A dragonfly perched on a reed at dusk is not just waiting for the light to fade. It is hunting in wavelengths you cannot see.
Researchers at Osaka Metropolitan University have discovered that dragonflies detect deep red light at wavelengths around 720 nanometers, well past the roughly 680-nanometer boundary where human red-cone vision drops off. The study, published in Cell Reports in early 2025 and gaining renewed attention as of June 2026, represents the first confirmed case of near-infrared light detection in a land-dwelling animal through an opsin-based visual system, the same class of light-sensitive proteins that power color vision in humans, rather than a heat-sensing organ like those found in pit vipers. (It is worth noting that certain deep-sea fish possess opsins sensitive to similarly long wavelengths, but among terrestrial animals this is unprecedented.)
What makes the finding especially striking is how the dragonflies pull it off. The molecular mechanism is virtually identical to the one mammals use to see red.
Two lineages, one solution, hundreds of millions of years apart
The team, led by biophysicists Mitsumasa Koyanagi and Akihisa Terakita, isolated long-wavelength opsins from dragonfly compound eyes, expressed them in cultured cells, and measured their absorption spectra. The proteins peaked near 720 nanometers, placing them squarely in the deep red to near-infrared zone.
The critical discovery was pinpointing the amino-acid position responsible for that extreme red shift. When the researchers mutated that single residue, the sensitivity peak moved in predictable directions. When they introduced the same mutation into mammalian red opsins, the shift was comparable. Two evolutionary lineages that diverged hundreds of millions of years ago had independently landed on the same spectral-tuning trick at the same spot in the protein.
“This convergent evolution at the molecular level is remarkable,” Koyanagi said in the university’s research announcement, which describes how dragonflies and humans share a biochemical strategy for pushing vision toward the red end of the spectrum. Dragonflies, however, pushed it further, past the edge of what we can perceive.
Why dragonflies were already suspects
Dragonflies have long stood out among insects for the sheer complexity of their visual hardware. A 2015 genomic survey documented an unusually diverse opsin repertoire in dragonfly eyes, with some species expressing more than a dozen distinct opsin genes. Most insects get by with three or four. That genetic richness, distributed across specialized regions of the compound eye, made dragonflies strong candidates for expanded color sensitivity.
Reviews of dragonfly visual ecology have described how different zones of the compound eye map to specific tasks: tracking fast-moving prey overhead, scanning the horizon, detecting mates. But until the Osaka Metropolitan team’s work, no study had shown that any insect opsin could reach wavelengths past the conventional long-wavelength boundary for arthropods. The genetic potential was there. The functional proof was not.
What 720 nanometers actually means
To put the number in perspective: 720 nanometers sits at the extreme fringe of what humans can perceive under ideal conditions. Most people experience it as a faint, barely-there reddish glow, if they notice it at all. For dragonflies, that same wavelength falls within the working range of a dedicated photoreceptor class.
This matters ecologically because long-wavelength light behaves differently than shorter wavelengths at twilight. As the sun drops below the horizon, blue and green light scatter and fade faster than red. A visual system tuned to deep red could, in theory, give a predator like a dragonfly a meaningful advantage during the low-light hunting window at dusk, when many small insects are still active but harder to spot against a dimming sky.
That dusk-hunting hypothesis has been discussed in published reviews of dragonfly color vision and crepuscular foraging behavior, and Koyanagi and Terakita’s team has flagged it as a plausible ecological application of the 720-nanometer sensitivity. But it has not been tested directly. No published behavioral experiment has isolated the contribution of near-infrared perception to prey capture or any other dragonfly activity under natural conditions. The researchers have been careful not to overclaim on this point.
Open questions the study does not answer
Several significant gaps remain between the molecular proof and a full understanding of what dragonflies do with near-infrared vision.
Filtering in the living eye. The study measured opsin properties in cultured cells, not in intact eyes. In a living dragonfly, incoming light must pass through cuticle, pigment cells, and crystalline cones before reaching the photoreceptor membranes. Each layer can absorb or scatter specific wavelengths. Earlier research has shown that screening pigments in dragonfly eyes can shift or narrow the effective sensitivity of underlying photoreceptors. How much 720-nanometer light actually reaches these opsins in vivo is an open question.
Neural processing. The study did not trace how signals from near-infrared-sensitive photoreceptors are handled in the dragonfly brain. Dragonflies are famous for their sophisticated motion-detection and target-tracking neural circuits, but whether near-infrared input feeds into those pathways, or supports entirely different tasks like horizon detection or mate recognition, remains unknown.
How widespread is this? The researchers examined a limited number of dragonfly species. Comparative sequence data hint that similar red-shifted opsins exist in other lineages within the order Odonata, but comprehensive alignments across the group have not been published. Whether near-infrared sensitivity is a common dragonfly trait or a specialization tied to particular habitats and lifestyles will require broader sampling.
Engineered opsins and the biotech angle
Beyond the evolutionary story, the study opens a practical door. The Osaka Metropolitan team did not stop at characterizing the natural 720-nanometer opsin. They engineered variants with additional substitutions at the same tuning site and pushed the sensitivity peak even deeper into the infrared.
That result matters for optogenetics, a technique that uses light-activated proteins to control specific neurons. One of the persistent challenges in optogenetics is that the blue and green light most commonly used does not penetrate tissue well. Red and near-infrared wavelengths pass through skin and muscle far more effectively. An opsin that responds reliably to light beyond 720 nanometers could, in principle, allow researchers to activate neurons deeper in the body without invasive light delivery.
The gap between a lab-engineered protein and a clinical tool is vast. But the molecular proof of concept is now published, and it comes with a clear mechanism: a single amino-acid position that acts as a tunable dial for spectral sensitivity. That gives protein engineers a specific target to optimize rather than a black box to screen randomly.
What dragonfly vision reveals about the limits of sight
For readers who follow sensory biology, the core takeaway is precise. Dragonflies and mammals converged on the same molecular trick for seeing red, yet dragonflies extended that strategy past the boundary of human perception. A small change in protein sequence unlocked a new sensory channel, and evolution found that solution independently in lineages separated by an enormous span of time.
The discovery also reframes how biologists think about the “visible spectrum.” That familiar rainbow from violet to red is not a fixed property of light. It is a description of what human photoreceptors happen to detect. Dragonflies, with their 720-nanometer opsins, are a vivid reminder that other animals inhabit a wider visual world, and that the molecular machinery for accessing it may be simpler to build, and to borrow, than anyone expected.
More from Morning Overview
*This article was researched with the help of AI, with human editors creating the final content.
