The Predictive Models For The Net Energy Value Of Rapeseed Meal And Cottonseed Meal For Avian Broilers Aged From 1 To 21 Days

This study was conducted to establish reliable prediction models for net energy (NE) of rapeseed meals(RSM) and cottonseed meals(CSM) based on chemical composition and apparent metabolic energy(AME) combined with chemical composition and Fourier near infrared spectroscopy (FNIRS) for Avian broilers aged from 1 to 21 days, and to compare the predictive models of the two samples separately and together.NE value of RSM and CSM was measured as the sum value of NE for maintenance (NEm) and NE for deposition (NEp). The NEm were measured by regression method with four treatments, which were group ad libitum to food, groups restricted feeding by 30%,50%,70%, respectively. The NEp of 15 RSM and 15 CSM were measured by the method of substitution, each of RSM and CSM was assigned as a treatment. Chicks in both trails were 7-d-old, with an average weight of of 97.3±4.0 g and randomly allotted into 36 treatments with six replications of two chicks. The experiments lasted 7 days. Proximate compositions of RSM and CSM were measured. Analyses of linear regression were carried out between NE and AME values, and chemical composition based on the two samples separately and together. Each sample with measured NE value was divided into five parts and the moisture contents of RSM and CSM were adjusted to 9%,10%,11%,12% and 13%, respectively, and the models of FNIRS were established based on the two samples separately and together The results showed as follows:the NE value of RSM and CSM for broilers aged from 1 to 21 days were from 4.72 to 7.22 MJ/kg DM and from 4.73 to 7.08 MJ/kg DM respectively.the R2 of the best regression equations for two samples separately and together based on the AME combined with chemical composition were 0.995,0.998 and 0.995, respectively, and the relative standard deviations (RSD) were were 0.052,0.033 and 0.052 MJ/kg, respectively, the R2 of the best regression equations based on chemical composition were 0.973,0.985 and 0.973, respectively, and the RSD were 0.123,0.100 and 0.123 MJ/kg, respectively, the reliable regression equations based on AME combined with chemical composition were better than the reliable regression equations only based on chemical composition; The coefficient of determination in calibration (R2cal) and root mean square error of calibration (RMSEE) of the predictive models of two samples separately and together were 0.99/0.042, 0.99/0.068,0.96/0.139 MJ/kg; The coefficient of determination in cross validation (R2CV) and root mean square error of cross validation (RMSECV) of the predictive models of two samples separately and togethe were 0.98/0.089,0.99/0.082, 0.89/0.219 MJ/kg; and the root mean square error of prediction (RMSEP) were 0.091, 0.089 and 0.377 MJ/kg respectively.The results indicated that the NE value of RSM and CSM were accurate and the reliable prediction models for NE of RSM and CSM were established based on chemical composition and AME combined with chemical composition and FNIRS, the reliable regression equations based on AME combined with chemical composition were better than the reliable regression equations only based on chemical composition, The accuracy of the predictive NE model by NIRS is similar to the best equation from AME combined with chemical composition,the prediction models for NE of the two samples separately and together established based on chemical composition and AME combined with chemical composition both were reliable,but the prediction models for NE of the two samples separately were better than their together established based on FNIRS.