Optimising insemination strategies in pigs

INTRODUCTIONReproductive efficiency shows large variation between farms. The origin of the variation between farms, with respect to these reproduction results, is very complex. Factors like health status, husbandry system, management and breed can have an influence on reproduction results. One of the management factors is timing of insemination, which influences reproduction results by affecting fertilisation.The research described in this thesis deals with the possibility of developing a method to optimise insemination strategies for individual farms. Therefore three objectives were formulated: the first objective is increasing insight in the effects of the interval between insemination and ovulation on fertilisation results. The second objective is increasing knowledge on the possibilities of predicting the moment of ovulation of sows at a farm. The final objective is developing a method which can be used for optimising insemination strategies at commercial farms.FERTILISATION IN RELATION TO INSEMINATION AND OVULATIONIn Chapter 2 the sensitivity of the relation of the insemination to ovulation interval (IO) and fertilisation results is studied. Fertilisation results are not very sensitive to variation in the number of inseminated sperm cells in the range of 1 x 10 9 to 6 x 10 9 sperm cells (Chapter 2.1). Sows with more than 4 ml backflow of semen during insemination had reduced fertilisation results when the sows were inseminated with 1 x 10 9 sperm cells, but this was not seen with an insemination dosage of 3 x 10 9 or 6 x 10 9 sperm cells (Chapter 2.2). Backflow of semen after insemination did not affect fertilisation results. It could be concluded that sub-optimal circumstances like a combination of a low dosage and loss of sperm cells due to backflow during insemination, lead to sub-optimal fertilisation results.Fertilisation is a complex process, resulting in no, partial or complete fertilisation of the oocytes. The variation in conception (at least one oocyte fertilised) and fertilisation rate between sows is high, but a large part of the variation is related to the interval between insemination and ovulation. A mathematical model for conception and fertilisation is described in Chapter 2.3. The data used for estimating the parameters in the model were derived from multiparous sows that were inseminated once with a commercial sperm dose of 3 x 10 9 sperm cells of proven quality which was stored for less than 48 h and with sperm cells. In the model, the probability of conception is maximal (98%), when insemination is performed between 29 and 3 h before ovulation. The probability of complete fertilisation (all oocytes fertilised) is maximal when insemination was performed at 9.6 h before ovulation. At this optimal fertilisation point, the probability of partial fertilisation is 21% which increases beyond this point.PREDICTION OF OVULATIONFertilisation results are related to the interval between insemination and ovulation. Therefore, the moment of ovulation is a crucial moment for timing of insemination. Many potential ovulation predictors have been studied, but only oestrus duration is a reasonable estimate (retrospectively) for ovulation. Ovulation takes place at on average twothirds of oestrus. Unfortunately oestrus duration is very variable.The average oestrus duration is different between farms ranging between 31 and 64 h (Chapter 3.1). Moreover, oestrus duration is consistent from month to month within a farm with a repeatability of 86%. Furthermore, oestrus duration is negatively related to the weaning to oestrus interval. This relation differs among farms. These specific farm parameters can be used to predict the oestrus duration and from that the ovulation can be predicted. These farm parameters (average oestrus duration and the relation of weaning to oestrus interval and oestrus duration) can be used to define a specific insemination strategy for each farm.DEVELOPMENT OF A MODEL FOR INSEMINATION STRATEGIESThere are a variety of factors influencing the reproduction process. The complexity of this reproduction process makes a modelling and simulation approach valuable because effects of the underlying processes can be controlled. A PIG Simulation model for Insemination strategies (PIGSIS) was developed which consists of two parts: (1) the reproduction events from the number of ovulated oocytes until the number of piglets at farrowing and (2) timing of insemination relative to ovulation based on the farm parameters (weaning to oestrus interval, oestrus duration, etc.). PIGSIS simulates the reproduction results at day 1, 5, 10, 15, 35 and 110 of pregnancy. Many physiological processes are included in PIGSIS e.g. fertilisation, embryonic mortality (degeneration, maternal recognition of pregnancy, embryonic uterine capacity) and foetal mortality (foetal uterine capacity). After verification and validation it could be concluded that PIGSIS is a robust model that reasonably simulates reproduction results. Under the basic situation (average oestrus duration of 47 h and average parity of 4.2) and when insemination was applied between 0 and 24 h before ovulation PIGSIS simulates 12.9 total born piglets and a farrowing rate of 94.9%. Under these conditions the average embryonic and foetal mortality of the conceptuses was 34.9% and 3.0%, respectively. The effect of insemination to ovulation interval on fertilisation results is clear, but the effect becomes less clear as gestation proceeds resulting in a more pronounced effect on litter size than on farrowing rate.In the General discussion the results of the studies are discussed and an illustration of the usability of PIGSIS is given. Verification and partial validation gave confidence in the model. However, a further validation is required to evaluate the model as a whole. Therefore PIGSIS is still in its developing stage and reservations has to be taken into account at this stage by using PIGSIS for defining optimal insemination strategies on farms