Multi-element analysis for determining the geographical origin of mutton from different regions of China

Multi-element analysis for determining the geographical origin of mutton from different regions of China

Food Chemistry 124 (2011) 1151–1156 Contents lists available at ScienceDirect Food Chemistry journal homepage: www.elsevier.com/locate/foodchem Ana...

314KB Sizes 0 Downloads 1 Views

Food Chemistry 124 (2011) 1151–1156

Contents lists available at ScienceDirect

Food Chemistry journal homepage: www.elsevier.com/locate/foodchem

Analytical Methods

Multi-element analysis for determining the geographical origin of mutton from different regions of China Shumin Sun a,b, Boli Guo a, Yimin Wei a,*, Mingtao Fan b a

Key Laboratory of Agriculture Product Processing and Quality Control, Ministry of Agriculture, Institute of Agro—Food Science and Technology, Chinese Academy of Agricultural Sciences, P.O. Box 5109, Beijing 100193, PR China b College of Food Science and Engineering, Northwest Sci-Tech University of Agriculture and Forestry, No. 28, Xinong Road, Yangling 712100, PR China

a r t i c l e

i n f o

Article history: Received 23 October 2009 Received in revised form 23 June 2010 Accepted 11 July 2010

Keywords: Mutton Multi-elements Geographical origin Authentication

a b s t r a c t Elemental fingerprints were investigated for their potential to classify mutton samples according to their geographical origin. The concentration of 25 element contents in 99 mutton samples from three pastoral regions and two agricultural regions of China were analysed by ICP-MS. Multivariate statistical analysis including principal component analysis (PCA) and linear discriminate analysis (LDA) were used for this purpose. Twenty-one elements (Be, Na, Al, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Ag, Sb, Ba, Tl, Pb, Th and U) in de-fatted mutton showed significant differences (p < 0.05). LDA gave an overall correct classification rate of 93.9% and cross-validation rate of 88.9%. Furthermore, mutton samples from agricultural regions and pastoral regions were differentiated with 100% accuracy. These results demonstrate the usefulness of multi-element fingerprints as indicators for authenticating the geographical origin of mutton in China. Ó 2010 Elsevier Ltd. All rights reserved.

1. Introduction Increasing consumer demand for nutritional balance and food safety has led to mutton becoming more popular as a source of protein that has lower fat content and a reduced risk of contamination. As the leader of mutton production and consumption, two major mutton-producing groups have been formed in China, pastoral regions in northern China and agricultural regions in middle and southwestern China. However, the traceability system of sheep husbandry in China is not well developed. This lack of traceability and control for food safety could restrict mutton export trade and bring some potential safety hazards (Regulation (EC) 178, 2002). Therefore, technologies for checking, identifying and tracing the origin of mutton are required to reassure consumers, protect geographical indication, ensure fair competition, reinforce governmental supervision, and permit the implementation of product recall. In recent years, research efforts have focused on the potential of analytical techniques for the determination of agricultural products according to geographical origin (Kelly, Heaton, & Hoogewerff, 2005; Peres, Barlet, Loiseau, & Montet, 2007; Reid, O’Donnell, & Downey, 2006). Trace element analysis is considered to be an

* Corresponding author. Tel.: +86 10 62815956; fax: +86 10 62895141. E-mail addresses: [email protected] (S. Sun), [email protected] com.cn (B. Guo), [email protected] (Y. Wei), [email protected] (M. Fan). 0308-8146/$ - see front matter Ó 2010 Elsevier Ltd. All rights reserved. doi:10.1016/j.foodchem.2010.07.027

effective tool, because the element compositions in the local environment (soil, drinking water, etc.) can be reflected into agricultural products, and differences in the distribution of these trace elements among different geographical locations can give various element signatures in the organic tissues (Schwägele, 2005). The method of trace element analysis has been applied for geographical origin assignment of some plant-derived products such as wine (Baxter, Crews, Dennis, Goodall, & Anderson, 1997; Coetzee et al., 2005), honey (Arvanitoyannis, Chalhoub, Gotsiou, Simantiris, & Kefalalas, 2005; Hernández, Fraga, Jiménez, Jiménez, & Arias, 2005), tea (Moreda, Fisher, & Hill, 2003), coffee (Anderson & Smith, 2002), orange juice (Simpkins, Louie, Wu, Harrison, & Goldberg, 2000), and other crops (Anderson, Magnuson, Tschirgi, & Smith, 1999; Yasui & Shindoh, 2000) with different degrees of success. However, trace elements in meat products have been primarily studied for quality and safety purposes (Boccia, Lanzi, & Aguzzi, 2005; Sacco, Brescia, Buccolieri, & Jambrenghi, 2005). A few studies were related to geographical origin, mainly focusing on beef and poultry (Franke et al., 2007, 2008; Guo, Wei, Pan, & Li, 2007). The major reason for the limited use for animal products is that the element composition in animal tissues is influenced by both natural deposits and element supplements in animal feed, which can induce uncertainty. It is less complicated in sheep compared to poultry and fattened beef, because sheep are mostly kept outside and mainly fed on local pasture or forage, few supplements are used and little feed is imported (Franke, Gremaud, Hadorn, & Kreuzer, 2005). As a result, trace elements in sheep tissues would be more

1152

S. Sun et al. / Food Chemistry 124 (2011) 1151–1156

related to the environment in which they are reared, and therefore could provide a better means to determine the geographical origin of mutton. A problem worthy of noting is that animal tissues have different capabilities of elements accumulation. Chessa, Calaresu, Ledda, Testa, and Orrù (2000) reported that heavy metal elements Pb, Zn and Cd in muscle tissues of sheep were not different between polluted and unpolluted regions. They explained that these heavy metal elements probably just accumulated in inner organs (e.g. kidney and liver) and not preferentially in muscle tissues. Therefore multi-elements should be analysed to offset these limitations and screen more promising elements for the authentication of the geographical origin. The objectives of this paper were to investigate the multi-element contents of mutton from five major mutton-producing regions in China and to assess their feasibility for classifying mutton according to geographical origin. 2. Materials and methods 2.1. Sampling A sample set of 99 right hindquarters from animals at the age of 7–10 months was collected at the slaughter house from five major mutton-producing regions in China. These regions can be classified as pastoral or agricultural regions according to the combination of breed and the feeding practice. The pastoral regions where people consume large quantities of sheep meat include three regions in northern China, which are Alxa League, in the west of Inner Mongolia Municipality; Xilin Gol League, in the middle of Inner Mongolia Municipality; and Hulunbuir City, in the northeast of Inner Mongolia Municipality. The agricultural regions which are the major places for goat meat production include Chongqing City, located in southwestern China; and Heze City, located in the middle of China. Table 1 contains detailed information of the origin. To ensure samples were representative, three counties or towns were chosen in each region and six or seven authentic local mutton samples were collected from each site. 2.2. Sample preparation Meat samples (250 g each) were kept frozen at 20 °C prior to processing. A 50 g (fresh weight) sub-sample was removed from each sample and cut into small pieces, then freeze-dried for 24 h before being pulverised in a ball mill. The crude fat of muscle powder was extracted with petroleum ether in a soxhlet apparatus, and residue, mostly de-fatted protein was used for further analysis. 2.3. Elemental analysis The samples were analysed after microwave digestion using MARS (CEM Company) microwave digestion system. A 0.2 g of de-fatted mutton sample, 8 mL of 65% HNO3 and 3 mL of 37% HCl were added into a PTFE digestion tube and digested for 40 min by increasing the power to 1600 W and the temperature to 210 °C in a stepwise fashion. The digested solution was diluted to

50 mL with ultra pure water (MX > 18 M) and stored in plastic tube before analysis. Twenty-five elements (Be, Na, Al, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Ag, Sb, Ba, Tl, Pb, Th ,U, Mg, K, Cd and Mo) were measured by Inductively Coupled Plasma Mass Spectrometry (ICP-MS, 7500a, Agilent, America). Optimisation of the instrument was done for higher sensitivity and lower detection limits. The optimised operation conditions for analysis of the diluted samples were as follows: radio frequency power 1200 W, plasma gas flow rate 1.12 L/min, auxiliary gas flow rate 0.5 L/min, nebulisation chamber temperature 2 °C, the oxide indices 0.45%, and dual current indices 1.01%. The standard matter of soil (GBW07401) supplied by the Institute of Geophysical and Geochemical Exploration of China was used for calculating the recovery and accuracy. After above digestion process and ICP-MS analysis, the recovery and the relative standard deviation (RSD) of each element in standard matter of soil (GBW07401) (0.1 g) were higher than 90% and lower than 10% (measured in triplicate), respectively, which indicated the whole analysis method was validated for elemental analysis. Analysis of each sample was done in triplicate and quantified using external standards analysis. All the results were expressed as the average of triplicate measurements. The Environmental Calibration Standard (Part# 5183-4688) supplied by Agilent company was used as a standard solution and the determination coefficient of standard curve was higher than 0.99. The internal standards Li, Ge, Y, In, Tb and Bi were used to ensure the stability of the instrument. The samples were remeasured whenever the RSD of internal standards was higher than 3%.

2.4. Statistical analysis The statistical analysis of the data was performed using the SPSS 16.0 package for windows. ANOVA was carried out for each element. Duncan’s multiple comparison was performed to determine the significant difference between the individual regions when the F value was significant in ANOVA. In order to reduce the dimensionality of the data set and to describe all the variability of the system using a smaller number of variables, principal component analysis (PCA) was used. The first principal component (PC) describes the maximum possible variation and the second PC accounts for the second most and so on. Unsupervised classification was performed with cluster analysis (CA) to measure the similarity between objects. The objects were grouped into clusters in terms of their nearness or similarity. The clustering distance used was Mahalanobis distance and the measurement of the similarity was based on the squared Euclidean distance. Linear discriminant analysis (LDA) using the stepwise method was carried out to evaluate whether mutton from different regions could be mathematically distinguished on the basis of elements which had significant differences among the regions. The robustness of the classification model was evaluated by a cross-validation test, using the ‘leave-one-out’ procedure. In this test, each case was classified by the functions derived from all cases other than that

Table 1 The region information of mutton samples. Region

Denote by production system and breed

Staple feed species

Sampling time

Number of samples

Xilin Gol League Inner Mongolia Hulunbuir City Inner Mongolia Alxa League Inner Mongolia Chongqing City Heze City Shandong Province

Pastoral region Pastoral region Pastoral region Agricultural region Agricultural region

Grasses Grasses Grasses Leguminosae Maize, Leguminosae

2008.09 2008.09 2008.10 2008.12 2008.12

19 20 20 20 20

1153

S. Sun et al. / Food Chemistry 124 (2011) 1151–1156

case, and the success of the classification was recalculated by comparison with the known. The sample separation in the discriminant space was illustrated by plotting the first three discriminant function scores. 3. Results 3.1. Differences in element concentrations of de-fatted mutton among the regions The ANOVA test showed that 21 of 25 elements (Be, Na, Al, Ca, V, Cr, Mn, Fe, Co, Ni, Cu, Zn, As, Se, Ag, Sb, Ba, Tl, Pb, Th, and U) in de-fatted mutton were significantly different among the regions. Their mean values and standard deviations are presented in Table 2. Duncan’s multiple comparison results indicated that each region had a characteristic element contents profile. The element contents in the samples from the agricultural regions were generally higher than those from the pastoral regions. Chongqing samples had the highest contents of Ag, Be, Ni, Th, Tl and U when compared with the other regions. Heze samples were generally characterised by the highest contents of Ca, Ba, Al, Cr, Cu, As, Se and V. In the pastoral regions, Alxa League samples had the highest contents of Na and Fe. Hulunbuir samples could be clearly separated from the others based on higher Sb and Co contents, while Xilin Gol League samples had the lowest contents of Mn, Zn and Pb. From Table 2, large standard deviations of elements reflected the great variability among samples from the same region. This could possibly be related to the samples in the same region originating from subregions with a wide variation in conditions. 3.2. Principal component analysis (PCA) Twenty-one elements in de-fatted mutton originating from the five regions for which ANOVA detected significant differences, were analysed using PCA. The first seven factors explained 76% of

the total variability. The contents of U, Th, Ag, Be and Tl had the highest weight on the first PC (explaining 27.3% of the variability) and Chongqing samples could be separated from other regions. V, Cr, Ca and Ba dominated the second PC (explaining 15.7% of variability) and Heze samples could be distinguished from the remainder. Co and Sb contents showed the highest weight on the third PC (explaining 9.5% of variability) and Hulunbuir samples were differentiated from other regions. It was more difficult to deduce explanations for the groupings of elements in the remaining PC, where the dominated elements were Fe and Mn in PC4, Ni in PC5, Na and Cu in PC6, and Pb in PC7. 3.3. Cluster analysis (CA) In order to better visualise the relative distribution of the defatted mutton samples according to their geographical origin, CA was performed using the first seven principal component normalisation scores. Samples from different regions were separated into seven clusters based on the dendrogram cut at a distance of 11.61. As can be seen in Fig. 1, the first cluster was composed of samples from the pastoral regions, mainly from Xilin Gol League (n = 11) and Hulunbuir City (n = 11), as well as Alxa League (n = 3). The second and third clusters were also composed of samples from the pastoral regions, mainly from Alxa League (n = 17), as well as Xilin Gol League (n = 6) and Hulunbuir City (n = 1). The fourth cluster mainly comprised Hulunbuir samples (n = 8) and only one Xilin Gol League sample, all from the pastoral regions. The fifth and sixth clusters were mostly composed of Chongqing samples (n = 19) from an agricultural region. Finally, the seventh cluster was composed of samples from the agricultural regions, almost entirely from Heze City (n = 20) with just one Chongqing sample. The cluster analysis makes it clear that the samples from agricultural regions and pastoral regions typically formed independent groups. Two agricultural regions (Chongqing and Heze) also clustered separately. Some of the samples from different pastoral regions were

Table 2 Mean elemental concentrations and standard deviations for 25 elements in de-fatted mutton samples. Element

AL

XL

HL

CQ

HZ

P-value

Na Mg K Ca Fe

3.47 ± 0.67a 1.20 ± 0.12a 18.82 ± 1.42a 0.31 ± 0.07b 0.12 ± 0.02a

2.68 ± 1.02b 1.18 ± 0.38a 20.35 ± 0.90a 0.24 ± 0.05b 0.09 ± 0.02c

2.76 ± 0.70b 1.30 ± 0.09a 19.62 ± 1.35a 0.26 ± 0.07b 0.09 ± 0.02c

2.58 ± 0.45b 1.26 ± 0.07a 18.98 ± 0.95a 0.28 ± 0.04b 0.0.11 ± 0.02ab

3.45 ± 0.78a 1.21 ± 0.06a 17.76 ± 1.07a 0.48 ± 0.22a 0.10 ± 0.02bc

0.001 0.265 0.080 0.001 0.001

Zn Ba Al As Co Cr Cu Mn Ni Pb Sb Se V

0.11 ± 0.02a 0.41 ± 0.15c 11.81 ± 5.07b 0.65 ± 0.57b 0b 3.22 ± 3.29bc 5.07 ± 1.11c 1.49 ± 0.77ab 0.21 ± 0.11b 0.12 ± 0.04ab 0c 0.55 ± 0.40abc 3.73 ± 1.22b

0.08 ± 0.01b 0.37 ± 0.14c 9.81 ± 7.76b 0.92 ± 0.82b 0.28 ± 0.26b 4.75 ± 3.29bc 5.18 ± 0.901bc 0.62 ± 0.14c 0.21 ± 0.11b 0.09 ± 0.04b 1.82 ± 2.19b 0.44 ± 0.38bc 2.89 ± 0.512c

0.09 ± 0.01b 0.56 ± 0.27bc 8.55 ± 3.80bc 0.19 ± 0.33b 0.93 ± 0.91a 7.08 ± 3.98b 5.73 ± 0.59ab 1.32 ± 0.57b 0.24 ± 0.15b 0.21 ± 0.22a 3.72 ± 3.60a 0.76 ± 0.55ab 2.43 ± 0.63c

0.12 ± 0.01a 0.72 ± 0.24b 18.26 ± 7.29a 2.32 ± 1.86a 0.08 ± 0.14b 0c 5.96 ± 0.81a 1.79 ± 0.34a 1.50 ± 1.08a 0.15 ± 0.07ab 0.33 ± 0.73bc 0.33 ± 0.41c 2.99 ± 1.21c

0.10 ± 0.01a 1.11 ± 0.51a 19.86 ± 13.6a 2.36 ± 1.09a 0b 38.08 ± 19.35a 6.06 ± 1.32a 1.62 ± 0.74ab 0.58 ± 0.52b 0.19 ± 0.12ab 0c 0.88 ± 0.49a 7.70 ± 1.39a

0.001 0.001 0.001 0.001 0.001 0.001 0.004 0.001 0.001 0.016 0.001 0.001 0.001

Ag Be Th Tl U Mo Cd

3.22 ± 4.03bc 0.75 ± 0.19b 6.54 ± 6.06a 8.38 ± 5.57b 3.32 ± 5.06b 43.10 ± 17.41a 23.01 ± 6.82a

2.20 ± 2.34bc 0.60 ± 0.78b 4.30 ± 3.45b 6.60 ± 5.73bc 3.14 ± 2.73b 45.00 ± 15.93a 16.04 ± 8.61a

1.03 ± 1.28c 1.42 ± 2.45b 4.01 ± 4.15b 5.04 ± 2.35c 2.07 ± 1.59b 46.61 ± 34.80a 20.82 ± 7.10a

8.60 ± 5.42a 6.15 ± 4.01a 10.3 ± 6.35a 13.0 ± 4.86a 7.23 ± 5.71a 40.23 ± 17.03a 24.11 ± 33.42a

3.60 ± 2.77b 1.52 ± 1.36b 5.42 ± 4.37b 8.82 ± 5.40b 4.25 ± 4.50b 38.91 ± 19.92a 15.79 ± 10.32a

0.001 0.001 0.002 0.001 0.003 0.808 0.066

Note: The unit for most abundant element (Na, Ca, Fe, K, and Mg) contents was g/kg; the unit for abundant element (Zn, Al, Ba, As, Co, Cr, Cu, Mn, Ni, Pb, Sb, Se, and V) contents was mg/kg; the unit for less abundant element (Ag, Be, Tl, Th, U, Mo and Cd) contents was lg/kg. Values followed by the same lowercase letters are not significantly different (p < 0.05) based on the ANOVA test. AL, Alxa League; XL, Xilin Gol League; HL, Hulunbuir City; CQ, Chongqing City; HZ, Heze City.

1154

S. Sun et al. / Food Chemistry 124 (2011) 1151–1156

Fig. 1. Dendrogram of cluster analysis. AL, Alxa League; XL, Xilin Gol League; HL, Hulunbuir City; CQ, Chongqing City; HZ, Heze City.

clustered together, notably some of those from Xilin Gol League and Hulunbuir City because of similar element contents. Overall, the cluster results were generally in agreement with the actual origin of samples, which implied that multi-element information could be suitably utilised to classify mutton samples from the different regions or region types. 3.4. Linear discriminant analysis (LDA) For a better understanding of the discriminating efficiency of each element, linear discriminant analysis was carried out on the basis of 21 element compositions of de-fatted mutton samples found to be significant difference among the regions by ANOVA. 12 elements (Be, Cr, Mn, Fe, Cu, Zn, As, Sb, V, Ba, Ni and Na) were selected to establish a classification model using a stepwise discriminant procedure. A cross-validation procedure was used to evaluate this model. A satisfactory classification was obtained with an overall correct classification rate of 93.9% and a cross-validation rate of 88.9% (Table 3). This classification model could clearly discriminate samples from agricultural region and pastoral region. In addition, the model could discriminate between agricultural region samples better than between pastoral region samples. Classification was absolutely accurate for the two agricultural regions with

a cross-validation rate of 100%. Among three pastoral region samples, 100% of the Alxa League samples were also correctly classified, while large classification errors were associated with Hulunbuir and Xilin Gol League samples. However, there was no misclassification between Alxa League and Hulunbuir samples. The separation among regional groups in the discriminant space was checked by plotting the first three discriminant functions (Fig. 2). The first three functions accounted for 97.8% of the total variance. It can be clearly seen that the samples from different regions were plotted in different spaces. The agricultural region samples were entirely separated from the pastoral region samples. Samples from the two agricultural regions Chongqing and Heze were also clearly distinguished from each other. Among the pastoral regions, a few of Alxa League samples intermixed with Xilin Gol League, while overlap mainly existed between samples from Xilin Gol League and Hulunbuir City. These results were consistent with those obtained by PCA and CA, and reconfirmed the feasibility of multi-element analysis for mutton geographical origin traceability. 4. Discussion This study showed that each region had a typical profile of elemental compositions in mutton. The elemental composition of

Table 3 Classification of mutton samples in different regions and percentage of observations correctly classified. Predicted group membershipa

Original

Count

XL

HL

AL

CQ

HZ

Total

XL HL AL CQ HZ

17 4 0 0 0 89.5

1 16 0 0 0 80

1 0 20 0 0 100

0 0 0 20 0 100

0 0 0 0 20 100

19 20 20 20 20 93.9b

XL HL AL CQ HZ

16 4 4 0 0 84.2

2 16 0 0 0 80.0

1 0 16 0 0 80.0

0 0 0 20 0 100

0 0 0 0 20 100

19 20 20 20 20 88.9c

% Cross-validated

Count

%

AL, Alxa League; XL, Xilin Gol League; HL, Hulunbuir City; CQ, Chongqing City; HZ, Heze City. a The number of correctly classified observations are tabulated diagonally. b 93.9% of empirical grouped observations correctly classified. c 88.9% of cross-validated grouped observations correctly classified.

Fig. 2. 3D scatter plot of individual de-fatted mutton sample scores on the first three discriminant functions. AL, Alxa League; XL, Xilin Gol League; HL, Hulunbuir City; CQ, Chongqing City; HZ, Heze City.

S. Sun et al. / Food Chemistry 124 (2011) 1151–1156

meat can be influenced by the element profile of environment (soil, feed, drinking water, air, etc.) through the food chain. Trace element profile of soils are different based on several environmental and geological factors such as soil type, soil parent material, soil pH, and climate conditions (Cresser, Killham, & Edwards, 1993). Compared with agricultural regions, the pastoral regions typically had poor quality soils due to serious desertification and salinisation, and many element contents in the herbage were lower attributed to the low contents in the soil (Gao, Wang, Zhang, & Li, 2007; Qu, 1999). Heze City is located on the alluvial plains of a section of the Yellow River with fertile soil. Certain element contents (Cu, Cr, As, and Mn) in the soil were higher than the national average (Wei, Zheng, Chen, & Wu, 1991). In addition, since this region is a major crop producer, fertilisers were often used during periods of crop growth, enabling certain essential elements such as Ca, Ba, Al and Se to accumulate in the soils and the plants eaten by sheep. In the other agricultural region, Chongqing City, covered with purple stony soils and abundant in rare elements such as Tl, Ge and the radioactive contaminant elements, Th and U were high because of ore smelting (Yang, 2008). Furthermore, the element contents in vegetation were also influenced by plant species, since the element absorption capacity of vegetation is species dependent. Average contents of Ca, Mg, V, Se and Sr of legumes have been reported to be markedly higher than those of grasses (Du & Du, 2003). Based on these differences, the multi-elements were used to classify mutton according to their different regions of origins in China and satisfactory results were obtained. The total correct classification of mutton samples was 93.9%, and the classification was more successful for agricultural samples (100%) than for pastoral ones (90%). In the pastoral regions, some samples from Hulunbuir City and Xilin Gol League were misclassified to a certain degree, probably because these regions are close to each other and characterised by similar element profile within the environment, which led to little difference between element contents. The results of PCA, CA and LDA all apparently indicated that ‘‘fingerprints” of elemental compositions would be effective as unique markers to trace the geographical origin of mutton. In addition, animal characteristics (the breed, age, etc.) may also influence the elemental composition of meat. In this study, though the age and breed of animals at sampling were known by inquiring of the feeders, the lack of enough samples restricted the investigation on the difference of elemental content related to these two factors, and a further study should be considered to answer this question. In a previous study, Franke et al. (2007, 2008) investigated the suitability of element signatures for authenticating poultry meat and dried beef samples of different origins. The results showed that elements were significantly different and could discriminate samples from five different countries. Furthermore, the characteristic elements of the different countries and their discrimination power had some difference according to the type and number of samples. The element contents were also found to be different in beef samples from four regions of China, and Se, Sr, Fe, Ni and Zn were suggested as good tracers for beef origin assessment resulting in 98.4% correct classification for all samples (Guo et al., 2007). These results and our own suggest that the useful element indicators and classification efficacy are different depending on the meat producing species, the number of samples, and the range of investigative regions. The number of samples in the database should be sufficient to establish practical and stable models to authenticate the geographical origin of mutton. Multivariate statistical analysis will be a powerful tool to deal with the greater amount of data and achieve this goal. Moreover, some studies have attempted other promising approaches such as stable isotope analysis (Guo, Wei, Pan, & Li, 2010; Schmidt et al., 2005) and chemical composition analysis (Franke et al., 2006) for meat authentication. Among these methods, trace elements combined with stable isotopes has been

1155

considered as the most successful way (Heaton, Kelly, Hoogewerff, & Woolfe, 2008), but still need further study to confirm this proposal. In addition, to supply more powerful support for better utilisation of this traceability technology, a more detailed study on how environmental signatures are transferred into food systems also should be considered in the future.

5. Conclusion Multi-element analysis, combined with multivariate statistical analysis is confirmed to be an effective method for authentication of mutton originating from different regions of China. Acknowledgements This work was funded by subject of technology of Food Contaminant Traceability, under Key Projects (2006BAK02A16)—‘‘Food Safety Key Technology”, within the National Science and Technology Pillar Program during the Eleventh Five-year Plan Period and National Natural Science Foundation of China (30800862). We would like to thank our co-workers and the staff of various slaughter houses, butchers and farmers for assistance in obtaining authentic mutton samples and associated production information. References Anderson, K. A., Magnuson, B. A., Tschirgi, M. L., & Smith, B. (1999). Determining the geographic origin of potatoes with trace element analysis using statistical and neural network classifiers. Journal of Agricultural and Food Chemistry, 47, 1568–1574. Anderson, K. A., & Smith, B. W. (2002). Chemical profiling to differentiate geographic growing origins of coffee. Journal of Agricultural and Food Chemistry, 50, 2068–2075. Arvanitoyannis, I. S., Chalhoub, C., Gotsiou, P., Simantiris, L., & Kefalalas, P. (2005). Novel quality control methods in conjunction with chemometrics (multivariate analysis) for detecting honey authenticity. Critical Reviews in Food Science and Nutrition, 45, 193–203. Baxter, M. J., Crews, H. M., Dennis, M. J., Goodall, I., & Anderson, D. (1997). The determination of the authenticity of wine from its trace element composition. Food Chemistry, 60, 443–450. Boccia, L. G., Lanzi, S., & Aguzzi, A. (2005). Aspects of meat quality: Trace elements and B vitamins in raw and cooked meats. Journal of Food Composition and Analysis, 18, 39–46. Chessa, G., Calaresu, G., Ledda, G., Testa, M. C., & Orrù, A. (2000). Lead, zinc and cadmium in biological tissues of sheep bred in a polluted region. In B. Markert & K. Friese (Eds.), Trace elements—Their distribution and effect in the environment (pp. 479–483). Amsterdam: Elsevier. Coetzee, P. P., Steffens, F. E., Eiselen, R. J., Augustyn, O. P., Balcaen, L., & Vanhaecke, F. (2005). Multi-element analysis of south African wines by ICP-MS and their classification according to geographical origin. Journal of Agricultural and Food Chemistry, 53, 5060–5066. Cresser, M. S., Killham, K., & Edwards, T. (1993). Soil chemistry and its applications. Cambridge: Cambridge University Press. Du, Z. C., & Du, J. Y. (2003). A comparative study on the mineral element contents of some tame fine grasses. Journal of Sichuan Grassland, 1, 26–28 (in Chinese). Franke, B. M., Gremaud, G., Hadorn, R., & Kreuzer, M. (2005). Geographic origin of meat – Elements of an analytical approach to its authentication. European Food Research and Technology, 221, 493–503. Franke, B. M., Haldimann, M., Gremaud, G., Bosset, J. O., Hadorn, R., & Kreuzer, M. (2008). Element signature analysis: Its validation as a tool for geographic authentication of the origin of dried beef and poultry meat. European Food Research and Technology, 227, 701–708. Franke, B. M., Haldimann, M., Reimann, J., Baumer, B., Gremaud, G., Hadorn, R., et al. (2007). Indications for the applicability of element signature analysis for the determination of the geographic origin of dried beef and poultry meat. European Food Research and Technology, 225, 501–509. Franke, B. M., Ziolko, T., Luginbuhl, W., Gremaud, G., Hadorn, R., Bosset, J. O., et al. (2006). Investigations on the determination of the geographic origin of dried beef using NIR spectroscopy. Mitteilungen aus Lebensmitteluntersuchung und Hygiene, 97, 345–347. Gao, H. X., Wang, X. K., Zhang, Q., & Li, S. B. (2007). Characteristics of soil background value in Hetao regions, Inner Mongolia. Geology and Resources, 16, 209–212 (in Chinese). Guo, B. L., Wei, Y. M., Pan, J. R., & Li, Y. (2007). Determination of beef geographical origin based on multi-element analysis. Scientia Agricultura Sinica, 40, 2842–2847 (in Chinese).

1156

S. Sun et al. / Food Chemistry 124 (2011) 1151–1156

Guo, B. L., Wei, Y. M., Pan, J. R., & Li, Y. (2010). Stable C and N isotope ratio analysis for regional geographical traceability of cattle in China. Food Chemistry, 118, 915–920. Heaton, K., Kelly, S. D., Hoogewerff, J., & Woolfe, M. (2008). Verifying the geographical origin of beef: The application of multi-element isotope and trace element analysis. Food Chemistry, 107, 506–515. Hernández, O. M., Fraga, J. M. G., Jiménez, A. I., Jiménez, F., & Arias, J. J. (2005). Characterization of honey from the Canary Islands: Determination of the mineral content by atomic absorption spectrophotometry. Food Chemistry, 93, 449–458. Kelly, S. D., Heaton, K., & Hoogewerff, J. (2005). Tracing the geographical origin of food: The application of multi-element and multi-isotope analysis. Trends in Food Science and Technology, 16, 555–567. Moreda, P. A., Fisher, A., & Hill, S. J. (2003). The classification of tea according to region of origin using pattern recognition techniques and trace metal data. Journal of Food Composition and Analysis, 16, 195–211. Peres, B., Barlet, N., Loiseau, G., & Montet, N. (2007). Review of the current methods of analytical traceability allowing determination of the origin of foodstuffs. Food Control, 18, 228–235. Qu, Y. L. (1999). A comparative study on the contents of mineral elements in soil and herbage from two desert regions. Acta Prataculturae Sinica, 8, 69–72 (in Chinese). Regulation (EC) 178 (2002). Laying down the general principles and requirements of food law, establishing the European Food Safety Authority and laying down procedures in matters of food safety. Official Journal of the European Communities, L31/1–L31/24.

Reid, L. M., O’Donnell, C. P., & Downey, G. (2006). Recent technological advances for the determination of food authenticity. Trends in Food Science and Technology, 17, 344–353. Sacco, D., Brescia, M. A., Buccolieri, A., & Jambrenghi, A. C. (2005). Geographical origin and breed discrimination of Apulian lamb meat samples by means of analytical and spectroscopic determinations. Meat Science, 71, 542–548. Schmidt, O., Quilter, J. M., Bahar, B., Moloney, A. P., Scrimgeour, C. M., Begley, I. S., et al. (2005). Inferring the origin and dietary history of beef from C, N and S stable isotope ratio analysis. Food Chemistry, 91, 545–549. Schwägele, F. (2005). Traceability from a European perspective. Meat Science, 71, 164–173. Simpkins, W. A., Louie, H., Wu, M., Harrison, M., & Goldberg, D. (2000). Trace elements in Australian orange juice and other products. Food Chemistry, 71, 423–433. Wei, F. S., Zheng, C. J., Chen, J. S., & Wu, Y. Y. (1991). Study on the background contents on 61 elements of soil in China. China Environmental Science, 12, 12–19 (in Chinese). Yang, M. Z. (2008). Division, distribution pattern of the disperse element ore deposits and the ore-searching and comprehensive utilization direction. Contributions to Geology and Mineral Resources Research, 23, 1–10 (in Chinese). Yasui, A., & Shindoh, K. (2000). Analytical chemistry for environmental and human health. Determination of the geographic origin of brown-rice with traceelement composition. Bunseki Kagaku, 49, 405–410.