gupta et al. 2012; Cheriaghi et al. 2013; Luis et al. 2014).

Metal monitoring in taros is a field of research that yields information that is necessary to assess safety, develop regulatory guidelines, and to determine nutritional values and essential elements in the diet. This study is intended to quantify the essential and trace elements in taros in order to create a database encompassing both food composition and mineral intake. Therefore, the aims of this study were (i) to determine the levels of ten essential elements (Na, K, Ca, Mg, Cu, Fe, Mn, Zn, Cr, and Ni) and two toxic metals (Cd and Pb) in taros sold on the island of Tenerife (Canary Islands, Spain), (ii) to estimate the contribution of essential elements present in taros to the current recommended daily intake (RDIs), and (iii) to evaluate the contribution of toxic metals to the current tolerable weekly intake (TWIs) and tolerable dairy intake (TDIs).

Materials and methods

Sampling

A total of 42 samples of taros were randomly obtained from supermarkets, vegetable markets, and farmer’s plots on the island of Tenerife (Canary Islands, Spain). All samples were harvested between September 2011 and May 2012.

Sample treatment and analytical procedure

Fresh taros were processed to separate the mesocarp (pulp) and the exocarp (skin). The edible portion (pulp) was the only part considered for analysis. Each sample of taro was thoroughly washed under tap water for 20–30 s to eliminate dirt and soil particles and then washed again with Milli-Q (Millipore, Milford, USA) quality water for 15–20 s. Finally the samples were peeled. Twenty grams of each sample was placed in a porcelain crucible and then oven-dried at 80 ± 10 °C for 48 h (Nabertherm Inc, USA). The dried samples were then subjected to pyrolysis in a muffle oven at 450 ± 10 °C for 50 h. The white ashes obtained were dissolved in 5 % nitric acid (HNO3) up to a volume of 50 mL (Luis et al. 2011, 2014). Before sample preparation, all laboratory materials used were washed with Acationox laboratory cleaning agent (Merck, Darmstadt, Germany) to avoid contamination and to remove any possible trace metals and kept in 5 % HNO3 for 24 h before being washed with Milli-Q quality water (Millipore, Bedford, USA).

Na, K, Ca, Mg, Cu, Fe, Mn, and Zn were determined using a Perkin-Elmer flame atomic absorption spectrometer (FAAS) model 2100, equipped with hollow cathode lamps (Wellesley, MA, USA). Cr, Ni, Cd, and Pb contents were analyzed by graphite furnace atomic absorption spectrometry (GFAAS), using a Perkin Elmer model 4100 ZL Zeeman spectrometer (Wellesley, MA, USA) equipped with a graphite furnace tube with an automatic AS70 sampler (Wellesley, MA, USA). In the analysis of Cr, Ni, Cd, and Pb, a mixture of NH4H2PO4 and Mg(NO3)2 was used as matrix modifier. All analyses were carried out in duplicate.

Quality control

Measurement quality controls were performed using blank samples and the following certified reference materials (SRM) from The National Institute of Standard and Technology (NIST) (Gaithersburg, MD USA): SRM 1515 apple leaves, SRM 1570a trace elements in spinach leaves, and SRM 1573a tomato leaves. Recovery rates obtained from the reference materials were over 94 % (Table 1). Instrumental detection and quantification limits in terms of reproducibility, calculated as 3 and 10 times the standard deviation (SD) resulting from analysis of 15 targets of acid digest (IUPAC 1995), are detailed in Table 2.

Table 1 Quality control study
Full size table
Table 2 Detection and quantification limits
Full size table

Statistical analysis

The statistical analyses of the metals being studied were performed using the data analysis program SPSS Inc., version 19.0. For the normality data assessment, all the results were tested with the Kolmogorov-Smirnov and Shapiro-Wilk tests (Xu et al. 2002). For variance homogeneity, results were tested with the Levene test (Pan 2002). Because our data showed an unusual distribution, the following statistical tests were used: a nonparametric test, the Kruskal-Wallis test, which allows discrimination of individual samples with significantly different results, and the Mann-Whitney U test, to establish whether there were significant differences between the groups or not (Choy et al. 2001). Probability values below 0.05 were considered statistically significant.

The dietary intake of each element was calculated by multiplying the concentration of the element in taros by the mean consumption for this food (10.41 g/person/day) established by the Food and Agriculture Organization (FAO) (FAOSTAT 2009; Luis et al. 2011).

Results and discussion

In the present study, 4 macroelements (Na, K, Ca, Mg), 6 microlements (Cu, Fe, Mn, Zn, Cr, Ni), and 2 toxic metals (Cd and Pb) were analyzed in 42 samples of taro (Table 3).

Table 3 Range, mean, and standard deviation of essential and toxic metals in taro samples (mg/kg)
Full size table

From a quantitative point of view, K turns out to be the most important macromineral with levels ranging between 1375 and 4340 mg/kg. The order of the concentrations for the rest of macroelements analyzed was (mg/kg) Na > Mg > Ca (Table 3). Significant differences were observed between the mean levels of Na, Ca, and Mg (p < 0.05). As for microelements, for all samples, the results showed that Fe was the most abundant trace element and the range obtained for this metal was 2.116–6.200 mg/kg. The mean concentrations of microelements (mg/kg) for all taro samples, listed in decreasing order, were Zn > Mn > Cu > Cr > Ni (Table 3). Statistical analysis revealed that significant differences were observed between the mean levels of Fe, Cu, and Cr (p < 0.05).

With respect to toxic metals, mean contents of Cd in the analyzed taro samples ranged between <0.001 and 0.009 mg/kg. Only 11.9 % of the samples analyzed had Cd levels lower than the quantification limit (0.001 mg/kg). The European Commission Regulation (EC) No. 1881/2006 has set a maximum level for Cd in tubers of 0.1 mg/kg wet weight (EC 2006). So, mean contents of Cd in the analyzed taro samples did not exceed the legal limit in any of the analyzed samples. Pb contents in the taro samples ranged from <0.001 to 0.007 mg/kg. In this study, 45 % of the samples analyzed had Pb levels lower than the quantification limit (0.001 mg/kg). In all samples, Pb levels were lower than the maximum limit set by the European regulation (0.10 mg/kg weight) (EC 2006).

Mg, Na, and Cd levels were higher, to a statistically significant degree, in the samples gathered during the rainy season than in the dry season. For the other metals, the differences between the samples gathered in the two different seasons were statistically insignificant. The results partly confirm that metal absorption in tubers is higher during the rainy season (Ekpetsi Bouka et al. 2013).

The relationships between the levels of individual metals in the taros analyzed were calculated using the Spearman rank correlation analysis. Statistical significance of coefficients of correlation was tested at the levels of p < 0.05 and p < 0.01. High positive correlations (p < 0.05) were found between the concentrations: Na-Mg (r = 0.709), Mg-Cu (r = 0.632), Mg-Cr (r = 0.588), Cu-K (r = 0.759), Cu-Cr (r = 0.671), Fe-Cr (r = 0.526), Fe-K (r = 0.636), and Fe-Mn (r = 0.564). Moreover, there were statistically significant positive differences (p < 0.01) in the levels of K-Mn (r = 0.457), Mg-Mn (r = 0.504), Cu-Fe (r = 0.519), Cu-Mn (r = 0.464), Cr-K (r = 0.505), Cr-Mn (r = 0.467), and Fe-Cd (r = 0.467), and significant negative correlations between the concentrations of Ca-Ni (r = −0.487) and Na-Pb (r = −0.503) were noted (Table 4).

Table 4 Spearman rank correlations between taro concentration of different metals
Full size table

The levels of metals in taros can vary significantly and may be influenced by many factors such as the taro variety, the production area, the soil and climate, agricultural practices (the addition of fertilizers and metal-based pesticides), transportation, harvesting processes, and storage and commercialization conditions (Luis et al. 2011, 2014). Table 5 shows a comparison between the metal levels obtained from this study and previous results published in the existing literature. Taro’s mineral element contents vary widely among different countries (Bhandari et al. 2003; Leterme et al. 2006; Nabulo et al. 2006; Onianwa et al. 2000; Orech et al. 2007; Englberger et al. 2008; Zhuang et al. 2009). In general, the mean concentrations of macro- and microelements in this study were within the concentration ranges found in Nepal (Bhandari et al. 2003), although the contents of Na and Mg (565.6 and 364.5 mg/kg, respectively) found in this study were considerably higher. Among the various studies, the highest toxic metal contents were found in Uganda (Cd) and in China (Pb) (Nabulo et al. 2006; Zhuang et al. 2009), possibly due the industrialization of the areas near the plantations of these tubers (Yan et al. 2007; Zhuang et al. 2009) (Table 5).

Table 5 Comparison of macro-, micro-, and toxic elements in taros from different countries
Full size table

In order to assess the contribution of taro consumption to the RDIs of essential metals for the population of the Canary Islands, a figure for the average daily consumption of taros (10.41 g/person/day) was drawn from the data published by the FAO (FAOSTAT 2009). In addition, default body weights of 60 kg for women and 70 kg for men were assumed.

Table 6 shows the average daily essential metal intakes from taro consumption and their contributions to the RDIs. Among the macrominerals, the highest contribution to the intakes was observed for Mg (1.265 % in adult women and 1.084 % in adult men), followed by K, Na, and Ca. Among microelements, the highest contribution to the intakes was observed for Cu, followed by Cr, Mn, Fe, and Zn. For Ni, the estimated intake was 0.219 μg/day for adults. Based on a TDI of 8 μg Ni/kg body weight/day as set by the EFSA (EFSA 2005), taro consumption contributes 0.045 % of the TDI of Ni for adult women and 0.039 % for adult men.

Table 6 Mean dietary intake (mg/day) of taro and contribution (%) to the daily requirements of macro- and microelements for adults in the Canary Islands
Full size table

The FAO’s figure of 10.41 g/person/day of taro was also used to evaluate the contribution of taro consumption to the intake of toxic metals. In 2009, the EFSA’s Panel on Contaminants in the Food Chain (CONTAM Panel) established a TWI of 2.5 μg/kg body weight for Cd (EFSA 2011). Taro consumption is estimated to contribute 0.031 μg/day to the intake of Cd in the adult population of the Canary Islands. This Cd intake represents 0.144 % of the EFSA reference value for adult women and 0.124 % for adult men.

The estimated intake of Pb from taro consumption is 0.062 μg/day. In 2010, the EFSA concluded that the current TDI of Pb of 0.5 μg/kg body weight (b.w.) should be maintained in order to ensure a high level of consumer protection, including subgroups of the population such as children, vegetarians, or people living in highly contaminated areas (EFSA 2010). Taro intake in the Canary Islands is estimated to contribute 0.207 % to this Pb intake limit for adult women and 0.177 % for adult men.

Conclusions

This study confirms that taros are a source of essential dietary elements, particularly K and Fe. This study adds to the current body of knowledge on the content of essential and toxic metals in taros, providing data that is useful from the point of view of confirming their safety and quality. The results obtained in the present study suggest that exposure to toxic metals (Cd and Pb) through taro consumption is low, and that there is negligible risk of their respective TDI/TWIs being exceeded as a result of taro consumption by Canarian consumers.