Fig. 3 | Nature Communications

Fig. 3

From: A retrieval-specific mechanism of adaptive forgetting in the mammalian brain

Fig. 3

Forgetting is independent of the retrieval cue. a Experiments 3 and 4 include an additional session during encoding in which object B is also presented in arena context 2. During practice, we also interleaved sessions in arena context 2 to match the amount of additional exposure to the arenas. The final test was done in arena context 1 or 2. b The assumed retrieval dynamics during practice. Efforts to retrieve object A in arena 1 should trigger control processes to inhibit B, but not a non-competitor object F presented only in arena context 2. Object F is only included in Experiment 4 (additional encoding phase for F is not represented in a, but see Supplementary Figure 5B). c (Left) Discrimination index means ± SEM during test in arena 1 (left) or arena 2 (right). Individual values used to calculate the mean and SEM are presented as dots. Arena 1: ***p = 0.0004 (t20 = 4.52), d = 1.96 (RP vs. IC); p = 0.0003 (t20 = 4.72), d = 2.04 (RP vs. TC). Arena 2: ****p < 0.0001 (t22 = 6.31), d = 2.36 (RP vs. IC); ****p < 0.0001 (t22 = 6.49), d = 2.85 (RP vs. TC). Bonferroni post hoc comparisons after a repeated-measures ANOVA, n = 11 animals (arena 1) and n = 12 animals (arena 2). (Right) Discrimination index means ± SEM during test in arena 1 (left) or arena 2 (right). We found significant forgetting of B when tested in either arena context (cue-independent forgetting), but no forgetting of F. Arena 1: ***p = 0.001 (t14 = 4.17), d = 3.02 (RP vs. IC); ***p = 0.0004 (t14 = 5.16), d = 2.22 (RP vs. TC), Bonferroni post hoc comparisons, n = 8 animals. Arena 2: p = 0.53, F(2,14) = 0.65, one-way ANOVA