Evaluating pen-allocation strategies for uniform weights in finishing pigs estimating age at 120 kg

Variable performance of finishing pigs within pens causes losses to producers through delayed availability of pen space and deviation from target market weight. We explored alternative pen allocation strategies to assess their potential to reduce losses. We estimated retrospectively the maximum benefit gained if individual growth to target weight could be predicted at allocation. If a strategy’s retrospective benefit is significant there is a case to invest resources to implement it. We used a dataset with three routine weight measurements per pig during finishing (n=240, range (kg): 14.8-31.8, 43.0-86.5, 100.0-138.0, CV: 0.12, 0.12, 0.06) at points where age differed considerably and increasingly (range (d): 53-66, 103-128, 144-194 (0.03, 0.04, 0.07)), and comprising six cohorts of contemporary pigs. Three pen-allocation strategies were compared: the actual strategy used by producers (PS) against an ‘optimal’ (OS) and a ‘random’ ‘strategy (RS), both applied to contemporaries. Bayesian modelling was used to predict individual growth and age at 120 kg (A120) due to the deviations of end weights from 120 kg. In the OS, cohort pens were filled with pigs of similar predicted A120. In the RS, sorting was random. The economic loss of each strategy quantified revenue lost in unused pen space and in deviation from target weight in each pen until empty, and for each cohort. We estimated, based on current market values, that OS can generate up to +28% in profit per pig in relation to the PS (cohort average +15%); i.e. a gain of up to €29,000 per year in a 4,000-pig farm. The RS lead more often to considerably greater loss than profit compared to the PS. Therefore, the producer’s strategy is clearly better than the RS but can be improved through direct or proxy estimation of A120. The level of improvement suggests it is worth using or collecting further data to allow this estimation at the start of finishing. Precision management will become more relevant in future systems where optimised pen sorting is applied and animals are given designed feed and treatment on a pen (or individual) basis. This study was supported by the EU’s H2020 program (agreement no. 633531) and UK’s BBRSC.