Effects of management, environmental and temporal factors on mortality and feed consumption in integrated swine fattening farms

Article history: Received 10 September 2008 Received in revised form 13 November 2008 Accepted 15 November 2008 The purpose of this study was to assess the effects of management, environmental and temporal factors of farms and batches on mortality and feed consumed per market hog. Information of management, environmental and temporal factors from 158 all-in all-out finishing batches completed in 27 integrated fattening farms in Galicia (Spain) were collected from 1999 to 2002. Multilevel random effects models were used to determine the factors associated with the mortality and the average amount of feed provided per marketed hog. The level 1 unit was the finishing batch (batch level). At the level 2 the fattening farm where the batch of pigs was developed was introduced as random effects. A bad level of care provided by the farmer to the pigs increased the risk of mortality (P=0.03). The mortality of the fattening batch was associated with the supplier sow farm (Pb0.01). The level of mortality of the batches of piglets entered the fattening farms both between January and April and October and December was higher than for batches incorporated from May to August (Pb0.01). The year when the piglets entered the farm was significantly associated with mortality (Pb0.01). The introduction of such variables in the multilevel model for mortality reduced the total unexplained variation by 41.06%. The proportion of the variation of the mortality attributable to the fattening farms was small (3.65%). The variability of mortality was higher when the placement of piglets was from September to December (P=0.01). Higher consumption of feed was recorded when the feed was provided in a non-pellet form (Pb0.01). The introduction of piglets between September and December was associated with a higher consumption of feed (Pb0.01). The averagefinalweight of the hogswas positively associated with the feed consumed per hog of the fattening batch (Pb0.01). The amount of feed provided per hog was statistically different among the years and these effects were modified by the level of mortality recorded in the batch and by the average weight of the marketed hogs. After introducing all the fixed effects in the multilevel model for feed provided per hog the unexplained variation at the farm and batch levelwas reduced by 47.72% and 85.35%, respectively. A considerable amount of variation of feed provided per hog resided at the fattening farm (32.50%). When the piglets were placed on September–December the variability of feed consumedwas higher than in the other two periods of the year (P=0.02). In conclusion, our study shows that specific factors of the batches are the main contributors to explain mortality. On the other hand, by improving permanent farm characteristics a substantial profit is expected in the long run as a result of lower feed costs. © 2008 Elsevier B.V. All rights reserved.