Meat samples (n = 100) from the neck-region were randomly collected from semi-domesticated reindeer in 10 different grazing districts located in northern and middle Norway in the period from October–December 2008 and September–December 2009. The samples were collected from four different counties distributed as follows: Finnmark County (seven districts), Troms County (one district), Nordland County (one district) and South-Trøndelag County (one district). The selection of the 10 districts was based on obtaining a broad geographical range and the susceptibility of certain districts to pollution from mines, smeltery and military activities (Figure 1). The selection of seven districts from Finnmark County which is the biggest and northernmost Norwegian county was based on the fact that this county has the largest number of semi-domesticated reindeer and 50% of the total number of reindeer grazing districts in Norway. The districts of Eastern Sør-Varanger, Pasvik/Sør-Varanger and Varanger Peninsula are located near the Norwegian-Russian border, where the contamination from mines and smeltery activities is taking place. Spierttagáisá is located in a military activity area, whereas Karasjok West is a neighbouring district. There is also mining work taking place in Ábborašša, which is close to Fávrrosorda a neighbouring district.

Our focus was on young animals (1.5 years old), which represented 77% of the total samples. However, calves (approximately10 months old) and older animals (>2 years old) with the respective proportions of 12% and 11% had to be selected due to the limited availability of 1.5 year olds in some districts (n = 4). There were 52 males and 48 females of the 100 selected reindeer.

All samples were collected directly after the slaughter/dressing process and carcass weighing in acid-rinsed glasses. The glasses were labelled with sample type, carcass number, district name/number and date of sample collection. The samples were kept cool in a cooling box (at approximately 4 °C) and then moved the same day to a −20 °C freezer until they were shipped frozen to the laboratory for analysis. All of the animals from the collected samples were healthy, i.e., had passed the veterinary meat inspection.

The samples were separately decomposed using a microwave oven technique, with concentrated supra-pure HNO₃ (5 mL) and H₂O₂ (3 mL) being added to the decomposed sample (0.6–0.7 g) before undergoing microwave oven treatment. The following temperature regimes were subsequently used in the microwave oven: 20–50 °C (5 min), 50–100 °C (10 min), 100–180 °C (5 min) and 180 °C (15 min). After cooling the heated decomposed sample, the solution was diluted to 50 ml. The sample solution was then analysed using an inductively coupled plasma high resolution mass spectrometer (ICP-HRMS), Bremen, Germany. All standards and calibration solutions contained 1 ppb Rhenium (Re) as the internal standard, together with 1% nitric acid (HNO₃). The calibration curve was verified by standard quality control (QC) sample, National Institute of Standards and Technology (NIST, USA). The resolutions used for the elements were low (at 10) for Cd and Pb, middle (at 20) for Cu, V and Ni, and high (at 30) for As. The lens adjustment was optimized daily for maximum intensity and top separation. The analyses were done by the NILU (Norwegian Institute for Air Research) Laboratory (Kjeller, Norway). The laboratory is accredited for the methods used in the analyses according to NS-EN ISO/IEC 17025, No. TEST008. The limits of detections (LODs) for the studied toxic elements were three times standard deviation (SD) of the laboratory blanks, whereas the limits of quantifications (LOQs) were 10 times the SD of the blanks, decomposed simultaneously with the meat samples.

STATA/SE 11.0 for Windows (STATA Corp. College Station, TX, USA) was used for the statistical analyses. Laboratory results for elements below the limit of detection (LOD) were given a numeric value at half the LOD (LOD/2) according to accepted statistical practice [31]. All toxic element concentrations, except for Cu which was normally distributed, were positively skewed (skewed to the right). Therefore, all concentrations were log transformed to obtain a normal distribution before statistical evaluation. A standardized residuals test was conducted prior to the log transformation in order to check for possible outliers (observations that were more than three standard deviations from the mean). Consequently, all outliers were removed (n = 10, details in results and discussion).

An analysis of variance and covariance (ANOVA), with the specific element as a dependent (outcome) variable and the district, age and sex as independent (explanatory) variables, was used to test for the effect of the independent variables on toxic element concentrations. Bonferroni multiple comparison tests were used to test for significant differences in toxic element concentrations among the 10 grazing districts, using the specific element as an outcome and the districts as a factor variable. A Welch test was used whenever the homogeneity of variance was violated. The level of statistical significance was set at p < 0.05 for all the statistical analyses.

Cadmium was detected in 99% of the meat samples, and had the second largest geographical variation after arsenic. The geographical variations in cadmium concentrations are presented in Table 2 and Table 3.

The presence of older animals in some districts revealed a significant district × age interaction, therefore stratified data were given for Cd concentrations in districts with mixed age groups (Table 3). Calves and young animals demonstrated no significant difference in Cd concentration across geography (GM = 0.9 and 1.5, for calves and young animals respectively), while older animals (>2 years old) showed a higher Cd concentration (GM = 4.3, p < 0.01). The increase in Cd concentration with age was in accordance with previous studies [15,16,32]. A Finnish study on toxic metals in reindeer from four districts in Lapland has shown comparable muscle Cd concentrations ranging from 1 to 6 ng/g ww [15]. Moreover, Cd concentrations from the present study were up to 100 times lower than those found in reindeer muscle (mean = 580 ng/g ww) originating from Karelia in the Russian Federation [33].

Cd concentrations in some districts from the present study were higher than those from reindeer muscles collected from three Greenlandic districts, whereas concentrations from the districts of Essand/Røros, Ábborašša, Kanstadfjord and Fávrrosorda were comparable to that reported from one district (GM = 3 ng/g ww) [17]. The high Cd concentration in Kanstadfjord was due to presence of some older animals (n = 4). Furthermore, high Cd concentration in Fávrrosorda might be due to atmospheric deposition from the neighbouring district Ábborašša with its gold mining activity. The high concentration in Essand was probably a result of acid rain due to atmospheric transport of industrial pollution from Europe which has previously been reported to affect southern Norway to a greater extent than northern part of the country [34,35,36].

Pb was detected in 92% of the meat samples. Concentrations of the Pb among the various grazing districts did not vary to the same degree as for As and Cd (Table 2).

The study on Finnish reindeer by Rintala and colleagues had also reported little differences in concentrations of Pb among different Lapland areas [15]. Even so, the Pb concentrations detected in this study were much lower (10 to 20 times) than those reported in reindeer meat by the same reference above. The Pb concentration from Fávrrosorda in the present study (GM= 7.4, AM= 8.6 ng/g ww) was an exception and could be compared to the concentration from the southern, western and eastern parts of Finnish Lapland, in which the Pb concentration was half of that detected from northern Lapland [15]. The mean Pb concentration of 2.14 µg/g ww previously measured from the Karelian reindeer muscle in Russia was 289 times higher than the greatest level of GM= 7.4 ng/g ww measured in the present study [33].

The Pb concentration was five times higher (p < 0.01) in Fávrrosorda (a neighbouring district) compared to Ábborašša (gold mining facilities). This suggests that the deposition of lead occurs in the neighbourhood of the mining area rather than the mining area itself, which is based on the wind direction and is in accordance with results presented elsewhere [19]. The Pb concentration measured from Fávrrosorda in this study was in good agreement with those from unknown point sources in Greenlandic caribou and reindeer muscles from Isortoq (GM = 7 ng/g ww) in the north and Akia (GM = 6 ng/g ww) in the middle part of the country [17]. However, Pb concentrations in the remaining districts in the present study were two to four times lower than those reported by the Greenlandic study.

As for Cd, high Pb concentration in Fávrrosorda might be due to atmospheric deposition from the neighbouring district Ábborašša with its gold mining activity.

As was detected in all meat samples (100%). The As was the element that showed most of the geographical variations (Table 2) among the studied elements. An east-west downward geographical gradient was observed for As, with the highest concentrations measured in the east (the three districts in the vicinity of the Russian border). However, a north-south trend (highest in the south) for As, Cd and Pb has previously been found in Norwegian surface soils, coniferous forest ecosystems and some herbivorous animals [2,26,37,38,39].

Due to the wind frequency towards the districts of Eastern Sør-Varanger and Pasvik/Sør-Varanger, the high As concentration in these districts could be explained in this study by the As being released from the smelter activity in the Russian town of Nikel, which was further reported to be higher during summer as compared to winter [20]. Furthermore, the wind from the east (E) brings waste from the town of Nikel towards the direction of Svanvik/Passvik, while the wind from the north (N) and north-east (NE) brings waste from Zapoljarny during summer. The dominating wind in winter from the south (S) and south-west (SW) brings waste from Nikel to Jarfjord. However, the elevated As and Ni concentrations below the LOD in these two districts indicates additional As sources such as mining and geogenic sources in this area. The As concentrations in meat from Eastern Sør-Varanger (GM = 106.1 ng/g ww) and Pasvik/Sør-Varanger (GM = 47.9 ng/g ww) from this study were in agreement with those formerly revealed in liver samples from reindeer in the same area when compared with samples from other areas in the County [40]. By comparison, Bernhoft and colleagues reported median As concentration of 0.035 µg/g ww in reindeer liver collected from north western Russia [32].

The district of Ábborašša (gold mining activity) displayed As concentration (GM) three times higher than that found in the neighbouring district of Fávrrosorda, which could be explained by pollution from the mining work in the district, as As has been reported to be associated with gold mineralization [41,42]. In addition, the district of Spierttagáisá (military activity) revealed a lower As concentration (GM = 5.9 ng/g ww) than that (GM = 8.8 ng/g ww) found in the neighbouring district of Karasjok West, although the difference was not statistically significant. The As concentration in samples from Essand/Røros was similar to those detected in Ábborašša (gold mining activity) and Spierttagáisá (military activity). This could be due to the long-range atmospheric pollution from Europe, as a previous study has demonstrated that southern Norwegian areas are more prone to atmospheric pollution from Europe than the northern areas [43]. Contribution from soil due to geogenic sources could also lead to elevated As level in the surrounding environment. In accordance with previous study, the reindeer’s age and sex had no effect on the arsenic concentration [32].

Cu was detected in all meat samples (100%) and had the highest concentration among all the studied elements (GM = 1,439 ng/g ww). Cu concentrations did not vary much among the districts (Table 2). Study on Finnish reindeer had also reported little differences in concentrations of Cu among different Lapland areas [15].

Cu concentrations detected in this study were in agreement with those reported from Finnish Lapland and Russian Karelia, and were half of those reported from Greenlandic reindeer muscle [15,17,33]. The results for the effect of age and sex on Cu concentrations from this study stand in contradiction to those reported by Bernhoft and colleagues, in which hepatic Cu concentration was higher in reindeer calves than in adult females, and higher in adult males than in adult females [32].

Districts that displayed relatively high Cu concentrations could be due to contamination from local point sources (gold mining in Ábborašša) and atmospheric transportation in Pasvik and Eastern Sør-Varanger (Russian towns of Nikel with its nickel smeltery and Zapoljarny with its briquette industry) [16].

Ni was detected in samples from five of the 10 districts (Fávrrosorda, Tromsdalen, Varanger Peninsula, Karasjok West and Essand/Røros) and in 20% of the total meat samples (n = 100). The Ni detection percentage within these five districts varied as follows: 80% (Fávrrosorda), 60% (Tromsdalen), 40% (Varanger Peninsula) and 10% in the districts of Karasjok West and Essand/Røros.

The districts of Varanger Peninsula, Fávrrosorda and Tromsdalen had geometric/arithmetic mean Ni concentrations above the LOD (Table 2), in which the Ni concentrations were comparable.

Previous studies on human Ni exposure along the Norwegian-Russian border have shown that urinary Ni concentrations in this area were no higher than the ones exhibited in other populations [44]. No data were available on Ni and V from reindeer muscle other than that of the Ni concentration from the Karelian Russian reindeer, which was reported to be below the LOD [33]. Nonetheless, the Ni concentrations formerly reported in liver and kidney samples from reindeer originating from the Norwegian-Russian border (Sør-Varanger, north eastern Norway and Rybatsjy Ostrov, north western Russia) had exhibited geographical variations and were much higher in levels than those documented in this study due to tissue differences [16,32].

V was detected in 21% of the total meat samples (n = 100) and in seven of the 10 districts (Kanstadfjord, Tromsdalen, Pasvik, Spierttagáisá, Karasjok West, Eastern Sør-Varanger and Ábborašša). The detection percentages within these seven districts varied as follows: 50% (Spierttagáisá), 40% (Kanstadfjord), 30% (Eastern Sør-Varanger, Pasvik and Ábborašša) and 10% in Tromsdalen and Karasjok West. The district Spierttagáisá was the only one that had geometric/arithmetic mean V concentration (0.2/0.3 ng/g ww) above the LOD.

V has been described as a useful environmental pollution marker for the potential release of toxic elements from fossil fuels and oil refinery processes [45]. Results on V from Canadian (Yukon) caribou kidney reported an average concentration of 0.42 µg/g dry weight, 79.9% moisture [3].

Reindeer meat is consumed as fresh, smoked or dried products in Norway, with the highest consumption among the indigenous Sami people, particularly reindeer herders and their families, compared to ethnic Norwegians [46]. The average consumption is generally low compared to other meat types and constitute approximately 23 g and 70 g/week for low and high consumers in areas with both Sami and ethnic Norwegians [47,48]. The estimated human toxic elements intake from reindeer meat based on the high consumption (70 g meat/week) were 0.01 and 0.01 µg/kg human body weight for Cd and As, respectively. These estimations (monthly for Cd and weekly for As) were much lower than permissible tolerable monthly intake (PTMI) for Cd (25 µg/kg human body weight) and weekly intake (PTWI) of 15 µg/kg human body weight for As [49,50]. The FAO/WHO-JECFA has recently withdrawn the PTWI limit of 25 µg/kg human body weight for Pb due to its association with a decrease of at least three intelligence quotient (IQ) points in children and an increase in systolic blood pressure of approximately 3 mmHg in adults [49]. No new PTWI limit was established for Pb. Nevertheless, weekly human Pb exposure from reindeer meat of 0.002 µg/kg human body weight was 12,500 times lower than the previous PTWI limit. The estimated daily Cu intake from reindeer meat in this study was about 0.0003 mg/kg body weight which was well below the acceptable daily intake (ADI) of 0.5 mg/kg body weight [51]. There are no established PTWI/PTMI limits for Ni and V. Nevertheless, concentrations of Ni and V detected in the present study were considerably lower than tolerable upper intake levels of 1 mg and 1.8 mg per day that have been reported elsewhere for Ni and V, respectively [52].

Estimation of human exposure to toxic elements through reindeer meat was previously done based on dietary data from a questionnaire on the Population-based Health and Living Conditions in areas with Sami and Norwegian populations–The SAMINOR Study and the equation described in one of our previous studies [47,48]. The dietary data from the questionnaire and calculations based on the equation mentioned above have revealed considerably low human exposure to toxic elements through meat and other edible tissues from reindeer.

The presence of individual animals with elevated toxic element concentrations (outliers) was investigated further by relating these concentrations to concepts such as maximum levels (ML), acceptable daily intakes (ADI), professional tolerable intakes (PTI) and healthy animals’ parameters [12,49,50,53]. For instance, the elevated concentrations of outliers in cases of cadmium (13 and 15 ng/g ww) and lead (23 and 28 ng/g ww) have been estimated to constitute 20% and 40% of the maximum levels (ML) set for Cd, and 20% and 30% of the ML set for Pb [53]. Consequently, the elevated concentrations measured in this study should not be an issue of concern to consumers.