Interfacing mechanistic and breeding scheme simulation to predict selection response on lactation efficiency in dairy cows under different nutritional environments.

backgroundPredicting selection response for lactation efficiency in dairy cows is challenging, as the expression of this complex trait depends on dynamic interactions between the ability of cows to acquire nutrients and allocate them to different life functions. Moreover, the relative emphasis of these components may change due to energetic trade-offs between life functions when kept in limiting environments. The objective of this study is to present a new approach combining mechanistic and breeding scheme simulations to predict selection response on components of lactation efficiency of dairy cows under a non-limiting nutritional environment and when transferred to a limiting environment with a moderate feed restriction. These predictions were compared to the ones obtained with the conventional method used in quantitative genetics considering a typical dairy cattle breeding scheme and several breeding goals (BG) aiming at improving milk production, lactation efficiency and fertility.

resultsIn the non-limiting environment, selection responses predicted by the two methods differed for both milk production and fertility. The sign and magnitude of differences depended on BGs. Selection response predictions were consistent only for BGs that did not change much the body reserve mobilization patterns of cows, and hence their conception probability. Indeed, pregnancy status impacted energy allocation of cows and consequently milk production, more energy being allocated to lactation in case of reproductive failure. Differences in selection response between modelling approaches were slightly increased when cows were reared in the limiting environment. Overall, the choice of prediction method led to substantial BG reranking with respect to selection response on milk production and fertility. Mechanistic-based predictions of selection response for lifetime efficiency were also sensitive to the nutritional environment and BG.

conclusionsCombining mechanistic and genetic modelling is a promising approach to benchmark breeding strategies of dairy cow lactation efficiency and better anticipate the impact of changes in energetic trade-offs induced both by selection and changes in the nutritional environment. Moreover, the simulations of phenotypic trajectories over cow lifetime before and after selection enabled a better understanding of the mechanisms underlying genetic improvement.