Generation of models from existing models composition: An application to agrarian sciences

Area of Science:

• Animal Science
• Mathematical Modeling
• Ruminant Nutrition

Background:

• Mathematical models are crucial for estimating rumen degradation, digestibility, and kinetics.
• Accurate modeling of gas production is essential for understanding feed utilization in ruminants.

Purpose of the Study:

• To propose a novel mathematical model by combining existing ones.
• To compare the proposed von Bertalanffy-Gompertz two-compartment model with the logistic two-compartment model for gas production kinetics.
• To identify the best model for describing gas production curves in pinto peanut cultivars.

Main Methods:

• A semi-automated in vitro technique was employed to measure gas production over 96 hours.
• Data from pinto peanut cultivars grown at the Far South Animal Sciences station (Essul) were utilized.
• Model parameters were estimated using the least squares method with the Gauss-Newton iterative process in R software (version 3.4.1).

Main Results:

• The von Bertalanffy-Gompertz two-compartment model demonstrated a superior fit for cumulative gas production.
• Model performance was evaluated using adjusted coefficient of determination, residual mean squares, mean absolute deviation, Akaike information criterion, and Bayesian information criterion.
• The proposed model effectively described the kinetic curve of gas production under the study's conditions.

Conclusions:

• The von Bertalanffy-Gompertz two-compartment model is the most suitable for describing cumulative gas production kinetics in pinto peanut cultivars.
• This study provides a refined method for developing and evaluating mathematical models for feed evaluation.
• The findings contribute to more accurate estimations of rumen degradation and digestibility in ruminant nutrition.