Environment-, health-, performance- and welfare-related parameters in pig barns with natural and mechanical ventilation

A multifactorial approach using environmental, performance, health and welfare parameters was used to investigate the numerous associations of ventilation throughout three consecutive fattening batches (08/2015 to 12/2016) in a farrow-to-finish commercial pig farm in Belgium. Two fattening pig units were used, unit A (1256 pigs) with mechanical ventilation and unit B (1264 pigs) with natural ventilation. Animal genetics, nutrition, stocking density and health management were the same for both units. Key environmental indicators were monitored in real-time (temperature, humidity, CO2 and NH3) and the daily prevalence of respiratory disease cases was recorded to monitor the temporal expression of disease over time within a farm environment. The welfare status of the animals was assessed twice per production round (batch) with a simplified version of the Welfare quality (R) protocol. Serological tests for the most prevalent respiratory infectious agents (Mycoplasma hyopneumonniae, swine influenza virus (subtypes H1N1, H1N2 and H3N2), Actinobacillus pleuropneumoniae serotypes 1-2-9-11 and porcine circovirus type 2) were performed during the first, second and final third of each fattening period. Finally, key performance parameters were calculated (average daily growth, feed conversion ratio) and upon slaughter lungs from pigs from unit A (n: 782) and from unit B (n: 544) were assessed for the presence of lung lesions, pleurisy and fissures. To identify the associations of ventilation on the aforementioned parameters, statistical models were run that also included other factors (when applicable) namely production batch, season, age and sampling. Overall, the use of natural ventilation was associated with a less optimal environment with regards to thermal comfort (p < 0.001), CO2 (p < 0.001) and NH3 (p < 0.001). A higher daily prevalence of respiratory disease cases was seen in the naturally-ventilated unit (p < 0.001). Concerning the sem-prevalence of the infectious agents tested, the odds to have a positive H1N1 sample were 3.17 higher in the naturally-ventilated unit (p = 0.003). From the visual assessment of the lungs no statistically significant associations were seen between ventilation type and the presence of lesions, fissures or pleuritis. Yet, the lung lesion score was expected to be lower in the naturally-ventilated unit (p = 0.010). Regarding performance parameters, feed conversion ratio and average daily gain were overall better in the mechanically-ventilated unit (descriptive results). Finally, a better welfare score was seen in the mechanically-ventilated unit in all three production batches (descriptive results). In conclusion, the mechanically-ventilated farm was associated with better environmental conditions for the fattening pigs. Yet, further research is needed to reach definite causal claims.