Numerical Cognition in Zebrafish (Danio rerio)

Potrich, Davide (2018) Numerical Cognition in Zebrafish (Danio rerio). PhD thesis, University of Trento.

[img]PDF - Disclaimer
Restricted to Repository staff only until 23 March 2020.

1174Kb
[img]PDF - Doctoral Thesis
Restricted to Repository staff only until 23 March 2020.

5Mb

Abstract

The use of numerical information is widespread throughout the animal kingdom, providing adaptive benefits in several ecological contexts, including foraging, anti-predatory strategies and mating. Given the importance to possess numerical abilities, it is plausible that similar selective pressures in favor of processing numerical information would have acted in different species, even in those more distantly related to humans, such as fish. The aim of this work was to investigate several aspects of numerical abilities in zebrafish (Danio rerio). In the first part, discrimination of quantity (magnitude) was investigated. Zebrafish were tested in free-choice experiments for their preference for different numerosities of conspecifics, taking advantage of shoaling behavior. Zebrafish chose to approach the location previously occupied by the larger in number between two groups of conspecifics (no longer visible at test) in sets of 1 versus 2 items and 2 versus 3 items, but failed at 3 versus 4 items. Similarly, when tested with larger numbers, zebrafish succeeded with 2 versus 4, 4 versus 6 and 4 versus 8 items, but failed with 6 versus 8 items. The results suggest that zebrafish rely on an approximate number system to discriminate memorized sets of conspecifics of different magnitudes and the degree of precision in recall is mainly dependent on the ratio between the sets to be discriminated. The aim of the second part was to investigate, for the first time, the use of proto-arithmetic addition abilities in zebrafish. Fish were tested in a spontaneous choice-test paradigm in which sets of conspecifics disappeared one-by-one behind one of two opaque screens, forming two groups that differed in number. Fish preferred to inspect the screen occluding the larger group of fish in sets of 1 versus 2 items and 2 versus 4 items, but failed at 2 versus 3 items. When tested under controlled conditions for continuous variables (overall time of the stimuli presentation in the two groups equalized) zebrafish were affected by the motion of the stimuli, showing a preference for the group of conspecifics that moved faster. Although results suggest that fish possess proto-arithmetic addition capacities, further studies seem to be needed to clarify in which circumstances zebrafish use numerical or non-numerical features. In the third part, ordinal numerical competencies were investigated. Fish learned to identify the second element in a series of five identical elements arranged sequentially. To assess whether zebrafish used ordinal information rather than non-numerical information, such as spatial distances, fish underwent a series of tests. When the length of the apparatus (exp. 1) and the inter-element distance (exp. 2) varied at test, creating a potential conflict with ordinal information, fish selected the correct ordinal position over spatial distance. Fish showed however difficulty when the set of elements changed, such as when the number of elements almost doubled (from 5 to 9 elements, exp.3). The aim of the fourth part was to study the possible link between number and space in the mapping of numerosities. Such an ability has been observed in primates and in birds, resembling a human mental number line. Zebrafish learned to associate a target number with a reward and then were tested in a preference choice test between two identical numerosities, but different to the training one, placed on the right and the left side of the experimental apparatus. Results suggested that zebrafish spontaneously associated smaller numbers with the left space and larger numbers with the right space, although a potential limitation of the use of mapping strategies is probably related to the ratio between the numerosities presented during training and testing. Control conditions confirmed that the overall perimeter and the overall area did not strongly influence the orientation of the supposed mental line. However, zebrafish were not completely unaffected by changes in surface areas of the stimuli. Given its widespread use in the field of genomics, zebrafish may provide a useful model organism in the study of the genetic bases of numerical cognition.

Item Type:Doctoral Thesis (PhD)
Doctoral School:Cognitive and Brain Sciences
PhD Cycle:30
Subjects:Area 11 - Scienze storiche, filosofiche, pedagogiche e psicologiche
Repository Staff approval on:26 Mar 2018 10:22

Repository Staff Only: item control page