Incidence of Fusarium Mycotoxins in Wheat and Maize from Albania

In this study, ten Fusarium toxins were analysed in wheat and maize commodities from Albania. In total, 71 samples of wheat and 45 samples of maize were collected from different producing regions. The analytical procedure consisted of a simple one-step sample extraction followed by the determination of toxins using liquid chromatography coupled with tandem mass spectrometry. Fusarium toxins were found in 23% of the analysed wheat samples and in 78% of maize samples. In maize samples, most often fumonisins B1 (FB1) and B2 (FB2) were found. They were present in 76% of samples. They were detected in all positive samples except in one with concentrations ranging from 59.9 to 16,970 μg/kg. The sum of FB1 and FB2 exceeded the EU maximum permitted level (4000 μg/kg) in 31% of maize samples. In wheat samples, the only detected Fusarium mycotoxin was deoxynivalenol (DON), present in 23% of samples. In one sample with the concentration of 1916 μg/kg, the EU maximum permitted level (1250 μg/kg) was exceeded. This is the first report on the presence of Fusarium toxins in wheat and maize grains cultivated in Albania.

Mycotoxins are relatively stable, and prevention methods to prevent mycotoxin contamination on the field before harvest, during harvesting, or storage have been developed, while their complete elimination from the contaminated material is difficult if not impossible [6][7][8][9][10]16,17,19,20]. However, sorting, cleaning, debranning, and thermal processing can significantly reduce mycotoxin concentrations in cereal products, whereas milling processes result in the repartitioning of mycotoxins in different milling fractions leading to a reduction in mycotoxin burden in the fractions used for human consumption [21,22]. The fate of different mycotoxins during these processes have been studied by several authors [22][23][24][25][26][27]. However, the results presented by Griessler et al. [20] reporting high contamination levels of FBs in complementary and complete feeding stuff prove that mycotoxins burden the fractions commonly used as animal feed [21,22,26]. Mycotoxin contamination in different food and feed commodities is an ongoing global threat and will be more significant due to the climate changes and increased exchanges in the food and feed global market [11,15,19]. Exposure to these compounds is a significant threat to human and animal health and is associated with different acute or chronic mycotoxicoses [1,6,7,9,15,16,19,28].
This study aimed to get an insight into the occurrence of ten Fusarium mycotoxins (DON, 3-AcDON, 15-AcDON, ZEA, FB1, FB2, T-2, HT-2, DAS, and NEO) in two primary cereal commodities produced in Albania, maize, and wheat, both used as food and feed [51]. The study is the first-ever report on the presence of these mycotoxins in crop commodities from Albania. The data will expand the information on the occurrence of these mycotoxins in different grains from Southern Europe. We expected similar occurrence rates and concentrations as reported from the Southern European and Central European countries.

Occurrence of Fusarium Toxins in Samples
The samples containing one or more individual Fusarium toxins at concentrations above limit of quantification (LOQ) were considered as positive. Altogether, 44% of the analysed wheat and maize samples were contaminated. In the year 2014, the contamination rate was 58%, while in 2015 it was 26%.
In winter wheat samples from all regions, DON was the only detected Fusarium toxin, while the concentrations of the other nine Fusarium toxins were below LOQ (50 µg/kg). The number and the percentage of positive samples, the mean value of positive samples, the median, the minimum, and maximum determined concentrations are presented in Table 1. The maximum permitted level of 1250 µg/kg given in Regulation (EC) No 1126/2007 [29] for unprocessed cereals other than durum wheat, oats, and maize intended for human consumption was exceeded in only one of the wheat samples from the year 2015 with a DON concentration of 1916 µg/kg.
The contamination rate in maize was much higher than in wheat. In Table 2, the incidence, mean value of positive samples, median, minimum, and maximum concentrations of Fusarium mycotoxins are presented for maize commodities. The main mycotoxins detected were FB1 and FB2. In the year 2015, they were the only mycotoxins found in maize. FB1 was present in all contaminated samples but one. In two-third of cases, both FB1 and FB2 were present, while in one-third of samples, only FB1 was detected. No sample contained only FB2. In the samples collected in 2014, DON, ZEA, and T-2 were also detected. ZEA was found in two samples from Kruja, and the T-2 toxin was detected in one sample from the Korça region. The results were in line with the global risk maps presented by Battilani and Logrieco [52]. According to the maps, Albania is among the countries with a low/absent risk regarding the DON contamination of wheat at harvest and among the countries with a high global risk of fumonisin contamination of maize. However, in the samples of barley and rye, no Fusarium toxins were detected. Concerning the sum of FB1 and FB2, the maximum permitted level of 4000 µg/kg laid down in Regulation (EC) No 1126/2007 [29] for unprocessed maize was exceeded in 12 samples (39%) in 2014 and two samples (14%) in 2015, altogether in 31% of maize samples. In six samples from the year 2015, the concentration was even higher than 10,000 µg/kg. The concentrations of DON, ZEA, and T-2 toxin in maize samples were lower than the maximum permitted levels and the EU indicative level for the sum of T-2 and HT-2 in unprocessed maize intended for human consumption (1750 µg/kg, 320 µg/kg, and 200 µg/kg, respectively) given in Regulation (EC) No 1126/2007 and Commission Recommendation 2013/165/EU [29,31].
The data on the presence of DON in winter wheat and maize according to the regions are shown in Tables 3 and 4, respectively. The incidence in wheat was similar in two western regions, Fieri and Lushnja, situated along the Adriatic Sea and the Elbasan region in the inner part of the country, but it was considerably lower in the eastern region Korça with the temperate continental climate (Table 3). The concentrations were below the maximum permitted level of 1250 µg/kg, except in one sample from Elbasan, where the concentration was 1916 µg/kg. The incidence of DON in the maize varied from 0% (Fieri and Lushnja) to 64% (Kruja) ( Table 4). All the concentrations were far below the maximum permitted level of 1750 µg/kg. However, the incidence and concentrations of FBs were higher. In Fieri, Lushnja, and Kruja, the mean concentrations (sum of FB1 and FB2) were higher than the maximum permitted level (4000 µg/kg). Concerning the median values of positive samples, the order of the regions was Fieri > Lushnja > Kruja > Elbasan > Korça, situated from the west of the country with a hot-summer Mediterranean climate to the east of the country belonging to the continental climate. However, the highest concentration was determined in a sample from Korça. As mentioned before, the maximum permitted concentration was exceeded in 14 samples. Most of these samples were from Lushnja and Kruja, where 50% and 36% of samples, respectively, were non-compliant. However, the incidence of contaminated maize samples containing at least one mycotoxin was similar in the Fieri, Lushnja, and Korça regions (57%, 70%, and 64%, respectively), while it was higher in the Kruja and Elbasan regions (91% and 100%, respectively).
The co-occurrence of two or more Fusarium toxins was detected only in maize samples. Of 34 contaminated samples, five samples (15%) contained one mycotoxin, 19 samples (56%) contained two mycotoxins, nine samples (26%) contained three mycotoxins, and one sample (3%) contained four mycotoxins. This finding is consistent with the results presented by Jakovac-Strajn et al. [34], Ibáñez-Vea et al. [38], Stanković et al. [40], Juan et al. [43], Alkadri et al. [46], and Kirinčič et al. [47], who reported a high percentage of samples containing more than one mycotoxin. FB1 and FB2 co-occurred most often. They were present together in 18 of the 19 samples containing two mycotoxins. While in eight of the nine samples containing three mycotoxins, DON, FB1, and FB2 co-occurred, in one sample, DON, ZEA, and FB1 were present. In the sample with four mycotoxins, DON, ZEA, FB1, and FB2 were present. However, the most common co-occurrence of DON, 3-AcDON, and 15-AcDON reported by Ibáñez-Vea et al. [38] and Van Der Fels-Klerx et al. [41] was not recognized in our study.
The incidence rate and the mean concentrations of mycotoxins were higher in the year 2014 than in 2015 in both commodities. However, the mycotoxin pattern was different between wheat and maize, which could be dependent on Fusarium species that infect these two cereals [4]. Both the production of mycotoxins and Fusarium profile are dependent on several factors, primarily climatic conditions, particularly rainfall and temperature at the flowering stage, but also agronomic factors, such as tillage, nitrogen fertilization, use of fungicides, crop rotation, and host genotype [4,16,17]. These data were not collected in the study and therefore no conclusions can be drawn on the correlation of the contamination rate and mycotoxin pattern on geographical and climatic conditions. However, the available meteorological data show that the annual average temperatures and precipitations were above the long-term average in all regions in both years. Furthermore, the incidence of mycotoxins in the samples from the Korça region seems lower than with samples from the other investigated western plain regions close to the Adriatic Sea. Considering that the level of agricultural development was similar in all investigated regions, the lower incidence can be explained by the climate differences between the temperate climate Korça and the typical Mediterranean climate regions of the western part of the country.
The maximum level of DON in the wheat commodity in our study was much lower than in a significant number of reports from other countries. The highest reported levels were 10,000 µg/kg in a sample from the Netherlands [41], 5865 µg/kg and 5510 µg/kg in samples from Finland [18,41], 3306 µg/kg in a sample from Serbia [40], and 3700 µg/kg [34] and 3070 µg/kg [47] in samples from Slovenia. However, in some studies from Poland, Croatia, Serbia, and Spain, respectively, the highest reported levels were at 100 µg/kg [48], 278 µg/kg [44], 309 µg/kg [37], and 437 µg/kg [35].
The incidence rate of DON in maize samples (24%) was comparable with incidence data for maize commodity from Slovenia (35.3%) [47] or Romania (42.9%) [49], but lower than the incidence rates of 71%-87.9% in this crop commodity in other studies reporting data from Slovenia [34], Spain [35]), Northern Europe [41] and Croatia [44] ( Table 6). The highest determined level of DON (799 µg/kg) was comparable to the highest reported level in the study from Spain (580 µg/kg) [35], but lower than in all other studies in the range of 1269.9 to 14,420 µg/kg.
The incidence rate of ZEA in maize samples was 4.4%, similar to data from Romania (7.1%) [49], but lower than in all other studies where the occurrence rate of 13.3%-78% was given ( Table 6). The highest determined level of ZEA in our study (263 µg/kg) was higher than the data reported by Manova and Mladenova [33] and Gagiu et al. [49], but much lower than in all other studies given in Table 6 (611-1000 µg/kg). The contamination rate of T-2 was similarly low, as in the reports of Jakovac-Strajn et al. [34] and Cano-Sancho et al. [35]; however, Pleadin et al. [44] reported a contamination rate of 57%.
The incidence of FBs (sum of FB1 and FB2) in our study (76%) was found to be similar with the published data on the incidence from Croatia (90%) [44] but was higher than in the other studies [34,47]. The highest determined level (16,970 µg/kg) is comparable to the concentration reported by Kirinčič et al. [47]. However, both values are much higher than those reported elsewhere [33,34,44].

Sample Collection
Samples were collected from different regions, taking into account the country's geography and production. According to the Food and Agriculture Organization (FAO) of the United Nations database [51], which provides data relating to food and agriculture for countries worldwide, the yearly production of wheat, maize, barley, and rye in Albania is around 275,000 tonnes, 380,000 tonnes, 7500 tonnes, and 3000 tonnes, respectively. Correspondingly, mainly wheat and maize samples were collected, but also a few samples of barley and rye. Winter wheat and maize were sampled after their respective harvesting seasons from five main agriculture regions in Albania: Fieri, Lushnja, Kruja, Elbasan, and Korça. The sampling of wheat commodity in the regions of Fieri, Lushnja, Elbasan  Acetonitrile, methanol, acetic acid (Sigma-Aldrich, Steinheim, Germany), and ammonium acetate (Merck, Darmstadt, Germany) were p.a. or LC-MS grade purity. Deionized water was prepared using a Milli-Q system (Millipore, Bedford, MA, USA).

Sample Preparation
For the simultaneous determination of mycotoxins (DON, 3-AcDON, 15-AcDON, ZEA, FB1, FB2, T-2, HT-2, DAS, and NEO), a procedure described in detail by Topi et al. [54] was used. The procedure consisting of the extraction of mycotoxins from ground cereal samples and liquid chromatography-tandem mass spectrometry (LC-MS/MS) was based on the analytical procedures of Rasmussen et al. [55], Lattanzio et al. [56] and Schenzel et al. [57]. Samples were ground to a particle size of 1 mm using a laboratory mill Retsch ZM 100 (Haan, Germany). Ten grams of a sample were shaken with 100 mL of an acetonitriledeionised water mixture (84 + 16) for 1 h using an IKA HS 501 digital linear shaker (IKA Labortechnik, Staufen, Germany). A total of 4 mL of the filtered extract was evaporated under vacuum to dryness using a Syncore Polyvap system (Büchi, Flawil, Switzerland). For mycotoxin concentrations above the calibration range, the filtered extracts were diluted for further work. The dry residue was reconstituted in 0.5 mL of a methanol-deionised water mixture (20 + 80). An aliquot-10 µL of the solution-was injected into the UPLC-MS/MS system (Acquity UPLC H Class system) coupled with a triple-quadrupole mass spectrometer (Xevo TQ MS) equipped with an electrospray ionization (ESI) interface and MassLynx software for data collection and processing (Waters, Milford, MA, USA). The vials were kept in the autosampler at 15 • C. For the matrix-matched calibration, 4 mL portions of the filtered extracts were spiked with the appropriate amounts of standard solutions and prepared along the samples.

LC-MS/MS Analysis
For the LC-MS/MS determination, the conditions reported by Topi et al. [54] were applied. Chromatographic separation was performed on a Zorbax Eclipse Plus C18 Rapid Resolution HD column, 2.1 × 100 mm, 1.8 µm (Agilent, Santa Clara, CA, USA). The mobile phase consisted of two components mixed in gradient mode. Component A was deionized water and component B was methanol, both containing 0.5% acetic acid and 2.5 mM ammonium acetate. The starting composition of the eluent was 95% A and 5% B. The portion of component B was linearly increased to 40% within 4 min and further increased to 70% within the next 8 min. This latter composition was held for 4 min, and then component B was increased to 90% in 1.5 min. The proportion of component B was held at 90% for 2.5 min and then returned back to 5% in 1 min. The final composition was held for 4 min. The mobile phase flow rate was 0.3 mL/min, and the column temperature was 40 • C. MS/MS analysis was carried out in multiple reaction monitoring (MRM) mode switching between positive and negative ionisation mode during a single run. The capillary voltage was 3.4 kV (ESI+) and 3.0 kV (ESI−), the desolvation temperature was 500 • C, the ion source temperature was 150 • C and the collision cell voltage was 20 V. Specific MS/MS parameters related to determined mycotoxins (retention times, ionisation mode, and monitored transitions) are presented in Table 7.

Method Validation
The limit of detection (LOD) of the single analytes was determined at a signal-to-noise ratio of 3:1. A value 3.3 times the LOD was selected as the LOQ. The recoveries and precision were tested using maize and wheat samples spiked with Fusarium toxins at the concentration levels of 50, 100, and 500 µg/kg.
The linearity of the method was tested in the concentration range of 50-500 µg/kg using matrix-matched standard solutions analysed in triplicates. Good linearity was proven for all analytes with correlation coefficients higher than 0.997. The accepted limit of detection (LOD) and the limit of quantification (LOQ) of all single Fusarium toxins were 15 µg/kg and 50 µg/kg, respectively. The mean recoveries of single toxins determined in maize at the tested concentration levels were between 90% and 117%. The recoveries were between 87% and 112% in wheat, except for DON, which was 124%. The reproducibility expressed as RSD R was less than 16% for all Fusarium toxins in maize and ≤30% for all Fusarium toxins in wheat. The reproducibility and mean recoveries of the toxins were in line with the criteria given in Commission Regulation (EC) No 401/2006 and its amendments [53] except the recovery of DON which slightly exceeded the required value.

Conclusions
In the study, the results of determination of ten Fusarium toxins in 125 samples from two seasons were obtained. They represent the very first insight into their occurrence in cereal commodities from Albania and a contribution to the knowledge on the issue in southern Europe.
Relevant Fusarium toxins in the region seem to be DON and FBs. Other toxins were detected in only a few samples (ZEA, T-2) or not at all (3-AcDON, 15-AcDON, HT-2, DAS, and NEO). The incidence was comparable with those reported in the neighbouring countries, but the FB concentrations in maize were significantly higher than reported elsewhere.
The incidence of mycotoxins in the samples from the Korça region seem lower than with the samples from the other investigated western plain regions close to the Adriatic Sea. Considering that the level of agricultural development is similar in all investigated regions, the lower incidence can be explained by the climate differences between temperate climate Korça and typical Mediterranean climate regions of the western part of the country.
A significant difference between the data from the years 2014 and 2015 indicates that data from further harvesting years need to be provided to adequately characterize the occurrence of Fusarium toxins in cereal grains in Albania. However, with regard to the incidence rates and the concentrations of DON and FBs, farmers should consider all principles of good agricultural practices including tillage, crop rotation, cultivar selection, planting date, irrigation and fertilisation regimes, insecticide/fungicide treatments, harvest timing, as well as drying, cleaning, segregation, and the storage of cereals under controlled conditions in order to reduce mycotoxin contamination and to ensure safe food and feed.