Open AccessAvaliação de modelos de crescimento com erros heterocedásticos de frangos de corte de linhagem caipira alimentados com sorgo
Author(s) -
Kewin Marcel Ramirez Ferreira,
AUTHOR_ID,
E. M. de Souza,
Henrique Jorge de Freitas,
Geisson de Araújo Negreiros,
AUTHOR_ID,
AUTHOR_ID,
AUTHOR_ID
Publication year - 2021
Publication title -
revista agrária acadêmica
Language(s) - English
Resource type - Journals
ISSN - 2595-3125
DOI - 10.32406/v4n6/2021/49-63/agrariacad
Subject(s) - gompertz function , mathematics , zoology , statistics , biology
This study aimed to evaluate growth curves of broilers (pedres caipiria lineage) fed sorghum (40% replacement for corn) in a basal ration, in the rearing period from 1 to 70 days. Weight and average consumption accumulated over time were analyzed with the following regression functions: Gompertz and Logistic, on four different models for residual variance. The results showed that the logistic model (with exponential-type variance) had the lowest values for the AIC and BIC, in addition to fitting well to the data. The models that used the Gompertz function presented several problems related to the basic assumptions.