Multidimensional tolerance landscapes reveal antibiotic-environment interactions affecting population dynamics of...

City sewers harbor diverse bacterial communities that are continuously exposed to a myriad of antibiotic residues resulting from human consumption and excretion. Despite their sub-inhibitory concentrations in sewage, these pharmaceutical residues affect the growth rate and the yield of susceptible wastewater-associated bacteria. Moreover, environmental conditions in sewers are complex, including variations in temperature and, in many coastal city sewers, salinity. These variables can modulate antibiotic tolerance and therefore affect the dynamics of microbial populations. To explore such interactions between antibiotics and abiotic environmental factors, we built continuous multivariate tolerance landscapes for three bacterial species commonly detected in sewage: Escherichia coli, the emerging pathogen Streptococcus suis, and a typical sewer dweller, Arcobacter cryaerophilus. We projected their intrinsic growth rate and carrying capacity onto a complex environment including temperature, salinity, and a range of concentrations of two antibiotics frequently measured in urban wastewater (ciprofloxacin and azithromycin). We revealed that antibiotic tolerance was maximal at salinities close to seawater for both E. coli and S. suis, and that the direction of the interaction between antibiotics and temperature is species dependent. In E. coli, we additionally observed a third-order interaction among salinity, temperature and antibiotics, highlighting the limits of predicting field dynamics of bacterial populations using standard laboratory measures. We projected these tolerance curves onto time series data of temperature and conductivity measured in the sewers of Barcelona. Our model highlights that low concentrations of antibiotics could exclude the most sensitive species, while interactions between antibiotics, temperature, and salinity substantially affected the dynamics of the more tolerant ones.