Can Fish See Water? The Science of Fish Vision

Fish do not see water as a visible object because water is the medium through which their visual system operates, not an object to be perceived. The question of whether fish can see water is best answered by examining how fish eyes process light, how the underwater environment shapes visual perception, and what practical limits exist for fish vision in aquaculture and research settings. This article explains the science of fish vision for students, researchers, life-science professionals, and informed general readers, with attention to what these principles mean for fish husbandry, monitoring, and welfare.

What Fish Eyes Are Designed to Detect

Fish eyes share the basic vertebrate plan with human eyes, including a cornea, lens, iris, and retina, but they are adapted for an aquatic environment where light behaves differently than in air. The cornea of a fish eye has almost no refractive power underwater because water and the corneal tissue have similar refractive indices. The spherical lens of a fish eye does most of the focusing work, and many fish species have lenses that are nearly spherical to compensate for the reduced corneal refraction.

The retina of a fish contains photoreceptor cells called rods and cones. Rods support vision in low light, while cones support color vision and visual acuity in brighter conditions. The balance of rods and cones varies widely among species, reflecting the light conditions of their natural habitats. Fish that live in dim or turbid water tend to have more rods, while fish in clear, well-lit surface waters tend to have more cones.

Color vision circuits in fish evolved underwater for hundreds of millions of years, long before animals moved onto land. The organization of modern visual systems, including human vision, is fundamentally shaped by these aquatic beginnings. Research on zebrafish has shown that their color vision circuits are linked to the statistics of natural light in the underwater world, and studying these circuits can help explain vision in general, including human vision. The NCBI Literature Resources provide access to the broader biomedical literature on visual system evolution and function.

Why Fish Do Not Perceive Water as a Visible Object

Water is transparent to fish because their visual systems are adapted to process light that has passed through water. A fish does not see water itself any more than a human sees air. What a fish sees are objects, prey, predators, and conspecifics that reflect or emit light within the water column.

The perception of a medium depends on a contrast between the medium and something else. Humans notice air when it contains dust, fog, or smoke because those particles scatter light. Fish may notice water when it contains suspended particles, bubbles, or temperature gradients that create visible distortion, but they do not perceive the water itself as a discrete object.

The practical implication is that fish rely on visual cues that are degraded by the very medium they inhabit. Water absorbs and scatters light, reducing contrast and limiting the distance at which fish can see. Turbidity, depth, and dissolved substances all affect how far a fish can see and what it can distinguish. These factors matter for feeding, predator avoidance, and social behavior in both wild and farmed fish.

The Underwater Light Environment

Light underwater is fundamentally different from light in air. Water absorbs light selectively by wavelength, with red light absorbed most quickly and blue-green light penetrating deepest in clear water. This means that at depth, the available light is shifted toward the blue-green part of the spectrum, and objects that reflect red light appear gray or black because the red wavelengths have already been absorbed.

Turbidity further reduces visibility by scattering light. Suspended particles such as sediment, plankton, and organic matter create a veiling effect that reduces contrast between an object and its background. A fish in turbid water may only see a few centimeters or meters, depending on the particle load. This has direct consequences for feeding behavior, since many fish are visual predators that need to see their prey to strike accurately.

The PubMed database includes a bibliographic record for a 1976 article titled "Underwater vision" from the Proceedings of the Royal Society of Medicine, which addresses the optical challenges of seeing underwater. The title and publication metadata indicate that the topic of underwater visual perception has been studied for decades, and the principles of light absorption and scattering remain central to understanding fish vision.

How Fish Use Vision in Their Environment

Fish use vision for a range of behaviors that are essential for survival and reproduction. These include locating prey, avoiding predators, recognizing conspecifics, selecting mates, navigating their habitat, and maintaining position within a school. The relative importance of vision compared to other senses such as olfaction, hearing, and the lateral line system varies by species and environment.

For many fish species, vision is the primary sense for feeding. Visual predators need adequate light and water clarity to detect and pursue prey. In aquaculture, this means that feeding behavior can be affected by water quality, stocking density, and lighting conditions. Farmers who observe reduced feed intake may need to consider whether poor visibility is a contributing factor.

Vision also plays a role in social behavior. Many fish species use visual signals for communication, including color patterns, fin displays, and body postures. These signals can be important for establishing dominance hierarchies, courtship, and schooling behavior. In farmed fish, the ability to see conspecifics can affect stress levels and social dynamics within the population.

The Limits of Fish Vision

Fish vision has several inherent limits that are important to understand. The first is the limited range of visibility imposed by water clarity. Even in clear water, fish cannot see as far as terrestrial animals can see in air because of light absorption and scattering. The second limit is the spectral range of fish vision, which varies by species. Some fish can see ultraviolet light, while others have more limited color vision. The third limit is the visual field, which depends on the position of the eyes on the head.

The arrangement of eyes on a fish head determines the field of view and the degree of binocular overlap. Fish with eyes on the sides of the head have a wide monocular field but limited binocular vision, while fish with forward-facing eyes have more binocular overlap and better depth perception. Research on fish-like binocular vision systems for underwater robots has shown that biological fish exhibit a remarkably broad-spectrum visual perception capability, and the eye arrangement of biological fish has inspired the design of wide-field vision systems for underwater robots. A fish-like binocular vision system developed for underwater robots achieved a horizontal field of view of 306.56 degrees, demonstrating the wide-field perception that fish eye placement enables. This work is reported in A Fish-like Binocular Vision System for Underwater Perception of Robotic Fish in Biomimetics.

At a Glance: Fish Vision and Practical Implications

Practical Demonstration: Testing Fish Visual Acuity

A simple experiment can illustrate the limits of fish vision and help students or researchers observe how fish respond to visual stimuli. This demonstration is designed for a controlled aquarium or tank setting and requires no specialized equipment beyond basic materials.

Materials Needed

• A clear glass or acrylic aquarium with established water conditions
• A target species of fish that is accustomed to the tank
• Two identical feeding rings or targets, one colored and one neutral
• A ruler or measuring tape
• A notebook for recording observations

Procedure

1. Place the two feeding rings at opposite ends of the tank, at the same depth and distance from the observation point.
2. Observe the fish for a baseline period of 10 minutes and record their swimming patterns and any approaches to either ring.
3. Introduce a small amount of feed near one ring and record how quickly the fish locate the food.
4. Repeat the observation with the feed placed near the other ring.
5. Record the time to first feeding response and the number of fish that approach each ring.

What to Record

• Time to first feeding response in seconds
• Number of fish approaching each ring within 5 minutes
• Water temperature and turbidity at the time of the test
• Lighting conditions in the room and any reflections on the tank glass

Interpreting the Results

If fish respond more quickly to one ring than the other, this may indicate that they can distinguish between the two visual targets. If there is no difference in response time, the fish may not be able to perceive the color difference, or the lighting conditions may not support color discrimination. Repeat the test under different lighting conditions to see whether the response changes.

This demonstration is not a controlled scientific experiment, but it provides a practical way to observe fish visual behavior and to understand the factors that affect what fish can see. For more rigorous approaches, researchers can consult the literature on fish vision and behavior available through PubMed.

How Researchers Study Fish Vision

Scientists study fish vision using a combination of behavioral experiments, anatomical examination, and physiological recording. Behavioral experiments present fish with visual stimuli and measure their responses, such as approaching a target or choosing between two options. Anatomical studies examine the structure of the eye and retina to determine the types and distribution of photoreceptors. Physiological studies record the electrical activity of retinal cells in response to light of different wavelengths and intensities.

The study of color vision circuits in zebrafish has provided detailed insights into how the fish brain processes spectral information. Research published in Circuit mechanisms for colour vision in zebrafish in Current Biology explains that the use of spectral information in natural light is one of the oldest and most fundamental abilities of visual systems, and that circuits for color vision evolved underwater for hundreds of millions of years. The study notes that comparatively little is known about the aquatic counterparts of terrestrial visual circuits, and that studying fish can help us understand vision in general, including human vision.

Vision in Aquaculture Monitoring

Fish vision is a practical concern for aquaculture in addition to being a topic of basic scientific interest. Many modern aquaculture systems use computer vision to monitor fish behavior, estimate biomass, and detect health problems. These systems rely on cameras that capture images underwater, and the same optical principles that affect fish vision also affect camera performance.

Underwater cameras face challenges from light absorption, scattering, and turbidity. Researchers have developed various methods to improve underwater image quality, including image enhancement algorithms and knowledge-based approaches that incorporate biological information about fish species. A study on Underwater fish image recognition based on knowledge graphs and semi-supervised learning feature enhancement in Scientific Reports describes a framework that integrates a Fish Multimodal Knowledge Graph with deep visual recognition to improve fish species identification under degraded underwater conditions. The study reports that this approach outperforms state-of-the-art underwater image enhancers and recognizers, particularly under low signal-to-noise and severe blur conditions.

Stereo Vision for Fish Measurement

Stereo vision systems use two cameras to capture images from slightly different angles, allowing the calculation of depth and three-dimensional structure. This technology has been applied to measure fish body length, estimate biomass, and monitor fish behavior in aquaculture settings.

A study on In-Water Fish Body-Length Measurement System Based on Stereo Vision in Sensors describes a non-contact method that uses binocular stereo vision to measure fish body length underwater. The system captures RGB and depth images, segments the fish using a contrast-adaptive algorithm, and corrects for the errors caused by water refraction. The experimental results indicate a mean relative percentage error of 0.9% for fish-length measurement, meeting the accuracy requirements for aquaculture applications.

Another study on TOF-assisted binocular vision accuracy improvement method for underwater fish size inspection in the Journal of the Optical Society of America describes a method that combines time-of-flight depth sensing with binocular stereo matching to improve the accuracy of fish size measurement. The study reports that the TOF-assisted system improves measurement accuracy compared to single binocular vision and reduces measurement error when the fish body has a significant inclination along the depth axis.

Computer Vision for Fish Detection and Classification

Computer vision systems are also used to detect and classify fish in underwater environments. These systems face challenges from water turbidity, fish camouflage, degraded image quality, and the similarity between fish and their background. A study on BiOLO-Wave: a bio-inspired chromatic-deformable attention network for camouflaged fish detection in turbid underwater environments in Scientific Reports describes a deep learning architecture that detects and classifies fish under degraded visual conditions. The study reports an accuracy of 93.88% and a mean average precision of 88.90% at an intersection over union threshold of 0.5, with an inference latency of 12.54 milliseconds.

The DeepFish dataset described in Scientific Reports provides approximately 40,000 images collected underwater from 20 habitats in tropical Australia. This dataset includes classification, point-level, and segmentation labels that enable models to learn to monitor fish count, identify locations, and estimate sizes. The study notes that while models pre-trained on ImageNet have performed successfully on this benchmark, there is still room for improvement in this challenging domain of underwater computer vision.

Open-Source Systems for Fish Monitoring

Open-source computer vision systems are making fish monitoring more accessible for small-scale aquaculture operations. The Kraken system described at the Latinoware 2025 conference is an open-source computational system for automated fish detection, counting, and size estimation in turbid freshwater environments. The system integrates computer vision frameworks including YOLO, OpenCV, and TensorFlow with open hardware platforms such as Arduino and Raspberry Pi. Experimental evaluation demonstrates that Kraken achieves an average size estimation at a fixed distance of 30 cm and a classification accuracy of up to 90%.

Robotic Fish and Vision Research

Robotic fish provide a platform for testing vision algorithms and understanding fish behavior. A study on Global Vision-Based Formation Control of Soft Robotic Fish Swarm in Soft Robotics describes a soft robotic fish swarm system with global vision positioning. The study shows that soft robotic fish can shift their formations to mimic three typical swarming behaviors of natural creatures: highly parallel group, encircling, and torus. This research may guide future work on soft robots and robotic swarms for underwater applications.

Another study on Vision-Based Obstacle Avoidance and Formation Control for Underwater Robotic Fish in IEEE Robotics and Automation Letters describes an adaptive image enhancement module that improves recognition performance under low-light and high-interference underwater conditions. The study validates the synergistic efficacy of the visual and control systems through experiments including target tracking, formation maintenance, and obstacle avoidance.

Practical Workflow for Assessing Fish Vision in a Farm or Lab Setting

For fish farmers, researchers, or educators who want to assess the visual environment of their fish, the following workflow provides a structured approach.

Step 1: Characterize the Water Environment

Measure water clarity using a Secchi disk or turbidity meter. Record the depth at which the Secchi disk disappears from view. This measurement gives an estimate of the visible distance for fish in the water. Also record water temperature, pH, and any visible color or suspended particles.

Step 2: Observe Fish Behavior

Spend at least 15 minutes observing the fish during a normal feeding period. Record the following observations:

• How quickly do fish locate food after it is introduced?
• Do fish approach food from a distance or only when it is close?
• Do fish respond to visual cues such as a hand passing over the tank or a moving object?
• Are there any fish that appear to have difficulty finding food?

Step 3: Test Visual Response

Use the simple demonstration described earlier to test whether fish respond to visual targets of different colors or sizes. Record the results and compare them across different lighting conditions.

Step 4: Review Lighting Conditions

Assess the lighting in the tank or pond. Is there adequate light for the fish species being kept? Are there areas of bright light and deep shadow that may create visual stress? Consider whether the lighting spectrum matches the species' visual capabilities.

Step 5: Document and Monitor

Keep a record of water clarity, fish behavior, and any changes in feeding response. This record can help identify trends over time and alert you to changes in water quality or fish health that may affect vision.

Records and Measurements for Fish Vision Assessment

Maintaining accurate records is essential for tracking changes in fish vision and the factors that affect it. The following table provides a template for recording key measurements.

Common Failure Patterns in Fish Vision Assessment

Several common problems can affect the assessment of fish vision and the performance of vision-based monitoring systems.

Turbidity Interference

Turbidity is the most common factor that degrades underwater visibility. When turbidity increases, fish may reduce feeding activity, and camera-based monitoring systems may fail to detect or classify fish accurately. If turbidity is high, address the underlying cause, which may include overfeeding, inadequate filtration, algal blooms, or sediment resuspension.

Lighting Artifacts

Reflections on the water surface or tank glass can create glare that interferes with both fish vision and camera systems. Surface reflections can also cause fish to perceive false visual cues. Position lights to minimize reflections and use diffusers to create even illumination.

Camera Calibration Errors

Stereo vision systems require careful calibration to produce accurate depth measurements. Refraction at the water-air interface and at the camera housing window can introduce errors. Regular calibration checks are necessary to maintain measurement accuracy.

Species-Specific Differences

Not all fish species have the same visual capabilities. A monitoring system or behavioral test that works for one species may not work for another. Always consider the species-specific visual system when interpreting results.

Welfare and Safety Context

Understanding fish vision has direct implications for fish welfare. Fish that cannot see well may experience stress from an inability to locate food, avoid predators, or interact with conspecifics. In aquaculture, poor water clarity can reduce feed intake and growth rates, and may contribute to uneven growth within a population.

The study on non-invasive characterization of rainbow trout responses to ammonia nitrogen stress in Biology demonstrates how computer vision can be used to assess fish welfare non-invasively. The study exposed rainbow trout to four ammonia nitrogen concentrations and used stereo vision to track locomotor behavior and optical-flow analysis to quantify ventilation frequency. The results showed that with increasing ammonia nitrogen concentration, average swimming speed decreased from 3.83 cm/s in the control group to 1.03 cm/s in the highest concentration group, while ventilation frequency increased from 84.91 breaths/min to 133.43 breaths/min. This research shows that vision-based monitoring can detect physiological stress in fish without handling or invasive procedures.

When working with fish, always follow established welfare guidelines and consult with a veterinarian or aquatic animal health professional if you observe signs of distress, disease, or abnormal behavior. If fish show a sudden loss of visual response or a marked change in feeding behavior, investigate water quality parameters immediately and escalate to a professional if the cause is not apparent.

Limitations of Current Knowledge

While much is known about fish vision, significant gaps remain. The study of color vision circuits in fish is still developing, and comparatively little is known about the aquatic counterparts of terrestrial visual circuits. Research on circuit mechanisms for colour vision in zebrafish notes that the computational strategies of fish color vision are linked to the statistics of natural light in the underwater world, but many details remain to be explored.

The practical application of fish vision research to aquaculture is also evolving. While computer vision systems have shown promise for fish monitoring, they face ongoing challenges from environmental variability, species diversity, and the complexity of underwater scenes. The DeepFish dataset study notes that current datasets for fish analysis tend to focus on classification tasks within constrained environments that do not capture the complexity of underwater fish habitats, and that there is still room for improvement in this challenging domain.

Professional Escalation Criteria

If you observe any of the following signs, escalate to a qualified professional such as a fish health veterinarian, aquatic biologist, or aquaculture extension specialist:

• A sudden and unexplained loss of feeding response across a fish population
• Fish swimming erratically or showing signs of disorientation that may indicate visual impairment
• Water clarity that deteriorates rapidly and does not respond to standard management interventions
• Camera-based monitoring systems that consistently fail to detect fish in conditions where they previously worked
• Any signs of eye disease, including cloudiness, swelling, or visible lesions on the eye

Frequently Asked Questions

Can fish see water?

Fish do not see water as a visible object because water is the medium through which they see. Just as humans do not see air, fish do not perceive water itself. What fish see are objects, prey, predators, and other fish that reflect or emit light within the water column. Water becomes visible to fish only when it contains particles, bubbles, or temperature gradients that scatter light.

How far can fish see underwater?

The distance fish can see depends on water clarity, which is affected by turbidity, dissolved substances, and light availability. In clear water, some fish may see several meters, while in turbid water, visibility may be limited to a few centimeters or less. The Secchi disk depth provides a practical estimate of visible distance in natural waters.

Do fish have color vision?

Many fish species have color vision, but the extent of color discrimination varies by species. Fish retinas contain cone photoreceptors that support color vision, and some species can see ultraviolet light. The color vision circuits in fish evolved underwater over hundreds of millions of years and are linked to the statistics of natural light in the underwater world.

Why do fish eyes look different from human eyes?

Fish eyes are adapted for an aquatic environment where the cornea has little refractive power. The lens of a fish eye is often nearly spherical to compensate for this, and the position of the eyes on the head varies by species to support different visual needs. Fish with side-set eyes have wide fields of view but limited depth perception, while fish with forward-facing eyes have more binocular overlap.

How does turbidity affect fish vision?

Turbidity reduces visibility by scattering light and reducing contrast between objects and their background. In turbid water, fish may have difficulty locating food, avoiding predators, and interacting with conspecifics. Turbidity also affects camera-based monitoring systems used in aquaculture.

Can fish see in the dark?

Some fish can see in very low light conditions because their retinas contain a high density of rod photoreceptors. However, no fish can see in complete darkness. Fish that live in dim or deep water rely more on other senses such as the lateral line system, olfaction, and hearing.

How is fish vision studied?

Scientists study fish vision using behavioral experiments, anatomical examination, and physiological recording. Behavioral experiments present fish with visual stimuli and measure their responses. Anatomical studies examine the structure of the eye and retina. Physiological studies record the electrical activity of retinal cells in response to light.

Why does fish vision matter for aquaculture?

Fish vision affects feeding behavior, social interactions, and stress levels in farmed fish. Understanding the visual environment of fish can help farmers optimize feeding schedules, lighting, and tank design. Computer vision systems that monitor fish behavior and estimate biomass also depend on the same optical principles that affect fish vision.

Related Articles

• Genome Science Uw
• Genome Science Uw
• Genome Science Uw
• Veterinary Science
• Health Science Degree

References and Further Reading

• NCBI Literature Resources. National Center for Biotechnology Information.
• PubMed. National Library of Medicine.
• A Fish-like Binocular Vision System for Underwater Perception of Robotic Fish.. Biomimetics (Basel, Switzerland), 2024.
• Underwater vision.. Proceedings of the Royal Society of Medicine, 1976.
• TOF-assisted binocular vision accuracy improvement method for underwater fish size inspection.. Journal of the Optical Society of America. A, Optics, image science, and vision, 2023.
• In-Water Fish Body-Length Measurement System Based on Stereo Vision.. Sensors (Basel, Switzerland), 2023.
• A realistic fish-habitat dataset to evaluate algorithms for underwater visual analysis.. Scientific reports, 2020.
• Underwater fish image recognition based on knowledge graphs and semi-supervised learning feature enhancement.. Scientific reports, 2025.
• Global Vision-Based Formation Control of Soft Robotic Fish Swarm.. Soft robotics, 2021.
• Circuit mechanisms for colour vision in zebrafish.. Current biology : CB, 2021.
• An Overview of the Research Status and Advances in Precision Feeding Technology and Equipment in Aquaculture. 2026.
• BiOLO-Wave: a bio-inspired chromatic-deformable attention network for camouflaged fish detection in turbid underwater environments.. 2026.
• A Freshwater Fish Dataset for Visual Recognition with Manually Localized ROIs and SAM-Derived Instance Masks.. 2026.
• From Beauty to Protection: How Phenotypic Traits Influence Conservation Perceptions of Freshwater Fish.. 2026.
• Non-Invasive Characterization of Locomotor and Ventilatory Responses in Rainbow Trout Under Acute Ammonia Nitrogen Stress.. 2026.
• An integrated method for non-intrusive underwater fish measurement based on keypoint detection and stereo vision. Aquaculture International, 2025.
• Vision-Based Obstacle Avoidance and Formation Control for Underwater Robotic Fish. IEEE Robotics and Automation Letters, 2025.
• Kraken: Open-Source Computer Vision for Underwater Fish Detection and Measurement. Anais do XXII Congresso Latino-Americano de Software Livre e Tecnologias Abertas (Latinoware 2025), 2025.
• Underwater swimming fish mass estimation based on binocular vision. Aquaculture International, 2024.
• A Method for Estimating Fish Biomass Based on Underwater Binocular Vision. Chinese Control and Decision Conference, 2024.
• DP-FishNet: Dual-path Pyramid Vision Transformer-based underwater fish detection network. Expert Systems with Applications, 2024.
• An integrated ROV solution for underwater net-cage inspection in fish farms using computer vision. SN Applied Sciences, 2020.
• Fine-Grained Underwater Visual Search Using a Hybrid-Driven Robotic Fish With Ultrawide FoV Vision. IEEE Transactions on Industrial Electronics, 2026.

This article is educational and does not replace institutional policy, professional advice, or applicable safety and regulatory requirements.