Fig. 6

a Light-sheet microscope combined with a behavioral rig used for PC imaging. Inset shows location of two example ROIs shown in d. Scale bar: 100 µm; ro rostral direction, l left, r right, c caudal. b Modified long-term adaptation protocol. c First bout duration in each trial in normal-reafference control larvae (N = 8), lag-trained non-adapting larvae (N = 8), and lag-trained adapting larvae (N = 9), mean across larvae is shown. d Imaging data processing flow shown for two example ROIs (see details in text). d_i Z-scored fluorescence in five trials. Vertical shaded bars indicate first swimming bout in each trial. d_ii First-bout-triggered fluorescence. d_iii First-bout-triggered responses in each trial. Thick lines represent box-filtered responses, with a filter length of nine trials. d_iv Four criteria that represent change in bout-triggered responses during important transitions of the experimental protocol. d_v Criteria converted into 4-digit ternary barcodes. d_vi First-bout-triggered fluorescence averaged across respective blocks of ten trials. e Clustering of ROIs using barcodes. e_i Barcodes of all ROIs pooled from imaged larvae. e_ii Within-fish fractions of ROIs assigned to different clusters. Only clusters containing, on average, at least 2% of ROIs in at least one experimental group are shown. Magenta rectangles indicate 0-0 + cluster, which was the only cluster that was significantly enriched in lag-trained adapting larvae (*p = 0.011; Kruskal–Wallis test). f_i (top), First-bout-triggered responses of all 0-0 + ROIs pooled from lag-trained adapting larvae. f_i (bottom), First-bout-triggered responses of 0-0 + ROIs, averaged within each lag-trained adapting larva. Solid line and shaded area represent mean ± SEM across larvae. f_ii First-bout-triggered fluorescence of 0-0 + ROIs, averaged across respective blocks of ten trials and across ROIs within each lag-trained adapting larva. Colored lines and shaded areas represent mean ± SEM across larvae.