Microbiological and mycotoxicological analyses of processed cereal-based complementary foods for infants and young children from the German market.

This study investigated several food safety criteria in 38 different commercial products of processed cereal-based foods (PCF) from the German market. Microbiological assessment, followed by 16S RNA gene sequencing of suspect colonies, included aerobic mesophilic bacteria, moulds, Enterobacteriaceae, Cronobacter spp., and presumptive Bacillus cereus. Mycotoxin analyses were performed by enzyme immunoassays for deoxynivalenol (DON), zearalenone (ZEN), T-2/HT-2 toxins (T-2/HT-2; oat containing products only), ergot alkaloids (EA), and alternariol (AOH). No violative result above existing European Union regulations or international guidelines was obtained. Most samples had very low aerobic mesophilic cell counts (<2.0 × 101 CFU/g), the maximum was 9.6 × 102 CFU/g. A few samples contained low numbers of opportunistic pathogens, most notably Cronobacter sakazakii, Acinetobacter spp., Pantoea spp., and enterotoxigenic Bacillus wiedmannii. Levels of mycotoxin contamination were very low, well below European Union maximum limits. DON was found in 10 samples, at levels of 9-35 µg/kg. T-2/HT-2 were found in all 15 oat-based products (1-8 µg/kg). All samples were negative for ZEN and EA. A high number (n = 25) of samples yielded weakly positive results for the nonregulated AOH (0.4-2 µg/kg), but just three samples exceeded a level of 1 µg/kg. No relationship between cereal composition and analytical findings for microbiological parameters and mycotoxins could be found. As long as PCF meals are freshly prepared and consumed immediately after preparation, the risk from sporadically occurring opportunistic bacteria appears to be minimal.

[Investigations on the excretion of antibiotics in goats' milk following therapeutic application]. / Untersuchungen zum Ausscheidungsverhalten von Antibiotika in Ziegenmilch nach therapeutischer Applikation.

OBJECTIVE: Data on the excretion of antibiotic residues following therapeutic drug dosages in lactating goats with clinical signs of bacterial infections are currently lacking. Therefore, this study aimed at monitoring the drug residue excretion of a subset of frequently used antibiotics in the milk of dairy goats following their therapeutic administration. MATERIAL AND METHODS: Over a period of 4 months, milk samples (udder halves) were collected in 2 goat milk farms from animals treated with antibiotics in routine practice based on the diagnosis of a bacterial infection. The samples were examined up to 3 days following the withdrawal time point. The animals were classified in 3 groups depending on their clinical symptoms and treatment. Goats in group 1 (afebrile goats with various bacterial infections excluding the udder) were treated with intramuscular amoxicillin injection (n = 5). Animals in group 2 (mastitis catarrhalis) were treated with intramammary injection of oxacillin and ampicillin (n = 6). Group 3 consisted of a single goat diagnosed with mastitis. This individual was treated with cefquinome in accordance with the results of the antibiogram. Milk samples were examined qualitatively by using a receptor assay (Betastar®) as well as a microbiological inhibitor assay (Brilliant black reduction test, BRT). The latter assay was also used to semiquanti-tatively analyse drug residue levels in samples from group 2. RESULTS: Following intramuscular treatment with amoxicillin, drug residue levels were estimated to be very similar in both udder halfs. Elimination was complete 3 days after the end of the treatment period. Animals in group 2 showed significant differences between treated and untreated udder halves. However, the untreated halves still exhibited residue levels exceeding the maximum residue limits during the treatment period. In both group 2 and 3, all milk samples were tested negative for drug residues before the end of the withdrawal period. CONCLUSION: In the present study, no evidence of prolonged residue excretion into milk of goats following therapeutic administration of antibiotics was detected. Both the receptor test and the BRT represent suitable methods for an efficient antibiotic drug residue testing in goat milk. Reliable testing was merely not attainable in cases of milk samples possessing highly altered organoleptic characteristics.

Detection of Total Ergot Alkaloids in Cereal Flour and in Bread by a Generic Enzyme Immunoassay Method.

Four sets of polyclonal antibodies against ergot alkaloids ergometrine, ergotamine, α-ergocryptine, and ergocornine were produced and characterized in a competitive direct or indirect enzyme immunoassay (EIA). Standard curve LODs were 0.03 ng/mL (ergometrine EIA) to 2.0 ng/mL (ergocornine EIA). Three EIAs were highly specific, whereas the ergometrine EIA had a broad specificity pattern and reacted, albeit weakly, with all seven major ergot alkaloids and their epimeric forms. Using the ergometrine EIA, a generic test system was established in which total ergot alkaloids are quantified by a standard curve for a toxin mixture composed of three alkaloids that matched the ergot alkaloid composition in naturally contaminated rye and wheat products. Sample extraction with acetonitrile-phosphate-buffered saline at pH 6.0 without further cleanup was sufficient for EIA analysis. The LODs for total ergot alkaloids were 20 ng/g in rye and wheat flour and 14 ng/g in bread. Recoveries were 85-110% (RSDs of 0.1-11.7%) at a concentration range of 50-1000 ng/g. The total ergot alkaloid EIA was validated by comparison with HPLC-fluorescence detection. Although some under- and overestimation by the total ergot alkaloid EIA was observed, it was suitable for the reliable identification of positive samples at 10-20 ng/g and for the determination of total ergot alkaloids in a concentration range between 100 and 1000 ng/g.

Immunochemical Analysis of Paxilline and Ergot Alkaloid Mycotoxins in Grass Seeds and Plants.

Limited availability of toxin standards for lolitrem B and ergovaline impedes routine control of grasses for endophyte toxins. This study aimed at assessing the applicability of an enzyme immunoassay (EIA) for the indole-diterpene mycotoxin paxilline, in combination with a generic EIA for ergot alkaloids, as alternative parameters for screening purposes. Analysis of grass seeds and model pastures of four different grass species showed that both EIAs yielded highly positive results for paxilline and ergot alkaloids in perennial ryegrass seeds. Furthermore, evidence for natural occurrence of paxilline in grass in Germany was obtained. High performance liquid chromatography-tandem mass spectrometry analysis qualitatively confirmed the paxilline EIA results but showed that paxilline analogues 1'-O-acetylpaxilline and 13-desoxypaxilline were the predominant compounds in seeds and grass. In the absence of easily accessible reference standards for specific analysis of some major endophyte toxins, analysis of paxilline and ergot alkaloids by EIA may be suitable substitute parameters. The major advantage of this approach is its ease of use and speed, providing an analytical tool which could enhance routine screening for endophyte toxins in pasture.

Detection of the tremorgenic mycotoxin paxilline and its desoxy analog in ergot of rye and barley: a new class of mycotoxins added to an old problem.

A newly developed enzyme immunoassay (EIA) for the detection of the tremorgenic indole-diterpene alkaloid paxilline (PAX) and closely related analogs was used to analyze ergot sclerotia collected from rye and barley fields. The mean EIA standard curve detection limit was 0.47 ± 0.14 ng/mL; relative cross-reactivity of toxin standard solutions was found for 11-hydroxy-paspaline (terpendole E, 1.1%) but not for lolitrem B or ergot alkaloids. Sclerotia from all fields were positive in the PAX-EIA at concentration levels of 620 ± 200 and 160 ± 37 µg/kg in ergot of rye and 130 ± 47 µg/kg in ergot of barley. Confirmatory analyses of sclerotia by liquid chromatography-tandem mass spectrometric detection identified PAX and its analog 13-desoxypaxilline. To the best of our knowledge, this is the first report on the natural occurrence of tremorgenic indole-diterpene alkaloid mycotoxins in ergot sclerotia from rye and barley. Along with details on the analytical methodology developed in this study, particularly PAX-antibody production, the relevance and implications of these findings for food and feed safety are discussed. Presence or absence of elevated levels of tremorgenic mycotoxins, along with the ergot alkaloids, would help in explaining the difference between the two distinct manifestations of historic ergotism, the convulsive and the gangrenous form. Further method development for paxilline and other tremorgenic mycotoxins in cereals used for food and feed is a prerequisite for a comprehensive risk assessment, which seems to be necessary in light of the findings reported here. Paxilline in ergot of rye.

Reassessment of Cronobacter spp. originally isolated as Enterobacter sakazakii from infant food.

Cronobacter spp. cause infant disease, several cases have been associated with powdered infant formulae (PIF). In the early 2000s, contamination of German PIF with these opportunistic pathogens was quite common. Before 2008, all isolates Cronobacter spp. had been classified as Enterobacter sakazakii, therefore little is known about species diversity within such isolates. Genetic, serologic, and biochemical traits of 80 Cronobacter isolates, originally obtained 2003-2006 within infant food surveys in Germany, were reassessed in this study. By sequencing of the fusA gene, all isolates were unambiguously assigned to two species, C. sakazakii (n = 73) and C. malonaticus (n = 7). PCR serotyping identified five C. sakazakii serotypes and two C. malonaticus serotypes, biochemical profiling yielded five biogroups. PFGE analysis also showed high heterogeneity in both species. Multilocus sequence typing of 26 selected isolates yielded 16 different sequence types (ST), including C. sakazakii ST 1 (n = 6) and the highly virulent ST 4 (n = 2). The results suggest that just two, but highly heterogeneous species were responsible for the Cronobacter contamination problem which challenged the German PIF industry in the beginning of this century. This fact may have influenced the success of efforts to identify and eliminate sources of contamination.

Extended-spectrum ß-lactamase producing Enterobacteriaceae in bulk tank milk from German dairy farms.

Although the dairy farm environment is a known source of extended-spectrum ß-lactamase (ESBL)-producing bacteria, surveillance data on ESBL in the milk production chain are still scarce. This study aimed at estimating the dimensions of the problem for public health and animal welfare by surveying ESBL-producing Enterobacteriaceae in raw bulk tank milk in Germany. Samples from 866 dairy farms, comprising about 1% of the total number of dairy farms in Germany, were first screened for presence of cefotaxime-resistant bacteria by selective enrichment. Suspect colonies were identified phenotypically and further characterized by biochemical and molecular methods, including analysis of resistance genes and clonal diversity in ESBL-producing isolates. Bulk tank milk from 82 (9.5%) farms yielded Enterobacteriaceae with confirmed ESBL-production. The most frequent ESBL-producing species was Escherichia coli (75.6%), followed by Citrobacter spp. (9.6%), Enterobacter cloacae (6.1%), and Klebsiella oxytoca (3.7%), a few isolates belonged to other species within the genera Hafnia, Raoutella and Serratia. The majority of isolates (95.1%) harbored the ß-lactamase blaCTX-M gene, which has gained increased importance among ESBL-producing strains worldwide; the CTX-M group 1 was found to be the dominating (88.4%) phylogenetic group. All ESBL-positive Escherichia coli isolates were clonally heterogeneous, as determined by pulsed-field gel electrophoresis. The results from this survey demonstrate that ESBL-producing bacteria are distributed widely in the dairy farm environment in Germany. Therefore, raw milk is a potential source of exposure for the consumer, which is of increasing importance considering the trend of farmer-to-consumer direct marketing. Furthermore, dairy farm staff have an increased likelihood of exposure to ESBL-producing bacteria. Finally, ESBL-producing bacteria may also be transferred via waste milk to calves, thus further spreading antibiotic resistance in the farm environment.

CTX-M producing Escherichia coli isolated from cattle feces in Bogor slaughterhouse, Indonesia

Objective: To determine the occurrence of CTX-M producing Escherichia coli (E. coli) from cattle feces in Bogor slaughterhouse, Indonesia. Methods: A total of 220 cattle feces samples were collected from Bogor slaughterhouse from March to April 2015. Presence of extended-spectrum beta-lactamase (ESBL) producing E. coli was detected by disc diffusion test based on the recommendation from Clinical and Laboratory Standards Institute (2014). Bacterial strains which were confirmed as producing ESBLs were further analyzed for the presence of bla genes of the ESBL by PCR. Results: The results showed that CTX-M producing E. coli isolates were detected in 19 samples from 220 samples (8.6%). The b-lactamase genes detected were CTX-M-1 (n = 10) and CTX-M-9 (n = 9). All of the CTX-M producing E. coli isolates showed multidrug resistance phenotypes to at least four antibiotics. The highest incidence of an-tibiotics resistance was showed to ampicillin (100.0%), cefotaxime (100.0%), and cef-podoxime (100.0%), followed by streptomycin (84.3%), trimethoprim-sulfamethoxazole (73.7%), erythromycin (52.6%), kanamycin (26.3%), doxycycline (10.5%), and ceftazi-dime (0.0%). Conclusions: Detection of CTX-M-producing E. coli in cattle feces raises important questions as they can represent a potential risk factor to public health.

Microbiological Quality of Raw Dried Pasta from the German Market, with Special Emphasis on Cronobacter Species.

The microbiological quality of 132 dried pasta products available on the German market, originating from 11 different countries, was studied. Sample materials included soft or durum wheat products, some of which produced with other ingredients such as eggs, spices, or vegetables. Parameters included hygiene indicators (aerobic plate count, mold count, the presence of Enterobacteriaceae) and pathogenic/toxinogenic bacterial species (Salmonella spp., Staphylococcus aureus, presumptive Bacillus cereus, and Cronobacter spp.). The overall results of hygiene parameters indicated a satisfactory quality. Salmonella was not found in any sample. Three samples were positive for S. aureus (10(2) to 10(4) colony forming unit (CFU)/g). Presumptive B. cereus at levels of 10(3) to 10(4) CFU/g were detected in 3 samples. Cronobacter spp. were isolated from 14 (10.6%) products. Of these, 9 isolates were identified as C. sakazakii, 2 each as C. turicensis and C. malonaticus, and 1 as C. muytjensii. The isolates were assigned to 9 multilocus sequence typing (MLST) sequence types and to 14 different PFGE profiles. Although pasta products are typically cooked before consumption, some consumers, and children in particular, may also eat raw pasta as nibbles. Raw pasta seems to be a relevant source of exposure to dietary Cronobacter spp., although health risks are probably restricted to vulnerable consumers. High numbers of presumptive B. cereus as found in some samples may be a risk after improper storage of cooked pasta products because toxinogenic strains are frequently found within this species.

Immunochemical analysis of fumigaclavine mycotoxins in respiratory tissues and in blood serum of birds with confirmed aspergillosis.

The ergoline alkaloid fumigaclavine A (FuA) is one of the major mycotoxins produced by Aspergillus fumigatus, the main causative fungal agent of avian aspergillosis. To study in situ production of FuA, post-mortem respiratory tissues of various avian species, as well as blood samples of falcons (Falco sp.), were analysed by enzyme immunoassay (EIA). At a detection limit of 1.5 ng/ml, FuA EIA positive results were obtained for tissue samples from seven (64%) out of 11 birds with confirmed aspergillosis, with a maximum concentration of 38 ng/g, while all controls (n = 7) were negative. No FuA could be detected in blood serum (detection limit 0.7 ng/ml) of 15 falcons, experimentally inoculated with A. fumigatus conidia. Fungal mycelium material from tissue of clinical aspergillosis cases, cultured on malt extract agar, was highly positive in the FuA EIA in milligrams per gram range. Chromatographic analysis of mycelium extracts revealed the co-presence of FuA and the structurally related fumigaclavine C (FuC). Alkaline hydrolysis of FuA and FuC yielded the corresponding deacetylation products, FuB and FuE. This is the first report showing that fumigaclavine alkaloids are produced by A. fumigatus in situ during the course of clinical aspergillosis in birds; however, the role of these compounds in the pathogenesis of this disease is still unknown.

Extended-Spectrum ß-Lactamase (ESBL)-Producing Klebsiella pneumoniae in Bulk Tank Milk from Dairy Farms in Indonesia.

Bulk tank milk from 80 dairy farms located in the West Java Region of Indonesia was analyzed for the presence of extended-spectrum ß-lactamase (ESBL)-producing Enterobacteriaceae. Isolates from seven dairy farms were ESBL positive, and all were identified as Klebsiella pneumoniae. The isolates showed ESBL-characteristic antibiotic resistance patterns. Further analysis revealed that all K. pneumoniae isolates harbored the blaSHV gene, and two isolates were additionally positive for the blaTEM-1 and blaCTX-M-15 genes. Isolates from different farms were clonally diverse according to macrorestriction analysis. The results indicate that the relatively high frequency of ESBL-producing K. pneumoniae in bulk tank milk implies the risk that milk is both a source of local exposure and a vector contributing to the supraregional spread of antibiotic-resistant bacteria by trade.

A broadly applicable approach to prepare monoclonal anti-cephalosporin antibodies for immunochemical residue determination in milk.

A simple, efficient and rapid method for the synthesis of cephalosporin-protein conjugates was established. These conjugates were used as immunogens to produce monoclonal antibodies (mAbs) and as solid phase antigens in competitive indirect enzyme immunoassays (EIAs). With this generic approach, a novel set of monoclonal antibodies for cephalosporins was prepared, including ceftiofur and cephalexin as well as, reported here for the first time, cefoperazone, cefquinome and cephapirin. All 5 EIAs were highly sensitive, with standard curve IC(50) values of 0.7 (ceftiofur), 1.1 (cefquinome), 5.2 (cephalexin), 13.8 (cefoperazone) and 40.3 ng mL(-1) (cephapirin). Detection limits (IC(30)) ranged from 0.3 (ceftiofur mAb 1D7) to 17.2 ng mL(-1) (cephapirin mAb 2F10). Specificity studies revealed that cephalosporin-antibody binding was strongly determined by the side chain residues of the cephem nucleus. Therefore all mAbs, to some extent, recognized other beta-lactam antibiotics with similar side chain residues. Within the group of cephalosporins approved for use in veterinary medicine, however, the final EIAs were highly selective for their respective antigen, except for the ceftiofur EIA which showed cross-reactions with cefquinome. The applicability of the five assays for drug residue testing in milk was demonstrated. In each EIA the target drug could be determined in milk with high accuracy and precision at concentrations far below the European Union maximum residue limits.

Enzyme immunoassay for tenuazonic acid in apple and tomato products.

The Alternaria mycotoxin tenuazonic acid was derivatized with succinic anhydride and conjugated to keyhole limpet hemocyanin (KLH) and to horseradish peroxidase (HRP), respectively. The KLH conjugate was used to produce polyclonal antibodies in rabbits. A competitive direct enzyme immunoassay (EIA) for tenuazonic acid was established, which was moderately sensitive for tenuazonic acid [50% inhibition concentration (IC(50)): 320 ± 130 ng/mL] but strongly reacted with tenuazonic acid acetate (IC(50): 23.3 ± 7.5 ng/mL). Therefore, an optimized EIA protocol was established, which employed acetylation of standard and sample extract solutions. The mean standard curve detection limit (IC(30)) for tenuazonic acid acetate was 5.4 ± 2.0 ng/mL, enabling detection limits for tenuazonic acid in apple and tomato products of 25-50 ng/g (150 ng/g in tomato paste). Recoveries in a concentration range of 50-2000 ng/g were 60-130% in apple juice and tomato juice and 40-150% in other tomato products. Tenuazonic acid was detected in apple juice and tomato products from German retail shops at levels of 50-200 ng/g. In conclusion, this novel EIA for tenuazonic acid could be useful within a screening program for Alternaria mycotoxins in food.

Widespread occurrence of low levels of alternariol in apple and tomato products, as determined by comparative immunochemical assessment using monoclonal and polyclonal antibodies.

This study investigated the production of polyclonal (pAB) antibodies and the first time production of monoclonal (mAB) antibodies against the mycotoxin alternariol, and their implementation in enzyme immunoassay (EIA) for the rapid determination of alternariol in foods. Both EIAs were highly sensitive, with detection limits (IC20) of 35 ± 6.9 pg/mL (mAb EIA) and 59 ± 16 pg/mL (pAb EIA). Food products (n = 109; apple and tomato products, white wine) from German retail shops were analyzed. At a detection limit of 1-2 µg/kg, alternariol at 1-13 µg/kg was found with high frequency in apple (67%) and tomato (93%) products. Tomatoes with visible signs of Alternaria infection, stored at room temperature for up to 4 weeks, contained alternariol at levels up to 50 mg/kg, as determined by EIA and HPLC-FLD. It is concluded that the alternariol immunoassays present a versatile screening tool which could facilitate food control for Alternaria toxins.

A coagulase-negative variant of Staphylococcus aureus from bovine mastitis milk.

Bacteriological analysis of milk samples from quarters of a dairy cow suffering from subclinical mastitis yielded two isolates of Staphylococcus aureus which gave a negative reaction in the standard coagulase test. Both isolates were also clumping factor and thermonuclease negative, and gave a negative reaction in the Staphaurex® test. The isolates were identified by using commercial biochemical systems, and by PCR analysis of different staphylococcal cell surface protein and exoprotein genes. Further molecular identification of the isolates, which included sequencing of the 16S rRNA gene and RT-PCR of coagulase (coa), clumping-factor (clfA) and thermonuclease (nuc) genes, was consistent with the diagnosis phenotypically 'coagulase-negative variant of Staph. aureus'. The fact that coagulase-negative Staph. aureus variants can occur in the context of intramammary infections in cattle may result in the incorrect diagnosis 'coagulase-negative staphylococci (CNS)' in routine mastitis diagnostic, at least in rare cases. To fully ensure correct species diagnosis, sequencing of the 16S rRNA gene and amplification of specific genes such as coa is necessary in these cases.

Development of an enzyme immunoassay for the antibiotic cefquinome and its application for residue determination in cow's milk after therapeutical mastitis treatment.

The aim of this study was to develop and evaluate an enzyme immunoassay (EIA) for the cephalosporin antibiotic in milk, in combination with a new microbiological test system (brilliant black reduction test, BRT-P). Polyclonal antibodies against cefquinome were produced in rabbits, using cefquinome-keyhole limpet hemocyanine as the immunogen. These antibodies and a cefquinome-glucose oxidase conjugate were used in a competitive indirect EIA. The detection limit for cefquinome in milk was 1.5 ng ml(-1), recoveries were 80-128% at 4-40 ng ml(-1). Cross-reactivities with other cephalosporins/penicillins were all <1%. The EIA was used to determine cefquinome in incurred raw milk, the BRT-P (detection limit ≈ 20 ng ml(-1)) and a receptor assay (ßeta-s.t.a.r., detection limit ≈ 15 ng ml(-1)) were used in parallel. Five lactating cows, suffering from clinical mastitis, were treated with cefquinome by simultaneous intramammary and intramuscular injection. Cefquinome residues (maximum 10-27 µg ml(-1)) were most exclusively found in the udder quarter which was treated intramammary, residue levels in the other three quarters were low (<20 ng ml(-1)). Even in milk from intramammary-dosed quarters, residue levels fell below European Union maximum residue level (MRL, 20 µg kg(-1)) 2 days before the end of the withdrawal period. EIA, BRT-P, and ßeta-s.t.a.r. results showed acceptable agreement for milk samples, but the newly developed EIA is superior in aspects of sensitivity. In conclusion, this is the first one description of immunoassay and microbiological tests capable to determine cefquinome in milk at the MRL in incurred sample material.

Mycotoxins in horse feed.

A total of 62 samples of commercial horse feed preparations (complementary feeds) containing cereal mixtures ("muesli" or mash, n = 39; pelleted feeds, n = 12), and plain horse feed grains (maize, n = 5; oats, n = 4; barley, n = 2) were purchased from 21 different producers/distributors from the German market. All samples were analysed by competitive enzyme immunoassays (EIA) for six different mycotoxins (mycotoxin groups). Analytes (detection limit, mean recovery) were: deoxynivalenol (DON, 10 µg/kg, 84%), zearalenone (ZEA, 5 µg/kg, 93%), fumonisin B1 (FB1, 2 µg/kg, 113%), T-2 toxin (T-2, 0.1 µg/kg, 71%), sum of T-2 + HT-2 toxin (T-2/HT2, 0.2 µg/kg, 97%), ochratoxin A (OTA, 0.2 µg/kg, 67%), and total ergot alkaloids (Generic Ergot Alkaloids "GEA", 30 µg/kg, 132%). All samples contained DON (16-4,900 µg/kg, median 220 µg/kg), T-2/HT-2 (0.8-230 µg/kg, median 24 µg/kg), and T-2 (0.3-91 µg/kg, median 7 µg/kg). ZEA was detected in 98% of the samples (7-310 µg/kg, median 61 µg/kg). Most samples (94%) were positive for FB1 (2-2,200 µg/kg, median 27 µg/kg). Ergot alkaloids were detected in 61% of samples (28-1,200 µg/kg, median 97 µg/kg), OTA was found in 42% of samples (0.2-4 µg/kg, median 0.35 µg/kg). The results demonstrate that a co-contamination with several mycotoxins is very common in commercial horse feed from the German market. The toxin concentrations were in most cases well below the levels which are usually considered as critical or even toxic. The highest mycotoxin concentrations were mostly found in single-grain cereal feed: the maximum values for DON and FB1 were found in maize, the highest T-2/HT-2 toxin concentrations were found in oats, and the highest concentration of ergot alkaloids was found in barley. In composed feeds, no correlation between cereal composition and mycotoxin levels could be found.