18th PhD Symposium Agricultural and Natural Resource Economics

Decisions on slaughter weight and timing of slaughter might be as old as pigs production itself. Hence, these questions have repeatedly been studied by researchers in the past. It has been a side result of studies focusing on profit maximisation per unit of time from batch production (Heady et al., 1976; Kawaguchi and Kennedy, 1989), and it received more attention in studies by ( Chavas et al., 1985; Jolly et al., 1980; Boland et al., 1993; Niemi, 2006; Boys et al., 2007 Niemi et al., 2010; Kristensen et al., 2012). The research questions on optimal slaughter weight mainly focused on 1) differences in optimal delivery weight and optimal feeding decisions provoked by animal performance,2) optimal slaughter weight under various market circumstances and 3) losses from suboptimal delivery of finisher pigs for the farmer when pig herds are heterogeneous. While these historic studies answered these questions starting from different premises and assumptions on the technical and economic context in which the slaughter decisions were to be made, some general insights into the questions ,outlined above, could be derived.

Still, renewed requests come from the industry to investigate the optimal delivery weight problem. This revival might be due to several reasons. First, the decision context is dynamic in time and spatially situated. For example, pricing schemes can differ between different regions and have changed with time. Similarly, spatial-temporal dynamics in market conditions might require the information to be updated. Moreover, the production process has changed; genotypes have been further improved by genetic selection and alternatives for surgical castrated males, in the form of intact and GnRH-vaccinated males are being used increasingly as a response to welfare concerns.
Secondly, some issues for practical use of optimization models arise. Optimizing delivery weight is a relatively complex problem and as for all hard-science based optimisation models, the application of these model in practise is troublesome (Martin, 2015). Historical optimisation models might not have reached the intended decision makers. In the past, computational capacity of computers (Glen, 1983), might have been a burden. Nowadays this issue should be greatly overcome. Still, the theoretical and mathematical complexity (Tanure et al., 2013), combined with a reliance on large informational needs and parameters which are difficult to estimate on farms (Black, 2014), still pose burdens on practical implementation of optimisation models. Lastly, the historic optimization models might have included too much decision variables, over which farmers do not have (full) control or information to be able to include these variables in their decision set. This might have limited the applicability of these model’s and their results in a practical context.

To acquire a clear understanding of the practical decision context and to be able to develop a model close aligned with the end-users needs, a participatory problem analysis (see Leen et al., 2017) of the delivery weight decision was executed. In this process stakeholders were involved in reframing the decision problem, i.e. listing the crucial factor and processes to be modelled and the questions to be solved. The industry members demanded insight into i) the evolution of losses due to suboptimal delivery in addition to the mere optimization results, ii) the effect of sex and differences in animal performance on the optimal delivery results and iii) the value of having separate delivery moments for different sexes in the same batch (split-harvesting). However, the experts did put more emphasis on the learning potential of the model simulations for the farmers than on the normative value of the model producing the simulations (Leen et al., 2017). While the proposed model outlook aligned with the historical studies, they urged to limit the model complexity to enhance the conceptual accessibility and applicability and prevent the user of being discouraged to use the model. Therefore, the objective of this study was to provide insights into questions raised by stakeholders concerning the optimization of finisher pig delivery weights at the tactical decision level. For this aim, a simulation model was developed, in line with the expectations and needs of the pig production stakeholders.