Use of NIRS technology for predicting the nutritional value of silage made from tropical grasses enriched with corn ethanol co-products and intercropped with corn

Abstract

This study aimed to evaluate the chemical composition of tropical grass silage with added co-products from the production of corn ethanol, dried distillers’ grains (DDG) and wet distillers’ grains (WDG), and grass intercropped with corn. The estimation of the chemical composition was performed with near-infrared reflectance spectroscopy (NIRS). In Experiment I (elephant grass), the experimental design was completely randomized with four replicates, and treatments were arranged in a 2×6 factorial scheme with two factors (additives: DDG and WDG, and application levels: 0, 5, 10, 15, 20, and 30%). In Experiment II (Tanzania grass), the experimental design was completely randomized with four replicates, and the treatments included five DDG levels (0, 5, 10, 15, and 20%). In Experiment III, the experiment was conducted in randomized blocks with five replicates, and the treatments were organized in a 2×4 arrangement. Factor 1 included two cultivation methods, monocropped corn and corn intercropped with ruziziensis grass, and factor 2 involved four parts of the corn plant: whole plant, half plant, cobless plant, and cob with husk. The dry matter (DM), crude protein (CP), neutral detergent fiber (NDF), acid (ADF), ash, and estimated total digestible nutrient (TDN) contents were evaluated. Reference values were added to the spectra of forage samples. Data preprocessing and chemometric model building, that is, calibration curve development, were performed using the Opus 7.5 software employing partial least squares (PLS) regression. The calibration model was selected based on the lowest root mean square error of the cross-validation (RMSECV) and the highest coefficient of determination (R2cv). The nutritive value of elephant grass and Tanzania grass silage improved with the use of DDG when compared with that of in natura silage. The NDF and ADF contents were lower, and DM was higher in ruziziensis grass silage intercropped with corn, highlighting the importance of adopting integrated production systems. Estimates by NIRS presented high R2cv values (>0.95), demonstrating the potential of this technology for routine analysis of tropical grass silages for CP, NDF, ADF, and ash.

Keywords

• Elephant grass
• Near-infrared spectroscopy
• Ruziziensis grass
• Tanzania grass
• Zea mays.