A meta-analysis of the effects of high ambient temperature on growth performance of growing-finishing pigs

International audience

anglais. High ambient temperature (T) is one of the most important climatic factors influencing pig performance. Increased T occurs sporadically during summer heat waves in temperate climates and year round in tropical climates. Results of published experiments assessing the effects of high T on pig performance are surprisingly variable. Thus, a meta-analysis was performed to aggregate our knowledge and attempt to explain differences in the results across studies on the effect of increased T on ADFI and ADG in growing-finishing pigs. Data for ADFI and ADG were extracted from 86 and 80 trials, respectively, from articles published in scientific journals indexed in PubMed, Science Direct, and from proceedings of scientific meetings through November 2009. Data on ADFI and ADG were analyzed using a linear mixed model that included the linear and the quadratic effects of T and BW, and their interactions as continuous, fixed effects variables, and the trial as a random effect factor (i.e., block). In addition, the effects of housing type (2 levels: individual and group housing) and the year of publication (3 levels: 1970 to 1989, 1990 to 1999, and 2000 to 2009) on the intercept and the linear regression term for T (i.e., the slope) were also tested. Results showed that high T had a curvilinear effect on ADFI and ADG and that this effect was more pronounced in heavier pigs. Across T, ADFI was less when pigs were group-housed. The intercept and the regression coefficient (slope) for T were significantly affected by the year of publication. The effect of increased T was greater in more contemporary works, suggesting that modern genotypes could be more sensitive to heat stress than older genotypes of lesser growth potential. In conclusion, pig performance decreases at an accelerating rate as T is increased. The large between-study variability on the effects of high T on pig performance is partially explained by differences in pig BW and to a lesser extent by the year the study was published.