1. We analyzed the rapid adaptation of elbow movement to unexpected changes in external load conditions at the elbow joint. The experimental approach was based on the lambda model, which defines control variables (CVs) setting the positional frames of reference for recruitment of flexor and extensor motoneurons. CVs may be specified by the nervous system independently of the current values of output variable such as electromyographic (EMG) activity, muscle torques, and kinematics. The CV R specifies the referent joint angle (R) at which the transition of flexor to extensor activity or vice versa can be observed during changes in the actual joint angle, theta, elicited by an external force. The other CV, the coactivation (C) command, instead of a single transition angle, defines an angular range in which flexor and extensor muscles may be simultaneously active (if C > 0) or silent (if C < 0). Changes in the R command result in shifts in the equilibrium state of the system, a dynamic process leading to EMG modifications resulting in movement or isometric force production if movement is obstructed. Fast movements are likely produced by combining the R command with a positive C command, which provides movement stability and effective energy dissipation, diminishing oscillations at the end of movement. 2. According to the model, changes in the load characteristic (e.g., from a 0 to a springlike load) influence the system's equilibrium state, leading to a positional error. This error may be corrected by a secondary movement produced by additional changes in R and C commands. In subsequent trials, the system may reproduce the CVs specified after correction in the previous trial. This behavior is called the recurrent strategy. It allows the system to adapt to the new load condition in the subsequent trials without corrections (1-trial adaptation). Alternatively, the system may reproduce the CVs specified before correction (invariant strategy). If the movement was perturbed only in a single trial, the invariant strategy allows the system to reach the target in the subsequent trials without corrections. 3. To test the assumption on the dominant role of the recurrent strategy in rapid adaptation of movement to new load conditions, we performed experiments in which subjects (n = 6) used a pivoting manipulandum and made fast 60 degrees movements to a target. After a random number of trials (5-10) with no load, we introduced opposing (experiment 1), assisting (experiment 2), or randomly varied opposing or assisting loads (experiment 3) for 5-10 trials before unexpectedly switching loads again (14-18 switches in total). The opposing or assisting torque was created by position feedback to a torque motor and was a linear function of the displacement of the manipulandum form the initial position (springlike load). Subjects were instructed to correct positional errors as soon as possible to reach the target. The EMG activity of two elbow flexors (biceps brachii and brachioradialis) and two elbow extensors (triceps brachii and anconeus), elbow position, velocity, and torque were recorded. Kinematic and EMG patterns were compared with those obtained in similar experiments in which subjects were instructed not to correct errors. 4. In 94% of the trials in which a change in the load occurred, the primary movement was in error and was followed by a corrective secondary movement. In primary movements, both the phasic and tonic levels of EMG activity as well as the kinematics were load dependent, implicating reflex and intramuscular mechanisms in the adaptation of muscle forces counteracting external loads. These mechanisms, however, were not sufficient to eliminate positional errors. 5. An undershoot error occurred in trials with an opposing load after those with no load or in trials with no load after those with an assisting load. After adaptation to a new load condition, a sudden return to the previous load condition resulted in an error of the oppo
Genome sequences reveal global dispersal routes and suggest convergent genetic adaptations in seahorse evolution