Fermented food ecosystems are composed of a rather low number of microbial species compared to most other ecosystems. While their taxonomic diversity has been well described, a better understanding of their functional diversity and rules of assembly is still needed. In the last two years, a consortium of INRA teams has developed the Florilege database, to gather microbial phenotypes of food interest, from the literature. In parallel, an Italian initiative has developed FoodMicrobionet database, to gather bacterial taxa from fermented foods and their interaction networks. These two databases represent an opportunity for the scientific international community to better understand food community assembly and functioning. They also provide information to design synthetic microbial consortia to study microbial interactions or to develop innovative fermented foods. However, these two databases still require some improvements. In addition, the methods used for analyzing interaction networks have been developed for complex communities and their power for studying simpler food communities need to be further investigated.
In this project, we propose to improve tools dedicated to the study of food microbial ecology and to illustrate their interests for consortium design. The project is divided into 4 tasks including improvement of both databases and of statistical methods to study species interactions followed by an experimental consortium design for milk and cereals mix fermentation.