Discriminant Canonical Analysis as a tool to determine traces of endangered native hen breed introgression through egg hatchability phenomics.

Abstract

The main objective of this study was to develop a pipeline to detect phenogenomic introgression across different multivariety breeds and to validate such classification focusing on external egg and hatchability-related traits using a discriminant canonical analysis approach. For this, 1368 eggs belonging a flock of 94 endangered Spanish autochthonous breed breeding hens (Andalusian Tufted, Blue Andalusian, Spanish White-Faced, and Utrerana) and a control outgroup comprising 32 eggs belonging to 4 Araucana hens were considered. Multicollinearity analysis of hatchability-related traits revealed embryonic mortality, embryonic mortality in the second stage of incubation, viable hatching chick, major diameter, and minor diameter should be discarded from the analysis (VIF≤5) given they did not significantly contribute to variability explanation potential of the discriminant model. A stepwise discriminant canonical analysis was developed and egg weight, shape index, hatchability, and fertility variables reported the highest discriminant power (Wilks' Lambda values of 0.7861, 0.7871, 0.8076, and 0.9457, respectively). The first two functions explained 85.25% intergroup variability. Interbreed and varieties proximity was evaluated using Mahalanobis distances representation and data mining cross-validation allowed to detect genetic introgression between different genotypes. Easily collectable traits as egg weight and shape index must be considered for the development of breeding programs as a measure to ensure breed protection. The model may be translatable to other endangered breeds to optimize avian breeds conservation plans worldwide.