The reinforcement mountain: allocation of behavior as a function of the rate and intensity of rewarding brain stimulation.

The single-operant matching law has been used to describe the relationship between time allocated to pursuit of brain stimulation reward (BSR) and the obtained rate of reinforcement. We generalize this relationship to a third dimension by including the strength of the stimulation (the number of pulses per train) as an independent dimension, and we dub the resulting 3-dimensional structure "the reinforcement mountain." The validity of generalizing the single-operant matching law in this way was assessed by determining the changes in the position of the mountain produced by increasing the stimulation current or the train duration. Most of the predictions were supported, and the mountain model fitted the data closely. It is argued that application of this model can remove ambiguity inherent in 2-dimensional descriptions of operant performance and can reveal whether lesions, drugs, or physiological manipulations that alter performance for BSR act before or after the output of the ("reward-growth") function that translates the electrically induced impulse flow into the intensity of the BSR.