Microbial interactions in theory and practice: when are measurements compatible with models?

Most predictive models of ecosystem dynamics are based on interactions between organisms: their influence on each other's growth and death. We review here how theoretical approaches are used to extract interaction measurements from experimental data in microbiology, particularly focusing on the used, we argue that the gLV model should be avoided for estimating interactions in batch culture - the most common, simplest and cheapest in vitro approach to culturing microbes. Fortunately, alternative approaches offer a way out of this conundrum. Firstly, on the experimental side, alternatives such as the serial-transfer and chemostat systems more closely match the theoretical assumptions of the gLV model. Secondly, on the theoretical side, explicit organism-environment interaction models can be used to study the dynamics of batchculture systems. We hope that our recommendations will increase the tractability of microbial model systems for experimentalists and theoreticians alike.

Automated summary

This review article discusses how theoretical approaches are used to extract interaction measurements from experimental data in microbiology. The authors suggest avoiding the use of the gLV model for estimating interactions in batch culture and propose alternative approaches such as the serial-transfer and chemostat systems. This is expected to increase the tractability of microbial model systems for both experimentalists and theoreticians.