Circumnutations drive embodied mechanical sensing and support selection in twining plants
Ohad, A.; Porat, A.; Meroz, Y.
Show abstract
Climbing plants use self-generated oscillatory movements called circumnutations to search their environment for supports to attach to. Yet little is known about what information these movements provide. Here we show that circumnutations enable climbing plants to actively assess the mechanical stability of a newly encountered support and determine whether to initiate twining. Analogous to whisking in mammals, circumnutating shoots generate predictable mechanical loading that probes support resistance. Force measurements of freely circumnutating bean shoots reveal that contact forces follow a characteristic sinusoidal pattern. We develop a minimal physical model of this system, and experimentally informed simulations recover the measured force trajectories. We find that the stem-support interaction is captured by a simple torque balance between external loading and the intrinsic bending moment of the stem, equivalent to a cantilever beam with a rotating load. Analysis of force trajectories, supported by experimentally informed simulations, shows that force amplitude is set by stem stiffness and geometry, whereas the characteristic timescale is governed by the circumnutation rate. Twining occurs only after the stem reaches a critical torque threshold, corresponding to a threshold deformation of the stem that likely serves as the mechanical trigger for twining initiation, reflecting both sufficient support stability and a minimal geometric overshoot required for grasp. Motorized-stage experiments further demonstrate that increasing the effective circumnutation rate accelerates twining initiation to minutes, whereas reducing it can suppress twining despite prolonged contact. Together, these results establish embodied mechanical sensing in plants and show how morphology and self-generated motion enable support selection without centralized control.
Matching journals
The top 3 journals account for 50% of the predicted probability mass.