Mantis Shrimp Vision Preview

I haven’t been able to post any hard science this week since I’ve been working on a presentation for my department’s annual symposium. So far, this is the largest audience I have presented my work to.

I thought I might quickly share one of my slides from that presentation as a preview for a much larger future post about the ridiculously complicated mantis shrimp visual system.

Click to embiggen. Stomatopod Photo: Roy Caldwell

This is a comparison of photoreceptor classes in human and mantis shrimp retinas. Each photoreceptor class has a distinct wavelength sensitivity curve. On the human plot, you can see our three cone photoreceptor classes; blue, green, and red. These receptors cover the electromagnetic light spectrum between 400 nm (violet) and 700 nm (red). Our brains are able to process relative stimulation between the three cone photoreceptor classes, allowing us to differentiate many colors.

Mantis Shrimp don’t have the advantage of a large brain for downstream processing, so they take another approach to seeing many colors: They have 16 distinct photoreceptor classes, packed via optical filtering into tight slivers of the spectrum. Of these, five classes are sensitive to UV light, below our visual range (these are the receptor classes that I am attempting to characterize). In addition, not shown in this slide, mantis shrimp can discriminate linearly and circularly polarized light.

Stay tuned for an in depth mantis shrimp vision post at some point.


  1. […] are often adorned with attention grabbing color and polarization patterns that stand out to other visually adept mantis shrimp. Attenuation of light in water. Adapted from Levine and MacNichol, 1982However, the deep ocean is […]

  2. […] of marine awesomeness.  But that’s not fall it brings to the table: The mantis shrimp has 16 (!) color receptors in its eyes (compared to a normal human’s paltry 3), enabling it to see both ultraviolet and […]

  3. Daniel SImon says:

    Why does the graph for the mantis shrimp have 13 peaks? There are 12 cone classes and the other 4 channels are polarization based. Where does the other peak come from?

  4. Michael Bok says:

    One of the polarization receptors in included here, it is the one that peaks sharply around 350 nm. There are three primary polarization channel, this 350 nm one plus two that I excluded that peak very broadly in the blue-green. That brings the total to 15, but there is also one receptor cell that we don’t have spectral recordings from (it is a tiny UV-sensitive photoreceptor) which is also excluded.