Trees are gigantic and long living organisms where gravitropic movements are governed by cambial growth spread along organs. By confronting simple biomechanical models to observations and experiments, we will try to synthetize how size and time constrain the ability of trees to control their posture. Basic dimensional effects of length and diameter will be stated using a now widely used model describing how stem curvature changes with time. This model suggests that diameter and cambial growth strongly limit the curvature rate. Previous investigations in greenhouse experiments where a strong initial perturbation of lean was applied will be used to discuss how the capacity of gravitropic response differs between individuals according to species autecology, growth conditions and ontogenic stage. Then, we will present new results about movements of saplings observed in natural conditions and try to demonstrate that i) the posture control is a relevant ecological process contributing to explain how saplings respond to the disturbance of their environment, ii) this biomechanical function could be a reason why resource allocation is preferentially directed to cambial growth after gap opening. We will also investigate the scaling of curvature changes among a wide range of size and competition environments in a beech silvicultural experiment, and show that the control of reaction wood formation does not balance dimensional effects. Lastly we will discuss how allometric relationships between diameter and height, modified by environment and ontogeny, could be derived from these considerations. The conclusion emphasizes that allometric laws of posture control must be discussed with respect to other constraints such as the self-weight and long-term consequences on the tree development over its whole life.
