Early detection of diarrhoea in weaned piglets from individual feed, water and weighing data

Abstract. This study, part of HealthyLivestock project, aims to analyse individual water and feed consumption related to weight of weaned piglets and their link to diarrhoea. Data were collected from 15 batches of 102 piglets using specific automata (connected feeders and drinkers, automatic weighing scales, RFID ear tags) and observations of piglets’ health status made by specialized technicians. Analyses were carried out every week on the 138 healthy animals compared by weight category (light, medium, heavy). The average feed consumption had no significant difference between categories whatever the week and was close to 4% of the live weight. For the average water consumption according to weight, it was close to 10% and there was no significant difference between groups. However, at the end the water consumption of one heavy pig gave significant difference with the light group. These overall stable averages in healthy pigs, were promising for distinguishing deviant behaviours of sick pigs. However, they shaded the high intra-individual variabilities, around 40% at the beginning of post-weaning for both feed and water consumption and almost 16% for feed and 25% for water at the end. Then, the comparison between healthy and diarrheic piglets showed no statistical difference for average water consumption on the day of the first clinical signs and even 1 and 2 days before. In contrast, the average feed consumption had a very significant difference (P≤0.001) for days 5-7 after the weaning and a significant difference for day 8 (P≤0.05). These differences were also significant for data collected 24-48h before first clinical signs. This means either those diarrheic piglets decreased their feed consumption the first days after weaning or that it is because they eat less when they become diarrheic. As feed consumption seemed to be an interesting indicator to detect early diarrheic weaned animals, we continued the study by using machine learning methods. But they all failed in detecting individually diarrheic animals from water and feed consumption related to weight, probably because of considerable individual variability of feed consumption, even in healthy pigs. To improve these results, new data are collected, like location of piglets in the pen by image analysis.