Octopus bimaculoides’ arm recruitment and use during visually evoked prey capture

Octopus’ limb hyper-redundancy complicates traditional motor control system theory due to its extensive sensory inputs, subsequent decision-making, and arm coordination. Octopuses are thought to reduce flexibility control complexity by relying on highly stereotypical motor primitives (e.g., reaching^1,2,3,4 and crawling⁵) and multi-level processes to coordinate movement,^6,7 utilizing extensive peripheral nervous system (PNS) processing.^2,8,9 Division of labor along the anterior-posterior axis¹⁰ and limb specialization of the four anterior arms in T-maze food retrieval¹¹ further simplify control. However, specific arm recruitment and coordination during visually guided reaching behavior remains poorly understood. Here, we investigated visually evoked Octopus bimaculoides’ prey capture capabilities^12,13 by eliciting and examining prey-specific arm recruitment. When striking crabs, octopuses preferred synchronous arm recruitment, while sequential arm recruitment with a characteristic swaying movement is employed for shrimp. Such behavioral selection aligns with specific prey escape strategies and the octopus’ flexible arm biomechanical constraints. Although side bias existed, we found significant bilateral symmetry, with one side being functionally a mirror of the other rather than anterior arm use being functionally equal and differing to posterior arm use. Among arms, the second limb is unequivocally dominant for goal-directed monocularly driven prey capture. Although the eight arms share gross anatomy and are considered equipotential,^10,14 such arm use for specific actions could reflect subtle evolutionary adaptations. Finally, we quantitatively show, corroborating earlier observations,^10,15 that octopuses employ a dimension reduction strategy by actively deciding to recruit adjacent arms over other available arms during either sequential or synchronous visually evoked prey attack.