Chemical assessment of mycotoxin contaminants and veterinary residues in Costa Rican animal feed

Abstract Background When present in animal feedstuff, mycotoxins contaminants and antibiotic residues can have negative implications for animal production and Public Health, including the transmission of carcinogenic compounds and the selection of antibiotic resistant bacteria, respectively. So far there are no available methods in which both mycotoxins and antibiotic residues are analyzed using a parallel extraction approach. To address this issue, we developed a LC/MS methodology with high sensitivity (0.005 to 6.42 and 24.55 to 132.73 μg kg− 1 for mycotoxins and antimicrobials) and specificity (unique target ion mass/charge) that allows the detection of 26 mycotoxins and 23 antibiotic residues in animal feedstuff and validated it through the determination of these analytes in 294 animal feed and feed ingredient samples in the framework of a country-wide surveillance program. Two hundred and five of these samples were analyzed for mycotoxins and 89 for antibiotics. Findings Fumonisin was the most frequently toxin found, with FB1 and FB2 presenting prevalences of 50 and 52% and maximum concentrations of 14,927.61 and 8646.67 μg kg− 1, respectively. Other toxins, including diacetoxyscirpenol n = 4/101 (3.96%), fusarenon-X n = 2/101 (1.98%), citrinin n = 2 (1.98%), and patulin n = 1 (0.99%) were rarely found. Toxicologically relevant concentrations of toxin metabolites, such as HT-2 (6.38–485.49 μg kg− 1) and 3−/15-acetoxydeoxynivalenol (877.89–3236.56/5.44–1685.3 μg kg− 1), were also found. Few samples exceeded threshold mycotoxin concentrations defined in current EU guidelines. Dairy cattle and swine feeds included the higher number of samples exceeding guideline values (n = 6 and n = 5, respectively). From the total of samples analysed for antibiotics, 7.7% (n = 7/89) were classified as medicated for poultry and pigs. Unexpectedly, 57% of these medicated samples contained no detectable antibiotics (n = 4/7). The remaining 43% of the samples (n = 3/7) presented inconsistencies regarding the concentration of analytes declared on the labels or the antibiotics found. Likewise 74.6% (n = 50/67) of the non-medicated feed samples analyzed had antibiotic residues. Additionally, we analyzed commercial monensin standards for purity and evaluate batch-to-batch flushing feed industry practices. Conclusions Herein we report the results for a year-wide analysis for mycotoxins and antibiotics in feed samples. Mycotoxins, several metabolites, and the occurrence of these emerging contaminants were evaluated and antibiotic residues in non-medicated feed samples were found using a targeted MS-based LC approach. This validated multi-analyte method is expected to facilitate the monitoring and surveillance of contaminants, from natural and anthropogenic origin, in animal feed.