The objective of this work was to use the Bayesian method in the fitting of the Wood´s model for milk production of Saanen goats. Two groups of animals from first and second lactation were considered in the analysis. The posterior marginal distributions for each parameter and production functions, peak milk yield, time of peak yield, persistency and total milk production, were obtained via Gibbs Sampler algorithm. The inference was done for each population. The results showed differences in the slope of the curve after the peak and in persistency, indicating highest production for the second lactation. The data were simulated for evaluating Bayesian method under several covariance matrices structures. The simulation results indicate the efficiency of this method for lactation curves studies when the covariance matrices show high correlation for parameters.
Gibbs Sampler; covariance matrix; dairy production
