Microbial ecology has tended to use two approaches. The first approach is to make observations of natural ecosystems and try to infer causal relationships from these observational data. This approach is hampered by the difficulty in distinguishing cause and effect. The second approach is to conduct experiments under controlled conditions, usually using microcosms in the lab. This approach sacrifices realism for control, and is able to link cause and effect. However, it is often unclear whether the results have relevance for natural ecosystems.
Our research tries to bridge the gap between messy observations and simplified experiments by conducting lab and field experiments that are able to bring some real-world complexity without sacrificing control. Overall, the research is focussed on trying to answer 2 main questions:
- What processes are important in maintaining biodiversity in microbial communities?
- How do compositional changes to communities impact ecosystem functioning?
Many of these questions are complex, and require large, well-replicated experiments for definitive answers. We have therefore invested heavily in scaling up experiments to allow truly large-scale manipulations. Part of this process has involved automation of experiment design and assays using laboratory robots.