Abstract
Currently, there is no report on prediction models of standardized ileal amino acid digestibilities (SIAADs) in soybean meals (SBMs) for medium-growing yellow-feathered chickens. This study firstly analyzed the chemical compositions of 10 SBMs, then determined their SIAADs in chickens, and finally established and verified prediction models for SBM SIAADs based on their chemical compositions and amino acid (AA) profiles. A total of 276 55 d-old Tianluma roosters were selected and randomly divided by body weight into 11 treatment groups. On d 63, chickens were fed either a nitrogen-free diet (NFD) or one of 10 SBM diets for 5 d. On d 67, ileal chyme samples were collected to determine SIAADs. Data from nine SBM samples and stepwise regressions were employed to build prediction models, while one SBM sample was randomly selected to validate model accuracy. Different SBM sources affected (p ≤ 0.007) SIAADs in medium-growing yellow-feathered chickens. The standardized ileal digestibility (SID) of glutamic acid (Glu) was the highest (93.9%), whereas that of cysteine (Cys) was the lowest (81.7%). Fifteen prediction models (R(2) = 0.567-0.993, p < 0.03) for the SIDs of methionine (Met), isoleucine (Ile), leucine (Leu), phenylalanine (Phe), lysine (Lys), histidine (His), arginine (Arg), aspartic acid (Asp), serine (Ser), Glu, glycine (Gly), alanine (Ala), Cys, tyrosine (Tyr), and proline (Pro) in SBMs for medium-growing yellow-feathered chickens were effectively established based on chemical compositions and AA profiles. Among them, the prediction model for the SID of Cys showed the best fit (R(2) = 0.993, p = 0.002), while the model for the SID of Ala had the lowest fit (R(2) = 0.567, p = 0.019). Except for His and Pro, which exhibited poor predictive accuracy, all other models showed good accuracy. These prediction models thus provide a valuable reference for rapidly estimating the SIDs of key AAs in SBMs for medium-growing yellow-feathered chickens.