The ocean’s midwater is a uniquely challenging yet predictable and simple visual environment. The need to see without being seen in this dim, open habitat, has led to extraordinary visual adaptations. To understand these adaptations, we compared the morphological and functional differences between the eyes of three hyperiid amphipods – Hyperia galba, Streetsia challengeri, and Phronima sedentaria. Combining micro-CT data with computational modelling, we mapped visual field topography and predicted detection distances for visual targets viewed in different directions across mesopelagic depths. Hyperia’s eyes provide a wide visual field optimized for spatial vision over short distances, while Phronima and Streetsia’s eyes have the potential to achieve greater sensitivity and longer detection distances using spatial summation. These improvements come at the cost of smaller visual fields, but this loss is compensated for by a second pair of eyes in Phronima and with behaviour in Streetsia. The need to improve sensitivity while minimizing visible eye size to maintain crypsis has likely driven the evolution of hyperiid eye diversity. Our results provide an integrative look at how these elusive animals have adapted to the unique visual challenges of the mesopelagic.

Specimens were euthanized by being immersed in iced seawater or the addition of a few drops of 95% ethanol to their holding tank. The heads and anterior-most segments of Hyperia and Phronima were dissected from the body, then fixed in 4% seawater buffered formalin (Hyperia) or 2% cacodylate buffered glutaraldehyde (Phronima). Hyperia and Phronima specimens were run through a dehydration series then stained with a 0.5% phosphotungstic acid (PTA) solution in 70% ethanol for 30 days before scanning. Streetsia specimens were fixed in 4% phosphate buffered paraformaldehyde and stained with an aqueous 1% w/v iodine and 2% w/v potassium iodide solution for seven days. Specimens were mounted in 500 µl sealed Eppendorf tubes stabilized with either minute blocks of foam or 1% low temperature gelling agarose. Hyperia and Phronima specimens were scanned using a GE Phoenix v|tome|x M 240/180 kV Dual Tube micro-CT at the Smithsonian National Museum of Natural History at 80 – 90 kV and 2.65 – 4.26 W, with optical magnification of 70.8 and 36.6x. Streetsia specimens were scanned with a Versa 520 XRM (Zeiss, Pleasanton, CA, USA) at 70 kV and 6 W with 4x optical magnification at the University of Western Australia. Voxel sizes of 2.82 – 11.17 µm were achieved and the standard 0.7 kernel size reconstruction filter was used. Raw projection data from Hyperia and Phronima were reconstructed using datos|X (GE Sensing and Inspection Technologies GmbH) and visualized and exported using VG Studio (Volume Graphics), while XRM Reconstructor software (v.10.7.3679.13921, Zeiss) following a standard centre shift and beam hardening (0.1) correction was used for Streetsia.