New model of evaluation of sunflower and corn silages by the in vitro gas production technique

Abstract

This study aimed to propose a model called Two-compartment Logistic-von Bertalanffy (LVB) and to identify among the proposed and Two-compartment Logistic (TL) models the one that has the best goodness of fit to the kinetic curve of cumulative gas production (CGP) of sunflower and corn silages alone and combined using the in vitro semi-automated gas production technique. A random block split-plot experimental design was employed in which the inoculums were the blocks, the incubation times were the split-plots, and the experimental diets were: CS - corn silage, SS - sunflower silage (as single roughage), and their mixtures, i.e., 340SS (660 g kg-1 corn silage and 340 g kg-1 sunflower silage) and 660SS (340 g kg-1 corn silage and 660 g kg-1 sunflower silage). The parameters were estimated by the least squares method using the iterative Gauss-Newton process in the software R version 3.4.1. The criteria adopted were: adjusted coefficient of determination (R2adj.), residual mean squares (RMS), mean absolute deviation (MAD), Akaike information criterion (AIC), Bayesian information criterion (BIC), and relative efficiency (RE). The TL model had higher R2adj. values compared to LVB, however, such difference may be considered negligible. The LVB model had RE above one, which indicates it is superior to the TL model, in addition to the lowest RMS, MAD, AIC, and BIC values, The Two-compartment Logistic-von Bertalanffy model had the best fit to describe the CGP over time according to the methodology and conditions of the present study.

Modelos Matemáticos Para Estimativa Da Cinética De Fermentação Ruminal Do Pseudofruto Do Cajueiro Através Da Técnica in Vitro Semi-Automática De Produção De Gases / Mathematical Models to Estimate the Cashew Tree False Fruit Ruminal Fermentation Kinetic Through the Semi-Automatic in Vitro Gas Production Technique

“…O desenvolvimento de novas técnicas de avaliação de alimentos e a automatização das mesmas conduziu à necessidade de utilização de modelos matemáticos capazes de agrupar várias informações em poucos parâmetros biologicamente interpretáveis por modelagem (SANTOS et al, 2020;BAIA et al, 2020;CAVALCANTE et al, 2020;DOUGHERTY et al, 2019;BATEKI et al, 2019;THEODOROU et al, 1994;BEUVINK et al, 1992;FRANCE et al, 2000). Os parâmetros matemáticos obtidos por meio da curva de produção de gases permitem descrever e interpretar os perfis cumulativos da produção de gases, avaliar diferentes ambientes de fermentação, além de possibilitar a comparação de diferentes substratos e frações dos alimentos (OLIVO et al, 2017;NOGUERA et al, 2004).…”