Protist predation can select for bacteria with lowered susceptibility to infection by lytic phages

Consumer-resource interactions constitute one of the most common types of interspecific antagonistic interaction. In natural communities, complex species interactions are likely to affect the outcomes of reciprocal co-evolution between consumers and their resource species. Individuals face multiple enemies simultaneously, and consequently they need to adapt to several different types of enemy pressures. In this study, we assessed how protist predation affects the susceptibility of bacterial populations to infection by viral parasites, and whether there is an associated cost of defence on the competitive ability of the bacteria. As a study system we used Serratia marcescens and its lytic bacteriophage, along with two bacteriovorous protists with distinct feeding modes: Tetrahymena thermophila (particle feeder) and Acanthamoeba castellanii (surface feeder). The results were further confirmed with another study system with Pseudomonas and Tetrahymena thermophila.

We found that selection by protist predators lowered the susceptibility to infections by lytic phages in Serratia and Pseudomonas. In Serratia, concurrent selection by phages and protists led to lowered susceptibility to phage infections and this effect was independent from whether the bacteria shared a co-evolutionary history with the phage population or not. Bacteria that had evolved with phages were overall more susceptible to phage infection (compared to bacteria with history with multiple enemies) but they were less vulnerable to the phages they had co-evolved with than ancestral phages. Selection by bacterial enemies was costly in general and was seen as a lowered fitness in absence of phages, measured as a biomass yield.

Our results show the significance of multiple species interactions on pairwise consumer-resource interaction, and suggest potential overlap in defending against predatory and parasitic enemies in microbial consumer-resource communities. Ultimately, our results could have larger scale effects on eco-evolutionary community dynamics.