Using Near-Infrared Spectroscopy to Predict the Metabolizable Energy of Corn for Pigs

The aim of this study was to determine the chemical composition and metabolizable energy (ME) of different varieties of corn, and to validate mathematical models for predicting the ME of corn for pigs using near-infrared spectroscopy (NIRS). The dry matter (DM), mineral matter (MM), neutral detergent fiber (NDF), acid detergent fiber (ADF), ether extract (EE), crude protein (CP) and gross energy (GE) content of 99 different cultivars of corn were determined by conventional laboratory analyses (LAB) and by NIRS technology. Eighty cultivars were used to establish and calibrate the NIRS model, nine samples were used for external validation and ten cultivars were used in the metabolism assay. Samples were read in the spectrum range from 1100–2500 nm, and the model parameters were estimated by the modified partial least squares (MPLS) method. The results for the chemical composition of corn obtained by LAB and by NIRS technology were compared in a paired manner (observed and predicted values). The estimated metabolizable energy (EME) values were obtained from 11 predictive models in the literature, and inserted into the NIRS model. The observed metabolizable energy values (OME) were obtained in the metabolism assay. The metabolism assay was carried out using 44 barrows with an average initial weight of 25.05 kg, which were distributed into 11 treatment groups based on a randomized block design, with four replicates per treatment and one animal per experimental unit. Validation of the prediction equations was performed by adjusting the linear regression models of the first degree of EME values due to the OME, following a Bayesian approach. The linear relationship between the estimated and observed values was evaluated by detecting the significance of the estimated posterior parameters β₀ and β₁, recorded where the null (zero) did not belong to the 95% credible intervals for each parameter. Near-infrared spectroscopy was effective for determining the NDF, EE, CP and GE contents of corn when compared to the conventional LAB method. The prediction equations ME₁ = 4334 - 8.1MM + 4.1E - 3.7NDF, ME₂ = 4,194 - 9.2MM + 1.0CP + 4.1EE - 3.5NDF, ME₇ = 344.272 + 0.90886GE + 57.9377EE - 86.9320CP and ME₈ = 16.13 - 9.5NDF + 16EE + (23CP × NDF) - (138MM × NDF) were the most suitable (p < 0.05) for predicting the ME values of corn for pigs using NIRS compared to the LAB tests; however, neither method was accurate.