Deciphering cattle temperament measures derived from a four-platform standing scale using genetic factor analytic modeling

Abstract

Background: The animal's reaction to human handling, also known as temperament, is critical for work safety, productivity, and welfare. Subjective phenotyping methods, such as docility score, have been traditionally used in cattle production as a means for improving the safety, productivity, and welfare of animals. Even so, subjective scales rely on the evaluator's knowledge and interpretation of temperament, which may require substantial experience. With that being said, selection based on such subjective scores may not precisely impact temperament preferences in cattle.Results: We investigated the statistical relationships among subjective methods including docility score (DS), temperament score (TS), 12 qualitative behavior assessment (QBA) attributes along with a movement-based objective method (four-platform standing scale, FPSS) including the standard deviation of total weight on FPSS over time (SSD) and coefficient of variation of SSD (CVSSD) using 1,528 calves at weaning age. An exploratory factor analysis (EFA) identified two underlying latent variables contribute to TS and 12 QBA attributes that were named difficult and easy according to their characteristics. As DS is evaluated under constraint and other measures are not, inclusion of DS in EFA of subjective methods was not a good fit.A Bayesian confirmatory factor analysis inferred the factor scores of difficult and easy using the latent structure discovered in EFA. This was followed by fitting a pedigree-based Bayesian multi-trait model using difficult, easy, DS, SSD, and CVSSD to characterize the genetic interrelationships between subjective and FPSS measures. The estimates of heritability range from 0.17 to 0.4 with the posterior standard deviation averaging 0.06. The factors of difficult and easy exhibited a large negative genetic correlation of -0.92. DS displayed a moderate genetic correlation with difficult (0.36), easy (-0.31), SSD (0.42), and CVSSD (0.34).FPSS and DS were found to have a moderate genetic correlation with difficult (0.35 to 0.41) and easy (-0.39 to -0.31), indicating selection could be performed with either and have similar outcomes.Conclusions: The application of genetic factor analytic model for temperament measures provided a new approach to unravel the complexity of animal behaviors. In summary, we contend that FPSS measures could increase the efficiency of genetic selection by providing automatic, objective, and consistent phenotyping measures with a lower cost, and may potentially become an alternative of DS which has been widely used in beef production.3

Modeling multiple phenotypes in wheat using data‐driven genomic exploratory factor analysis and Bayesian network learning

“…This section closely follows the work of Yu et al. (2020). The aforementioned t = 45 phenotypes were analyzed using EFA by fitting Where Y is the t × n phenotypic matrix; Λ is the t × q matrix of factor loading indicating the relation between phenotypes and latent common factors; F is the q × n matrix of latent factor scores; and U is the t × n vector of unique effects that is not explained by q underlying common factors.…”

“…With the availability of large volumes of measured observations per individual because of recent advances in phenomics, it is critical to develop a phenotype‐centric statistical approach. Factor analysis is an effective method for handling many response variables in a quantitative genetic framework (Peñagaricano et al., 2015; Rocha et al., 2018; Runcie & Mukherjee, 2013; Yu et al., 2019, 2020). The central idea behind factor analysis is to model the observed phenotypes through unobserved latent factors by maximizing the common variance between correlated phenotypes.…”

“…(2019), Yu et al. (2020) using the default priors in the blavaan R package (Merkle & Rosseel, 2018). A Gaussian distribution with a mean of zero and variance of 100 was assigned to the factor loading term.…”

A Systematic Review of Genomic Regions and Candidate Genes Underlying Behavioral Traits in Farmed Mammals and Their Link with Human Disorders

“…Animal behavior can be assessed through a plethora of measurements (as reviewed by Brito et al [ 74 ]). Similar performance and outcomes for behavioral traits in cattle (i.e., docility, temperament score, and qualitative behavior assessments through four-platform standing scale; [ 113 ]) and in pigs (i.e., struggling bouts and feeding behavior; [ 114 ]) were observed. However, null genetic correlations were also found among other behavior-related traits, such as activity score (i.e., mild and excited scoring), struggling bouts, and feeding behavior [ 114 ].…”

Differences in evaluators and genetic parameter estimations using subjective measurements of beef cattle temperament

“…Objective temperament methods include exit velocity ( Burrow et al, 1988 ) and movement measuring devices ( Sebastian et al, 2011 ; Yu, 2016 ). Subjective methods include chute score ( Grandin, 1993 ), pen score ( Hearnshaw and Morris, 1984 ), and qualitative behavior attributes (QBA; Sant’Anna and Paranhos da Costa, 2013 ).…”