Fitting nonlinear models to the growth of New Zealand White rabbits

This study aimed to assess nonlinear models fit to the growth of New Zealand White rabbits. Rabbit weight was measured every five days between 35 (weaning) and 75 days old (slaughter), in 64 individuals allocated to eight groups and fed a reference diet (REF), simplified and semi-simplified diets consisting of different combinations of three fiber sources: alfalfa hay (AH), cassava leaf meal (CLM) and hay from the upper third of cassava roots (HUTCR). The Santos, Gompertz, Brody, Logistic, Richards and von Bertallanfy models were investigated to determine the best fit based on the model fitting criteria assessed, in addition to identity testing at 5% significance to assess diet feasibility and their effect on animal nutrition. In general, Santos' model performed best according to the fitting criteria analyzed, obtaining the highest adjusted coefficient of determination (Raj) and the lowest values for the Akaike (AIC) and Bayesian information criteria (BIC) and asymptotic index (AI). Analysis of the growth curves showed that REF exhibited the best growth performance in relation to the other diets, while largely CLM-based diets performed poorly, with low asymptotic weight estimates of 1815 to 1908 g. The semi-simplified diet based on HUTCR performed satisfactorily in terms of asymptotic weight estimates (2210 g), making it a feasible alternative for rabbit farming.