Quantile Regression for Longitudinal Functional Data with Application to Feed Intake of Lactating Sows.

Area of Science:

• Animal Science
• Statistical Modeling
• Environmental Physiology

Background:

• Understanding factors influencing feed intake in lactating sows is crucial for optimizing animal health and productivity.
• Environmental temperature is a key factor affecting animal metabolism and behavior, including feed consumption.
• Previous models may not fully capture the dynamic and quantile-specific effects of temperature on feed intake.

Purpose of the Study:

• To investigate the influence of daily temperature fluctuations on the lower quantiles of feed intake in lactating sows.
• To develop and apply a robust statistical framework for analyzing longitudinal data with functional covariates.
• To quantify the time-varying association between temperature and sow feed intake.

Main Methods:

• Utilized a quantile regression model with time-varying coefficients to analyze longitudinal data.
• Incorporated subject-specific intercepts to account for within-subject dependence.
• Employed spline representations for coefficient functions and bootstrap procedures for bias adjustment and standard error computation.

Main Results:

• The study identified a significant influence of temperature on the lower quantiles of daily feed intake in lactating sows.
• The impact of temperature on feed intake was observed to increase throughout the lactation period.
• The developed model framework effectively handles longitudinal data and functional covariates.

Conclusions:

• Quantile regression provides a powerful tool for understanding complex relationships between environmental factors and animal intake.
• Temperature management strategies may need to be adjusted dynamically throughout the lactation cycle to support sow feeding behavior.
• The findings contribute to improved precision livestock farming and animal welfare in swine production.