Abstract

Bacteria influence many important ecosystem processes such as decomposition. Two experimental approaches have generally been used to study these systems. The first is a reductionist “bottom up” approach, which isolates individual bacteria, and builds synthetic communities to understand inter-specific interactions impact functioning. The second is a “top down” approach, observing or manipulating communities in situ, measuring process rates, and building models to link community structure and ecosystem functioning. I will summarise a third intermediate approach: using large libraries of intact communities inoculated into standardised conditions. This approach preserves realism and allows more direct inferences about cause and effect. I will outline how we have used this approach to infer interactions among bacteria in complex microcosm communities, allowing us to identify the functional groups and ecological guilds that are maintaining decomposition-related processes within the microcosms. The approach generates highly repeatable dynamics, potentially providing a route toward domestication of intact, preserved communities.