NIRS Models Developing For Predicting Ergosterol In Alfalfa(Medicago Sativa L.) Hay

Near-infrared (NIR) spectroscopy technology is a rapid and practical method for the detection of organic compounds in matter. With the objective to assess the extent of fungal contamination in alfalfa hay, this work assesses the applicability of NIR for the rapid identification of ergosterol, which is the indicator of fungal contamination. Two set of samples were made, for one set of97artificial contaminated samples, and another of100nationwide samples. To each set of sample, three forth samples were used to calibrate and one forth samples were used to validate. Gas Chromatograph-Mass Spectrum (GC-MS) was used to determine the ergosterol content, and NIR spectrums were collected. After the spectra range selected, the chemical reference value and spectral value were matched to develop the NIR equation by Partial Least Square (PLS). The number of PLS components was determined by the method of leave-one-out validation and the best spectral pretreatment methods were selected by the parameters of coefficient of determination (R2), standard error of calibration (SEC) and standard error of prediction(SEP). Then the outliers were rejected and the best NIR equation was developed. The parameters of relative standard error of calibration (RSEC), relative standard error of calibration (RSEP), the ratio of performance to deviation (RPD), and the ratio of SEC and SEP (SEC/SEP) were used to evaluate the performance of NIR equation.The research found that the best spectral pretreatment method for the artificial contaminated samples is "Savitzky-Golay smoothing+Standard Normalized Variate (SNV)" or "Savitzky-Golay smoothing+SNV+lst order derivative", and for the nationwide samplies is "1st order derivative". For the best NIR equations of the two samples sets, the R2reaches0.98and0.90respectively, and the SEC and SEP for each samples are both less than10mg/kg, RSEP is at the range of "0.27-0.34mg/kg", and RPD is more than3. The result shows that the NIR methodology can be applied for broad ergosterol range of Alfalfa hay and varied type of alfalfa hay. By using NIR spectroscopy technology, ergosterol in the alfalfa hay can be predicted rapidly and the alfalfa hay with potential safety problems related to fungal contamination would be screened.