Shifts from cooperative to individual-based predation defense determine microbial predator-prey dynamics

Classical predator-prey theory predicts equilibrium or steady-state dynamics. However, the inclusion of prey defense strategies adds complexity. Bacteria employ various defenses against protozoan grazing, differing in mode (evasion or attack), reversibility (inducible or permanent), and scope (individual or cooperative). While individual defense mechanisms are well-understood, the long-term ecological and evolutionary impacts of multiple interacting defenses, particularly trade-offs and the role of evolution, remain largely unexplored. Most studies focus on short-term dynamics. This research investigates the long-term dynamics of a microbial predator-prey system to address this gap, focusing on the succession of defense strategies in *P. putida* under predation pressure from *P. lacustris*. The researchers aimed to understand the mechanisms driving the transition between cooperative and individual-based defenses, examining the ecological and evolutionary factors involved.

Existing literature demonstrates that bacteria use diverse strategies to defend against predation, including active defenses like morphological alteration and aggregation, and community defenses reliant on social cooperation, such as the production of extracellular toxic compounds. Community defenses, while highly effective at high densities, are vulnerable to cheating. Defenses also vary in reversibility, with inducible adaptations contrasting with permanently acquired protected genotypes arising from mutations or selection. Previous research primarily focused on singular defense mechanisms observed over short timescales, lacking insight into long-term dynamics, interactions between alternative defenses, and the evolutionary optimization of defense strategies. Few studies have traced grazing resistance to specific mutations.

The study used a microbial predator-prey system consisting of *Pseudomonas putida* (prey) and *Poteriospumella lacustris* (predator) in semi-continuous cultures over five weeks. Cultures were diluted daily to replenish resources. Sixteen replicates were run with varying initial predator and prey densities. Cell abundances were quantified daily using microscopy. Filamentous bacterial isolates were subjected to whole-genome sequencing to identify mutations responsible for the observed defense shifts. A mathematical model, implemented in R, was developed to simulate the predator-prey dynamics. The model incorporates seven state variables (resources, four bacterial phenotypes, flagellates, and toxin concentration) and nine processes (bacterial growth, grazing, metabolite production regulation, metabolite excretion, and filamentation). The model considers various factors, including fitness costs associated with metabolite production and filamentation, predator recognition and quorum sensing interactions, conditional mutation rates, and asymmetric filament division. The model's parameters were calibrated using experimental data, enabling a comparison between simulated and observed dynamics.

Across all replicates, a consistent succession of bacterial defenses was observed. Initially, *P. putida* employed a highly effective cooperative defense involving toxic metabolites, nearly driving predators to extinction. However, this was consistently replaced by a second, individual-based defense characterized by filamentation, arising via de novo mutations. Filamentation, while less effective at reducing predator abundance compared to the initial metabolite-based defense, persisted due to its genetic basis and heritability. Whole-genome sequencing revealed mutations in genes involved in septum formation (*ftsQ*, *ftsA*, and *minC*) in several filamentous isolates. Experiments showed that the recovery of flagellates after the initial predator collapse was not due to predator adaptation to the bacterial toxin. The mathematical model successfully reproduced the observed dynamics, supporting the hypothesis that the shift from cooperative to individual-based defense is driven by the maximization of individual, rather than population, benefits. The model demonstrated that the filamentous phenotype persists despite a fitness cost due to its genetic stability and a self-stabilizing effect related to resource consumption. The model also confirmed that the filamentation was not present in the initial inoculum but emerged during the experiment.

The findings demonstrate a clear shift from a cooperative, environmentally-induced defense mechanism to a more costly but genetically stable individual-based defense. The initial defense, likely involving pyoverdines (siderophores), exhibits a dual function: inhibiting flagellate growth and enhancing bacterial resource exploitation. However, this defense is vulnerable to the emergence of individual-based defenses, highlighting the inherent fragility of cooperative strategies. The model's success in replicating observed dynamics underscores the importance of considering both individual-level selection and the costs and benefits of different defense mechanisms. The irreversible nature of the filamentous defense, coupled with its self-stabilizing effect, proves advantageous in the long run. The study's results provide valuable insights into the evolutionary dynamics of predator-prey interactions and the role of cooperation and individual selection in shaping microbial community structure.

This study demonstrates a transition from cooperative to individual-based defenses in a microbial predator-prey system driven by the optimization of individual fitness. The initial cooperative defense, based on toxic metabolites, is highly effective but susceptible to cheating and the emergence of individual-based defenses through rapid evolution. The shift to a filamentous phenotype, although less effective in reducing predator numbers, persists due to its irreversible nature and self-stabilizing effects. Future research could investigate the genetic basis of filamentation in greater detail and explore the generality of this defense strategy across different predator-prey systems and environmental conditions.

The study focused on a specific predator-prey system under controlled laboratory conditions. The generalizability of the findings to other microbial systems and natural environments requires further investigation. The model, while successfully reproducing the observed dynamics, simplifies complex biological processes and interactions. Additionally, the identification of mutations responsible for filamentation was limited to a subset of isolates.