Using data from electronic feeding stations for investigating the effect of lameness on the feeding behaviour of sows

It is widely recognized that lameness is an important economic problem for farmers and a welfare problem for breeding sows. Electronic feeding stations (EFS) are installed in numerous farms and collect large amounts of data that enable warning the farmer when a sow fails to feed. These data are currently not exploited for other purposes although they could generate variables reflecting different aspects of feeding behaviour. Since there is a body of evidence that lameness status affects feeding behaviour, this study aimed at deriving variables related to feeding behaviour from EFS data and at investigating the effect of lameness on the latter variables. For this purpose, variables relating to feeding behaviour were derived from EFS raw data. These variables were then linked with gait scores collected using a 150mm tagged visual analogue scale (tVAS) and statistically analysed using two tests: (i) t-tests (SPSS) for which data were dichotomised between non-lame (tVAS < 60mm) and lame sows (tVAS > 60mm); and (ii) spearman correlation tests using continuous tVAS data. Feeding rank differed between lame (11.8 ±0.74 SEM) and non-lame sows (9.8 ±0.61) (P=0.040). The number of non-feeding visits (lame 4,18 ±0.26; non-lame 5.22 ±0.32; P=0.028) and maximum duration between visits (lame 15.1h ±0.48; non-lame 13.05h ±0.04; P=0.004) differed as well. Moreover, gait scores were negatively correlated with the duration of feeding visits (r=-0.228, P<0.001) and positively correlated with the maximum duration between visits (r=0.206, P=0.001). These results indicate that lame sows postpone their visits to the EFS, which may suggest that they are less able to compete and/or more reluctant to walk to the feeder. This study demonstrates that lameness status influences several aspects of feeding behaviour that can be derived routinely from EFS data. Hence, EFS data has potential to be integrated in automated lameness monitoring systems for sows.