Suitability of semen stress tests for predicting fertilizing capacity of boar ejaculates

Besides classical semen parameters, semen stress tests (SSTs) are helpful tools to assess the fertilizing capacity of sperm. However, valid studies on SSTs in relation to fertility are rare because several corrections of common fertility parameters for female and male effects are required. Therefore, over a one-year period, we analyzed semen parameters of 260 ejaculates obtained from 130 Pietrain boars aged between 8 and 9 months in one AI center as well as 1521 corresponding insemination records for these ejaculates. Two consecutive ejaculates (4th and 5th) were collected from each boar and extended in DiluPorc™ BTS. In addition to routine semen evaluation, sperm motility was assessed after heat-resistance test (300 min incubation at 38 °C after seven days storage at 16 °C, HRT) and cold-resistance test (10 min incubation at 38 °C after three days storage at 6 °C, CRT). Generalized linear models (GLMs) were applied to analyze effects for the following predictors of farrowing rate (FR), number of total born (NTB) and live born (NLB) piglets: farm (P = 0.013 [FR], P = 0.001 [NTB], P = 0.023 [NLB]), parity (P = 0.679, P = 0.01, P < 0.001), weekday (P = 0.012, P = 0.08, P = 0.009) and year × season (P < 0.001, P = 0.688, P = 0.574). On boar level, GLMs revealed significant effects on FR, NTB and NLB for the predictors sow (all P < 0.001), total sperm number per dose (P = 0.007, P = 0.002, P < 0.001), total sperm motility (P = 0.002, P = 0.2, P = 0.003) and mitochondrial activity (P = 0.004, P < 0.001, P = 0.002). Moreover, FR and NTB were influenced by membrane integrity (both P < 0.001), FR and NLB by cold-resistance (P < 0.001, P = 0.043), and NTB and NLB by sperm morphology (P = 0.001, P < 0.001) and boar (both P < 0.001). NLB was additionally influenced by heat-resistance (P = 0.004) and farm (P = 0.018) and solely NTB was influenced by sperm output (P = 0.03). Boar and semen related factors explained 9% of the total variation in NTB and 7% of the total variation in NLB. Only 14.2% (n = 37) of the samples were both cold- and heat-resistant (≥65% of motile sperm). Cold- and heat-resistance were dependent factors (Chi-square, P = 0.001) and sperm motility after CRT and HRT showed a moderate positive correlation (r_s = 0.40, P < 0.001, Spearman's rho). Finally, ROC curves demonstrated that neither SST can be used as a sole test for predicting the fertilizing capacity of boar ejaculates.