Determination and prediction of standardized mineral retentions in different sources of rapeseed meals for broilers

The aim of this study was to firstly assess the chemical compositions in 10 different sources of rapeseed meals (RSMs), then determine their standardized mineral retentions (SMRs) for broilers and establish predicting models for SMRs in RSMs based on their chemical compositions, and finally validate the accuracy of these models. A total of 308 (22 d-old) male Arbor Acres broilers with similar body weights were randomly assigned to 11 dietary treatments. On d 26, broilers were provided with either a mineral-free diet or different RSM diets formulated from 10 RSMs for 4 h. Meanwhile, feces were collected for 52 h to evaluate their SMRs. The predicting models for SMRs were developed through stepwise regression using data from 9 RSMs. One RSM was then randomly selected to evaluate the SMRs-predicting models' accuracy. The results showed that different sources of RSMs significantly affected (P < 0.02) their SMRs in broilers. The standardized retention was highest for copper (Cu, 35.3 %) and lowest for manganese (Mn, 12.9 %). Based on the chemical compositions, 8 predicting models (R² = 0.520∼0.982, P < 0.03) for standardized retentions (SRs) of calcium (Ca), phosphorus (P), magnesium (Mg), potassium (K), Cu, iron (Fe), Mn, and zinc (Zn) in RSMs for broilers were effectively established. The predicting model of standardized Ca retention showed the highest fitted goodness (R² = 0.982, P = 0.0146), whereas the predicting model of standardized Cu retention had the lowest fitted goodness (R² = 0.520, P = 0.0284). The predicting models for standardized Ca, P, K, Mg, and Zn retentions demonstrated good predicting accuracies of their predicted values, while the predicting models for standardized Cu, Fe, and Mn retentions performed poor predicting accuracies of their predicted values. In conclusion, the predicting models established in this study for the SRs of Ca, P, K, Mg, and Zn in RSMs for broilers demonstrated accurate predicted results, providing a valuable reference for the rapid prediction of the SRs of these minerals in RSMs for broilers.