Take care of the environment: housing conditions affect the interplay of nutritional interventions and intestinal microbiota in broiler chickens

BACKGROUND: The intestinal microbiota is shaped by many interactions between microorganisms, host, diet, and the environment. Exposure to microorganisms present in the environment, and exchange of microorganisms between hosts sharing the same environment, can influence intestinal microbiota of individuals, but how this affects microbiota studies is poorly understood. We investigated the effects of experimental housing circumstances on intestinal microbiota composition in broiler chickens, and how these effects may influence the capacity to determine diet related effects in a nutrition experiment. A cross-sectional experiment was conducted simultaneously in a feed research facility with mesh panels between pens (Housing condition 1, H1), in an extensively cleaned stable with floor pens with solid wooden panels (H2), and in isolators (H3). In H1 and H2 different distances between pens were created to assess gut microbiota exchange between pens. Feed with and without a blend of medium-chain fatty acids (MCFA) was used to create differences in cecal microbiota between pens or isolators within the same housing condition. Male one-day-old Ross broiler chickens (n = 370) were randomly distributed across H1, H2, and H3. After 35 days cecal microbiota composition was assessed by 16S ribosomal RNA gene amplicon sequencing. Metabolic functioning of cecal content was assessed based on high-performance liquid chromatography. RESULTS: Microbial alpha diversity was not affected in broilers fed +MCFA in H1 but was increased in H2 and H3. Based on weighted UniFrac distances, the nutritional intervention explained 10%, whereas housing condition explained 28% of cecal microbiota variation between all broilers. The effect size of the nutritional intervention varied within housing conditions between 11, 27, and 13% for H1, H2, and H3. Furthermore, performance and metabolic output were significantly different between housing conditions. The distance between pens within H1 and H2 did not influence the percentage of shared genera or operational taxonomic units (OTUs). CONCLUSIONS: The cecal microbiota of broilers was modifiable by a nutritional intervention, but the housing condition affected microbiota composition and functionality stronger than the diet intervention. Consequently, for interpretation of intestinal microbiota studies in poultry it is essential to be aware of the potentially large impact of housing conditions on the obtained results. ELECTRONIC SUPPLEMENTARY MATERIAL: The online version of this article (10.1186/s42523-019-0009-z) contains supplementary material, which is available to authorized users.