Figure 2.

Development and characterization of the novel one-trial place memory task (experiments 1 and 2). Stippled horizontal lines indicate chance values of performance measures. The rats in experiment 1 (gray symbols; n = 15) were previously trained on a flavor-place memory task, whereas the rats in experiment 2 (black symbols; n = 16) were only shaped to dig in sandwells and habituated to the arena before training on the place memory task. A, Task acquisition (experiments 1 and 2a). The percentage of correct first choices (percentage of rats digging first in the correct sandwell; left) and errors (number of novel wells in which rats dug before digging in the correct one; mean ± 1 SEM; middle) for the initial training trials conducted with a 5 min delay. Trial 17 was a probe trial, in which food was omitted during the retrieval phase and the rats' dig time was measured for 60 s to calculate the percentage of dig time at the correct sandwell and the average percentage of dig time at novel sandwells (mean ± 1 SEM; right). B, Delay dependence of one-trial place memory (experiment 2b). The rats were trained with 5, 20, 60, 180, and 360 min delays (six trials at each delay: four standard training trials and two probe trials). The dig-time measures for the probe trials at the four different delays are depicted (average of two probe trials; mean ± 1 SEM). C, Arena rotation between encoding and retrieval (experiment 2c). To demonstrate that performance did not rely on cryptic odor cues, rats were tested with a 20 and a 360 min delay, with the symmetric arena being rotated by 90 or 180° between encoding and retrieval while the configuration of the intra-arena landmarks and the extra-arena room cues was kept constant (six trials at each delay: four standard training trials and two probe trials). The dig-time measures for the probe trials at the two different delays are shown (average of two probe trials; mean ± 1 SEM). D, Darkness during retrieval (experiment 2d). To demonstrate the requirement of visuospatial cues, retrieval in darkness was compared with retrieval in light (one probe trial for each condition) at a 20 min delay. The dig-time measure (mean ± 1 SEM) is shown for both conditions. Only 12 rats dug under both conditions (main graph), whereas four rats did not dig in any sandwell during darkness but performed normally in light (inset). That the average percentage of dig time in the dark condition (right; 32.3 ± 10.1%) is numerically higher than chance is essentially attributable to a single rat that dug briefly (1.0 s) in the correct well, without touching the novel wells.