Repeatability, reproducibility and consistency of horse shape data and its association with linearly described conformation traits in Franches-Montagnes stallions.
Authors: Gmel Annik Imogen, Druml Thomas, Portele Katrin, von Niederhäusern Rudolf, Neuditschko Markus
Journal: PloS one
Summary
# Editorial Summary Conformation assessment in horses relies on linear description scoring to reduce subjectivity, yet these methods still suffer from convergence around mean values and limited discrimination. Researchers examined whether a geometric morphometric approach—analysing digitised photographs of 243 Franches-Montagnes stallions using landmark-based shape analysis and derived joint angles—could provide more objective, repeatable data on conformation traits. The digitisation process itself proved highly reliable (repeatability 0.72–0.99), though inter-digitiser variability in specific landmarks such as the shoulder point was concerning (intra-class correlation coefficients ranging 0.054–0.92); postural variation and age significantly influenced shape consistency. Shape data correlated meaningfully with several linear description traits including withers height, shoulder morphology, leg conformation and stride characteristics, whilst offering additional morphological detail beyond traditional scoring systems. For practitioners involved in breeding decisions or performance assessment, this geometric morphometric approach holds promise as a complement to existing evaluation methods, though standardised digitisation protocols and operator training would be essential before clinical application—particularly when precision across multiple assessors matters.
Read the full abstract on PubMed
Practical Takeaways
- •Digitised shape analysis is more objective and repeatable than traditional conformation scoring, offering potential for more standardised breeding decisions, though landmark identification requires careful training to minimise between-assessor error
- •Standardise age and posture during conformation photography to improve consistency and comparability of assessments across individual horses
- •Shape-based assessment can quantify relationships between body morphology and functional traits like stride length, helping breeders select for both appearance and movement quality
Key Findings
- •Digitised shape analysis showed high repeatability (Rep = 0.72-0.99) and reproducibility (ICC = 0.99), substantially more objective than traditional linear description scoring
- •Angle measurements between different digitisers showed variable reproducibility (ICC = 0.054-0.92), with shoulder point landmark being particularly prone to error
- •Age and posture significantly influenced shape consistency (p < 0.05), indicating these factors must be controlled in conformation assessment protocols
- •Shape data identified significant associations with withers height, shoulder length/incline, leg conformation, and stride length traits, providing additional information beyond conventional linear scoring