Ruggedness of ecological landscapes informs community optimization

Assembling optimal microbial communities is key for various applications in biofuel production, agriculture, and human health. The number of possible communities grows exponentially with the number of species, so an exhaustive search cannot be performed even for a dozen species. We investigate how the success of a heuristic search for the optimal combination of microbes depends on community ecology. Using consumer-resource models with and without cross-feeding, we show that search success depends on the ruggedness of the appropriately-defined ecological landscape. We identify ruggedness measures robust to noise and low sampling density that can be used to predict the performance of a heuristic search from realistic experimental data. More importantly, we report how the patterns of species interactions influence landscape ruggedness. This relationship can guide the choice of the species and their environment in biotechnology applications and allows one to assess the likelihood of finding a high-performance microbial community. We show the feasibility of our approach using experimental data from simple soil microbial communities.