Use of canonical correspondence analysis in animal science data

This study investigates the relationship between genotype and egg quality traits using Canonical Correspondence Analysis (KUA) and Principal Component Analysis (PCA), which are powerful multivariate statistical techniques. Two different genotypes of laying hens were evaluated based on six key egg quality parameters: egg weight, albumen index, Haugh unit, yolk index, yolk color score, and pH. The aim of the study is to assess how genotype affects overall egg quality variation and to visualize these effects through multidimensional data reduction and association models. The KUA results revealed that genotype significantly influences the combination of egg quality traits (F = 7.10, p = 0.0082), with clear separation observed in KUA1 scores between groups. PCA explained 53.9% of total variance across the first two components, with PC1 associated primarily with internal quality traits (e.g., albumen index, Haugh unit), and PC2 reflecting external and color-related parameters (e.g., yolk score). The PCA triplot showed that the genotypes formed distinct clusters, driven by opposing influences of yolk color and internal freshness indicators. Additionally, correlation analysis demonstrated strong positive associations among traits such as albumen index and Haugh unit, while yolk score appeared to be inversely related. Regression-based performance metrics indicated that multivariate models explained only limited variance individually (R² < 0.06), highlighting the need for holistic multivariate interpretation. Genotypic comparisons confirmed that Genotype 2 excelled in internal quality traits, whereas Genotype 1 had higher yolk coloration. These findings demonstrate that genotype plays a significant but complex role in shaping egg quality traits, and support the use of KUA and PCA for detailed phenotypic evaluation in poultry breeding and selection programs.