The late Ediacaran marks a pivotal interval in early animal evolution, when benthic organisms began engaging with increasingly heterogeneous and dynamic seafloor environments. These shifts signal the emergence of more sophisticated locomotory and sensory capacities, yet the perceptual capacities of early motile animals remain poorly resolved because body fossils from this interval rarely preserve diagnostic sensory organs, if such structures existed. However, trace fossils offer a complementary archive, directly recording the behavioral consequences of perception, but quantitative estimates of sensory mechanisms and effective perceptual ranges have remained out of reach. Here, we estimate perceptual/sensory distances in over 230 Ediacaran–Cambrian tracemakers by integrating persistent homology analysis of trace fossils with numerical simulations of cue-gradient-driven foraging. Persistent homology captures perceptual coverage across spatial scales, providing a topological analog for integrated sensing and space-filling navigation. Simulations based on Gaussian distance-weighted sensory integration reveal a strong linear relationship between perceptual distance and the centroid of the persistence curve, enabling direct inference of sensory distance from fossilized movement paths. Application of this framework to stratigraphically constrained looping trails reveals an exponential expansion of sensory distance from millimetric contact-based sensing, subcentimetric chemo-/mechanosensation to decimetric long-range visual-sensation by the terminal Ediacaran. This expansion transformed environmental cues from locally random noise into structured, exploitable signals, shifting animal–environment interactions from passive response to active selection. Such restructuring of early benthic information landscapes provided a biological and ecological foundation for intensified bioturbation, niche diversification, and the increasingly complex species interactions that foreshadowed the Cambrian Explosion.
Wang et al. (Mon,) studied this question.