Evaluating NET Energy Of Full-fat Rice Bran For Ducks

The diets for poultry are commonly formulated based on metabolizable energy(ME)system.But there are defects existing in ME system,which ignored the heat consumption(HI)and the energy deposition efficiency in the metabolism process.The net energy(NE)system solves these problems,which can accurately reflect the animals,energy needs and energy value of feedstuffs.As a high-quality non-grain type energy feed' full-fat rice bran is widely used in the diets of ducks.However,the net energy value of full-fat rice bran has not been reported yet.Considering that the determination of the net energy value of full-fat rice bran is cumbersome and laborious,a method for fast prediction of the net energy of full-fat rice bran is in need.In this study,the NE value of full-fat rice bran for ducks was determined by the metabolism experiment,feeding experiment and comparative slaughter experiment.The prediction models for the net energy of the full-fat rice bran were established based on their chemical composition and near infrared spectra.This study will provide an experimental basis for construction and application of net energy system.First,full-fat rice bran samples were collected from primary rice-cultivation regions in China,and their chemical composition was analyzed.24 kinds of full-fat rice bran samples were selected according to the "Rice bran for feedstuffs" national standard for animal experiments.The factorial method was used to evaluate the NE of rice bran substitution diets,and the NE was divided into NEm(Net energy for maintenance)and NEP(Net energy for production)determined,respectively.320 Pekin ducks with similar body weight(125±5)g were used in the experiment.And 20 ducks were slaughtered at the beginning and their body energy were measured and served as the control of the initial energy deposition.50 ducks were divided into five treatments group for NEm determination,which fed 100%,80%,60%,40%,20%of feed compared to Ad libitum,respectively.The remaining 250 ducks were divided into 25 treatment groups to measure NEp,fed either a basal diet or a basal diet substituted with 20%of Rice bran.The NE value was obtained by adding up the NEm and NEp.Correlation and multiple linear stepwise regression were conducted and the model for predicting NE of full-fat rice bran by chemical composition were established.In addition,the near-infrared prediction model was established by using near-infrared spectroscopy technology to collect data,optimize and process the data,establish the model and test the model.Results follows:1)The content of chemical components in 24 kinds of full-fat rice bran were obviously different.2)According to the regression formula IgHP = a + bMEI,with different dietary levels of MEI as independent variables,the corresponding IgHP as the dependent variables,the following regression equation was obtained:lgHP=0.191MEI-0.261(R2=0.980,RSD=0.000,P<0.05),the fast heat production(FHP)of Pekin ducks was 548.28 KJ/(kgBW0.75·d).3)The NEm and NEP of rice bran substitution diet were(2.26±0.05)MJ/kg and(6.19±0.12)MJ/kg,respectively.The NE of rice bran substitution diet was(8.45±0.13)MJ/kg.4)According to the substitution method,the NE was(8.39±0.65)MJ/kg,which was provided to Pekin ducks by full-fat rice bran.The AME of rice bran was(12.40±0.59)MJ/kg,and the conversion ratio of AME to NE was(67.62±3.39)%.5)The NE prediction model based on chemical composition with the highest fitting degree was NE = 3.378 + 0.116 EE-0.055 NDF + 0.396 AME-0.037 ADF(R2=0.961,RSD= 0.141 MJ/kg,P<0.05).6)The near infrared spectra of different full-fat rice bran were different,and measure value/the predictive value of the near-infrared prediction model established in this experiment was as follows:The coefficient of determination of calibration(R2cal)of the forecasting model was 0.916,root mean squared error of calibration(RMSEE)was 0.213 MJ/kg,the modeling condition was principal component dimension for 10,spectral range of 4601.5?4246.7 cm-1,pretreatment method was the vector normalization.The coefficient of determination of cross validation(R2cv)was 0.868,and the root mean squared error of cross validation(RMSECV)was 0.235 MJ/kg.The coefficient of determination of validation(R2val)was 0.846,and the root mean squared error of prediction(RMSEP)was 0.269 MJ/kg.The results of this study showed that the net energy range of full-fat rice bran for ducks was 7.54?9.28MJ/kg,and the average value was(8.39 + 0.65)MJ/kg.It's feasible to establish NE prediction model of full-fat rice bran based on its conventional chemical composition and NIRS spectra.