Introduction

Barbary sheep (Ammotragus lervia, Pallas 1777) is a species of Caprinae (goat-antelope) native to the mountains of Northern Africa (from Mauritania in the west to the Red Sea in the east) (Cassinelo 1998). Barbary sheep was also introduced for hunting or tourism purposes (in most cases illegally) in several parts of Europe (including Croatia) and North and South America (Cassinelo 1998; Cassinello et al. 2008; Mori et al. 2017). It does not adapt to all environments, preferring arid or semi-arid areas (Cassinello 2012). Human consumption and trade of its meat is confirmed on the African continent (Cassinello et al. 2008), though this has not officially confirmed elsewhere. Both the goat and sheep characteristics of this species (Cassinelo 1998) suggest that Barbary sheep meat might be consumed in local hunting communities worldwide. However, no studies are available on the standard nutritional or toxicological properties of this rare species’ meat.

Diet largely affects the intake and exposure to environmental pollutants (French et al. 2017; Goretti et al. 2018). Also, diet of this species and thus direct sources of pollutants, are scarcely investigated in Europe (reviewed in Mori et al. 2017) without any data for the small and isolated Croatian population. Recent studies have shown that Barbary sheep is primarily a grazer, selecting forbs and grasses when available (Cassinello 2012; Miranda et al. 2012; Mimoun and Nouira 2015; Cassinello 2018). However, feeding behaviour of this species is highly flexible and can, in specific periods/areas, feed with numerous species of trees, shrubs and climbing plant species (Cassinelo 1998; Mimoun and Nouira 2015).

With regard to inorganic pollutants, both stable and radioactive elements cycle through air, water and soil to enter plant, animal and human organisms. Some organisms bioaccumulate or even biomagnify metals such as 137Cs, Cd, Pb or Hg in their tissues, leading to the risk of exceeding threshold levels for adverse health effects in animals or their consumers (Ramanzin et al. 2010; Berzas Nevado et al. 2012; Patiño Ropero et al. 2016; Durkalec et al. 2018). Numerous studies have been performed using large wild mammals to characterise the element status in the environment inhabited by animals and humans and to estimate the risk for human health due to the similar trophic level (Jarzyńska and Falandysz 2011; Berzas Nevado et al. 2012; Bąkowska et al. 2016; Ertl et al. 2016; Larter et al. 2016; Lehel et al. 2017; Patiño Ropero et al. 2016; Skibniewski et al. 2017; Durkalec et al. 2018; Vukšić et al. 2018; Zacs et al. 2018). Elements in large herbivores have been studied more often than in sympatric carnivores, likely due to the wider distribution, larger populations, hunting status (easier sample collection) and direct link to humans in the food chain via consumption of their meats (Jarzyńska and Falandysz 2011; Dannenberger et al. 2013; Chiari et al. 2015; Ertl et al. 2016; Ferri et al. 2017; Larter et al. 2016; Lehel et al. 2017). Lately, game meat has found its market, not only in traditional communities, but also in urban cultures with growing awareness and demand for meat from non-intensively farmed animals (organic food), free of additives and pharmaceuticals (Kudrnáčová et al. 2018). However, free-ranging wild animals tend to have higher levels of toxic metals than farmed domestic animals (Kramárová et al. 2005; Forte and Bocca 2007; Durkalec et al. 2018), and their meats can occasionally exceed the maximum limits of Cd and Pb (Lazarus et al. 2014) set by the European Commission Decision No. 1881/2006 for contaminants in foodstuff (EC 2006). Unlike the marketed meat of wild animals that are subjected to tests for residual contaminants at the national level, non-marketed meat of common and rare species consumed directly by hunters is not controlled (Ramanzin et al. 2010; Lazarus et al. 2014). Therefore, the occurrence data obtained in this study would be the only available source of information for consumers about contaminants in an exotic species such as Barbary sheep. The most common contaminants of terrestrial wild animal meats are Cd and Pb, as their persistent environmental concentrations tend to accumulate with age in soft tissues (primarily kidney and liver as detoxifying tissues) of animals. Lead is often found even in muscle tissue in high amounts due to secondary contamination from lead ammunition (Tsuji et al. 2009). Other elements (e.g. As or Hg) are of minor toxicological relevance for the terrestrial food chain and large wild mammals, although in addition to the direct adverse effects in organ systems, many toxic elements demonstrate indirect damage via impairment of homeostasis of essential macroelements and trace elements (Jarzyńska and Falandysz 2011; Durkalec et al. 2018). Insufficient or excessive levels of Se (Flueck et al. 2012), Cu, Zn, Mo, Co (Frank 2004) and Ca (Bjorå et al. 2001) have been shown to be critical for the health or even life of wild herbivores, which, in turn, can negatively affect population size (Skibniewski et al. 2016).

The aim of this study was to assess (1) the dietary composition of Barbary sheep on Mt. Mosor, Croatia, to scan for possible pollutant hyperaccumulators; (2) the tissue levels of elements and radionuclides to estimate possible adverse effects that those could have on individual animals; (3) the potential of this species to reflect environmental contamination in this area by calculating the Toxic Contamination Index; and (4) the risk and benefit for consumers of Barbary sheep edible tissues.

Material and methods

Study area and species

Barbary sheep (21 individual in total; 15 males and 6 females) samples were collected on Mt. Mosor in the southern Dinarides (surface area 11.3 km2, Lat 43° 31′ 54.2573″, Lon 16° 38′ 29.1241″) with the highest peak Veliki Kabal (1339 m) (Fig. 1). Mesozoic limestone sediments dominate through the area (Redžić 2011). In the belt higher than 1200 m, the Mediterranean climate, with a mean annual temperature of 10.8 °C and mean annual rainfall of 1665 mm, shifts to a wet boreal climate (Zaninović et al. 2008). Topographically, the study area is highly heterogeneous, interrupted by ditches, bays and rocks. Scrublands and woodlands of sub-Mediterranean and Euro-Mediterranean vegetation represent the typical habitat. To date, 540 vascular plant taxa have been described on Mt. Mosor (Vladović and Ilijanić 1992). Here, Barbary sheep coexists with other large mammal species: European mouflon (Ovis aries musimon, also introduced), wild boar (Sus scrofa) and grey wolf (Canis lupus) (Anonymous 2012). Soil studies obtained in the Dinaric region, including Mt. Mosor, revealed higher levels of toxic Cd, Hg and Pb and lower K than elsewhere in Croatia (Halamić and Miko 2009). The only anthropogenic pollution source of this otherwise pristine mountain is a cement factory situated several kilometres southwest, in the suburbs of the city of Split (population 180,000; Census 2012). Along with long-range transported airborne particles, the factory is the most significant source of airborne As, Cd, Cr, Cu, Hg, Ni, Pb and Zn (Anonymous 2012), although minimal impacts to the mountain ecosystem are suggested due to prevailing northeaster wind (Bura) (Zaninović et al. 2008). Concerning the status of airborne radionuclides in the soil of this mountainous area, the higher 137Cs levels compared to the rest of the country were primarily attributed to a higher degree of precipitation, correlated to higher altitudes (Šoštarić 2017).

Fig. 1
figure 1

Study area of Mt. Mosor in the southern Dinarides of Croatia (11.3 km2)

Full size image

Based on the 2008 IUCN Red List, Barbary sheep is categorised as a vulnerable (VU) species in its natural distribution range (Cassinello et al. 2008). Currently in Europe, free-ranging populations of Barbary sheep are present only in Spain, Italy, Czech Republic and Croatia (Cassinelo 1998; Bartoš et al. 2010; Gančević et al. 2016). Barbary sheep in these countries is a non-native species. All these populations were introduced by hunters, confirming the hypothesis of Carpio et al. (2017) that hunting is one of the major pathways for the introduction of exotic species into Europe. In 2002, five Barbary sheep (three females and two males) of unknown origin were illegally released in the southern Dinaric region (Mt. Mosor) of Croatia, and the current population size is estimated at around 140 animals (unpublished data, P. Gančević). This illegal introduction of Barbary sheep has been the source of much controversy between hunters and environmental community. Hunters perceive the introduced animals as an attractive game species and potential source of game meat and income via hunting tourism, while environmentalists stress the possible negative effects of introduced exotic species on native flora and fauna and call for the enforcement of existing laws that prescribe the complete removal of alien animals.

Diet analysis

Four rumen contents of Barbary sheep were examined for their botanical composition according Klansek et al. (1995) during all seasons. Rumen contents were mixed, and at least 500 ml was collected by hunters immediately after culling and stored at − 18 °C until analysis. The rumen contents were thoroughly washed through three sieves with mesh sizes of 6.3 mm, 3.15 mm and 1.0 mm under running water. This enabled separation into differently sized plant particles and also of the unidentifiable microscopic fraction. Subsequently, for the analysis of the qualitative and quantitative composition of the rumen contents, the plant fragments remaining in the sieves with the largest and medium mesh sizes were separated according to their morphological characteristics (e.g. formation of the leaf margin, leaf top and bottom, veining, presence of hair and its morphology) into the following plant groups: (i) herbs and perennials, (ii) grasses, (iii) deciduous trees, (iv) conifers, (v) shrubs, (vi) dwarf shrubs, (vii) feed and (viii) cryptogams. The particles remaining in the sieve with the smallest mesh size or in the washing water were examined microscopically for determination of the type of different plants or concentrates (pollen, spores, starch, etc.). In order to determine the proportions of individual plant groups in the total dry weight of each rumen content sample, the test material was dried in evaporating dishes at a temperature of 65 °C in a desiccator until weight constancy was achieved. The dry matter was determined gravimetrically, and the proportion of the dry matter of the entire rumen content sample was calculated as the percentage of weight.

Radionuclide activity measurement

A total of 10 Barbary sheep muscle tissue samples (minimum of 500 g) were collected during the regular hunting season (hunting season open year-round pursuant to Croatian Hunting Act). Samples were cleaned of fat and connective tissues, placed in counting vessels (125 cm3) of known geometry and kept frozen at a temperature of − 18 °C until gamma-spectrometric analysis was conducted. Activities of 137Cs and 40K were measured by the gamma-spectrometric method using an HPGe semiconductor detector system coupled to a Canberra 8196 channel analyser. The counting time was 80,000 s. The obtained spectra were analysed by Genie 2000 Canberra software. 137Cs activities were calculated from the photo peak at 661.6 keV while activities of 40K were calculated from the photo peak at 1460.7 keV. The detector system was calibrated using mixed gamma standards supplied by Eckert & Ziegler (Analytics USA). Efficiency of the system was regularly checked during inter-comparison calculations. Precision and accuracy of the system were additionally checked by the simultaneous measurement of IAEA Reference Materials since the laboratory is accredited by the Croatian Accreditation Agency for gamma-spectrometric measurements (HRN EN ISO/IEC 17025:2007). It is noteworthy that efficiency was calculated as a function of energy and geometry at the base of the experimental data. The measured uncertainty, multiplied by the coverage factor k = 2, was calculated as the sum of net peak area uncertainty, efficiency uncertainty and background fluctuation uncertainty.

The effective equivalent dose that an individual receives due to radionuclide ingestion from the consumption of contaminated meat was calculated using following relationship:

$$ {H}_{\mathrm{E}}={A}_{\mathrm{k}}\times m\times {D}_{\mathrm{k}\mathrm{f}\left(\mathrm{k}\right)} $$

where HE is the effective equivalent dose (Sv), Ak is the average concentration of k radionuclide (Bq kg−1), m is the quantity of consumed meat (muscle tissue) (kg) and Dkf(k) is the dose conversion factor for k radionuclide. A dose conversion factor of 1.3 Sv Bq−1 × 10−8 Sv Bq−1 was used for 137Cs (IAEA 1996).

Element analyses

The outer layer of the semi-thawed muscle and liver of 15 individuals was removed with a ceramic knife to sample adhering tissue. The kidney cortex was sampled after the removal of the thick renal capsule protecting the tissue from secondary contamination. The percentage of moisture was calculated from sample weights before and after freeze-drying. Freeze-dried samples were digested in an UltraCLAVE IV (Milestone, Italy) microwave digestion system before the elements (As, Ca, Cd, Co, Cu, Fe, Hg, K, Mg, Mn, Mo, Pb, Se, Tl and Zn) were quantified by means of inductively coupled plasma mass spectrometry (ICP-MS; Agilent 7500cx, Germany) following the procedure detailed elsewhere (Vihnanek Lazarus et al. 2013). Ultrapure water (18 MΩ cm; GenPure system, TKA, Germany) and purified (duoPUR, Milestone, Italy) nitric acid (p.a. 65%; Merck, Germany) were used for sample preparation and dilution. To monitor the analytical quality of digestion and measurement, duplicate samples of certified reference materials (CRMs) BCR-185R bovine liver, BCR-186 pig kidney (Institute for Reference Materials and Measurements, Geel, Belgium) and DORM-2 dogfish muscle (National Research Council Canada, Ottawa, Ontario) and standard reference material (SRM) 1577a bovine liver (National Institute of Standards and Technology, USA) were included in each digestion series. The obtained and certified values of CRMs and SRM expressed on a dry mass basis are summarised in Table S1 as supplementary material together with the method detection limits for each element. All Barbary sheep tissue data were expressed on a wet mass (w.m.) basis. For the calculation of wet mass concentrations, we used the mean moisture percent for muscle, liver and kidney cortex (75%, 69% and 75%, respectively), corresponding to conversion factors (f) of 4.0, 3.2 and 4.0 (wet mass concentration = dry mass concentration/f), respectively.

Toxic Contamination Index

The Toxic Contamination Index (TCI) was calculated to estimate the contamination level of the mountainous area where Barbary sheep reside, as the TCI was reported to give more a robust and suitable overview of the biological response to environmental metal contamination than tissue metal levels or liver-to-muscle ratios alone (Goretti et al. 2018). Calculation of the TCI, as proposed by Goretti et al. (2018), is based on metal(loid) level (e.g. Pb) in liver (PbL) or kidney (PbK) as the tissues where the majority of toxic metals are detoxified and accumulated, metal level in muscle (PbM) as inert tissue concerning detoxification and human tolerable weekly intake/benchmark dose values proposed by the European Food Safety Authority (EFSA) for the intake of As (EFSA 2014), Cd (EFSA 2011), Hg (EFSA 2012) and Pb (EFSA 2010) from food. Human tolerable weekly intake (TWI)/benchmark dose lower confidence limit at 1% extra risk (BMDL01) was used, as similar safety margins for animals do not exist in the literature. The following equation is used to calculate the TCI (e.g. for Pb):

$$ {\mathrm{TCI}}_{\mathrm{Pb}}=\log \left({\left({\mathrm{Pb}}_{\mathrm{L}}\right)}^2/\left({\mathrm{Pb}}_{\mathrm{M}}\times {\mathrm{BMDL}}_{01\mathrm{Pb}}\right)\right). $$

Exposure assessment

To perform the nutritional and risk analysis of Barbary sheep meat consumption with regard to the intake of essential and non-essential elements, the most recent EFSA data for dietary reference values (DRVs) for essential elements (EFSA 2017) and TWI and benchmark doses (BMDL) for non-essential elements were applied (EFSA 2010, 2011, 2012, 2014). The same estimations were also performed for the consumption of Barbary sheep liver. Calculations were based on the mean and 95th percentile (worst-case scenario) element occurrence in the muscle and liver of Barbary sheep and the assumption that consumers eat only this game species. As in an earlier study (Lazarus et al. 2014), kidney consumption was assumed to be negligible. Consumption frequency of game mammals in the Croatian adult consumer was taken from the EFSA Comprehensive European Food Consumption Database, which published the results of a Croatian food consumption survey conducted in 2011–2012 on 2002 individuals. The estimated daily intake (EDI) of each element was calculated on the basis of both mean (43.4 g) and 95th percentile (80.2 g, for calculation of worst-case scenario) daily portions reported for the Croatian general population, game consumers only (average and P95 consumers).

Statistical procedures

Inter-tissue differences in element levels and TCI values were compared by the Wilcoxon matched-pairs test, as the condition of normality and homogeneity of variance were not met in all data. Statistical procedures were performed using Statistica for Windows 13.0 (StatSoft, Inc., Tulsa, USA).

Results and discussion

This study is one of the first studies to list the plant species consumed by Barbary sheep in Europe (see Miranda et al. 2012 for Spain) and the first one that quantified elements in Barbary sheep tissues transferred from plants and further to humans as game consumers. Element data from mountainous populations of wild animals are very rare due to the somewhat difficult sample collection but have the unique potential to reflect airborne pollutants that enter the soil-plant-animal-human food chain. Namely, certain anthropogenic toxic metals and gamma-emitting radionuclides of high relevance for humans, such as Pb and 137Cs, are found in higher concentrations in mountains due to the higher precipitation and high altitudes that stimulate dry and wet deposition of metal-bound particles (Halamić and Miko 2009; Šoštarić 2017). Contamination of this mountainous habitat and Barbary sheep was assessed using the Toxic Contamination Index.

Animal diet composition

The results of the botanical composition in Barbary sheep rumen contents revealed 19 plant species and 19 genera belonging to 14 families (Table S2). During the winter season, the proportion of dry mass was highest (57.3 g) and the diet was also most variable, consisting of 11 different plant species. The diet analysis showed that this species feeds mainly on grasses, as the quantitative proportion of different grasses amounted to more than 80% in spring, autumn and winter (Fig. 2). During summer, grasses are poorly present in the diet, as the high daily temperatures burn all the grasses in the area; during summer, the main diet of Barbary sheep is deciduous trees, i.e. Ostrya carpinifolia (71%) and Carpinus orientalis (22%).

Fig. 2
figure 2

Percentage of ingested plant groups according to seasons for Barbary sheep from Croatia

Full size image

In recent studies, Barbary sheep has been classified as grazer, primarily selecting forbs and grasses when available (Miranda et al. 2012; Mimoun and Nouira 2015; Cassinello 2018), so it might be suited to an empty niche in the Mediterranean environment (Cassinello 2012). Although we cannot infer the species feeding selection (we have no data on food availability), our data seems to align with this classification. It is, however, noteworthy that the number and spectrum of species and taxa detected in our food samples were very wide, indicating that the Barbary sheep may be extremely plastic in the utilisation of the available food resources, as previously reported in other studies (e.g. Gray 1980; Krysl et al. 1980; Rodríguez Piñero and Rodríguez Luengo 1992). None of the plant species identified as part of the Barbary sheep diet (Table S1) is known as a stabile element or radionuclide hyperaccumulator (Thijs et al. 2017).

Presence of essential elements

The levels of each essential element in muscle, liver and kidney tissues of Barbary sheep are summarised in Table 1. The highest level of macroelements Mg and K was found in muscle tissue, while Ca was the most abundant in the kidney, a critical organ for the regulation of Ca serum levels. Compared to levels from wild herbivorous ungulates from the recent literature, muscle Mg and K in Barbary sheep showed little variation, unlike Ca, which was up to five times lower than that reported for chamois, red deer and roe deer from Austria, Poland and Kazakhstan (Długaszek and Kopczyński 2013; Assenova et al. 2016; Ertl et al. 2016).

Table 1 Differences in radionuclide and stabile element levels in the muscle, liver and kidney of Barbary sheep (Ammotragus lervia) from Croatia
Full size table

Trace elements Cu, Mn, Mo and Zn were most abundant in the liver, Co and Fe in the liver and kidney and Se in the kidney of studied animals (Table 1). Barbary sheep seemed to have the lowest reported Co muscle levels compared to chamois, red deer and roe deer (Jarzyńska and Falandysz 2011; Ertl et al. 2016), although the hepatic and renal ranges fit the values for these species. The renal Cu of Barbary sheep was in accordance with the data reported for Croatian domestic sheep (Bilandžić et al. 2010). In contrast to similar muscle and renal levels, some variation in the hepatic Cu of Barbary sheep was seen in comparison with red deer from Croatia (Lazarus et al. 2008), Poland (Jarzyńska and Falandysz 2011), Italy (Chiari et al. 2015) and Scotland (French et al. 2017) and with roe deer (Wieczorek-Dąbrowska et al. 2013) and European bison (Durkalec et al. 2018). Two to 10 times higher Cu in the liver may be explained by the choice of plant species ingested by the Barbary sheep or due to the vicinity of the cement factory as a source of Cu emissions. Hepatic Cu would reflect the intake and status of the organism, as the liver is the primary organ for Cu metabolism (Davis and Mertz 1986). Fe in all tissues of Barbary sheep was in the same order of magnitude as measured in Croatian fallow deer (Vukšić et al. 2018), red deer (Lazarus et al. 2008), Polish bison (Kośla et al. 2013), chamois, red deer and roe deer (Dannenberger et al. 2013; Długaszek and Kopczyński 2013; Assenova et al. 2016; Ertl et al. 2016). Mn tissue levels were within the range of reported values for chamois (Ertl et al. 2016), red deer (Jarzyńska and Falandysz 2011; Ertl et al. 2016; French et al. 2017), roe deer (Długaszek and Kopczyński 2013; Ertl et al. 2016) and European bison (Durkalec et al. 2018), although the muscle of red deer from Poland (Jarzyńska and Falandysz 2011) and Caspian red deer (Assenova et al. 2016) was reported to contain 10 and 80 times higher levels, respectively, than the Barbary sheep. There is good accordance with liver and kidney levels of Mo in Barbary sheep from Croatia and in red deer from Poland, but as in the case with Mn and Co, muscle levels were the lowest reported. Selenium content in Barbary sheep tissues was in line or even higher than that in Se-supplemented Croatian fallow deer (Vukšić et al. 2018), but lower than that measured in red deer (Lazarus et al. 2008), both species hunted in eastern Croatia, distant from Mt. Mosor. We found similar muscle levels as in other European countries, though liver Se in Spanish red deer (Berzas Nevado et al. 2012; Patiño Ropero et al. 2016), Polish red deer (Jarzyńska and Falandysz 2011) and bison (Durkalec et al. 2018) was roughly half than that found here. Zinc levels showed a very narrow range of values among species and countries.

To the extent of our knowledge, there are no literature data for Barbary sheep trace elements, so the measured levels were compared with available data for wild herbivorous ungulates. As such, this variation in element levels may originate from inter-species or geographical differences. Mt. Mosor is a relatively pristine area with a cement factory in the foothills, with the prevailing wind carrying the only possible source of metal pollution in the area away from the mountain. Little influence from geographical variation is expected as the essential elements in mammals are regulated to maintain homeostasis in areas and conditions without notable contamination with elements (e.g. Cd or Hg). Namely, biological interactions between non-essential and essential elements are known to affect the distribution and status of essential elements in wild ungulates: Zn/Cu/Se-Cd (Lazarus et al. 2008; Reglero et al. 2009) and Hg-Se (Berzas Nevado et al. 2012).

The normal range of macroelements and essential trace elements has not been defined for wild ungulates, so we compared the measured concentrations in Barbary sheep with tissue values defined as adequate or deficient in domestic animals (goat/sheep). Hepatic and renal levels were in the range of adequate values reported for sheep and goat (Puls 1994). Nevertheless, mean Se in the liver (168 μg kg−1 wet mass) appeared to be marginal according to the reference range for sheep (150–250 μg kg−1 wet mass; Puls 1994), but even the animal with the lowest Se (108 μg kg−1 wet mass) had hepatic levels above those reported for deficient animals (10–100 μg kg−1 wet mass; Puls 1994). Farmed red deer affected by white muscle disease had < 35 μg kg−1 wet mass of Se in the liver (Wilson and Grace 2001). As such, the occurrence of white muscle disease or other adverse effect due to Se deficiency is unlikely in the examined Barbary sheep.

Presence of non-essential (toxic) elements

The content of toxic metals measured in the tissues of Barbary sheep is presented in Table 1, suggesting the kidney (and liver in case of As, Hg and Pb) is the target organ for the distribution of As, Cd, Hg, Pb and Tl, as noted before in other wild ungulate species (Jarzyńska and Falandysz 2011; Hermoso de Mendoza García et al. 2011; Berzas Nevado et al. 2012; Wieczorek-Dąbrowska et al. 2013; Lazarus et al. 2014; Chiari et al. 2015; Bąkowska et al. 2016; Patiño Ropero et al. 2016; Lehel et al. 2017; Skibniewski et al. 2017; Vukšić et al. 2018).

Barbary sheep had the lowest levels of metalloid As in comparison to Croatian fallow deer (Vukšić et al. 2018), chamois, red deer and roe deer hunted in Austria (Ertl et al. 2016). The literature data suggest that Cd is the most widely quantified toxic metal in wild herbivorous ungulates. Its presence in even unpolluted areas, accumulative nature and trophic transfer within the soil-plant-herbivore-human axis (Nordberg et al. 2015) prompted studies aiming at assessing food risk safety (Lazarus et al. 2008; Ramanzin et al. 2010; Jarzyńska and Falandysz 2011; Lazarus et al. 2014; Chiari et al. 2015; Ferri et al. 2017). The muscle tissues tested here contained the lowest levels of Cd, as expected, as Cd distributes mainly to the liver and kidney. Among the reported ungulates from Croatia, the muscle Cd of Barbary sheep was in the lower range and in strong agreement with data from Austria (Ertl et al. 2016) and Spain (Hermoso de Mendoza García et al. 2011) but was lower than that in red deer (Jarzyńska and Falandysz 2011; Chiari et al. 2015) and roe deer (Długaszek and Kopczyński 2013) from Poland and Italy. Bison (Durkalec et al. 2018), moose (Larter et al. 2016; Skibniewski et al. 2017) and mountain caribou (Larter et al. 2016) were species whose offal showed enhanced Cd accumulation compared to Barbary sheep, with 10–50 times higher renal Cd levels. These inter-species differences may be attributed to the affinity of bison and moose to browse on willow (Salix sp.) (Larter et al. 2016; Durkalec et al. 2018) and to the high intake of lichen in the caribou diet (Larter et al. 2016). A diet dominated by lichens known to readily absorb atmospheric contaminants, Cd among others, and willow as a known Cd hyperaccumulator plant (Tsuji et al. 2009), differs greatly in comparison with plants used by the Barbary sheep in Croatia (Table S2). Low renal Hg and hepatic and renal Pb levels found in Barbary sheep were closer to the values reported for Croatian domestic sheep (Bilandžić et al. 2010) than those for wild herbivorous ungulates from Croatia and other countries. Muscle Hg and Pb levels were up to 10 times lower than those in other species (Lazarus et al. 2008; Hermoso de Mendoza García et al. 2011; Jarzyńska and Falandysz 2011; Długaszek and Kopczyński 2013; Lazarus et al. 2014; Ertl et al. 2016; Lehel et al. 2017; Vukšić et al. 2018), while hepatic Hg content was comparable (Lazarus et al. 2008; Berzas Nevado et al. 2012; Lazarus et al. 2014; Patiño Ropero et al. 2016; Durkalec et al. 2018; Vukšić et al. 2018). Wild ungulate tissue results should be commented with caution regarding the source of Pb, as Pb incorporated into organ cells after environmental intake can be masked with Pb-based ammunition particles (Tsuji et al. 2009). Although toxic, Tl is, similar to hepatic levels of bison (Durkalec et al. 2018), present in low levels in Barbary sheep tissues, but its level varies compared to the data published for red deer from Poland (Jarzyńska and Falandysz 2011).

Liver and kidney As, Cd, Hg, Pb and Tl of Barbary sheep were in the range of normal values defined for sheep (Puls 1994) and below the terrestrial mammalian effects threshold level for Pb (liver > 30,000 μg kg−1 dry mass, kidney > 90,000 μg kg−1 dry mass; Ma 1996) and Hg (> 30,000 μg kg−1 wet mass; Thompson 1996). However, renal Cd levels in Barbary sheep exceeded the normal range values but were below the levels defined as high (4000–12,000 μg kg−1 wet mass) for domestic sheep (Puls 1994) and much lower than the nephropathy threshold of 100,000–200,000 μg kg−1 wet mass in mammals (Scheuhammer 1991). In general, Cd is more abundant in free-ranging wildlife offal than in farmed domestic animals, and this difference was confirmed even within the same species (e.g. free-ranging European bison [0.370 mg kg−1 wet mass] vs. captive [0.187 mg kg−1 wet mass]; Durkalec et al. 2018).

The Toxic Contamination Index applied previously on results for mustelids (Goretti et al. 2018) was calculated for Barbary sheep to test it as an indicator of environmental contamination (Fig. 3). TCI values from 0.7 for AsL to 5.7 for CdK in Barbary sheep were higher than values reported for stone and pine marten from Italy and Croatia (CdL < 3, PbL < 2.5; Goretti et al. 2018) and for species occupying a higher trophic niche and inhabiting more anthropised environments than Barbary sheep. Such TCI differences are mostly influenced by 30 (for Pb) to 60 (for Cd) times lower muscle levels and similar liver levels measured in Barbary sheep compared to mustelid species (recalculated from dry mass based on the conversion factor 3.18) due to analytical method limitations (detection limit 43 μg Cd kg−1 vs. 0.21 μg Cd kg−1, and 42 μg Pb kg−1 vs. 1.71 μg Pb kg−1 in Goretti et al. (2018) vs. this study, respectively). Thus, TCI (at least for Cd and Pb) is strongly dependent on the analytical properties of the method applied, due to very low levels of toxic metals in meat. Thereby, the TCI comparison between studies is applicable only when similar instrumentation conditions are met. In light of these facts, we suggest that the interpretation of lower than unit and negative TCI values as the one pointing at poorly contaminated site condition and those much higher than unit pointing at polluted sites (Goretti et al. 2018) should be taken with caution. In our case, TCI values for Pb (3.6–3.8) and Cd (4.2–5.7) in Barbary sheep tissues indicate certain pollution in the habitat, but comparison with the reported literature data from ungulates living in uncontaminated areas and the available threshold data for toxicity in domestic animals indicates that Barbary sheep is a species with a very low toxic metal burden and its habitat is a pristine area.

Fig. 3
figure 3

Mean Toxic Contamination Index (TCI) for As, Cd, Pb and Hg calculated for liver and kidney of Barbary sheep. Differences tested with the Wilcoxon matched-pairs test, *p < 0.05, **p < 0.001

Full size image

Risk/benefit assessment for consumers

The presented efforts to quantify toxic metal(loid)s in Barbary sheep meat are in line with the EFSA call for the collection of inorganic contaminant occurrence data in food (with a focus on Cd, Pb, As and Hg) to enable exposure assessment for consumers.

Cadmium in 40% of Barbary sheep kidneys (6/15) exceeded the legislative maximum level (1000 μg kg−1 wet mass) set for meat and offal of farmed animals (EC 2006), while Cd and Pb levels in other tissues complied with the legislation. As mentioned by Ferri et al. (2017), muscle and liver, but not kidney, are preferred food items among consumers of wild ungulates. As such, the most frequently consumed tissues are all within the safety margins regarding Cd and Pb.

A risk/benefit assessment was performed for the intake of essential and non-essential elements for consumers of Barbary sheep meat (Tables 2 and 3) and liver (Table S3 in the Supplementary material). In the earlier exposure assessment for Croatian consumers of game (Lazarus et al. 2014), official data of consumption frequency were not available. As such, three proposed consumption scenarios (Lazarus et al. 2014) considered the much lower daily intake of game meat (21.4 g/5 g/1.64 g) than the Croatian food consumption survey revealed (80.2 g/43.4 g). Thus, the amount of game meat eaten per day is the main methodological difference between this and the earlier study. Our estimation (Table 2) showed that the nutritional contribution of essential elements from Barbary sheep muscle consumption is low (0.1% of DRV for Mn and Mo) to significant (39% of DRV for Zn) in average and P95 game consumers in Croatian population. In addition to Zn, muscle is also a good source of Fe, Cu, Se and K, with a daily portion adding 5–27% to the recommended DRV for respective elements (Table 2). When compared to domestic animals, we can conclude that Barbary sheep meat is a similar source of Fe, Mg and Zn as Croatian bovine, sheep and horse meat, but a better source of Zn than pork meat (Bilandžić et al. 2013; Bilandžić et al. 2014). Other authors found red deer meat and offal as a good source of Cu, Zn and Se in human nutrition, and kidney as also a good source of Co, Cr and Mn (Jarzyńska and Falandysz 2011).

Table 2 Nutritional contribution of Barbary sheep muscle consumption regarding essential elements
Full size table
Table 3 Risk estimation of Barbary sheep muscle consumption regarding toxic elements
Full size table

Safe intake levels of non-essential (toxic) As, Cd, Hg and Pb from food considered free of risk of adverse health effects for consumers were not exceeded either with average or with P95 consumption frequency of Barbary sheep muscle (Table 3). With Cd and Pb as the most frequently reported priority contaminants in game meat, consumption of meat from this study would add 0.1 to 3% TWI/BMDL01. As such, the contribution of toxic metal(loid)s from Barbary sheep meat to the overall diet of Croatian consumers can be considered low and consumption safe, even in the worst-case scenario conditions predicted here (P95 consumer of meat which contains P95 percentile level of toxic elements). Ferri et al. (2017) recorded that liver consumption was often connected with that of game meat, although the consumed amounts of meat largely prevail over liver (Lazarus et al. 2014). In the lack of official data about game offal consumption frequency in Croatia, we used the same scenario as for average muscle consumption, although it should be stressed that the scenario of chronic daily consumption of 43.4 g liver is unlikely. Our findings (Table S3) show that the average consumers of Barbary sheep liver meet their dietary requirements for Cu and Mo (54–189% DRV). Although the adequate intake for Cu (1.3–1.6 mg day−1; EFSA 2017) would be surpassed with liver consumption (1.61–3.02 mg Cu portion−1), those amounts would be similar to the average daily intake assessed by EFSA for adults (1.15–2.07 mg Cu day−1; EFSA 2017) and thus can be considered safe for health. In addition, liver was shown to be a good source of Fe, Se and Zn (16–28% DRV) with also a significant amount of K, Mg and Mn added to their daily nutrition (2–5% DRV). Among the toxic metal(loid)s measured here, Cd intake through liver consumption is of highest concern (Table S3), with the average daily consumption adding 36–60% TWI to the amount ingested daily from other dietary or environmental sources. Exposure to Pb (7–16% BMDL01) can be considered low if no other significant occupational or dietary sources enhance the overall intake of Pb in the human body. Hg and As intake from Barbary sheep liver is very low (0.3–1% TWI/BMDL01).

Radionuclides

Radiocaesium in Croatia originates from nuclear weapon testing and the Chernobyl accident in the twentieth century. It is persistent, has high gastrointestinal uptake rate in herbivores and humans and, due to its chemical similarity, replaces naturally occurring potassium (40K), thereby elevating the radiological dose. Potassium availability can affect uptake and biomagnification of caesium in the food chain (Šprem et al. 2013) so 40K and 137Cs should be monitored simultaneously. The concentrations of 137Cs and 40K in muscle samples of Barbary sheep from Mt. Mosor are shown in Table 1. Activities of both elements in the analysed samples are found at levels significantly below the statutory values of 600 Bq kg−1. The average mass activity of 137Cs in this study was 3.5 Bq kg−1 (maximum 4.50 Bq kg−1 ± 0.78 Bq kg−1), which is almost the same activity obtained in other mountain ungulates, such as Alpine chamois, with reported activities of 2.93 Bq kg−1 by Šprem et al. (2013) and a few becquerels per kilogram by Pourcelot et al. (2003). European ungulate meat 137Cs contamination has declined since the Chernobyl fallout in 1986 (Ramanzin et al. 2010). 40K mass activity showed equal dispersal with average mass activity of 136.07 Bq kg−1 (maximum 163 Bq kg−1 ± 22 Bq kg−1), as previously reported for different ungulates (Šprem et al. 2013) and carnivore species (Šprem et al. 2016). The descriptive statistics indicated that the measured 40K mass activity was uniformly distributed among the samples. A similar and parallel trend of the two radionuclides was also confirmed with this study, as Barbary sheep with a lower 40K mass activity of 82.7 Bq kg−1 also has the lowest load of 137Cs of 2.11 Bq kg−1. The estimated effective equivalent dose (HE) for consumption of 1 kg of Barbary sheep meat was 0.046 μSv. This low HE value is similar to certain other ungulate species, such as Alpine chamois (0.038 μSv), red deer (0.032 μSv) and roe deer (0.030 μSv) (Šprem et al. 2013). Therefore, we can conclude that the meat of herbivore ungulate species can be excluded as being a general risk as a foodstuff regarding caesium concentrations.

Conclusions

As the first study reporting Barbary sheep tissue contents of radionuclide and stabile elements, it can be concluded that the measured levels reflected low environmental (likely atmospheric) pollution with As, Cd, Hg and Pb, with levels generally less than those reported for wild herbivorous ungulates. Methodological differences (lower detectable levels of elements in muscle of our animals) hampered the interpretation and comparison of TCI values with the previously published data (for mustelids). Our results suggest that there was no homeostasis disturbance of trace elements in Barbary sheep, either due to inadequate intake via food or as an adverse effect due to high toxic metal(loid) burden. Consumption of the muscle and liver of wild Barbary sheep can be considered safe for the health of adult consumers regarding toxic metal(loid)s and radioactive caesium, though the liver should be avoided as a food item in vulnerable population groups (children and child-bearing and lactating women) for the possible adverse effects of Cd and Pb. Otherwise, muscle and liver are shown to be a rich source of Cu, Fe, Se and Zn for consumers and, as such, can benefit the overall dietary intake of essential elements.