Related Experiment Video
Updated: May 23, 2026
07:34
Probing the Limits of Egg Recognition Using Egg Rejection Experiments Along Phenotypic Gradients
Published on: August 22, 2018
Canonical discriminant analysis applied to broiler chicken performance
M F Rosário1, M A N Silva, A A D Coelho
11Department of Genetics, 'Luiz de Queiroz' College of Agriculture, University of São Paulo, Av. Pádua Dias, 11, PO Box 83, 13400-970, Piracicaba, São Paulo, Brazil.
Animal : an International Journal of Animal Bioscience
|March 27, 2012
Summary
Canonical discriminant analysis revealed significant differences in broiler chicken performance between sexes and strains. This multivariate approach simplifies complex growth trait evaluation, identifying key factors for improved broiler production.
Area of Science:
- Animal Science
- Quantitative Genetics
- Poultry Production
Background:
- Chicken growth control involves complex, biologically correlated traits often oversimplified by univariate analysis.
- Multivariate statistical approaches are necessary to accurately assess and understand broiler chicken performance.
Purpose of the Study:
- To evaluate broiler chicken performance using a multivariate approach, specifically canonical discriminant analysis.
- To identify key performance traits and differentiate between sexes and broiler strains.
Main Methods:
- Utilized canonical discriminant analysis (SAS® CANDISC procedure) on data from 1920 broiler chickens across eight treatments (four strains, both sexes).
- Collected data on average feed intake, live weight, feed conversion, carcass, breast, and leg weights from day 1 to 42.
- Employed SAS® CONELIP macro for visualization of multivariate performance using canonical variables.
Main Results:
- Canonical discriminant analysis effectively reduced six performance traits into two canonical variables, explaining 96.49% of the total variation.
- Significant distinctions in performance were observed between sexes (males superior to females) and strains (Arbor Acres, AgRoss 308, Cobb 500 superior to RX).
- Average live weight and carcass weight were the primary discriminators (first canonical variable), while feed intake and feed conversion contrasted with live weight (second canonical variable).
Conclusions:
- Canonical discriminant analysis is a powerful tool for simplifying and visualizing complex broiler chicken performance data.
- The study highlights significant performance differences attributable to sex and genetic strain, offering insights for breeding and management.
- Recommends the adoption of canonical discriminant analysis for analyzing broiler performance datasets to better understand trait interrelationships.
