Modelling Methionine Requirements of Fast- and Slow-Growing Chinese Yellow-Feathered Chickens during the Starter Phase

SIMPLE SUMMARY: In poultry production, consuming diets with low or excessive methionine levels leads to negative effects on growth performance. The requirements of methionine may differ among the fast and slow-growing breeds; therefore, the optimal dietary methionine level should be estimated for each. In this study, six dietary methionine levels were evaluated to estimate the optimal level for fast and slow-growing yellow feathered chicken breeds. The quadratic polynomial and exponential asymptotic regression showed that the optimal methionine requirements for maximal growth performance were 0.50% and 0.53% in the fast-growing breed, and 0.48% and 0.52% in the slow growing breed. ABSTRACT: Two experiments were carried out to investigate the dietary methionine requirement for fast and slow-growing Chinese yellow-feathered breeds during the starter phase, based on growth variables and regression models. In Experiment 1, a total of 2880 one-day-old Lingnan chicks (fast growing breed) were used to test the methionine requirement from 1 to 21 days of age for males and females separately. Of each gender, 1440 birds were allocated into 6 dietary methionine levels (0.28%, 0.32%, 0.37%, 0.43%, 0.50% and 0.63%), each with 6 pen replicates of 40 chicks. Experiment 2 had the same design with Guangxi chicks (slow growing breed) from 1 to 30 d of age. Results indicated that significant nonlinear or quadratic responses to increasing dietary methionine levels were observed in body weight, daily gain, feed intake and feed conversion ratio of both breeds. In summary, the quadratic polynomial regression showed that the optimal methionine requirements for maximal growth performance of Lingnan chickens were 0.52–0.58% in males, 0.51% in females, and 0.53% in mixed genders. The corresponding values for Guangxi breed were 0.53% in males by quadratic polynomial regression and 0.43% in females, and 0.48% to 0.49% in mixed sexes by exponential asymptotic models.