Performance of goldfish trained in allocentric and egocentric maze procedures suggests the presence of a cognitive mapping system in fishes

The parallel operation of different spatial strategies could increase navigational efficiency and diminish the occurrence of flawed or inaccurate responses.

“the parallel operation of different spatial strategies could increase navigational efficiency and diminish the occurrence of flawed or inaccurate responses”

Fish can perform "place" responses, potentially denoting the use of an allocentric spatial coordinate reference system for navigation, likely based on map-like memory representations anchored to the spatial environment and independent of the subject's own position.

“fish can perform “place” responses, potentially denoting the use of an allocentric (“world-centered”) spatial coordinate reference system for navigation, likely based on map-like memory representations anchored to the spatial environment and independent of the subject’s own position [17,52,53]”

In Rodríguez et al. [17], goldfish trained in the mixed turn-place procedure used turn and place strategies concurrently and switched between them depending on task requirements and available information.

“the goldfish trained in the mixed turn-place procedure used turn and place strategies concurrently, and that they switched from one to another depending on the task requirements and the kind of information available”

In Rodríguez et al. [17], goldfish trained in a turn procedure learned to rely on a purely idiothetic, body-centered reference system to find the baited feeder and did not take into account allothetic information for task solution.

“These results denoted that the animals trained in the turn procedure learned to rely on a purely idiothetic, body-centered reference system to find the baited feeder, and that these animals did not take into account allothetic information for task solution.”

The cooperative use of separate but complementary spatial strategies could explain the better performance observed in animals in the turn-place procedure compared to animals in other groups that likely used one strategy alone.

“The cooperative use of separate, but complementary, spatial strategies could explain the better performance observed in the animals in the turn-place procedure compared to the animals in the other groups that likely used one of these strategies alone.”

In Rodríguez et al. [17], goldfish in the place procedure preferentially swam to the previously rewarded goal location irrespective of turn direction and were lost when extramaze visual cues were deleted, indicating a place strategy based on allothetic information.

“the goldfish in the place procedure preferentially swam to the previously rewarded goal location (place response), irrespective of turn direction, and they were lost during the probe tests in which the extramaze visual cues were deleted, indicating that they solved the task using a “place” strategy based on allothetic information”

Goldfish in the Rodríguez et al. [17] study seem able to make spontaneous shortcuts and detours without previous route-specific experience, accurately navigating towards the goal place and choosing the most direct trajectory during transfer tests from novel start locations, even when the maze was displaced to new positions in the room.

“the goldfish in the Rodríguez et al. [17] study seem to be able to make spontaneous shortcuts and detours without previous route-specific experience, i.e., they accurately navigated towards the goal place, choosing the most direct trajectory to the goal during the transfer tests in which they departed from novel start locations, and even when the maze was displaced to new positions in the room”

These results suggest that these animals used a map-like spatial memory representation that endowed them with the capability to infer direct pathways to the goal place from unfamiliar start locations.

“The results suggest that these animals used a map-like spatial memory representation that endowed them with the capability to infer direct pathways to the goal place from unfamiliar start locations.”

Shortcutting and detouring behavior has been proposed as key evidence to distinguish allocentric navigation based on cognitive maps or survey representations from more simple navigational strategies.

“shortcutting and detouring behavior has been proposed as a key evidence to distinguish allocentric navigation based on cognitive maps or survey representations from more simple navigational strategies”

In Rodríguez et al. [17], goldfish trained to navigate to a fixed location in a four-arm maze surrounded by distal visual cues were disoriented when the maze was completely encircled by curtains, evidencing reliance on extramaze visual cues.

“the goldfish trained to navigate to a fixed location in a four-arm maze surrounded by an array of distal visual cues (place task) were disoriented during the probe tests in which the maze was completely encircled by curtains, evidencing that they relied on the extramaze visual cues”

These goldfish were still able to successfully find the goal when extramaze cues were partially occluded, implying that the goal place was defined by its redundant spatial relationships with multiple cues.

“these goldfish still were able to successfully find the goal when the extramaze cues were partially occluded, which imply that the goal place was defined by its redundant spatial relationships with multiple cues”