Figure 4.
Three-gain experiment. (A) Relative bout duration, (B) relative interbout duration, (C) relative number of bouts, and (D) relative latency to initiate swimming as a function of trial number in the three-gain experiment. The gain varies across trials as shown in (E): periods 1 (trials 1–3) and 4 (trials 10–12) have a gain setting of 1, periods 2 (trials 4–6) and 6 (trials 16–18) have a gain setting of 1.5, and periods 3 (trials 7–9) and 5 (trials 13–15) have a gain setting of 0.5. In (A,B) single and double asterisks denote significant differences between values at the same gain setting. In (A), * denotes a significant difference in the bout durations between trials 1 and 10 (paired t-test, p < 4 × 10−4) and also in the bout durations averaged across trials with the same gain setting (i.e., trials 1–3 compared to trials 10–12; paired t-test, p < 3 × 10−5). Also in (A), ** denotes a significant difference between trials 4 and 16 (paired t-test, p < 4 × 10−4) and also between the values averaged across trials with the same gain setting (i.e., trials 4–6 compared to trials 16–18; paired t-test, p < 2 × 10−3). In (B), * denotes the corresponding significant difference in the interbout durations: between trials 1 and 10 (paired t-test: p = 0.034) and trials 1–3 compared to trials 10–12 (paired t-test: p < 0.01) whereas ** denotes a significant difference in the interbout durations between trials 4 and 16 (paired t-test: p = 0.012) and trials 4–6 compared to trials 16–18 (paired t-test: p < 0.01). For these experiments, all error bars denote SE of the mean and n = 28 larvae.