Multivariate cluster analysis regression procedures as tools to identify motile sperm subpopulations in rabbit semen and to predict semen fertility and litter size

Computerized motility analysis (CASA) shows that four separate subpopulations of spermatozoa with different motility characteristics co-exist in rabbit ejaculates. There were significant (p < 0.01) differences in the distribution of these subpopulations among separate genetic lines, total sperm abnormalities and the percentage of altered acrosomes. Furthermore, logistic and linear multivariate regressions among several parameters of rabbit semen quality analysis were tested for use as predictive tools for the fertilizing ability of a specific artificial insemination semen sample. Logistic regression analysis rendered two mathematical, significant (p < 0.01) models: one between sperm viability and conception rate and the other between total sperm abnormalities and conception rate. Multiple linear regression analyses also yielded some significant relationships between both fertility (p < 0.001) and litter size (p < 0.05), with respect to some semen characteristics. Our results support the hypothesis that the predictive in vivo fertility use of the standard rabbit semen quality analysis coupled with a CASA determination could be reasonably achieved by applying linear and logistic regression analyses among several parameters of rabbit semen quality analysis.