Modeling growth curve parameters in Peruvian llamas using a Bayesian approach.

Area of Science:

• Animal Science
• Quantitative Biology
• Agricultural Research

Background:

• Understanding llama growth patterns is crucial for effective livestock management and breeding strategies.
• Accurate modeling of body weight is essential for optimizing nutrition and health interventions in llamas.
• Previous studies have explored various growth models, but a comprehensive comparison using both frequentist and Bayesian approaches in llamas is limited.

Purpose of the Study:

• To compare the fit of four nonlinear growth models (Brody, von Bertalanffy, Gompertz, Logistic) to llama weight data.
• To estimate growth curve parameters using both frequentist and Bayesian methodologies.
• To identify the optimal model for describing llama growth from birth to 12 months of age, considering sex and type differences.

Main Methods:

• Collected 43,332 monthly body weight records from 3611 llamas (birth to 12 months) between 1998-2017 in Peru.
• Estimated nonlinear growth model parameters using frequentist and Bayesian (MCMC with Metropolis-Hastings) approaches.
• Utilized noninformative prior distributions in the Bayesian analysis for parameter estimation.

Main Results:

• All four nonlinear models closely fitted the observed llama body weight data.
• The Brody model demonstrated the best fit for describing llama growth in both frequentist and Bayesian analyses.
• Female llamas and K'ara-type llamas exhibited significantly higher asymptotic weights than males and Ch'accu-type llamas, respectively.
• The Brody model estimated an asymptotic body weight of 42 kg at 12 months for Peruvian llamas.

Conclusions:

• Nonlinear functions, particularly the Brody model, are effective for modeling the weight-age relationship in llamas.
• The Bayesian approach offers a robust framework for estimating growth parameters in livestock.
• Future research should incorporate prior knowledge to refine Bayesian models and address limitations of historical data for current growth pattern representation.