'Here, we show that the conclusions of a microbiome research study are greatly dependent on the time of sample collection, and that experimental and control groups undergo a cycle of diverging and converging microbiome composition depending on the nature and timing of experimental interventions.'