Predicting Net Energy Value Of Corn For Broiler Chicken By Near-Infrared Reflectance Spectroscopy And Regression Equation Established By Chemical Composition Of Corn

The objective of this experiment is to predicting the net energy value of corn for 0-3 weeks old broiler chicken by near-infrared reflectance spectroscopy and regression equation eatabished by chemical composition. In the study, net energy value of 15 corn were measured in 126 broiler chicken by an comparative slaughter trial and a fasting metabolism trial. Collecting animal's excreta all through the comparative slaughter experimental stage. Meanwhile, determinating the chemical composition of corn. Then, the calibration model for net energy value, appearance metabolized energy, chemical composition of corn were developed by near-infrared reflectance spectroscopy. Otherwise, the regression equation was eatabished by chemical composition using SPSS(13.0).The results are as follows:1. The NE, AME of corn for 3 weeks old broiler chicken which were determined by an comparative slaughter trial and a fasting metabolism trial were 2.56±0.049 Mcal/kg, 3.38±0.113 Mcal/kg,respectively, the efficiency of AME transform to NE (NE/AME) is 78.1%。2. The evaluating parameters of the NIRS predicting model for chemical compositon of corn are as follows;The coefficient of determination of calibration (R2cal)are CP:0.97, EE:0.98,ST:0.96, CF:0.96, respectively; The standard deviation of calibration (RMSEE) are CP:0.084, EE: 0.023, ST:0.244, CF:0.001, respectively; The relative standard deviation of calibration(RSD) are CP:1.047, EE:0.57,ST:0.389, CF:0.079 respectively. The coefficient of determination of validation (R2cv) are CP:0.94, EE:0.98, ST:0.95, CF: 0.95, respectively; The standard deviation of validation (RMSECV) are CP:0.001, EE: 0.023, ST:0.257, CF:0.001, respectively; The relative standard deviation of validation(RSD) are CP:0.01, EE:0.57, ST:0.41, CF:0.07 respectively.3. The evaluating parameters of the NIRS predicting model for NE and AME of corn are as follows:The coefficient of determination of calibration (R2cal) are NE:0.91,AME:0.91 respectively; The standard deviation of calibration (RMSEE) are NE:0.014, AME:0.034 respectively; The relative standard deviation of calibration(RSD) are NE:0.55, AME:1.1,respectively; The coefficient of determination of validation (R2cv) are NE:0.86,AME:0.87 respectively; The standard deviation of validation (RMSECV) are NE:0.018,AME:0.037 respectively; The relative standard deviation of validation(RSD) are NE:0.7,AME:1.6, respectively;4. The best regression equation and its'R2 and RSD of NE of corn established by chemical composition and AME is as follows:NE=10.851+0.421ME-0.011CP+0.259EE+0.005ST-8.267CF, R2=0.882, RSD=0.81Conclusion:Both regressing model and calibration model by chemical composition and Fourier Near-infrared reflectance spectroscopy,respectivly, could meet the needs of quality evaluation. But Fourier Near-infrared reflectance spectroscopy appears to be superior to chemical composition as a rapid tool in predicting the poultry net energy of corn.