Development of a Feed Conversion Ratio Prediction Model for Yorkshire Boars Using Cumulative Feed Intake

Area of Science:

• Animal Science
• Agricultural Engineering

Background:

• Feed conversion ratio (FCR) is crucial for pig productivity but difficult to measure.
• Efficient FCR estimation is needed to reduce swine industry costs and labor.

Purpose of the Study:

• To develop a predictive model for cumulative feeding intake (CFI) to estimate FCR.
• To improve the efficiency of FCR measurement in pigs.

Main Methods:

• Utilized data from 987 Yorkshire boars using automatic feeders.
• Developed a predictive model for CFI using the segmented R package and Bayesian ridge regression (BRR).

Main Results:

• The optimal body weight range for FCR prediction was identified as 80-110 kg.
• The BRR model achieved 80% accuracy for CFI prediction.
• FCR derived from predicted CFI showed 81.4% similarity to corrected FCR.

Conclusions:

• The BRR model demonstrates strong predictive potential for FCR, even with limited data.
• This approach can significantly reduce production costs and measurement time in the swine industry.
• Findings support reduced selection pressure on FCR traits, benefiting swine production.