# Visual opponent mechanisms and spectral responses in non-primate vertebrates: taxonomic distribution, sampling, and classification

## TL;DR

This paper explores how different types of opponent cells in non-primate vertebrates are distributed and classified, highlighting the need for better sampling and classification methods.

## Contribution

The paper introduces three classification frameworks for opponent cell types to address cross-taxa ambiguity in vertebrate vision research.

## Key findings

- Opponent cells are found in mammals, reptiles, amphibians, fish, and birds, but not all groups show the same photoreceptor contributions.
- Only 58% of studies report the number of cells sampled, leading to potential biases in understanding opponent cell abundance.
- Three new classification frameworks are proposed to improve the consistency of opponent cell type identification across species.

## Abstract

Understanding neural mechanisms of opponency is the next frontier of comparative color vision research. Opponent cell properties can be investigated from different perspectives: the type of cone inputs to a cell (cone opponency), and the wavelengths of light to which the cell responds (spectral opponency). Based on a recent database (DOI 10.1016/j.dib.2024.111166) on cone and spectral opponency across non-primate vertebrates, we identified major taxonomic trends in the distribution of opponent cells through image-forming brain regions, analyzed sampling metrics and experimental variables influencing their detection, and established the limitations of existing classification systems. Evidence of opponency is reported in mammals, reptiles (exclusively turtles), amphibians, fish, and birds. While single cones drive every cone opponent cell described, double/twin cones and rods can also contribute in all groups, but reptiles. From a sampling perspective, only 58% of studies reported the number of cells sampled to find cone opponent cells. There are no estimates of the relative abundance of opponent cells, and our best approximation (frequency of opponent cell encounters) may be heavily biased by experimental goals, design, and equipment. To address cross-taxa ambiguity in opponency classification, we developed three classification frameworks for opponent cell types based on the relative positions of photoreceptor peak sensitivities and null point(s). Transparent sampling and universal classification methods will be critical for making evolutionary and functional inferences about opponency in vertebrate vision.