A meta-analysis of the effect of Eimeria spp. and/or Clostridium perfringens infection on the microbiota of broiler chickens.

Coccidiosis in chickens is caused by Eimeria spp. The infection provides a growth advantage to Clostridium perfringens (CP), frequently leading to necrotic enteritis. One approach to alleviate the negative impacts of the diseases is to improve the bacterial composition in chickens, and many experiments investigating chicken enteric health in recent years include the characterization of the bacterial microbiota. This meta-analysis synthesized the data of studies investigating the intestinal microbiota after infection with coccidia and/or CP to provide a basis for future research. Inclusion criteria were that experiments contained a group infected with one or both pathogens and an uninfected control group, the use of 16SrRNA Illumina sequencing and the availability of raw data. A total of 17 studies could be included. Meta-analyses of 3 different data sets were performed: 1 on data of 9 experiments on chickens infected with coccidia only; the second on data of 4 studies on chickens infected with CP only; the third on raw data of 8 experiments with chickens infected with coccidia and CP. The meta-analysis of relative abundance and alpha diversity of the data sets was performed in R using the SIAMCAT and metafor packages. The number of families of interest identified by the analyses of experiments with infection with coccidia only, CP only and the combined infection were 23, 2, and 29, respectively. There was an overlap of 13 families identified by analyses of experiments with infection with coccidia only and of experiments with the combined infections. Machine learning was not able to find a model to predict changes of the microbiota in either 1 of the 3 analyses. Meta-analyses of functional profiles showed a more uniform reaction to the infections with the relative abundance of many pathways significantly altered. Alpha diversity was not affected by infection with either pathogen or the combination. In conclusion, the heterogeneity of these microbiota studies makes recognizing common trends difficult, although it seems that coccidia infection affects the microbiota more than an infection with CP. Future studies should focus on the bacterial functions that are changed due to these infections using metagenome techniques.