Trusted information & resources for animal nutrition.

Technical Resources

Correlation between data on proximal composition, protein quality indicators, and amino acid profile of commercial samples of soybean meals analyzed by wet chemistry or estimated by NIRS technology

Dardabou, L., H. Kadardar, L. Cámara and G. Mateos

A total of 26 commercial soybean meal (SBM) samples were randomly collected in different locations at European crushing plants by specialized personnel and analyzed by wet chemistry and near-infrared spectroscopy (NIRS) for DM, ash, CP, amino acids, ether extract after previous hydrolysis (EEh), crude fiber (CF), neutral detergent fiber (NDF), gross energy, phosphorus (P), and protein quality indicators [PDI, protein dispersibility index; KOH solubility, and TIA, trypsin inhibitor activity]. All wet analyses were performed by the same laboratory using official methods. The NIRS values of the SBM samples were obtained in a second laboratory using their own calibration data. As fed wet chemistry values, independently of the origin of the SBM, varied (% as fed bases) from 85.7 to 91.1 for DM, 6.1 to 6.9 for ash, 44.5, to 49.8 for CP, 1.40 to 3.01 for EEh, 4.1 to 5.2 for CF, 9.8 to 12.9 for NDF, and from 0.58 to 0.69 for total P. When the SBM were sorted by the origin of the beans, the CP as fed wet chemistry values, varied from 45.2 to 46.8 % for ARG, 44.7 to 49.8 % for BRA, and 44.45 to 46.9 % for USA meals. Similarly, the protein quality indicators varied from 8.2 to 19.0 % for PDI, from 59.6 to 85.7 for KOH, and from 1.10 to 2.80 mg/g for TIA.

The Pearson correlation (r) analyses using the CORR procedure of SAS (SAS Institute Inc., 1990) were used to study the correlations between the values obtained by the two methods (wet chemistry and NIRS technology), independently of the origin of the SBM effect. Sample of SBM was considered the replicate with a total of 26 samples (USA = 8, BRA = 9, and ARG = 9). The highest correlations (P < 0.001) between the two methods were recorded for DM (r = 0.930), CP (r = 0.860), CF (r = 0.610), NDF (r = 0.690), and gross energy (r = 0.740). Also, the correlation was significant for ash (r = 0.580; P < 0.01) and phosphorus (r = 0.490; P < 0.05). The most limiting AA in non-ruminants diets that showed significant correlation between wet chemistry and NIRS were Ile (r = 0.580), Leu (r = 0.590) with P < 0.01, and Lys (r = 0.380), Thr (r = 0.440), and Arg (r = 0.460) with P < 0.05. However, no correlations between wet chemistry and NIRS for EEh and protein quality indicators were found. The correlations found were lower than expected for many of the variables compared because the data on wet chemistry and the NIRS analyses were conducted in two different labs.

It is concluded that the NIRS technology could be an alternative for the estimation of the chemical constituent of commercial SBM, including that of certain amino acids. However, more studies are needed to improve the accuracy of values obtained by NIRS for EEh and protein quality indicators.