Active navigation in disordered media is governed by the interplay between self-propulsion and environmental constraints. Using the chemotaxis of E. coli in agar gels as a model system, we uncover a universal trade-off between persistence and obstacle avoidance that dictates optimal search strategies. We find that populations evolving under pressure for rapid expansion adapt by shortening their mean run time (τf), counter to the intuition that longer runs always favor faster migration. Controlled experiments with a tunable strain confirm a non-monotonic relationship between run time and chemotactic velocity, with a clear optimum that shifts with environmental trap density. At the single-agent level, we identify and characterize a key motility state: transient trapping in the gel's pores. A minimal theoretical model, integrating run-tumble and run-trap dynamics, explains the optimum as a consequence of the antagonistic scaling of the diffusion coefficient (increasing with τf) and the chemotactic bias coefficient (decreasing with τf). This work establishes a general principle for the optimization of active matter transport in complex and obstructed environments.
Yang et al. (Tue,) studied this question.