Supporting Information

Files in this Data Supplement:

Fig. 5. The network state changes coherently along multiple dimensions. In the experiments shown in Fig. 2, we measured, simultaneously with the B8 firing frequencies in protraction and retraction, two additional parameters of the feeding motor programs: protraction phase duration (red symbols) and retraction phase duration (blue symbols). As shown here, these parameters were also determined by the previous history of activity rather than the immediate stimulus. Thus, the network state changes coherently along all four of these dimensions of feeding motor programs as a function of the activity of the network. EN, esophageal nerve; CBI-2, cerebral-buccal interneuron 2. ***, P < 0.001.

Fig. 6. Conclusions of the B20–B8 analysis are robust. In our statistical analysis of the contribution of B20 to B8 firing and the B20–B8 functional synaptic efficacy, we had to choose a value for the length of the window after B20 spikes (Statistical Analysis of B20 Contribution of B8 Firing in Feeding Motor Programs). The crosscorrelation histogram in Fig. 4B suggested 70 ms, and we used this value throughout Fig. 4 C and D. To make sure that our conclusions were not strongly affected by this specific choice, we varied the window length. Although the absolute values of the B20 contribution and the functional synaptic efficacy were altered, all of the trends that were present with the 70-ms window were still present with 50-ms and 90-ms windows, as shown here. Furthermore, all of the comparisons that were statistically significant in Fig. 4 C and D remained significant (P < 0.05). Thus, the conclusions of the analysis in Fig. 4 are robust against changes in the only free parameter in our method. EN, esophageal nerve; CBI-2, cerebral-buccal interneuron 2.