Get access

Multivariate Analyses for Determining the Association of Field Porcine Fertility With Sperm Motion Traits Analysed by Computer-Assisted Semen Analysis and With Sperm Morphology


Author's address (for correspondence): FP Bortolozzo, Setor de Suínos, Faculdade de Veterinária, UFRGS, Av. Bento Gonçalves, 9090 – CEP 91540 000, Porto Alegre, Brazil. E-mail:


This study investigates the association of semen traits with boar fertility. The fertility outcome (farrowing rate – FR and total piglets born – TB) of 14 boars was obtained from a field trial conducted during 10 week of breeding period on a commercial farm using multiparous sows (n = 948) through single-sire mating with 2 × 109 motile sperm cells per artificial insemination (AI) dose. Sperm motion parameters, evaluated with computer-assisted semen analysis system in raw and stored semen at 17°C for 240 h, in addition to morphological sperm defects, measured on the collection day, were included in the analysis to determine which semen traits were important to discriminate the fertility potential of ejaculates from these boars. The data underwent multivariate cluster, canonical and discriminant analyses. Four clusters of boars were formed based on fertility outcome. One boar, with the lowest FR and TB values (89.7% and 11.98), and two boars, with the highest FR and TB values (97.8% and 14.16), were placed in different clusters. The other boars were separated in two distinct clusters (four and seven boars), including boars with intermediate TB (12.64 and 13.22) but divergent values for FR (95.9% vs 91.8%). Semen traits with higher discriminatory power included total motility, progressive motility, amplitude of lateral head displacement and cytoplasmatic droplets. Through multivariate discriminant analysis, more than 80% of the 140 ejaculates were correctly classified into their own group, showing that this analysis may be an efficient statistical tool to improve the discrimination of potential fertility of boars. Nevertheless, the validation of the relationship between fertility and semen traits using this statistical approach needs to be performed on a larger number of farms and with a greater number of boars.