The identification of reliable estimates of growth potential and resilience over the fattening period in large populations is a challenge in actual swine breeding conditions. To overcome this drawback, the study by Revilla et al. 2021 in the frame of precision livestock farming aimed to propose an innovative modelling approach, in addition to previous studies from the same group (Revilla et al. 2019), to enhance the quality of the quantification of pig resilience during the entire fattening period. 

The authors developed a model that quantifies an “individual pig resilience indicator” based on longitudinal data, for instance body weight, recorded routinely by a commercially available automatic feeding system. Revilla and co-workers considered in their study two mainly commercialised pure pig breeds these being Piétrain including Piétrain Français NN Axiom line (Pie NN) free from halothane-sensitivity (ryanodine receptor gene, RYR1) and Piétrain Français Axiom line positive to this gene and Duroc. Therefore, the authors investigated the potential of improving resilience of swine livestock through inclusion for the first time of an “individual pig resilience indicator” in breeding objectives. A database of 13 093 boars (approximately 11.1 million of weightings) belonging to Pie (n= 5 841), Pie NN (n = 5 032) and Duroc (n= 2 220) finished under ad libitum feeding, high sanitary level and controlled temperature was used to develop robust models.

The authors checked the three datasets (for each pig breed)​ independently to explore the variation and gaps (a data pre-treatment procedure) to ensure high quality data for the modelling approach. Then, they applied the Gompertz model and linear interpolation on body weight data to quantify individual deviations from the expected production, allowing the creation of the ABC index. For the modelling, the authors applied a two-step mathematical model approach by first establishing a theoretical growth curve of each animal, while the second step aimed to build the actual perturbed growth curve. The heritability of the index ranged from 0.03 to 0.04, with similar heritability between Piétrain and Duroc breeds. Moreover, moderate genetic relationships were computed between the proposed index and important phenotypic traits in swine production likely BF100: backfat thickness at 100kg; LD100: longissimus dorsi thickness at 100kg; ADG: average daily gain during control and FCR: feed conversion ratio.

Developing models able to capture perturbations during the fattening period is a challenge in swine breeding industry. The model and methodology proposed by the authors in this innovative work (although preliminary and with low heritabilities) would help overcome such limit and facilitate a real implementation at large scale in pig breeding system. The modelling approach further offers an opportunity to develop a selection criterion to improve resilience in swine breeding conditions. 

To explore the full potential of this modelling approach, a larger database and other factors such as breed, behaviour and feeding behaviour of the animals, rearing practices, management and environment conditions, age… etc. are worthy to consider. In the future, more in depth measurements of behaviour that can be computed for example using computer vision should be desirable to increase the robustness of the proposed model.

References

Revilla, M., Friggens, N.C., Broudiscou, L.P., Lemonnier, G., Blanc, F., Ravon, L., Mercat, M.J., Billon, Y., Rogel-Gaillard, C., Le Floch, N. and Estellé, J. (2019). Towards the quantitative characterisation of piglets’ robustness to weaning: a modelling approach. Animal, 13(11), 2536-2546. https://doi.org/10.1017/S1751731119000843 

Revilla M, Lenoir G, Flatres-Grall L, Muñoz-Tamayo R, Friggens NC (2021). Quantifying growth perturbations over the fattening period in swine via mathematical modelling. bioRxiv, 2020.10.22.349985, ver. 5 peer-reviewed and recommended by Peer Community in Animal Science. https://doi.org/10.1101/2020.10.22.349985 

In this paper, Madouasse et al. (2021) present a generalisable Bayesian method for calculating the probability that a herd is free from disease, based on its prior disease status, and using data (herd status over time over a sufficient number of herds to inform the model) and reasonable prior estimates of the sensitivity and specificity of tests being used to determine animal infection status.  Where available, the modelling approach can also include relevant additional risk factors. 

By bringing all these factors together, it allows for most countries to use the same analytical approach on their data, with differences across datasets expressed in terms of the uncertainty around the central estimates. 

Having a single methodology that generates both a central estimate of disease freedom, and uncertainty thus provides the opportunity (given typically available data) to compare the probability of freedom across different systems. This is relevant in terms of the context of trade (since international trade of livestock in many cases depends on disease freedom). It is also important when evaluating, for example, transnational burdens of disease - and with different regulations in place in different countries, this is invaluable and can be used, for example, to assess risks of zoonotic infection including for zoonotic infection emergence. In the BVD example provided, the point is made that, since regular testing would probably pick up infection rapidly, the addition of risk factors is most valuable where testing is infrequent. This emphasizes the advantages of direct incorporation of risk factors into a single modelling framework. 

From a technical point of view, the analysis compares two different packages for the Markov Chain Monte Carlo (MCMC) implementation necesary to run the model. They show that, while there are some slight systematic differences, the estimates provided by the two methods are similar to each other; as one method is approximate but substantially more stable and generally much more computationally efficient, this is an important outcome. Both implementations are freely available and  with relevant additional software made similarly available by the authors. This is extremely welcome and should encourage its general adoption across different countries. 

No single model can of course account for everything. In particular, the reliance on past data means that there is an implicit assumption common to all purely statistical methods that the underlying risks have not changed. Thus projections to altered circumstances (changing underlying risk factors or systematic changes in testing or test performance) cannot so easily be incorporated, since these factors are complicated by the dynamics of infection that lie outside the modelling approach. Of course the well known quote from George Box that "all models are wrong" applies here - the generality of approach, statistical robustness and open source philosophy adopted make this model very useful indeed.  

Madouasse A, Mercat M, van Roon A, Graham D, Guelbenzu M, Santman Berends I, van Schaik G, Nielen M, Frössling J, Ågren E, Humphry RW, Eze J, Gunn GJ, Henry MK, Gethmann J, More SJ, Toft N, Fourichon C (2021) A modelling framework for the prediction of the herd-level probability of infection from longitudinal data. bioRxiv, 2020.07.10.197426, ver. 6 peer-reviewed and recommended by PCI Animal Science. https://doi.org/10.1101/2020.07.10.197426

Scientific research can be seen by some as a competitive territory: competition of opinions, concepts, publications, competition for funding. Fortunately, it is above all a territory of sharing and cross-fertilization of ideas. It is gradually becoming a territory of productive interdisciplinary collaborations, despite persistent resistance to making borders more permeable [1]. At the crossroads of worlds, many challenges must be met for communities to understand each other, to be able to communicate with one another, and to benefit mutually from scientific interactions [2].

Delphine Destoumieux-Garzon and co-authors [3] propose to stimulate a single Animal Research Kinship (OneARK) to promote the crossing of the scientific communities in animal production and animal ecology. These two communities share many concepts and methods, which, while they are based on marked specificities (natural versus artificial systems), also and above all have common points that need to be explored more closely. Seven concepts of shared interest to improve the resilience and sustainability of animal population systems were explored by the authors: selection, system viability, system management, animal adaptability, inter-individual diversity in systems, agroecology, and animal monitoring.

This foundation stone paves the way for a finer integration between these two communities, which are close and yet distant, and which are slowly getting to know, understand, and recognize each other.

References

[1] Ledford, H. (2015). How to solve the world’s biggest problems. Nature, 525, 308–311. doi: 10.1038/525308a
[2] Knapp, B., Bardenet, R., Bernabeu, M. O., Bordas, R., Bruna, M., Calderhead, B., … Deane, C. M. (2015). Ten simple rules for a successful cross-disciplinary collaboration. PLoS Computational Biology, 11(4), e1004214. doi: 10.1371/journal.pcbi.1004214
[3] Destoumieux-Garzón, D., Bonnet, P., Teplitsky, C., Criscuolo, F., Henry, P.-Y., Mazurais, D., … Friggens, N. (2020). OneARK: Strengthening the links between animal production science and animal ecology. Ver 6 Peer-Reviewed and Recommended by PCI Animal Science. doi: 10.5281/zenodo.3632731