The Influence of Ingredients, Corn Particle Size, and Sample Preparation on the Predictability of the Near Infrared Reflectance Spectroscopy

The near infrared reflectance spectroscopy (NIRS) technique is a rapid and non-destructive used to evaluate the chemical composition of complete feed ingredients. accuracy its prediction relies upon calibration standards account for variations in material particle shape size. purpose this study was determine effect alternative ingredient inclusion corn size along with sample preparation method on NIRS using standard calibrations provided instrument. Treatments were arranged as 4 × 3 factorial diet type (soybean meal (SBM) + DDGS (SD); SBM fish (SFD); wheat bran (SFB); (SB)); (400, 600, 800 μm); analysis (laboratory, NIRS-ground, NIRS-unground). All samples evaluated crude protein (CP) content. Laboratory values from wet chemistry analyses obtained Dumas Combustion comparison results NIRS. Ground unground scanned Foss D2500 machine wavelength range 400 2,500 nm at log (1/R) 2 intervals each sample. There no interaction CP; however, there an (P ≤ 0.05) between analysis. When analyzing diets laboratory methods differences CP, but when NIRS, grinding prior improved compared not grinding, though they still lower than also CP content NIRS-ground similar within used, different sizes closer expected (20%) NIRS-unground samples. Results significantly However, intermediate trial indicate necessity proper biasing improve especially contain Grinding scanning will accuracy, may differ equipment calibrations, further emphasizing importance biasing.