Determination of standardized mineral availability in different corn sources for broilers and establishment of prediction models

To date, few prediction models have been reported for estimating the standardized mineral availability (SMA) in feedstuffs for broilers and other animals. The present study aimed to analyze the chemical composition of 10 different corn sources, and then determine their SMAs for broilers, and establish prediction models for SMAs in corn based on their chemical components. Finally, the accuracies of these models were verified. In total, 352 male Arbor Acres (AA) broilers (average body weight of 1.03 ± 0.01 kg) at 22 days of age were selected and randomly divided by body weight into 11 treatment groups, with 8 replicate cages of 4 broilers per replicate cage for each treatment. On d 26, the chickens were fed either one mineral-free diet or 10 tested corn diets for 4 h. Fecal samples were collected continuously for 52 h to determine their SMAs. The data from nine corn samples were employed with stepwise regressions to build SMAs prediction models. One corn sample was randomly selected to test the accuracies of SMAs prediction models. Different corn sources significantly affected SMAs in broilers ( P < 0.05). The standardized availability (SA) value of potassium (K) was the highest (77.2%), whereas that of manganese (Mn) was the lowest (50.7%). Seven prediction models ( R 2 = 0.711 to 0.957, P < 0.05) for the SA values of calcium (Ca), phosphorus (P), magnesium (Mg), K, copper (Cu), Mn, and zinc (Zn) in corn for broilers were successfully established based on chemical composition (crude protein, phytic acid, ether extract, acid detergent fiber, crude ash, and crude fiber). Among these, the prediction model for the SA value of P exhibited the highest goodness of fit ( R 2 = 0.957, P < 0.001), while that of Ca demonstrated the lowest goodness of fit ( R 2 = 0.711, P = 0.024). Except for the prediction models for SA values of Ca and Mg, which showed relatively low prediction accuracies, all other prediction models for SA values of P, K, Cu, Mn, and Zn displayed good prediction accuracies. The prediction models for the SA values of P, K, Cu, Mn, and Zn in corn for AA broilers as established in the present study demonstrated accurate predicted results, and provided a valuable reference for the rapid prediction of the SA of these minerals in corn for broilers.