Bacteria often live in diverse communities where the spatial arrangement of strains and species is considered critical for their ecology, including whether strains can coexist, which are ecologically dominant, and how productive they are as a community. However, a test of the importance of spatial structure requires manipulation at the fine scales at which this structure naturally occurs. Here we develop a droplet-based printing method to arrange different bacterial genotypes across a sub-millimetre array. We use this to test the importance of fine-scale spatial structure by printing strains of the gut bacterium Escherichia coli that naturally compete with one another using protein toxins. This reveals that the spatial arrangement of bacterial genotypes is important for ecological outcomes. Toxin-producing strains largely eliminate susceptible non-producers when genotypes are well-mixed. However, printing strains side-by-side creates an ecological refuge such that susceptible strains can coexist with toxin producers, even to the extent that a susceptible strain outnumbers the toxin producer. Head-to-head competitions between toxin producers also reveals strong effects, where spatial structure can make the difference between one strain winning and mutual destruction. Finally, we print different potential barriers between two competing strains to understand why space is so important. This reveals the importance of processes that limit the free diffusion of molecules. Specifically, we show that cells closest to a toxin producer bind to and capture toxin molecules, which creates a refuge for their clonemates. Our work provides a new method to generate customised bacterial communities with defined spatial distributions, and reveals that micron-scale changes in these distributions can drive major shifts in their ecology.

The value of total coliforms, faecal coliforms and faecal streptococci in predicting the presence of Salmonella spp. and the numbers of Staphylococcus aureus and Candida albicans in sewage polluted coastal water were assessed. All indicators had strong positive association with Salmonella and moderate positive correlations with Staph. aureus and C. albicans. Total coliforms correlated better with salmonellas and Staph. aureus than did the two faecal groups. Regression analysis revealed that total coliforms have a better value as predictors of the presence of Salmonella and Staph. aureus, while faecal coliforms are better predictors of C. albicans, in moderately polluted areas. The conclusion reached is that enumeration of total coliforms is sufficient to predict the presence of Salmonella spp. or Staph. aureus in sea water moderately affected by sewage pollution, without the additional measurement of faecal coliforms and faecal streptococci.