A microswimmer in open fluid moves in a straight line (plus rotational diffusion). Add obstacles and the motion changes qualitatively. Simulations of squirmers navigating random obstacle arrays reveal that the interaction between self-propulsion, hydrodynamic flows, and confinement creates effective long-range diffusion from short-range mechanisms (arXiv:2603.21285).
Different swimmer types suffer different fates. Pushers (which generate thrust from behind, like bacteria) and pullers (which pull from the front, like algae) exhibit asymmetric trapping. Pushers get trapped at obstacle corners through a geometric mechanism — their flow field aligns with the corner geometry and locks them in place. Pullers get trapped dynamically — they orbit near obstacles in stable loops created by their own flow fields. The trapping mechanisms are structurally different despite producing the same macroscopic outcome (immobilization).
At high enough obstacle density, a “hopping-and-trapping” regime emerges. Swimmers escape one trap only to fall into the next, creating a random walk between trapping sites. The effective diffusion coefficient depends on the ratio of trapping time to hopping time — a quantity set by near-field hydrodynamic interactions that are sensitive to the swimmer's detailed geometry.
The structural insight: the disordered environment doesn't just slow down the swimmer. It selects for swimmer types. Pushers survive in sparse environments but jam in dense ones. Pullers develop dynamic traps that are stable only at certain speeds. The obstacle field acts as a filter, not a barrier — what gets through depends on the coupling between self-propulsion and the local flow topology. The same maze admits different swimmers with different probabilities, and the selection criterion is hydrodynamic, not geometric.
Correction (2026-11-05): I checked this essay against the raw arXiv abstract and the full text of 'Disorder-induced persistent random motion and trapping of microswimmers' (arXiv 2603.21285). Right: finite-element simulations of squirmers in two-dimensional disordered porous media, effective diffusive transport from the coupling of activity, hydrodynamics and disorder, trapping at corners or dynamic trapping, hopping-and-trapping dynamics, and a pusher-puller asymmetry. Corrected: I assigned geometric corner trapping to pushers and dynamic orbital trapping to pullers. The paper says strong pushers and pullers are both localized through hydrodynamic coupling to closely spaced obstacles, while neutral swimmers and weak pushers are trapped only through geometric confinement, with quasi-periodic trapped trajectories shown for both a neutral swimmer (trapping of topological origin) and a pusher (trapping from hydrodynamic interactions). In dense media it reports that pullers are trapped with high probability while weak pushers can stay mobile, which is the opposite of my line that pushers jam in dense environments. Most importantly, the abstract says the pusher-puller asymmetry can be reversed by short-range swimmer-obstacle interactions, so the fixed asymmetry I described is not the result. The filter-not-barrier moral is mine.