We studied the chemical composition and quality of the protein fraction of 27 soybean meals (SBM) from Argentina (ARG), Brazil (BRA), and USA, collected at random at European crushing plants and feed mills during the 2022/2023 harvest season. These samples were analyzed for their physicochemical characteristics, including particle size, Minolta color, crude protein (CP), amino acids (AA), non-starch polysaccharides (NSP), crude fiber, neutral detergent fiber (NDF), α-galactosides, and minerals. In addition, urease activity (UA), protein solubility in KOH (KOH), protein dispersibility index (PDI), trypsin inhibitor activity (TIA), and reactive Lys (RLys) were determined. Data were analyzed as a completely randomized design using an ANOVA, with SBM origin as main effect. In addition, Pearson correlation (r) analyses were conducted to determine the relationships between chemical composition and protein quality indicators of the SBM. The BRA meals had higher a* (redness) and lower L* (luminosity) values than the ARG and USA meals (P < 0.001). Particle size distribution was similar for all SBM. On 88% DM basis, CP content was higher for the BRA and USA meals than for the ARG meals (46.6 and 46.2 vs. 45.1%; P < 0.001) whereas sucrose (7.26 and 7.93 vs. 6.08%; P < 0.001) and stachyose (5.41 and 4.96 vs. 4.01%; P < 0.001) were higher for the USA and ARG meals than for the BRA meals. The NSP (15.6, 14.4, and 15.0% P = 0.093) and NDF (10.0, 8.5, and 9.1%; P < 0.05) contents were higher for the BRA than for the ARG meals, with the USA meals being intermediate. Per unit of protein, Lys (6.16 vs. 6.05 and 5.93%; P < 0.001), total sulfur AA (2.77 vs. 2.71, and 2.65%; P < 0.01) and the sum of Lys, Met, Cys, Thr, and Trp (14.2 vs. 14.1, and 13.8%; P < 0.001) were higher for the USA meals than for the ARG and BRA meals. The values for all protein quality indicators (UA, PDI, KOH, TIA, and RLys) were higher for the USA meals than for the South American meals (P < 0.05). Among the minerals, the main difference observed was for Fe that was higher for the BRA meals than for the USA and ARG meals (172 vs. 108 and 113 mg/kg; P < 0.01). Independently of the origin of the soybeans, the most important correlations (P < 0.01) observed were between stachyose and sucrose contents (r = 0.814) and between TIA and PDI (r = 0.808), RLys (r = 0.702), and KOH (r = 0.657) values. In summary, the chemical composition, AA profile, and protein quality indicators of the SBM vary with the country of origin of the soybeans.
An update of the SBM matrix, based on laboratory data and the country of origin of the soybeans, might help to optimize the nutritional value of commercial SBM, reducing feed formulation cost.