Two on-farm data collection methods to determine dynamics of swine compositional growth and estimates of dietary lysine requirements

Modelling batch farrowing management within a farrow-to-finish pig herd: influence of management on contact structure and pig delivery to the slaughterhouse

Pathogen spread within pig host populations can vary depending on within-herd interactions among pigs also called the contact structure. The recommended batch farrowing management, allowing for a fixed-interval mating for groups of sows of equal size, called batches, leads to an all-in/all-out management of pigs in which animals in different batches have no contact. To maintain a profitable pig delivery, producers have to deliver groups of pigs at a given weight, what needs sometimes herd management adaptations. However, producers' adaptations that avoid delivering pigs below slaughtering weight (out-of-range pigs), result in increasing the contact between animals from different batches. To study the influence of herd management on contact structure and on pig delivery, a stochastic mathematical model representing population dynamics within a farrow-to-finish herd was elaborated. Sixteen management systems were represented combining or not the all-in/all-out management system with producers' decisions: batch mixing, use of an extra room, suppression of the drying period and sale of post-weaning batches. Two types of contact were considered: via the animals themselves, when batch mixing occurred; and via the room, when decontamination was not complete. The impact of producers' decisions on contact structure and on pig delivery, differed radically when pig growth was normal and when it was slow (i.e. mean age at slaughtering weight increased by 20%). When pig growth was normal, the all-in/all-out management prevented both contact via the animals and via the room but resulted in 9% of pigs delivered out of range. The use of an extra room or batch mixing decreased this percentage, the latter resulting in very frequent contact between batches via the animals. When pig growth was slow, the all-in/all-out management led to a very high percentage of pigs delivered out of range (almost 80%). The suppression of the drying period at the end of the finishing period and the sale of post-weaning batches induced a significant decrease in this percentage (down to 2% to 20%), the latter allowing to reduce the percentage of batches that made contact via the room (40% instead of 80%). This pig herd model helped to understand the compromise for producers between implementing internal biosecurity or maintaining a profitable pig delivery. Our results show that there was no unique optimal system and that efficient producers' decisions (for biosecurity and delivery) may differ, depending on pig growth.

Development of sustainable precision farming systems for swine: estimating real-time individual amino acid requirements in growing-finishing pigs

The objective of this study was to develop and evaluate a mathematical model used to estimate the daily amino acid requirements of individual growing-finishing pigs. The model includes empirical and mechanistic model components. The empirical component estimates daily feed intake (DFI), BW, and daily gain (DG) based on individual pig information collected in real time. Based on DFI, BW, and DG estimates, the mechanistic component uses classic factorial equations to estimate the optimal concentration of amino acids that must be offered to each pig to meet its requirements. The model was evaluated with data from a study that investigated the effect of feeding pigs with a 3-phase or daily multiphase system. The DFI and BW values measured in this study were compared with those estimated by the empirical component of the model. The coherence of the values estimated by the mechanistic component was evaluated by analyzing if it followed a normal pattern of requirements. Lastly, the proposed model was evaluated by comparing its estimates with those generated by the existing growth model (InraPorc). The precision of the proposed model and InraPorc in estimating DFI and BW was evaluated through the mean absolute error. The empirical component results indicated that the DFI and BW trajectories of individual pigs fed ad libitum could be predicted 1 d (DFI) or 7 d (BW) ahead with the average mean absolute error of 12.45 and 1.85%, respectively. The average mean absolute error obtained with the InraPorc for the average individual of the population was 14.72% for DFI and 5.38% for BW. Major differences were observed when estimates from InraPorc were compared with individual observations. The proposed model, however, was effective in tracking the change in DFI and BW for each individual pig. The mechanistic model component estimated the optimal standardized ileal digestible Lys to NE ratio with reasonable between animal (average CV = 7%) and overtime (average CV = 14%) variation. Thus, the amino acid requirements estimated by model are animal- and time-dependent and follow, in real time, the individual DFI and BW growth patterns. The proposed model can follow the average feed intake and feed weight trajectory of each individual pig in real time with good accuracy. Based on these trajectories and using classical factorial equations, the model makes it possible to estimate dynamically the AA requirements of each animal, taking into account the intake and growth changes of the animal.