Predicting Adaptive Behavior in the Environment from Central Nervous System Dynamics
Figure 3
The dynamics of the Aplysia feeding CPG.
Data from the experiments of Proekt et al. [39], already mapped onto our model variables. In the buccal feeding CPG preparation in vitro, Proekt et al. stimulated either CBI-2 alone (A) EN alone (B), or CBI-2 with an embedded period of EN stimulation (C) in the pattern shown here, represented by the stimulus variable S. From each motor program elicited by the stimulation, Proekt et al. measured the B8 firing frequencies in protraction and retraction (see Figure S1 and Text S1, Section 6.1), here mapped onto the single ingestive-egestive dimension of the variable B. Each filled or empty circle (CBI-2 or EN stimulation, respectively) represents the mean±SE of 6–18 programs; altogether the dataset contains 466 programs. The red curve is the best fit (see Text S1, Section 1.4) of the 2D model (equations summarized above; for complete model specification see Text S1, Section 1.3). The blue curve shows the corresponding behavior of the 1D model, not an independent fit to the data but rather simply the behavior of the 2D model with the memory M set to 0 (see Text S1, Section 2).