The objective of this study was to assess the relationship between the animal welfare conditions evaluated through the supply chain and pork quality variation. A total of 4,680 pigs from 12 farms-5 animal welfare improved raising system (AWIRS) and 7 conventional raising system (CON) farms-were assessed from farm to slaughter through a comprehensive audit protocol merging the European Welfare Quality, the Canadian Animal Care Assessment, and American Meat Institute audit guide criteria. At the abattoir, a subsample of 1,440 pigs (120 pigs/farm) was randomly chosen out of 24 loads (2 farms per wk) transported by 2 drivers (driver A and driver B) for the assessment of stunning effectiveness, carcass bruises, blood lactate levels, and meat quality traits. Meat quality was assessed in the longissimus lumborum (LL) muscle 24 h postmortem by measuring ultimate pH (pHu), color (L*, a*, and b*), and drip loss. Data were analyzed by the MIXED, GLIMMIX, and NAPAR1WAY procedures of SAS. Spearman correlations were calculated to determine the relationship between audit scores and meat quality traits. Better animal welfare conditions, as showed by greater final scores for good housing (GHo; P = 0.001) and good health (P = 0.006) principles, were recorded at AWIRS farms. Pigs from AWIRS farms handled by driver B displayed a greater percentage of turning back (P = 0.01) and slips (P < 0.001) during unloading and a greater (P = 0.02) frequency of falls in the stunning chute. A greater (P = 0.02) reluctance to move at loading was found in CON pigs loaded by driver A compared with driver B, whereas a greater (P < 0.001) reluctance to move was found in these pigs at unloading when they were unloaded by driver B. Drip loss was higher (P = 0.003) and pale, soft, and exudative pork percentage was greater (P < 0.001) in the LL muscle of the heavier AWIRS pigs. The GHO principle was best correlated with pHu (r = -0.75, P = 0.01) and Minolta L* value (r = 0.87, P < 0.001) of the LL muscle. Overall, drip loss variation in the LL muscle was correlated with the frequency of slips at unloading (r = 0.63, P = 0.001) and in the restrainer area (r = 0.74, P < 0.001). The results of this study showed that the quality of the raising system and truck driver skills as assessed by animal welfare audit protocols are important sources of variation in the behavioral response of pigs to preslaughter handling and may affect pork quality variation. However, the different live weight between CON and AWIRS pigs may have biased the meat quality results in this study.
