Horse welfare in semi-extensive system: establishing a welfare protocol and comparing pasture and stable farming systems

There is not a welfare protocol for horses reared for meat production in semi-extensive systems. The aims of the study were to develop a specific welfare protocol suitable to be applied at pasture and on stable; and to evaluate whether the welfare items were influenced by the farming system (pasture vs stable). 52 non-animal-based measures (N-ABMs) and 14 animal-based measures (ABMs), classified into 6 thematic areas (training, feeding, facilities, ABMs, biosecurity, health management) were selected by a focus group. The protocol was applied on a total of 429 Catria horses located across 26 pastures during the warm seasons and on 7 stables during the cold seasons. Differences obtained within each horse-unit were calculated by using the diversity index (VARNC) and the distance from the ideal (dfi) index. Chi-square test was used for comparing the relative frequencies (%) of the answers (pasture vs stable). Most the welfare items were classified as adequate in both pasture and stable, yet differences were found within 'training' (p = 0.02) and 'feeding' (p < 0.01) areas in relation to the welfare items 'inspection of the animals' and 'feeding management'. Weaknesses of both pasture and stable were represented by some welfare items in the 'health management' and 'biosecurity' areas. After proper validation, the welfare protocol developed in the present study could help to fill the existing gap of knowledge on horse welfare assessment for semi-extensive system systems, providing support for official control of veterinarians and enabling the identification of key weakness to address preventive interventions.