A neural path for visual discrimination of magnitudes in zebrafish

The capacity to handle continuous (spatial or temporal) and discrete (numerosity) quantity evolved from a conserved system for approximating non-symbolic numerical magnitude, a phenomenon well-documented across various species, including fish. While neuronal systems specifically tuned to quantity discrimination have been identified in the parietal and prefrontal cortex of both humans and non-human primates, as well as in analogous areas in crows and young chicks, little is known about subpallial circuits subtending this ability in vertebrates. Combining a behavioural paradigm of habituation/dishabituation with molecular biology assays, we have partially uncovered the neural network associated with quantity discrimination in the adult zebrafish brain. Briefly, the fish underwent habituation to sets of either 3 or 9 red dots over four consecutive days. Throughout this period, the dots underwent variations in density, position, and size, while maintaining their numerosity and total surface area. During the dishabituation phase, zebrafish faced changes in (i) the number of dots (from 3 to 9 dots or vice versa, while keeping the overall surface area constant), (ii) the shape (switching from 3 or 9 red dots to 3 or 9 red squares with the same overall surface), or (iii) the size (maintaining the same shape and number). A control group underwent testing with identical stimuli as in the habituation phase. Following the dishabituation test, zebrafish were euthanized thirty minutes later, and their brains were harvested and dissected to quantify immediate early genes expression by quantitative chain reaction(qPCR). Additionally, we used egr-1 in situ hybridization assays to pinpoint the positional identity of neuronal correlates of perceived changes in quantity (number or size) or shape. We found a role of the retina and optic tectum in the encoding of continuous magnitude (e.g., a change in stimulus size). Additionally, we identified a involvement of specific tectum layers, the habenula, the preglomerular complex, and the caudal regions of the dorsolateral and dorsocentral pallium in the encoding of discrete magnitude (e.g., a change in numerosity). The most rostral part of the dorsal pallium revealed a response to shape discrimination. In summary, our results suggest an initial involvement of thalamic and tectal areas in the encoding of both continuous and discrete quantities. In contrast, the pallial regions seem implicated specifically in the encoding of discrete quantity. This study may provide the foundation for using zebrafish as an experimental model to explore developmental dyscalculia, a human disability associated with learning and comprehending quantity.