Safety Assessment of Honeys from Northern and Southern Algerian Regions

Abstract

Although the EU is a major producer of honey, commercial production is often insufficient to meet market demand and, as a result, honey is often imported into the EU from extra-EU countries that lack regulatory standards for food safety and quality. Since honey is a matrix highly susceptible to contamination, monitoring the quality and safety of extra-EU honey is of significant importance to show potential safety gaps. Hence, aim of the study was to monitor the mineral profile of monofloral and multifloral honeys from different regions of North (provinces of Tiaret and Laghouat) and South Algeria (province of Tindouf). In almost all the samples, Mg, Fe, Zn, Cd and Pb were found at levels exceeding the limits set for honey by the Codex Alimentarius and European Regulation 915/2023. In addition, a PCA analysis pointed out that the analysis of the element profile was useful to discriminate Algerian honeys more on the basis of geographical than botanical origin. The dietary exposure assessment indicates that the investigated honeys can be safely consumed in quantities comparable to those considered in Europe (1.8 g/day) and North Africa (0.3 g/day). Hopefully, data from this study may solicit the Algerian government to set regulatory limits on inorganic elements in honey and align with other international standards, to create a harmonized network able to improve the safety of this food.

1. Introduction

The Farm to Fork (F2F) strategy is a 10-year plan developed by the EU Commission as part of the European Green Deal to guide the transition to a fair, healthy and environmentally friendly food system through diverse initiatives, including a better management and reduction of all sources of pollution, such as industrialization and urbanization, the use of chemical pesticides, fertilizers, and antibiotics in agriculture, and the leaching of contaminants during food processing, which contribute to the accumulation and transport of pollutants into the environment (air, water, and soil) and, consequently, into the food chain [1].

However, in the framework of F2F, the implementation of a fair and sustainable trade of non-EU food matrices toward EU countries must also be considered. Indeed, if on the one hand the EU has recently introduced numerous regulations to determine which substances are permitted in food, and in what amounts these are allowed, the same cannot be said for extra-EU countries, which often produce food that does not meet safety standards [2]. It would therefore be highly advisable to strengthen the reliability and frequency of monitoring procedures for all those food products imported from extra-EU countries. In this scenario, honey represents a highly imported food product from extra-EU countries and, as will be discussed, it is prone to contamination by a number of xenobiotics, including certain inorganic toxic and potentially elements [2].

The term “honey” denotes the sugary substance produced by bees (Apis mellifera L.) from the nectar of flowers (floral honey or nectar honey) or from the secretions of living parts of plants or from substances secreted by sucking insects on the living parts of plants (honeydew honey), which they forage, transform, combine with their own specific substances, deposit, dehydrate, store, and allow to ripen in the honeycombs of the hive [3,4]. Honey is a food matrix characterized by many organic (sugars, proteins, organic acids, vitamins, etc.) and inorganic (minerals) chemical compounds. Its chemical composition and therefore its properties are linked to the botanical species and the geopedoclimatic conditions in which the nectariferous plants grow [5,6]. In fact, the soil conditions (e.g., type, moisture, etc.) and the possible contamination of the environment (water, air, and soil) can influence the chemical composition of the honey [7].

The concentration of inorganic elements, including those harmful to environment and humans, in honey varies between 0.1 and 0.2% and, broadly speaking, it is notoriously influenced by the environment in which the plant grows, the foraging activities of bees and, not least, by honey processing and packaging practices [8]. More specifically, climatic agents (i.e., high temperatures, dry climates, and wind) and environmental factors (i.e., polluted water, air and soil) may favor the accumulation of toxic and potentially toxic elements in honey [2,9,10]. Lead (Pb), for example, is one of the most dangerous substances because it is mainly present in the air from car exhaust and can directly contaminate nectar and honeydew [7]; cadmium (Cd) from the metal industry and incinerators can reach the soil and plants and contaminate nectar and honeydew [9]. As a result, honey can be considered as a valid matrix to study environmental pollution and its monitoring can help to assess environmental quality and conditions [10]. Beyond that, the inorganic contamination of honey may be related also to inappropriate working practices and use of certain materials during the harvesting, production, preparation, and storage of honey [11,12]. Given the ease with which honey can be contaminated by these elements, regulatory and monitoring measures are necessary to set maximum levels for the most dangerous elements and, hence, check the quality of the honey that ends up on our tables. At EU level, the Regulation 915/2023 only regulates Pb in honey, setting the maximum permissible content at 0.10 mg/kg [13]. In addition to hazardous elements, other inorganic elements are considered excellent markers to study the traceability (geographical and botanical origin) of the honey and to guarantee its quality to the consumer [14,15,16].

In Algeria, beekeeping plays an important role in agricultural and rural activities. However, although it is practiced in different regions, the north of the country produces the largest quantity, thanks to favorable climatic conditions and, therefore, a high floristic biodiversity [17]. This differentiation is linked to the environmental and geological conditions of Algerian territory. In fact, Algeria is divided into two distinct zones: North Algeria, with the northern Tell (5%) and the steppe (15%), and South Algeria with the Sahara Desert (80%). The Tell includes a coastal zone (1600 km), plains and mountains of the Tellian Atlas [18], and it is characterized by a Mediterranean climate that allows for the presence of a very large number of flower species [18]. The steppe includes the high plateaus between 600 and 1200 m, where there is a dry Mediterranean climate and consequently fewer flower species. The Sahara, on the other hand, has an extremely dry climate, with less flora and is therefore less favorable for beekeeping [18].

This difference in climate allows for wide variety of flora [19,20,21] and consequently the production of different types of honey: in the north of Algeria, flower honeys such as Citrus, Eucalyptus, Salvia rosmarinus, Thymus are found; while in the south, other types of honey are harvested, such as Spurge, Ziziphus, Thistle, and Peganum Harmala [22,23].

However, Algeria does not have a reference standard for toxic or potentially toxic elements in honey. For this reason, the country takes into account the Codex Alimentarius, which sets maximum limits for certain elements in honey: 25 mg/kg for Mg, 15 mg/kg for Fe, 5 mg/kg for Cu and Zn, 0.05 mg/kg for Cd, and 0.01–0.5 mg/kg for As [3,9].

Within this context, the following study had several aims: (1) to determine the content of 20 mineral elements (As, Be, Ca, Cd, Co, Cr, Cu, Fe, Hg, Li, Mg, Mn, Mo, Na, Ni, Pb, Sb, Sn, Ti, and Zn) in monofloral and multifloral honeys from North and South Algeria; (2) to assess and verify their suitability with respect to the requirements of international standards; (3) to assess the potential toxicological risk to humans from honey consumption. Given the absence of Algerian regulations on the element content of honey -especially with respect to toxic and potentially toxic elements-, and the scarce monitoring measures adopted to control the safety and quality of honey, this work intends to point out the urgency of routinely monitoring the quality and safety of food products imported from extra-EU countries through reliable protocol of analysis, and solicit the establishment of maximum limits for these contaminants that are in line with the international safety standards as well.

2. Materials and Methods

2.1. Samples

The study was conducted on 54 honeys produced in 2023 by beekeepers located in different parts of Algeria: North Algeria (Laghouat and Tiaret provinces) and South Algeria (Tindouf province) (Figure 1). Tindouf is in the southern part of the Sahara, on the Hammada plateau, and is characterized by high temperatures (>50 °C) in summer and low temperatures (until −5 °C) in winter. Tiaret is located at an altitude of 1000 m, and has a Mediterranean climate with some continental and semi-arid features due to very cold winters [23] and very hot summers (around 40 °C). Finally, Laghouat has a subtropical desert climate with mild winters and very hot and sunny summers. The choice of analyzing different types of honeys from Tiaret and Laghouat is because these two provinces are among the most important in Algeria for honey production [2], while the analysis of honeys from Tindouf is due to the desire to compare the variation in mineral content of honeys produced in areas with different climatic conditions (i.e., northern and southern Algeria). The honey samples obtained from these areas of Algeria were collected in glass containers of approximately 150 g with labels indicating the botanical origin, composition, and provenance. The transportation and storage at our laboratory facilities always took place in the dark and at a constant temperature until analysis.

The code and number of samples, as well as their geographical and botanical origin, are given in

Table 1. Honey samples under study.

Area	Code Samples	N. of Samples	Geographical Origin	Botanical Origin
North Algeria	EST	3	Tiaret	Eruca sativa
	MT,L	9	Tiaret/Laghouat	Multifloral
	ZT,L	9	Tiaret/Laghouat	Ziziphus lotus
	BMT	3	Tiaret	Bunium mauritanicum
	ET	3	Tiaret	Echinops ssp.
	EGL	3	Laghouat	Eucaliptus globulus
	EOL	6	Laghouat	Euphorbia orientalis
	TEL	3	Laghouat	Tamarix L. and Euphorbia orientalis
South Algeria	ESTD	6	Tindouf	Eruca sativa
	EOTD	3	Tindouf	Euphorbia orientalis
	ETD	3	Tindouf	Echinops spp.
	PHTD	3	Tindouf	Peganum harmala
	Total	54

The honeys analyzed also differed in their botanical origin. Eight types of multifloral and monofloral honeys (Eruca sativa, Multifloral, Ziziphus lotus, Bunium mauritanicum, Echinops ssp., Eucaliptus globulus, Euphorbia orientalis and Tamarix L.) were analyzed from the northern part of Algeria. From the province of Tindouf (southern Algeria), four types of monofloral honeys were studied: Echinops spp., Eruca sativa, Euphorbia orientalis and Peganum harmala (Table 1).

.

2.2. Chemicals and Reagents

For the pretreatment and digestion of the samples, ultra-pure water, 69% concentrated nitric acid for trace metal analysis, and 30% hydrogen peroxide were used. All reagents were purchased from J.T. Baker (Milan, Italy). Stocks of single-element standard solutions of As, Be, Cr, Cu, Fe, Li, Mo, Ni, Pb, Sb, Sn, Ti, and Zn (1000 mg/L in 2% nitric acid) purchased from Fluka (Milan, Italy) and of Ca, Cd, Co, Mg, Mn, and Na from Merck (Darmstadt, Germany), were used to prepare multi-element standard solutions at a concentration of 100 mg/L for each element, which were used for calibration solutions and method validation. Specifically, seven-point calibration curves were constructed with a concentration between 0.5 and 20 µg/L for As, Be, Cd, Cr, Li, Mo, Ni, Pb, and Sb; between 10 and 200 µg/L for Co, Cu, Fe, Mn, Sn, Ti, and Zn; and between 0.1 and 5 mg/L for Ca, Na, and Mg. A starting standard solution of Re at 1000 mg/L in 2% nitric acid, purchased from Fluka (Milan, Italy), was used to prepare the standard at 0.5 mg/L. A stock standard solution of ⁴⁵Sc, ⁷³Ge, ¹¹⁵In and ²⁰⁹Bi at 1000 mg/L in 2% nitric acid (Fluka, Milan, Italy) was used to prepare an on-line internal standard solution at 1.5 mg/L, to correct for matrix deviation and instrumental drift. For the subsequent ICP-MS analysis, a solution containing 1 µg/L of ⁷Li, ⁵⁹Co, ¹³⁸Ba, ²⁰⁹Bi, ¹⁴²Ce, ¹¹⁵In and ²³⁸U in 2% nitric acid and 0.5% hydrochloric acid, purchased from Thermo Fisher Scientific (Bremen, Germany) and used for instrument calibration, was used. The gases used were argon with a purity of 99.9990% and helium at 99.9995%. A mercury solution (1000 mg/L in 3% hydrochloric acid) was purchased from Merck (Darmstadt, Germany). A 3% HCl solution prepared from concentrated HCl (37%), supplied by Merck (Darmstadt, Germany), was used to clean the DMA-80.

2.3. Inorganic Elements

The preliminary step of the ICP-MS analysis was the acid digestion of the samples by a closed vessel microwave digestion system (ETHOS 1, Milestone, Bergamo, Italy). This procedure followed the method proposed by Massous et al. [24]. In brief, approximately 0.5 g of each sample was first homogenized at 40 °C and then weighed in closed PTFE vessels. At this point, 1 mL of internal standard Re was added at a concentration of 0.5 mg/L. The reagents used for the mineralization were 65% HNO₃ (7 mL) and 30% H₂O₂ (1 mL). Instrumental parameters for mineralization were as follows: -Step 1 consisted of heating the samples in a temperature range between 0 and 200 °C at 1000 W microwave power for 15 min; -In step 2, the temperature was kept constant at 200 °C, with the same power and for the same time; -The samples were cooled in step 3 for 20 min. The samples were then diluted with ultra-pure water to a final volume of 25 mL and filtered through 0.45 µm filters. Both the solution blank, consisting of 1 mL of Re internal standard at 0.5 mg/L, 8 mL of 65% HNO₃ and 2 mL of 30% H₂O₂, and the samples spiked with known amounts of the analytical standard of every target analyte used for the subsequent calculation of the recovery percentage, were digested under the same mineralization conditions. The iCAP-Q ICP-MS (Thermo Scientific, Waltham, MA, USA), allowed for the determination of the mineral content in all honey samples. Specifically, the following elements were determined: ⁷Li, ⁹Be, ²³Na, ²⁴Mg, ⁴⁰Ca, ⁴⁸Ti, ⁵²Cr, ⁵⁵Mn, ⁵⁶Fe, ⁵⁹Co, ⁶⁰Ni, ⁶³Cu, ⁶⁶Zn, ⁷⁵As, ⁹⁸Mo, ¹¹⁴Cd, ¹²⁰Sn, ¹²¹Sb, and ²⁰⁸Pb. The operating conditions of ICP-MS are already reported in previous studies [25,26,27] and are shown in Table S1.

The Thermo Scientific QtegraTM Intelligent Scientific Data System software (Thermo Scientific, Waltham, MA, USA) was used for data acquisition. For the quantification, a seven-point calibration curve was constructed for quantitative analysis. All experimental analyses were performed in triplicate along with the analysis of the analytical blank.

2.4. DMA-80 Analysis

The mercury content in honey samples was determined by a direct mercury analyzer (DMA-80, Milestone, Bergamo, Italy). The method of analysis followed the EPA 7473 guidelines [28] and the operating conditions already reported in a previous study [29], with some modifications. Briefly, each honey sample (0.1 g) was first dried at 290 °C for 3 min and then thermally decomposed at 650 °C for 5 min. The mercury content was determined at the typical wavelength of 253.7 nm.

2.5. Statistical Analysis

The SPSS 19.0 software package for Windows (SPSS Inc., Chicago, IL, USA) was used to analyze the significance of the results. The non-parametric Mann-Whitney U-test, applied to log-transformed data, allowed for the evaluation of the differences between honey samples based on geographical origin, with a statistical significance at p < 0.05. Conversely, the non-parametric Kruskal–Wallis test was applied to log-transformed data to assess the differences between the honey samples based on botanical origin, with statistical significance at p < 0.05. Some elements, i.e., Co, Ti, As, Ni, Sb and Sn, were not included in the data set because their concentrations were below the limit of quantification (LOQ) in more than 50% of the total samples analyzed. When the concentrations were below the LOQ in only a few samples, a value of LOD/2 was assigned to the concentrations. In addition, a Principal Component Analysis (PCA) was carried out to explore the sample differentiation in relation to the botanical and/or geographical origin.

2.6. Assessment of Contaminant Exposure through Diet

To determine the quality of the Algerian honeys studied and, consequently, to assess whether their intake may pose a health risk to consumers, their Estimated Daily Intake (EDI) was calculated. The formulas used were the same as those used in earlier study [15].

Specifically, the EDI values for essential elements were obtained using the following Formula (1):

EDI = C × I

(1)

where C is the average concentration of inorganic contaminants found in each sample, expressed in mg/kg; while I is the amount of honey consumed daily (1.8 and 0.3 g/capita/day for Europe and North Africa, respectively), as reported by FAOSTAT data [30]. The obtained EDI values were then compared with dietary reference values for essential elements, based on Directive 1169/2011 and EFSA [31,32].

To calculate the EDI values for toxic and potentially toxic elements, the following formula was employed (2):

EDI = (C × I)/Kg_b.w.

(2)

where Kg_b.w. is the average body weight of the consumer (70 kg). In addition, the EDIs obtained were compared with international safety reference values for toxic and potentially toxic elements [33,34,35,36,37,38,39,40,41,42,43].

Moreover, to assess the plausibility of chronic non-carcinogenic risk, we calculated the Hazard Quotient (HQ) [25]. Precisely, the HQs were calculated using the Formula (3):

HQ = EDI/RfD

(3)

where EDI is the estimated daily intake and RfD is the oral reference dose proposed by the U.S. Environmental Protection Agency (US EPA) [44]: 0.3 µg/Kg_bw/day for As; 1 µg/Kg_bw/day for Cd; 3 µg/Kg_bw/day for Cr; 40 µg/Kg_bw/day for Cu; 9 µg/Kg_bw/day for Fe; 140 µg/Kg_bw/day for Mn; 20 µg/Kg_bw/day for Ni; 3.5 µg/Kg_bw/day for Pb; 300 µg/Kg_bw/day Zn. A value of HQ above 1 is synonymous with a high non-carcinogenic risk.

3. Results and Discussions

3.1. Validation of the ICP-MS and DMA-80 Methods

Both the ICP-MS and DMA methods were validated by determining the linearity, sensitivity, and recovery, according to Eurachem guidelines [45]. Standard solutions prepared at seven concentrations for each analyte (range: 0.5–50 µg/L, except for Hg, range: 1–100 µg/L) were analyzed by checking the linearity between concentrations and signal intensities. Each concentration level was analyzed six times (n = 6). The linearity of the calibration curves was verified by the correlation coefficients (R²). Sensitivity was assessed by establishing a limit of detection (LOD) and limit of quantification (LOQ) for each of the analytes, determined as 3.3 σ/S and 10 σ/S respectively, where σ is the standard deviation of the value of ten blanks and S is the slope of the calibration curve. Recovery tests were performed using three replicates of honey samples at three different fortification levels: 0.1, 0.2 e 0.5 mg/Kg for Pb, Ni, Cr, Sb, As, Cd, Sn, and Hg; 0.5, 2.0 e 5.0 mg/kg for Fe, Li, Be, Zn, Ti, Mn, Mo, Cu e Co; and 0.5, 1.0 e 2.0 g/Kg for Ca, Na e Mg. Each recovery was calculated in percentage terms and by considering the experimental concentration of the element derived from the sample analysis and the known amount of standard spiked to the sample. The results are reported as mean recovery values (%) between the three different concentration levels. The repeatability was evaluated by analyzing the spiked samples on the same day (intraday precision) and over a longer period (i.e., 1 week, interday precision).

The values of R², LOD, LOQ, recovery percentage, and the precision are shown in Table S2. R² values > 0.9990 were obtained for all analytes (Table S2), synonymous with optimal linearity. The method also showed good sensitivity, expressed as LOD and LOQ values, calculated as 3.3 σ/b and 10 σ/b, respectively, where σ is the standard deviation of the analytical blank (n = 6) and b is the slope of the calibration curve. Specifically, the LOD and LOQ ranged from 0.001 to 0.003 mg/kg (for As, Cd, and Hg) and from 0.105 to 0.347 mg/kg (for Ca) (Table S2). The lowest and highest mean recoveries were observed for Na (91.85%) and Pb (104.50%) (Table S2), respectively. The intraday and interday precision, expressed as relative standard deviation (RSD%), were less than 1.3% and 1.4%, respectively (Table S2).

3.2. Mineral Elements in Honey Samples

The non-parametric Mann-Whitney U-test and the results, reported in Table 2, Table 3 and Table 4, demonstrated that the mineral element profile was influenced by geographical origin. This is because plants absorb elements from the soil through their roots and, thus, transfer them to the nectar. This absorption capacity is directly connected with the type of plant, and the type of soil in which the plant lives [46]. So, the proximity to industries or traffic areas also plays a fundamental role in the contamination of the nectariferous plant and, thus, bee products. However, the contamination is not only related to the type of plant and the environment in which it grows because, during the foraging phases, the bees come into direct contact with the environment, and this causes a further transfer of mineral elements from the environment (soil, water, and air) to the hive. For this reason, the determination of the element content of honey is of fundamental importance for assessing the possible bioaccumulation of inorganic elements in the environment, and for assessing the possible damage caused by anthropic activities, such as industrial activities and agricultural chemicals, near the apiary. However, attention must also be paid to another factor that may contaminate this matrix, namely the entire production process, in which the use of inappropriate practices and tools may lead to contamination with toxic elements in honey.

3.3. Macro-Elements

The concentrations of macro elements (Ca, Mg and Na) in the honey samples are shown in Table 2. In most samples, calcium was the most abundant mineral (range: 73.23 ± 2.15–110.34 ± 10.42 mg/kg respectively in samples PH_TD sample TE_L) and this result was consistent with other studies [47]. For North Sahara honeys, those of Laghouat origin showed a higher content of Ca, followed by Mg and Na. This trend agreed with those reported in other studies on honeys from different Algerian areas [19,47]. The trend of Tiaret honeys was Ca > Na > Mg, instead. Only sample PH_TD from the Tindouf region showed a different behavior from all the honeys, with Mg being the most abundant, followed by Na and Ca.

Table 2. Macro-elements content (mg/kg) in honey samples. Bold p-values are statistically significant (p < 0.05).

Sample Code	Geographical Origin	Ca	Mg	Na
EST	North Sahara	108.16 ± 9.45 a	55.17 ± 3.45	85.89 ± 6.99
MT,L		100.85 ± 5.51 a,b,c	57.57 ± 10.89	67.26 ± 21.59
ZT,L		96.63 ± 4.79 a,b,d	68.12 ± 13.67	55.23 ± 15.42
BMT		105.27 ± 6.06 a	46.29 ± 3.39	88.13 ± 8.52
ET		100.33 ± 5.57 a,b,c,d	48.16 ± 3.45	82.24 ± 6.39
EGL		89.23 ± 6.19 b,c,d,f	72.13 ± 8.12	42.78 ± 7.02
EOL		93.33 ± 7.41 a,b,d,f	80.90 ± 5.85	43.93 ± 5.55
TEL		110.34 ± 10.42 a	49.88 ± 3.12	42.39 ± 5.82
ESTD	South Sahara	75.03 ± 5.52 e,g	54.75 ± 11.28	75.27 ± 2.89
EOTD		84.14 ± 2.53 d,f	51.17 ± 4.67	38.55 ± 1.62
ETD		101.34 ± 4.61 a,b,d	49.50 ± 1.94	76.10 ± 2.46
PHTD		73.23 ± 2.15 e,g	101.70 ± 3.16	92.11 ± 1.81
p-Value		0.000	0.493	0.302

a–g: Different superscript letters in the same column indicate significantly different values (p ≤ 0.05 by Mann-Whitney U-test); same super-script letters in the same column indicate non-significantly different values (p > 0.05 by Mann-Whitney U-test).

Table 2. Macro-elements content (mg/kg) in honey samples. Bold p-values are statistically significant (p < 0.05).

Sample Code	Geographical Origin	Ca	Mg	Na
EST	North Sahara	108.16 ± 9.45 a	55.17 ± 3.45	85.89 ± 6.99
MT,L		100.85 ± 5.51 a,b,c	57.57 ± 10.89	67.26 ± 21.59
ZT,L		96.63 ± 4.79 a,b,d	68.12 ± 13.67	55.23 ± 15.42
BMT		105.27 ± 6.06 a	46.29 ± 3.39	88.13 ± 8.52
ET		100.33 ± 5.57 a,b,c,d	48.16 ± 3.45	82.24 ± 6.39
EGL		89.23 ± 6.19 b,c,d,f	72.13 ± 8.12	42.78 ± 7.02
EOL		93.33 ± 7.41 a,b,d,f	80.90 ± 5.85	43.93 ± 5.55
TEL		110.34 ± 10.42 a	49.88 ± 3.12	42.39 ± 5.82
ESTD	South Sahara	75.03 ± 5.52 e,g	54.75 ± 11.28	75.27 ± 2.89
EOTD		84.14 ± 2.53 d,f	51.17 ± 4.67	38.55 ± 1.62
ETD		101.34 ± 4.61 a,b,d	49.50 ± 1.94	76.10 ± 2.46
PHTD		73.23 ± 2.15 e,g	101.70 ± 3.16	92.11 ± 1.81
p-Value		0.000	0.493	0.302

a–g: Different superscript letters in the same column indicate significantly different values (p ≤ 0.05 by Mann-Whitney U-test); same super-script letters in the same column indicate non-significantly different values (p > 0.05 by Mann-Whitney U-test).

Among the regions studied, honeys from North Algeria were characterized by the highest Ca content (mean concentration: 100.52 ± 7.29 mg/kg) compared to those from South Algeria (mean concentration: 83.44 ± 12.86 mg/kg). These differences in Ca content between honeys from northern and southern Algeria were highly significant (Table 2), as shown by the low p-value obtained for this element (0.000). Our samples showed a higher Ca content compared to other studies conducted on honeys from other areas of Algeria, which were not studied in the present work, such as in the case of the study conducted by Haouam et al. (2016) [19], where a range of Ca between 13.20 and 33.51 mg/kg was obtained for multifloral honeys from semi-arid regions of northeastern Algeria. In other cases, our results were comparable to those found in the literature, such as in the case of the study conducted by Bereksi-Reguig et al. (2020) [47] on monofloral and multifloral honeys from different areas of Tlemcen province in northwestern Algeria (mean content: 95.23 mg/kg) and that of Bereksi-Reguig et al. (2022) [48] on honeys from western Algeria (mean content: 113.92 mg/kg). Moreover, Zerrouk et al. (2017) [49] conducted a study on 27 unifloral honeys of Zizphus lotus from different regions of Laghouat and Djelfa in Algeria and obtained a mean Ca concentration of 136.6 mg/kg, higher than ours. Furthermore, when comparing our results with those of honeys from countries other than Algeria, the results were different depending on the geographical and botanical origin of the product: Massous et al. (2023) [24], in their study on honeys from Morocco, found a slightly higher Ca content in euphorbia honey (125.62 ± 1.00 mg/kg) than ours; in the study by Di Bella et al. (2021) [12] on Tunisian honeys, the authors found a higher calcium content in eucalyptus honey (0.17 ± 0.02 g/kg) than in the following study. Perna et al. (2021) [50] carried out a study to determine the mineral content of honeys from the Basilicata region (Italy), where they obtained a Ca content (44.09 ± 16.08 mg/kg) lower than in the Algerian honey studied by us.

The Mg content ranged from 46.29 ± 3.39 mg/kg (BM_T sample) to 101.70 ± 3.16 mg/kg (PH_TD sample). In general, the average Mg content was higher in South Sahara honeys (mean: 64.28 ± 25.04 mg/kg) than in those from North Sahara (mean: 53.13 ± 12.59 mg/kg). However, the Mg concentrations were well above the upper limit set by the Codex Alimentarius: 25 mg/kg in honey. This high concentration could be due to the possible presence of this element in Algerian soils, but also to the contamination of hives located near industrial areas [51]. Our Mg concentrations were significantly higher than those reported by different studies: by Zerrouk et al. (2017) [49] (mean concentration: 22.1 mg/kg) and Haouam et al. (2016) [19] (mean concentration: 2.10 ± 1.23 mg/kg), but lower than those obtained by Bereksi-Reguig et al. (2022) [48] (mean concentration: 100.83 ± 33.88 mg/kg). Furthermore, the Algerian honeys under study contained a significantly higher average amount of Mg than that obtained by Hungerford et al. (2020) [10] (average content: 28.7 ± 19. 6 mg/kg), who carried out a study on the mineral characterization of Australian honeys, by Spiric et al. (2019) [52], who studied different Serbian honeys (average concentration of multiflora honey: 28.71 ± 11.12 mg/kg), and by Perna et al. (2021) [50], (average concentration: 19.30 ± 6.43 mg/kg).

The geographical origin of the product also had a great influence on the Na content. The Na concentration ranged from 38.55 ± 1.62 mg/kg (sample EO_TD) to 92.11 ± 1.81 mg/kg (sample PH_TD). South Sahara honeys had the highest average Na content (70.51 ± 22.67 mg/kg), followed by those from the North Sahara area (63.48 ± 20.00 mg/kg). Habati et al. (2017) [53] reported the Na content of Ziziphus lotus and Peganum harmala honey in the Laghouat region, with concentrations of 28.67 ± 1.20 mg/kg and 136.20 ± 3.00 mg/kg, respectively. The Na level of Ziziphus lotus honey, in particular, was lower than the Na content found in our samples. In the case of Peganum harmala honey, the Na content was higher than in the same type of honey from the Tindouf region analyzed in our study. Di Bella et al. (2020) [9], on the other hand, found a mean concentration of Na in multifloral honeys from the Laghouat region (1.07 ± 0.07 g/kg), much higher than ours. Our mean values were higher than those reported by Spiric et al. (2019) [52] (mean concentration in multifloral honeys: 15.30 ± 14.60 mg/kg), but comparable with those reported by Perna et al. (2021) [50] (mean concentration: 49.15 ± 18.39 mg/kg), and by Sakac et al. (2019) [54] in a study of honeys from Vojvodina (Republic of Serbia) (mean concentration: <50 mg/kg).

3.4. Micro-Elements

Table 3 shows the concentration of trace elements in the honey samples analyzed. Lithium (Li), molybdenum (Mo) and beryllium (Be) were also analyzed in this study, but their concentrations are not shown in Table 3 because they were below the LOQ in all samples.

Table 3. Micro-elements content (mg/kg) in honey samples. Bold p-values are statistically significant (p < 0.05).

Sample Code	Geographical Origin	Fe	Co	Cr	Cu	Mn	Ti	Zn
EST	North Sahara	11.33 ± 2.92	<LOQ	<LOQ	<LOQ	<LOQ	0.94 ± 0.07	7.92 ± 0.40
MT,L		17.53 ± 8.33	0.04 ± 0.01	0.32 ± 0.11 a,b,c,d,e,f	0.72 ± 0.08	0.04 ± 0.01	<LOQ	14.57 ± 0.69
ZT,L		17.69 ± 5.51	<LOQ	0.17 ± 0.10 a,b,d,e,f	0.47 ± 0.07	0.06 ± 0.01	<LOQ	8.45 ± 0.96
BMT		10.76 ± 3.25	<LOQ	0.30 ± 0.06 a	1.03 ± 0.08	0.06 ± 0.02	<LOQ	10.31 ± 0.31
ET		12.45 ± 2.37	<LOQ	0.35 ± 0.07 a	0.80 ± 0.04	0.05 ± 0.01	<LOQ	6.88 ± 0.20
EGL		18.54 ± 4.59	<LOQ	0.19 ± 0.05 a,b,d,e	0.29 ± 0.08	0.05 ± 0.01	<LOQ	11.51 ± 0.20
EOL		20.29 ± 5.05	<LOQ	0.14 ± 0.04 a,b,d,e,f	<LOQ	<LOQ	<LOQ	8.77 ± 0.39
TEL		14.31 ± 3.14	<LOQ	0.04 ± 0.01 a,c	<LOQ	<LOQ	<LOQ	9.71 ± 0.17
ESTD	South Sahara	13.03 ± 3.60	<LOQ	<LOQ	<LOQ	<LOQ	0.52 ± 0.11	16.90 ± 2.60
EOTD		18.52 ± 1.32	<LOQ	0.10 ± 0.03 a,b,d,e,f	<LOQ	<LOQ	<LOQ	9.97 ± 0.37
ETD		13.93 ± 0.64	<LOQ	0.15 ± 0.04 a,b,d,e,f	0.46 ± 0.07	0.06 ± 0.02	<LOQ	8.11 ± 0.51
PHTD		19.19 ± 0.34	0.03 ± 0.01	0.08 ± 0.03 a,b,d,e,f	0.64 ± 0.04	0.28 ± 0.04	<LOQ	7.43 ± 0.27
p-Value		0.877	-	0.003	0.132	0.791	-	0.344

a–f: Different superscript letters in the same column indicate significantly different values (p ≤ 0.05 by Mann-Whitney U-test); same super-script letters in the same column indicate non-significantly different values (p > 0.05 by Mann-Whitney U-test).

Table 3. Micro-elements content (mg/kg) in honey samples. Bold p-values are statistically significant (p < 0.05).

Sample Code	Geographical Origin	Fe	Co	Cr	Cu	Mn	Ti	Zn
EST	North Sahara	11.33 ± 2.92	<LOQ	<LOQ	<LOQ	<LOQ	0.94 ± 0.07	7.92 ± 0.40
MT,L		17.53 ± 8.33	0.04 ± 0.01	0.32 ± 0.11 a,b,c,d,e,f	0.72 ± 0.08	0.04 ± 0.01	<LOQ	14.57 ± 0.69
ZT,L		17.69 ± 5.51	<LOQ	0.17 ± 0.10 a,b,d,e,f	0.47 ± 0.07	0.06 ± 0.01	<LOQ	8.45 ± 0.96
BMT		10.76 ± 3.25	<LOQ	0.30 ± 0.06 a	1.03 ± 0.08	0.06 ± 0.02	<LOQ	10.31 ± 0.31
ET		12.45 ± 2.37	<LOQ	0.35 ± 0.07 a	0.80 ± 0.04	0.05 ± 0.01	<LOQ	6.88 ± 0.20
EGL		18.54 ± 4.59	<LOQ	0.19 ± 0.05 a,b,d,e	0.29 ± 0.08	0.05 ± 0.01	<LOQ	11.51 ± 0.20
EOL		20.29 ± 5.05	<LOQ	0.14 ± 0.04 a,b,d,e,f	<LOQ	<LOQ	<LOQ	8.77 ± 0.39
TEL		14.31 ± 3.14	<LOQ	0.04 ± 0.01 a,c	<LOQ	<LOQ	<LOQ	9.71 ± 0.17
ESTD	South Sahara	13.03 ± 3.60	<LOQ	<LOQ	<LOQ	<LOQ	0.52 ± 0.11	16.90 ± 2.60
EOTD		18.52 ± 1.32	<LOQ	0.10 ± 0.03 a,b,d,e,f	<LOQ	<LOQ	<LOQ	9.97 ± 0.37
ETD		13.93 ± 0.64	<LOQ	0.15 ± 0.04 a,b,d,e,f	0.46 ± 0.07	0.06 ± 0.02	<LOQ	8.11 ± 0.51
PHTD		19.19 ± 0.34	0.03 ± 0.01	0.08 ± 0.03 a,b,d,e,f	0.64 ± 0.04	0.28 ± 0.04	<LOQ	7.43 ± 0.27
p-Value		0.877	-	0.003	0.132	0.791	-	0.344

a–f: Different superscript letters in the same column indicate significantly different values (p ≤ 0.05 by Mann-Whitney U-test); same super-script letters in the same column indicate non-significantly different values (p > 0.05 by Mann-Whitney U-test).

For the micro-elements, the highest contribution was shown by Fe. Its concentrations ranged from 10.76 ± 3.25 mg/kg (sample BM_T) to 20.29 ± 5.05 mg/kg (sample EO_L). In general, most of these values (more than 50% of the samples) were always higher or close to the maximum level allowed in honey by the Codex Alimentarius (15 mg/kg). A possible explanation may be related to the contamination of materials used in the honey extraction process and the preservation of honey in inappropriate containers [51]. The high Fe content was probably due to the presence of iron mines in the area. As a result, the soil on which the plants grow had higher levels of this element, which was then absorbed by the plant’s roots. The Fe content found in honeys from the North Sahara region (mean concentration: 15.36 ± 3.62 mg/kg) was lower than that reported by Di Bella et al. (2020) [9] on multifloral honeys from the same area, for which average Fe levels of 27.44 ± 0.76 mg/kg were found, but higher than the study of Habati et al. (2017) [53], who showed an Fe content between 0.11 ± 0.52 mg/kg and 0.36 ± 0.47 mg/kg in honey from the Laghouat region. However, all samples showed higher Fe concentrations than other honeys from different countries. For example, Spiric et al. (2019) [52] reported a range of Fe between 0.57 and 7.02 mg/kg in Serbian multifloral honeys; Malhat et al. (2019) [55] showed a range between 0.06 and 5.87 mg/kg in Egyptian honeys; and Perna et al. (2021) [50] showed a mean concentration of 1359.48 ± 543.46 µg/kg. This confirms that the environment has a significant impact on the mineral profile of honey.

Zn varied from 6.88 ± 0.20 mg/kg (sample E_T) to 16.90 ± 2.60 mg/kg (sample ES_TD). Thus, all the honeys had a Zn content above the maximum limit set by the Codex Alimentarius for honey (5 mg/kg). Again, the high levels of Zn found in the honey samples analyzed were related to the important mineral resources in Algeria, including zinc deposits. The Zn content of our samples was higher than that reported by Di Bella et al. (2020) [9] (2.69 ± 0.08 mg/kg) for multifloral honeys from the Laghouat region and by Bereksi-Reguig et al. (2022) [48] (1.44 mg/kg) for multifloral honeys from the Tiaret region. Also, the analyzed honeys had a higher amount of Zn than honeys from other geographical areas, such as Italy (1081.13 ± 350.51 µg/kg) [50], Serbia (3.43 ± 3.39 mg/kg) [52], Morocco (range: 1.41 ± 0.03–6.98 ± 0.51 mg/kg) [24], Tunisia (range: 1.69 ± 0.56–2.81 ± 0.61 mg/kg) [12], and Jordan (1.67 ± 0.785 mg/kg) [56].

The Cu concentration in the honey samples was very variable. Cu levels ranged from <LOQ to 1.03 ± 0.08 mg/kg. However, considering that the Codex Alimentarius established a maximum permissible concentration of 5 mg/kg, our results were well below this limit. The average Cu content was higher in honeys from the North Sahara region (0.66 ± 0.29 mg/kg) than those from South Sahara (0.55 ± 0.13 mg/kg). The Cu values were significantly lower than those obtained by Di Bella et al. [9] (mean concentration 1.56 ± 0.12 mg/kg) and by Bereksi-Reguig et al. (2022) [48] (mean concentration: 4.88 mg/kg), but higher than those shown by Habati et al. (2017) [53] (range: 0.09–0.18 mg/kg). Furthermore, the Cu contents of our honeys were higher than those found, in Italian honeys analyzed by Perna et al. (2021) [50] and Meli et al. (2018) [57], (mean Cu content of 236.65 µg/kg and 0.23 mg/kg, respectively), while they were lower than Polish honeys analyzed by Roman et al. (2011) [58], with a mean Cu content of 1.18 ± 0.56 mg/kg. However, our results were comparable to those of Serbian honeys analyzed by Sakac et al. (2019) [54], with Cu levels always below 1 mg/kg.

Other trace elements (i.e., Mn, Cr, Co, and Ti) were detected in Algerian honeys at concentrations ≤ 1 mg/kg and often below the limit of quantification. A low p-value (0.003) was obtained for Cr, demonstrating the significant difference in the content of this element between honeys from northern and southern Algeria (Table 3).

The Mn showed a range between <LOQ and 0.28 ± 0.04 mg/kg and these concentrations were much lower than those reported in the research literature for honeys of different geographical origins: Italian honey (mean concentration: 1.83 mg/kg) [57], Moroccan honey (concentration range: 0.57 ± 0.03–4.00 ± 0.11 mg/kg) [24] and Serbian honey (concentration range: 0.08–17.78 mg/kg) [52]. However, our results were also lower than other studies on Algerian honeys: for example, Di Bella et al. [9] determined a mean Mn concentration of 3.03 ± 0.09 mg/kg for multifloral honey from the Laghouat region. However, Latifa et al. (2013) [59], in their study on 27 honey samples from Laghouat, Djelfa, Medea and El Bayadh, determined a mean Mn concentration of 0.077 ± 0.047 mg/kg for multifloral honeys and 0.069 ± 0.038 mg/kg for Ziziphus lotus honeys, comparable to our results.

Only two samples had a low detected Co concentration, the others had a Co content below the <LOQ. However, the two quantified samples had concentrations comparable to those revealed by Latifa et al. (2013) [59] (0.027 ± 0.015 mg/kg for multifloral honeys and 0.032 ± 0.019 mg/kg for Z. lotus honeys) and by Bereksi-Reguig et al. (2022) [48] (mean concentration for Tiaret honeys: 16.64 µg/kg), but lower than those described by Di Bella et al. [9] (0.37 ± 0.07 mg/kg for Laghouat honeys). Moreover, our results were comparable with Serbian honeys (concentration range: 4.00–78.00 µg/kg) [52], Moroccan honeys (0.01 ± 0.01 mg/kg–0.24 ± 0.11 mg/kg) [24] and Italian honeys (<0.5 mg/kg) [57].

Only two honeys had a quantifiable concentration of Ti. These two honeys had the same botanical origin, i.e., Eruca sativa: 0.52 ± 0.11 mg/kg for sample ES_TD; 0.94 ± 0.07 mg/kg for sample ES_T). This result showed that this plant species had a higher capacity to accumulate this element than the others. However, the Ti level of Algerian honeys were lower than those obtained by Di Bella et al. [9], who obtained a mean Ti concentration of 1.72 ± 0.04 mg/kg for honey from Laghouat.

Finally, the Cr content ranged from <LOQ in ES_T and ES_TD samples (Eruca sativa) to 0.35 ± 0.07 mg/kg in E_T (Echinops ssp.). In general, the highest Cr concentration was obtained from North Sahara honeys (mean concentration: 0.21 ± 0.11 mg/kg), followed by South Sahara honeys (mean concentration: 0.11 ± 0.03 mg/kg). Our results were higher than those obtained by Bereksi-Reguig et al. (2022) [48], who found a mean Cr content of 37.61 µg/kg for honeys from the same region. Moreover, our Cr content was comparable to that determined by Di Bella et al. [9] for honey from the same region (0.21 ± 0.01 mg/kg). However, our Algerian honeys showed a higher Cr concentration than honeys from different countries, namely Italy (mean concentration: 17.57 ± 5.53 µg/kg) [50], Serbia (concentration range: 2.00–27.30 µg/kg) [52] and Australia (mean concentration: 0.008 mg/kg) [10].

3.5. Toxic and Potentially Toxic Elements

It is important to study the concentration of toxic elements in honey because this matrix can be used as a bioindicator to determine the presence, an eventually, the degree of environmental pollution [60,61]. In this respect, Table 4 shows the content of toxic and potentially toxic elements in the honey samples studied. Mercury (Hg) was also analyzed in all samples, but its level is not shown in the table because it was always below the LOQ.

Table 4. Toxic and potentially toxic elements content (mg/kg) in honey samples. Bold p-values are statistically significant (p < 0.05).

Sample Code	Geographical Origin	As	Cd	Ni	Pb	Sb	Sn
EST	North Sahara	<LOQ	0.29 ± 0.12 a	<LOQ	0.17 ± 0.01 a	<LOQ	<LOQ
MT,L		0.21 ± 0.05	1.15 ± 0.85 a,d,e,f,g	0.33 ± 0.03	0.67 ± 0.44 a,b,c,f	<LOQ	0.34 ± 0.05
ZT,L		0.01 ± 0.00	1.10 ± 0.81 a,d,e,f,g	0.27 ± 0.04	1.40 ± 0.12 d,e	<LOQ	0.10 ± 0.03
BMT		1.09 ± 0.16	3.03 ± 0.19 b,c	<LOQ	1.08 ± 0.08 a,b	0.53 ± 0.05	0.51 ± 0.07
ET		1.57 ± 0.20	3.11 ± 0.16 b,c	<LOQ	0.79 ± 0.06 a,c	0.55 ± 0.08	0.50 ± 0.04
EGL		<LOQ	1.25 ± 0.12 a,d	0.31 ± 0.04	1.38 ± 0.08 d,e	<LOQ	0.34 ± 0.09
EOL		<LOQ	0.89 ± 0.12 a,e,g	0.38 ± 0.07	0.17 ± 0.04 a	<LOQ	<LOQ
TEL		<LOQ	0.36 ± 0.11 a	<LOQ	<LOQ	<LOQ	<LOQ
ESTD	South Sahara	<LOQ	0.51 ± 0.06 a,f	<LOQ	0.16 ± 0.02 a	<LOQ	<LOQ
EOTD		<LOQ	0.38 ± 0.06 a	0.33 ± 0.03	0.21 ± 0.03 a,f	<LOQ	<LOQ
ETD		1.13 ± 0.09	0.94 ± 0.13 a,e,g	<LOQ	0.27 ± 0.05 a,f	<LOQ	0.26 ± 0.05
PHTD		0.02 ± 0.01	0.04 ± 0.01 h	0.17 ± 0.06	0.04 ± 0.01 g	<LOQ	<LOQ
p-Value		-	0.010	-	0.038	-	-

a–h: Different superscript letters in the same column indicate significantly different values (p ≤ 0.05 by Mann-Whitney U-test); same super- script letters in the same column indicate non-significantly different values (p > 0.05 by Mann-Whitney U-test).

Table 4. Toxic and potentially toxic elements content (mg/kg) in honey samples. Bold p-values are statistically significant (p < 0.05).

Sample Code	Geographical Origin	As	Cd	Ni	Pb	Sb	Sn
EST	North Sahara	<LOQ	0.29 ± 0.12 a	<LOQ	0.17 ± 0.01 a	<LOQ	<LOQ
MT,L		0.21 ± 0.05	1.15 ± 0.85 a,d,e,f,g	0.33 ± 0.03	0.67 ± 0.44 a,b,c,f	<LOQ	0.34 ± 0.05
ZT,L		0.01 ± 0.00	1.10 ± 0.81 a,d,e,f,g	0.27 ± 0.04	1.40 ± 0.12 d,e	<LOQ	0.10 ± 0.03
BMT		1.09 ± 0.16	3.03 ± 0.19 b,c	<LOQ	1.08 ± 0.08 a,b	0.53 ± 0.05	0.51 ± 0.07
ET		1.57 ± 0.20	3.11 ± 0.16 b,c	<LOQ	0.79 ± 0.06 a,c	0.55 ± 0.08	0.50 ± 0.04
EGL		<LOQ	1.25 ± 0.12 a,d	0.31 ± 0.04	1.38 ± 0.08 d,e	<LOQ	0.34 ± 0.09
EOL		<LOQ	0.89 ± 0.12 a,e,g	0.38 ± 0.07	0.17 ± 0.04 a	<LOQ	<LOQ
TEL		<LOQ	0.36 ± 0.11 a	<LOQ	<LOQ	<LOQ	<LOQ
ESTD	South Sahara	<LOQ	0.51 ± 0.06 a,f	<LOQ	0.16 ± 0.02 a	<LOQ	<LOQ
EOTD		<LOQ	0.38 ± 0.06 a	0.33 ± 0.03	0.21 ± 0.03 a,f	<LOQ	<LOQ
ETD		1.13 ± 0.09	0.94 ± 0.13 a,e,g	<LOQ	0.27 ± 0.05 a,f	<LOQ	0.26 ± 0.05
PHTD		0.02 ± 0.01	0.04 ± 0.01 h	0.17 ± 0.06	0.04 ± 0.01 g	<LOQ	<LOQ
p-Value		-	0.010	-	0.038	-	-

a–h: Different superscript letters in the same column indicate significantly different values (p ≤ 0.05 by Mann-Whitney U-test); same super- script letters in the same column indicate non-significantly different values (p > 0.05 by Mann-Whitney U-test).

Pb was the most abundant toxic element, and its amount was influenced by the geographical rather than the botanical origin. In fact, a p-value < 0.05 was obtained for this element. In general, the content of Pb varied between <LOQ in TE_L samples (Tamarix L. and Euphorbia orientalis L.) and 1.40 ± 0.12 mg/kg in Z_T,L sample (Ziziphus lotus). The average Pb level detected in honeys from North Sahara was 0.81 ± 0.52 mg/kg, while that from South Sahara was 0.17 ± 0.10 mg/kg. Our results showed that the Pb levels found in most of the honey samples (except for TE_L and PH_TD samples) exceeded the limit set by the EU, and more precisely by EU Regulation 915/2023 [10], which indicates a maximum permissible value of Pb of 0.10 mg/kg. The high Pb content could be related to the air pollution caused by industries and exhaust gases in these Algerian areas, which consequently damages the environment and the products of the neighboring apiaries [9]. The average Pb values obtained for honeys from North Algeria area were higher than those reported by Di Bella et al. [9] (average concentration: 0.43 ± 0.03 mg/kg), and by Bereksi-Reguig et al. (2022) [48] (average Pb content of 91.36 µg/kg). Moreover, the average Pb content of honeys from North Algeria (but not of South Algeria honeys, which were often comparable) was higher than that found in Lithuanian honeys (average Pb content of 0.17 mg/kg) [62], Moroccan honeys (range from 0.06 ± 0.01 mg/kg to 0.16 ± 0.06 mg/kg) [24], Polish honeys (0.190 ± 0.179 mg/kg) [63], and Italian honeys (11.07 ± 5.49 µg/kg) [50].

As was detected in only twelve samples, the others being below the limit of quantification. In general, the As content ranged from 0.01 ± 0.00 mg/kg (sample Z_T,L) to 1.57 ± 0.20 mg/kg (sample E_T). The Codex Alimentarius established a maximum acceptable level of 10–500 µg/kg. Hence, three samples (BM_T, E_T and E_TD) exceeded the limit. The high presence of As in these samples may be related to the contamination of the sampling areas, due to the employment of non-ferrous metallurgy and agrochemicals, such as fertilizers and arsenic-based pesticides [64]. Di Bella et al. [9] also found very low levels of As (mean: 8.23 ± 0.60 µg/kg) in Laghouat honeys and close to the relative LOQ from this study (0.001 mg/kg). For other countries, the data in the literature are quite variable: Perna et al. [50] and Meli et al. [57] reported low mean As concentrations of 0.96 ± 0.53 µg/kg and <0.1 mg/kg in Italian honey; Massous et al. [24] reported a concentration range between 0.06 ± 0.01 mg/kg and 0.11 mg/kg in Moroccan honey; Spiric et al. [52] reported a mean As content of 1.68 ± 0.96 µg/kg in Serbian multifloral honey.

In our study, Cd was found in all the honeys analyzed, with concentrations ranging from 0.04 ± 0.01 (PH_TD sample) to 3.11 ± 0.16 (E_T sample) mg/kg. Except for the PH_TD sample, these concentrations were higher than the limit of 0.05 mg/kg for Cd established by the Codex Alimentarius. The presence of Cd is common because this element is released into the environment through its use in various industrial processes or fertilizers, or because it is present in mines [65]. In addition, the availability of Cd in soil is influenced by several factors: soil properties, concentration and form of Cd, organic matter content and soil pH. Consequently, this element can easily enter the food chain through the uptake by plants from contaminated soil or water. North Sahara honeys had the highest mean Cd content (1.40 ± 1.09 mg/kg), followed by those from South Sahara (0.47 ± 0.37 mg/kg). These significant differences were demonstrated by the low p-value (0.010) obtained for this element. Our results were very high compared to those reported by Di Bella et al. [9] (22.57 ± 2.33 µg/kg), Bereksi-Reguig et al. (2020) [47] (0.0012 ± 0.0018 mg/kg) and Demaku et al. (2023) [59] (0.013 ± 0.010 mg/kg for multifloral honeys and 0.011 ± 0.005 mg/kg for Z. lotus honeys). Furthermore, our average Cd content was higher compared to Italian honeys (average content: 3.31 ± 2.33 µg/kg) [50], Serbian honeys (concentration range: 1.00– 21.00 µg/kg) [52], Moroccan honeys (<0.003 µg/kg) [24] and Polish honeys (average content: 0.025 ± 0.023 mg/kg) [63].

There are no limit values for Ni, Sb, Sn or Hg in honey. The content of Ni of half of the honeys from the North Sahara region was always below the LOQ, while for the other half of the samples the range of Ni concentration varied from 0.27 ± 0.04 mg/kg (Z_T,L sample) to 0.38 ± 0.07 mg/kg (EO_L sample). These results were comparable to those obtained by Di Bella et al. [9] (0.31 ± 0.03 mg/kg). For the South Sahara area, the mean concentration of Ni was 0.25 ± 0.12 mg/kg. Among the different honeys analyzed, Euphorbia orientalis had the highest mean Ni content of 0.36 ± 0.03 mg/kg. This study also showed that the Ni content of Algerian honey was comparable to that of multifloral Serbian honeys (range: 50.30–551.4 µg/kg) [52], but lower than Italian honeys analyzed by Meli et al. (2018) [57] (mean concentration: 0.56 ± 0.13 mg/kg) and higher than Italian honeys analyzed by Perna et al. (2021) [50] (mean concentration: 32.89 ± 16.02 µg/kg).

Only two samples (BM_T and E_T), both from the North Sahara region, were quantified for Sb: 0.53 ± 0.05 mg/kg and 0.55 ± 0.08 respectively. For the others, the levels were below the LOQ. This is not comparable with the study by Di Bella et al. [9], who found a mean Sb concentration of 0.09 ± 0.01 mg/kg for the Laghouat area.

Sn concentrations were quite variable (from <LOQ to 0.51 ± 0.07 mg/kg). According to the results obtained, this variation seems to be related to the different floral origins of the honeys. In fact, only a few types of honey showed a quantifiable Sn content: Eucaliptus globulus (EG_L sample), 0.34 ± 0.09 mg/kg; Ziziphus lotus (Z_T,L sample), 0.10 ± 0.03 mg/kg; Multifloral (M_T,L sample), 0. 34 ± 0.05 mg/kg; Echinops spp. (E_T and E_TD samples), 0.50 ± 0.04 mg/kg and 0.26 ± 0.05 mg/kg; and Bunium mauritanicum (BM_T sample), 0.51 ± 0.07 mg/kg.

3.6. Statistical Analysis

3.6.1. Geographical Origin

The results, presented in Table 2, Table 3 and Table 4, showed that significant differences were observed for only four variables: Ca, Cd, Cr and Pb. As a result, the data set subjected to PCA included all the experimental data coming from these four variables, and it was normalized to achieve independence of the element concentrations from the scale factors of the different variables.

First, the data eligibility for PCA was verified: the determinant value was 0.317; the Bartlett’s test of sphericity had a significance of 0.000, confirming that there was a patterned relationship among the variables; and the Kaiser–Meyer–Olkin measure (KMO) of sampling adequacy was 0.539, showing the sampling adequacy.

Two principal components (PCs) with eigenvalues exceeding one (2.060 and 1.110) were extracted according to the Kaiser Criterion, and they explained up to 79.25% of the total variance (i.e., 51.49%, and 27.76%, respectively). Figure 2 illustrates the bidimensional score (Figure 2a) and loading plots (Figure 2b). Overall, the score plot showed a satisfactory clusterization of honey samples in relation to the different geographical origin, with some exceptions. In fact, a clear distinction was observed between most honeys from the north and south of Algeria. However, for some samples this geographical distinction was not observed, and it was speculated that this could be due to the same botanical origin of the honeys analyzed (see Section 3.6.2).

However, by superimposing the bidimensional score and the loading plots, it could be observed that almost all the honeys from southern Algeria, located in the third quadrant of the bidimensional score, were characterized by a lower Ca and Cr content than the honeys from northern Algeria. On the other hand, honeys from northern Algeria did not show a homogenous distribution in the bidimensional space, being present in the first, second and fourth quadrants. This allowed us to observe how the concentration of Pb and Cd, and, therefore the possible contamination of the northern Algerian honey samples, varied according to the geographical origin of the product.

3.6.2. Botanical Origin

The significant differences in mineral concentrations values among the honey samples from different botanical origin were estimated by the Kruskal-Wallis test with significant p-level below 0.05. The results showed that significant differences were observed for ten variables: Fe, Mg, Zn, Cr, Mn, Pb, Cu, Cd, Ca, and Na. Four principal components (PCs) with eigenvalues exceeding one (3.081, 2.254, 1.563 and 1.339) were extracted according to the Kaiser Criterion, and they explained up to 82.37% of the total variance (i.e., 30.81%, 22.54%, 15.63 and 13.39%, respectively). In Figure 3, the bidimensional score (Figure 3a) and loading plots (Figure 3b) are shown. Botanical discrimination was not as satisfactory as geographical discrimination. However, from the superimposition of the bidimensional score and the loading plots, the distribution of the investigated honeys in relation to the element contents could be observed. For example, it may be argued that the honey from the botanical species Euphorbia orientalis (EO) was the one with the highest content of Mg and Fe, together with the multifloral honey (M). Multifloral honey (M) was also characterized by the highest concentrations of Na, Ca, Cd and Cu, together with Echinops ssp. (E) and Bunium mauritanicum (BM) honeys. In addition, the highest contribution of Zn was obtained for Eruca sativa (ES), Tamarix L. and Euphorbia orientalis (TE) and Ziziphus lotus (Z) honeys. The latter honey also had the highest concentrations of Cr, Mn and Pb, together with Eucaliptus globulus (EG) and Peganum harmala (PH) honeys.

3.7. Uptake of Elements by Honey Samples

Another aim of this study was to assess the potential toxicological risk to humans from honey consumption and, consequently, to determine the quality of these Algerian honeys. This was possible thanks to the calculation of estimated daily intake for essential elements (

Table 5. Experimental EDIs, RDA% (or AI%), and UL% calculated for the different Algerian honeys with respect to the reference values for essential elements.

		EDI (mg/d or µg/d *)		RDA [31] (mg/d or µg/d *)	AI [32] (mg/d)	AR [32] (mg/d)	PRI [32] (mg/d)	UL [32] (mg/d)	% RDA or AI	% UL
		North Sahara	South Sahara
Mg	Europe	9.54 × 10−2	1.15 × 10−1 a	375 a	350			250 a	0.03	0.05
	North Africa	1.59 × 10−2	1.92 × 10−2 a						0.01	0.01
Ca	Europe	1.82 × 10−1a	1.49 × 10−1	800 a		860	1000	2500 a	0.02	0.01
	North Africa	3.03 × 10−2a	2.49 × 10−2						0.00	0.00
Na	Europe	1.15 × 10−1	1.28 × 10−1 a		2000 a				0.01
	North Africa	1.92 × 10−2	2.13 × 10−2 a						0.00
Fe	Europe	2.70 × 10−2	2.88 × 10−2 a	14 a		6	11		0.21
	North Africa	4.50 × 10−3	4.80 × 10−3 a						0.03
Cu	Europe	1.19 × 10−3a	9.90 × 10−4	1 a	1.6			5 a	0.12	0.02
	North Africa	1.98 × 10−4a	1.65 × 10−4						0.02	0.00
Cr	Europe	3.78 × 10−1a*	1.98 × 10−1 *	40 a*					0.95
	North Africa	6.30 × 10−2 a*	3.30 × 10−2 *						0.16
Mn	Europe	9.00 × 10−5	3.06 × 10−4 a	2 a	3				0.01
	North Africa	1.50 × 10−5	5.10 × 10−5 a						0.00
Zn	Europe	1.76 × 10−2	1.98 × 10−2 a	10 a		7.5-9.3-11-12.7 b	9.4-11.7-14-16.3 b	25 a	0.20	0.08
	North Africa	2.93 × 10−3	3.30 × 10−3 a						0.03	0.01
Mo	Europe	n.d.	n.d.	45 * [66]				2 [66]
	North Africa	n.d.	n.d.
Li	Europe	n.d.	n.d.	1 [67]
	North Africa	n.d.	n.d.

^a RDA (or AI) and UL reference value used for the calculation of experimental RDA% and UL%; ^b ARs and PRIs for Zn are provided for four levels of phytate intake (LPI): 300, 600, 900, and 1200 mg/day. Abbreviations: EDI, estimated dietary intake; RDA, recommended dietary allowance; AI, adequate intake; AR, average requirement; PRI, population reference intake; UL, tolerable upper intake level; n.d., not detected. EDI and RDA in µg/d are indicated with *.

) and toxic and potentially toxic elements (

Table 6. Experimental EDIs, TDI% (or TWI%, BMDL01%, PTWI%), and UI% (or PMTDI) calculated for the different Algerian honeys with respect to reference values for toxic and potentially toxic elements.

		EDI (µg/kgb.w./d)		TDI (µg/kgb.w./d)	TWI (µg/kgb.w./w)	BMDL01 (µg/kgb.w./d)	PTWI (µg/kgb.w./w)	UI (µg/kgb.w./w)	PMTDI (µg/kgb.w./d)	% TDI or TWI or BMDL01 or PTWI	% UI or PMTDI
		North Sahara	South Sahara
As	Europe	2.57 × 10−2 a	1.54 × 10−2			0.3 a–8 [33]	15 [40]			8.57
	North Africa	4.30 × 10−3 a	2.60 × 10−3							1.43
Be	Europe	n.d.	n.d.	no reference values
	North Africa	n.d.	n.d.
Cd	Europe	3.60 × 10−2 a	1.21 × 10−2		2.5 a [35]		7 [42]			10.08
	North Africa	6.00 × 10−3 a	2.01 × 10−3							1.68
Co	Europe	1.03 × 10−3 a	7.71 × 10−4					1.6 a [37]			0.45
	North Africa	1.71 × 10−4 a	1.29 × 10−4								0.07
Cu	Europe	2.57 × 10−2 a	1.54 × 10−2				3500 a [68]		500 a [68]	0.005	0.005
	North Africa	4.29 × 10−3 a	2.36 × 10−3							0.001	0.001
Mn	Europe	1.29 ×10−3	4.37 × 10−3 a				2500 a [68]		360 a [68]	0.001	0.001
	North Africa	2.14 × 10−4	7.29 × 10−4 a							0.0002	0.0002
Ni	Europe	8.23 × 10−3 a	6.43 × 10−3	22 a [43]				2.8 a [38]		0.04	2.06
	North Africa	1.37 × 10−3 a	1.07 × 10−3							0.01	0.34
Pb	Europe	2.08 × 10−2 a	4.37 × 10−3			0.5 a [34]	25 [41]			4.16
	North Africa	3.47 × 10−3 a	7.29 × 10−4							0.69
Sb	Europe	1.39 × 10−2 a	n.d.					6 a			1.62
	North Africa	2.31 × 10−3 a	n.d.								0.27
Sn	Europe	9.26 × 10−3	6.69 × 10−3	no reference values
	North Africa	1.54 × 10−3	1.11 × 10−3
Ti	Europe	2.42 × 10−2	1.34 × 10−2	no reference values
	North Africa	4.03 × 10−3	2.23 × 10−3
Zn	Europe	2.52 × 10−1	2.73 × 10−1 a				7000 a [68]		1000 a [68]	0.03	0.03
	North Africa	4.20 × 10−2	4.54 × 10−2 a							0.005	0.005
Hg	Europe	n.d.	n.d.		4 [36]		5 [39]
	North Africa	n.d.	n.d.

^a TDI (or TWI, or BMDL01, or PTWI) and UI (or PMTDI) reference values used for the calculation of experimental TDI% (or TWI%, or BMDL01%, or PTWI%) and UI% (or PMTDI%). Abbreviations: EDI, estimated daily intake; TDI, tolerable dietary intake; TWI, tolerable weekly intake; BMDL01, benchmark dose lower confidence limit 01; PTWI, provisional tolerable weekly intake; UI, tolerable upper intake level; PMTDI, provisional maximum tolerable daily intake; n.d., not detected.

). The former was calculated considering FAOSTAT data [30], according to which the amount of honey consumed in Europe is 1.8 g/day, while in North Africa it is 0.3 g/day, and compared with dietary reference values for essential elements, while the latter was calculated by considering international safety reference values for toxic and potentially toxic elements. To calculate the EDIs, we considered a daily honey consumption for an adult of 70 kg.

For all the elements analyzed, the average exposure levels were very low in relation to the reference value. This result was probably due to the low amounts of honey consumed in European and North African diets (1.8 and 0.3 g/capita/day, respectively). The calculated EDIs (Table 6) were well below the intake levels of the regulated inorganic contaminants, indicating that Algerian honey can be safely consumed in the intended dietary amounts. However, two possible factors must be considered: -people who consume larger amounts of honey per day than reported in the FAOSTAT data; -and a regular consumer of honey, who consumes other contaminated foods in their diet, may exceed the exposure limits. Therefore, the continuous monitoring of mineral content in food is of paramount importance to protect the health of the consumer.

For the non-carcinogenic risk assessment, the HQ did not exceed the threshold of 1 for any contaminant potentially ingested through honey by adults in both European and North African diets, indicating that non-carcinogenic health effects from the consumption of these Algerian honeys are not significant.

4. Conclusions

In this study, the monitoring of the element profile of Algerian honey highlighted that most of the honeys analyzed exceeded the Codex Alimentarius limits for Mg, Fe, Zn, and Cd. In addition, the Pb content was above the EU regulatory limit for honey in almost all samples analyzed. Interestingly, the univariate and multivariate analysis of obtained data pointed out that honey samples could be discriminated more effectively in relation to the geographical provenance rather than the botanical origin.

Considering the risk assessment, the average intake or exposure to the elements resulting from the consumption of these Algerian honeys did not exceed the reference limits. Additionally, the HQ calculated for the non-carcinogenic risk did not exceed the threshold of 1 for any potentially toxic element. This result indicates that the non-carcinogenic health effects from the consumption of these Algerian honeys are negligible.

Hopefully, this study can provide greater impetus for the establishment of new international policies aimed at the harmonization of food safety standards, especially for toxic and potentially toxic elements. Furthermore, the continuous monitoring of not only inorganic xenobiotics but also other types of contaminants is fundamental and necessary to assess the level of contamination of honey and, consequently, of the environment in which this matrix is produced. This would allow not only to combat environmental pollution, but also to prevent a possible risk to the health of consumers.