Fusarium-damaged kernel (FDK) is a commonly at-harvest measure used as an indicator of disease intensity and, in some cases, predictive of DON in the harvested kernels ( Beyer, Klix, & Verreet, 2007). FDK, determined by inspecting a sub-sample of 200 kernels, was defined as the proportion of visually scabby kernels in a sample of harvested grain, e.g. discoloured, shrivelled or pinkish white kernels. A liquid chromatography–mass spectrometry (LC–MS/MS) system was used to simultaneously

determine and quantify DON and NIV in the samples. Stock solutions of DON and NIV standards (Sigma Chemical Company, USA) were prepared by dissolution in benzene:acetonitrile (95:5) at a concentration of 100 μg/ml. The work solution was obtained by dilution to a concentration of 50 and 10 μg/ml, respectively, estimated Fulvestrant by the w/v relation and confirmed by a procedure utilizing molar absorptivity of the standard. Toxin extraction used 10 g of samples in acetonitrile:water Selumetinib nmr (70:30), shaken for 30 min. The analytical interferants were carried out with hexane by liquid–liquid partition, by drying under reduced pressure and in nitrogen at 30 °C. First, it was solubilized with chloroform:methanol

(9:1) and, secondly, in acetonitrile before injection. A Shimadzu High Performance Liquid Chromatograph with degasser DGU-20A3/DGU-20A5, system controllers CBM-20A/20Alite and UV–VIS detector SPD-20A/20AV was used. The manual injection utilized the loop standard

of the 20 μl of volume. The compounds were detected at λ = 220 nm and evaluated by elution in the reverse phase using column Waters Spherisorb 5 μ ODS2 (4.6 × 150 mm) with flow rate at 0.6 ml/min in water:acetonitrile (93:7). The retention time was 1022 and 1837 min for DON and NIV, respectively. The detection and quantification limits were determined by successive dilutions of the standard solution, until generating detection signal three and nine times superior to the standard BCKDHA deviation at the same time of the retention of mycotoxins when injecting the derivation control. Detection limits were 0.25 and 0.28 μg/g, quantification limits were 0.75 and 0.84 μg/g, mean recovery percentages were 96% and 94% (variation coefficient 3% and 7%), regression coefficients 0.996 and 0.986, linearity from 0.75 to 15 and 0.84 to 16.8 μg/g, all respectively for DON and NIV. Exploratory and descriptive statistics were used to summarize and map the occurrence, concentration levels and spatial distribution of the mycotoxins across the geographic region. Non-parametric tests (Kruskal–Wallis and Wilcoxon) were used to compare toxin concentration levels among years and between toxin types.