Multivariate structure of semen quality and growth performance in Limousin beef bulls: a CoSTATIS approach

In beef bull evaluation, semen quality and growth performance are routinely assessed, yet their multivariate relationship has not been systematically characterized. The present work aimed to characterize the multivariate organization of semen quality and growth performance in Limousin bulls and to assess the extent of shared structure between these domains using a multi-year analytical framework. Semen quality, including sperm kinematics and sperm concentration (CON), and growth performance data were collected from 30 Limousin bulls evaluated over three consecutive years (2021-2023) under standardized performance testing conditions. Multivariate analyses were conducted using STATIS DUAL to identify stable internal structures within each block of variables and CoSTATIS to evaluate co-structure between reproductive and productive traits across years. STATIS DUAL revealed internally consistent multivariate structures for both semen quality and growth performance traits despite differences in annual animal cohorts. Semen quality traits were organized into coherent functional groups dominated by sperm kinematic parameters, whereas quantitative ejaculate traits such as CON formed a distinct structural component. CoSTATIS analysis showed that the co-structure between semen quality and growth performance was captured within a low-dimensional space, with the first two co-inertia axes explaining 94.94% of the total co-inertia (axis 1: 80.07%; axis 2: 14.87%). The shared structure was driven mainly by CON and curvilinear velocity (VCL), while most semen quality traits contributed weakly to the common space. These results indicate that semen quality and growth performance in Limousin bulls represent largely independent multivariate biological structures with only partial overlap, highlighting the importance of evaluating reproductive and productive traits as complementary but non-redundant components in breeding and performance testing programs.