Evaluation of direct and maternal responses in reproduction traits based on different selection strategies for postnatal piglet survival in a selection experiment

Abstract

Background: Postnatal piglet survival is important both in economic and animal welfare terms. It is influenced by the piglet's own direct genetic effects and by maternal genetic effects of the dam, associated with milk production and mothering abilities. These genetic effects might be correlated, affected by other non-genetic factors and unfavourably associated with other reproduction traits such as litter size, which makes the development of optimal breeding strategies a challenge. To identify the optimum selection strategy for piglet survival, a selection experiment was carried out to compare responses in survival and reproduction traits to selection on only direct, only maternal, or both genetic effects of postnatal survival. The data of the experiment were recorded from outdoor reared pigs, with first- and second-generation sires selected based on their estimated breeding values for maternal and direct effects of postnatal survival of indoor reared offspring, respectively, with the opportunity to identify potential genotype-by-environment interaction.

Results: A Bayesian multivariate threshold-linear model that was fitted to data on 22,483 piglets resulted in significant (Pr(h² > 0) = 1.00) estimates of maternal and direct heritabilities between 0.12 and 0.18 for survival traits and between 0.29 and 0.36 for birth weight, respectively. Selection for direct genetic effects resulted in direct and maternal responses in postnatal survival of 1.11% ± 0.17 and - 0.49% ± 0.10, respectively, while selection for maternal genetic effects led to greater direct and maternal responses, of 5.20% ± 0.34 and 1.29% ± 0.20, respectively, in part due to unintentional within-litter selection. Selection for both direct and maternal effects revealed a significant lower direct response (- 1.04% ± 0.12) in comparison to its expected response from single-effect selection, caused by interactions between direct and maternal effects.

Conclusions: Selection successfully improved post- and perinatal survival and birth weight, which indicates that they are genetically determined and that genotype-by-environment interactions between outdoor (experimental data) and indoor (selection data) housed pigs were not important for these traits. A substantially increased overall (direct plus maternal) response was obtained using selection for maternal versus direct or both direct and maternal effects, suggesting that the maternal genetic effects are the main limiting factor for improving piglet survival on which selection pressure should be emphasized.

Fig. 1

Differences in estimated breeding values between selected and control boars at mating and phenotypic selection responses of piglet survival during the nursing period estimated at the sow and piglet levels. Estimated breeding values (EBV) of Landrace and Large White boars at the time of mating were based on survival (SURV) and mortality (MOR) for maternal and direct effects, respectively. $\text{H}$ and $\text{C}$ represent high and control breeding groups and subscripts $\text{D}$ and $\text{M}$ denote direct and maternal genetic effects; *: significantly different from 0 (P < 0.05), ns: non-significant

Fig. 2

Differences in estimated breeding values of piglet survival during the nursing period between selected ($\mathbf{H}$) and control ($\mathbf{C}$) boars based on performances realised within the selection experiment and the selection responses in direct ($\mathbf{D}$) and maternal ($\mathbf{M}$) effects of the same trait in crossbred sows and in third-generation piglets originating from different selection scenarios for direct (${\mathbf{H}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$) or maternal (${\mathbf{C}}_{\mathbf{D}}{\mathbf{H}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$) effects only or their combination (${\mathbf{H}}_{\mathbf{D}}{\mathbf{H}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$)

Fig. 3

Differences in estimated breeding values of piglet survival at birth of boars selected on postnatal survival and the correlated selection responses estimated as differences between the high ($\mathbf{H}$) and control ($\mathbf{C}$) groups in direct ($\mathbf{D}$) and maternal ($\mathbf{M}$) effects of the survival at birth in crossbred sows and in third-generation piglets originating from different selection scenarios for direct (${\mathbf{H}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$) or maternal (${\mathbf{C}}_{\mathbf{D}}{\mathbf{H}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$) effects only or their combination (${\mathbf{H}}_{\mathbf{D}}{\mathbf{H}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$)

Fig. 4

Differences in estimated breeding values of piglet individual birth weight of boars selected on postnatal survival and the correlated selection responses estimated as differences between the high ($\mathbf{H}$) and control ($\mathbf{C}$) in direct ($\mathbf{D}$) and maternal ($\mathbf{M}$) effects of birth weight in crossbred sows and in third-generation piglets originating from different selection scenarios for direct (${\mathbf{H}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$) or maternal (${\mathbf{C}}_{\mathbf{D}}{\mathbf{H}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$) effects only or their combination (${\mathbf{H}}_{\mathbf{D}}{\mathbf{H}}_{\mathbf{M}}-{\mathbf{C}}_{\mathbf{D}}{\mathbf{C}}_{\mathbf{M}}$)