Matching the hatch and why it can't be done.

Historically, some fly-tyers have gone to great lengths to match the colour of hatching flies with the greatest possible precision. Whilst there is certainly much value in experimenting with colours, there is a reason why we could never know whether the trout sees our colour mixture as an accurate match to the natural, as we do. The reason is technically known as 'metamerism'.

The word simply means 'changing colour'. It sometimes happens that two surfaces which appear the same colour in one condition of lighting become obviously different when the light changes; as, for example, from daylight to artificial light. It happens as a result of the two surfaces being coloured with different pigment mixtures. It is particularly a problem for commercial dyers who are commonly required to accurately match a provided sample. The dyer does not know what mixture of dyes were used for the original but must find a mixture which gives a visually similar result. This he can easily do, usually in daylight. But it often occurs that viewing the new match in artificial light shows them to be quite different.

The reason is that our colour vision is not such that we can recognise different mixtures of light wavelengths, unlike our ears which do indeed recognise different mixtures, or 'chords', in terms of music. And for any one colour mixture there are many others which can look the same - under one condition of lighting but not others. Imagine an object painted with green paint where the pigment in the paint is a true green, that is to say it reflects those wavelengths of the visible spectrum which give rise to the colour green. Then imagine that we mix blue and yellow paints to match the green colour, while viewing it in daylight. Then we view the result under ordinary tungsten light bulbs. This light is strong in yellow wavelengths, moderate in green but quite deficient in blue. So the original green paint reflects the greeen element of the light but although our matching reflects the yellow portion, there is no light which the blue paint component can reflect. So the balance of the yellow and blue is greatly altered and no longer appears the same as the green paint.

Since the way that our eyes filter the components of a colour can differ between individuals in the comparative strength of the three filters, red , green and blue, not everyone would have agreed that we had a good match in the first place. 'Colour blindness' is an extreme case, and any person's eyes are likely to change with age to the extent that an older person often disagrees with a younger as to what is or is not a good colour match. To this extent the result of looking with 'different eyes' is similar to the effect of changing the light source.

A trout's colour vision works in very much the same way as ours, with a small but significant difference. The red colour filter of the trout' eyes covers a slightly different and wider range than ours. So we can be pretty sure that when viewing a contrived colour match between two entirely different media, say the body of an insect compared with a mixture of coloured fibres, the trout will disagree with us as to the accuracy of the match. Most likely green is still green, red is still red, but probably not with the same differences of shade.

And after all, if the trout were so very picky about the precise shade of the Iron Blue, for example, they would be hard put to it to get a decent meal.


The importance of colour in fishing flies and lures is a subject which gives rise to a great diversity of opinion. In spite of living in a 'high tech' age there is still no widely accepted analysis of the situation and understanding is scattered due to being based on partial or faulty information; and on individual experience which may or may not have been objectively assessed. I would like to try to clarify the situation by examining scientific evidence relating to colour vision in trout and thereby equip ourselves with a more realistic basis for understanding the probable importance of colour in the trout's world.



To begin with we must understand the nature of colour vision in humans because this is the yardstick by which we measure the rest of the world. Colour is not a property of light. It is a physiological effect - a sensation occurring in the brain in response to neural signals arriving from the retina of the eye. Light falling on the retina causes a reaction in two types of nerve receptor, "rods" and "cones". The former register the presence of light but not in colour, like a monochrome photograph. It is the cones which are responsible for our seeing colour. But it is not correct to imagine that they respond to coloured light since there is no such thing. There is no colour outside our heads. (Note that in the following passages where reference is made, for example, to "red light" or "red wavelength", this is loose wording in order to avoid long phraseology)

There are three sub-types of cone receptor in the retina of the human eye. Each responds to a different narrow band of light wavelengths and each sends its own characteristic signal to the brain. The brain replies by causing us to experience the sensations of red, green and blue. Combinations of two signals give the remaining colours of our interpretation of the spectrum. The fact that we see the spectrum or rainbow as a graduated series of colour is because the wavelength ranges of the three cone types overlap so that a single wavelength causes response in at least two of the types, the ratio of the signal strengths determining the shade of the colour. However, no single light wavelength can cause a strong signal in the "red" and "blue" receptors simultaneously because they do not overlap to any extent, so that if we see a combination of short and long wavelengths, giving a strong response in both those receptors, we see colours which do not exist in the rainbow, that is the shades of purple and magenta. It is this phenomenon which allows us to bridge the gap between red and blue and create the colour circle, which is an analogy of our colour experience and is at variance with the linear form of the light spectrum.

It follows, of course, that neither is colour a property of surfaces that we see as coloured. Feathers, flowers, pigments, dyes all contain chemical compounds which have the property of absorbing most wavelengths of visible light and reflecting a narrow band. This selective reflection is perceived by us as colour via the mechanism explained above.



In asking whether a trout or any other creature can see colour we must not confuse ourselves. Since we have already established that colour does not exist in the material world, the real question is whether a trout has a neural mechanism similar to ours which causes selected bands of light wavelength to be perceived as having distinct character rather than simply assessing the total light energy.

It has been deduced that vision must have developed in early life forms as the ability to detect movement as a result of change from light to dark and vice versa. The rod receptors of our own eyes without the aid of a focussing device would do this. Research has indicated that the first evolutionary receptor was a cone with sensitivity to the longer wavelengths of the Sun's radiation, i.e. red. This does not mean that those creatures saw the world in red since a single cone response can only show differences between light and dark without reference to colour. Distribution of a number of such light sensitive spots would reveal direction of movement. Evolutionary modification of cells to form a focussing lens made it possible for later life forms to pin-point the precise position of all the light/dark differences around them - that is, to see objects. Two such optical devices makes it possible to accurately judge distance.

The usefulness of this mechanism is, of course, to detect both prey and danger; but where the vision is monochrome, camouflage is easy. Two materials of entirely different constitution - say the shell of a crab and the pebbles surrounding it - might to our eye be different colours but if they both reflect the same total intensity of light as assessed by the eye which has only one receptor, then the two will be indistinguishable.

Evolution overcame this limitation by introducing another cone, one which was sensitive to the shorter wavelengths of daylight, i.e. blue-violet. With two characteristically different sets of neural data arriving at the brain, the brain had to be able to superimpose the two images and assess the quality of the difference between them which is additional to the light/dark differential. This quality is that which we refer to as "hue", and its combination with light and dark results in colour. The combined siganl of two cone types does not give rise to the experience of the colour circle. It gives a linear colour graduation with no means of bringing its "ends" together to form a circle. About 2% of the human population is deficient in one cone type and has this kind of colour vision which also occurs normally in some mammals - dogs and cats, for example. The 3-cone system, referred to as trichromatic vision, is common to primates. So the relevant question is - do salmonids have cones in their retinas and if so how many types?



The three types of cone which occur in the human eye are defined by the chemical compound they carry, called a "diopsin", which is a photo-sensitive pigment. Each of these has the capacity to absorb a range of light wavelengths and is characterised by the wavelength at which maximum absorption occurs. The three are, each for the colour sensation it creates:
erythrolabe; peak absorption at 565nm; red
chlorolabe; peak absorption at 535nm; green
cyanolabe; peak absorption at 440nm; blue
Light wavelength is measured in nanometres and the visible spectrum ranges from 700nm (red) to 400nm (blue-violet).

Analysis of the visual receptors of Salmo trutta has shown that there are four peaks of cone response -
600nm, 535nm, 440nm, 355nm.
The second and third conform to the green and blue cones in humans. The first is analogous to the human red, but its sensitivity extends to longer wavelengths than its human counterpart. The fourth is outside the band of wavelengths visible to primates and is active in the region we refer to as "ultra-violet". However, it is found that this last class of cones disappears by the time a trout is two years old.

The significance of the fourth cone is difficult to deduce. If it interacted with the other three as they themselves interact then the richness of the trout's experience of colour would be as impossible for us to imagine as is the concept of shapes in a fourth dimension of space. It has been suggested that the small fauna on which the immature trout feeds reflect u.v. radiation and are therefore more visible in that range. Sensitivity to u.v. exists also in other fauna: it is of great benefit to some insects, for example, because many flowers reflect strongly in the u.v. range. It is also suggested that u.v. cones re-grow annually in mature trout as spawning time approaches and are used to track polarised light as a means of navigation. This would appear to be more significant for salmon and sea trout. In any case it seems unlikely that a mature trout in which these cones do not persist should use them for prey detection. It is reasonable to suppose in the absence of positive evidence that the neurological response of the u.v. cones is treated separately from that of the trichromatic system and does not assist in prey detection. It should be mentioned at this point that some people have made the statement that trout respond to fluorescent colours because they see in the u.v. range. This is an erroneous assumption and will be more fully explained later.

We should not overlook the shift of the long wave (red) cone response of the trout as compared with ours. It has its peak response at a point where the response of our "red" receptor is tailing off. This means that where we see a red colour, which though a rich hue we regard as a darkish colour compared with orange or yellow, the trout sees a much brighter colour, visible in lower light conditions than we can see red. Research has indicated that the trout's ability to distinguish between small differences in shade is highest in blue, second but much lower in red and lowest in green.

An important feature of the trout's visual mechanism is that the rods and cones physically swap places at the start and end of daylight. In the evening the cones which give rise to colour response but need high light levels to operate are withdrawn into the surface of the retina and the rods obtrude. At dawn the reverse action occurs.

So it is clear that trout possess the mechanism for full colour vision as we know it and with a somewhat wider range. How the trout brain assesses the combined cone response and the resulting subjective experience of the trout is perhaps impossible to prove; but then, any sceptic might claim that we can not prove common subjective experience between any two human subjects. In fact the subjective experience of colour is of no importance. It should be realised that when we talk about possession of colour vision we are talking about possession of a mechanism which enables qualitative assessment of received light in addition to quantitative; and it is well established that trout do in fact possess such a mechanism. If you opened a trout's head and found it full of a coil spring and little meshing cog wheels, you would not deny that it was driven by clockwork simply because you could not "feel" the wheels running for yourself.



Rationalisation of the view that trout possess full colour vision does not provide sufficient information to allow us to judge how it actually affects their mode of life. They might well have colour vision similar to our own but there are major differences in their environment most particularly with regard to the available light. The usefulness of their mode of vision must be limited by the quality of light which enters the underwater world, i.e. the full potential of their 4-cone system can be effective only if the full spectrum of sunlight from infra-red to ultraviolet (so named because they are invisible to us - but not to the trout) is available to them. On the other hand it must be borne in mind that natural selection does not favour useless mutations.

Colour vision underwater is more complicated than in air. A given colour can be perceived only if light of the appropriate wavelength is present, or to put it another way a surface which reflects a given narrow waveband must appear black if wavelengths in that band are not present, and the situation is that all visible light is absorbed and turned to heat as it passes through water, the distance required for full absorption of any waveband depending on the condition of the water at the time. 

In perfectly clear water short wavelengths - blue to ultraviolet - are scattered, causing the whole background to appear pale blue. This is indeed what happens in the atmosphere giving rise to a blue sky, but it is much enhanced in water. This occurs because the short wavelengths are small enough to find that water molecules present an obstacle to their passage and so any blue light which is reflected by, say, the shiny scales of a fish, arrive at the eye of an observer by a circuitous route from all directions. At short distance the image of the fish is blurred and at a greater distance invisible. In water which is not pure the scattering is exaggerated by suspended particles and animal and plant organisms. 

You might therefore expect directional sunlight passing through water to tend toward red (as it does in atmosphere at sunrise and sunset) and becoming more red with increased distance. However this is opposed by another phenomenon, which is the absorption by water of long light wavelengths. That is to say, the energy of the longer wavelengths - red - is absorbed and converted to heat. In clear water the red end of the spectrum is completely absorbed in about 12 feet. At about 60 feet only the blue and ultraviolet wavelengths remain, even though they are scattered in all directions. Note that these distances apply not only to depth but also to horizontal viewing. So this is why, when sunlight is reflected back from white sand, clear sea water looks green in the shallows and bluer with increased depth.

So much for clear water. However, the relative penetration of different wavelengths can be completely altered by impurities which might consist , for example, of mineral particles, algae and/or peat staining. In the first instance vision can be completely obscured in a few inches. A given strain of alga is likely to filter out selected wavelengths. Peat staining, when the water looks like weak tea, is interesting in that it offers no obvious particle suspension but acts as a colour filter, removing the ultraviolet and blue in a short distance and allowing long wavelengths to penetrate the farthest - but no farther than the clear water limit of (about) 12 feet. 

It is therefore apparent that the colour vision mechanism can be fully implemented only in relatively clear, shallow water and at relatively short distances, though this does not take into account the significance of luminescent organisms at greater depths.



The ultimate aim here is to ascertain the extent to which our perception of colour is potentially significant to the trout in its environment. How the trout's behaviourism is governed by such potential is outside the current scope and is more a question to be answered by anglers' experience than by science. Nevertheless it is worth considering that the reason for trout preferentially taking a particular colour one day and a different one on another occasion might be due to water conditions making that colour more visible rather than any fickleness on the part of the trout.

In rivers and most of the time in still water, we are fishing near the surface in clear water. At close quarters there can be no doubt that a trout can see full detail of a lure, including colour. At longer range absorption of the long wavelengths and blurring of the image due to scattering become significant and it at this point it would be useful to examine more closely the limitations of viewing horizontally, close to the surface.

Consider a red object in the water. At a depth of more than 12 feet no long wavelengths remain so it can reflect no light. Therefore even at close quarters it appears black. Near the surface the full sun spectrum reaches it and it reflects the red portion. This is visible at close range and as already deduced is probably a brighter colour to a trout than it is to us but as distance increases the reflected light is absorbed and at some distance approaching 12 feet the object is no longer visible - only the blue scattered light remains. If the object were white, i.e. had the property of reflecting all incident wavelengths, then it would remain visible at longer distances.

It is important to realise that the attenuation of light over a distance is on a logarithmic scale, i.e. is not directly proportional to the distance. Therefore if we observe that an image has finally disappeared at 12 feet, then at 6 feet there might be as little as 10% of the light energy left - not 50%.

It appears then that the full colour spectrum will be significant at close range near the surface. So what is likely to grab the fish's attention at greater distance? Of course, notwithstanding the hypothetical distances given for light absorption, the stronger the light is the longer it will persist. The flash of specular reflection from a shiny surface such as metallic tinsel or the scales of a fish will carry a much greater distance than body class="normal" colour.

It is in this respect also that fluorescent colours are important, because they increase the intensity of reflected light - usually of a particular colour. The increased visibility that we observe in our own environment is also valid underwater given that the same modes of attenuation apply as for "natural" light. A point needs to be explained here. It has been stated by other authors that fluorescence is important because "trout see in the ultraviolet". This is a misunderstanding. Fluorescence occurs where a surface has the property of absorbing ultraviolet radiation and converting its energy to be reflected as a lower wavelength which is within our visible range. This converted reflection is added to the reflection of normally visible light wavelengths and thus appears more intense than "should" be possible. The point about the trout's possession of a receptor which is sensitive to ultraviolet is that it can see unconverted ultraviolet reflection, which has nothing to do with fluorescent surfaces. Obviously if it were necessary to have ultraviolet sensitivity in order to appreciate fluorescence, then we would not be able to appreciate it.

It is interesting to note the experience of divers with regard to colour visibility in various water conditions:
In turbid water: fluorescent red, orange, yellow are best.
In clear water: any fluorescent paint
At long distance fluorescent green and yellow-green are more important (though this probably applies to deep water)

An important principle is involved here - since uv penetrates deeply (deeper than the visible blue wavelengths) all fluorescent colours are visible to the uv limit, way beyond the depth at which their "natural" couterparts have become invisible.

But the situation is very different in peat-stained water. The red-brown colour of the dissolved tannins is a result of their efficiency in filtering out the short wavelengths which causes a complete reversal of the clear-water filtering sequence. Ultra-violet and blue are filtered out first, possibly in the first few inches, whereas the distance that red penetrates is not greatly altered. Beyond its limit, however, no light remains. Fluorescence is therefore useless in peaty water at any depth but it should be borne in mind that where a lure is in use near the surface such that it does receive uv rays then red/orange fluorescence will be visible at greater distance than shorter wavelength colours.

It is a matter of common observation that a trout closely examines a relatively static or slow moving lure, for example an emerger, before taking or rejecting it. Add to this the (elsewhere) suggested probability that trout are sensitive to contrast and it can be seen that a strong feature of the fly dressing such as ribbing might be highly significant. At depth a fluorescent or shiny rib might have a marked effect. 

Insects sometimes carry a bubble and it is to be expected that this would have high visual impact. Of course, its visibility is due not to colour but to a difference in optical density between water and gas. Such a difference can be provided by transparent pearly mylar ribbon.

A dry fly is seen by the trout mainly as a silhouette broken by its "footprint" in the surface tension. It seems reasonable to suppose that colour has less importance than in the case of submerged lures but common experience suggests that it does have some importance. It is worth considering, then, that translucent colour is much more likely to be visible from below than an opaque mass of colour. For example, tightly wound red floss forming the body class="normal" of a Royal Coachman might be less effective than a winding of dyed red goose feather herl or dubbed wool or seal's fur which has been fluffed out. Particularly in this context, it is difficult to reconcile the trout's reported lack of acuity where shades of green (olive) are concerned with the lengths that people sometimes go to achieve accurate colour matching. It seems probable that a pattern of contrasts (body class="normal"/thorax?) is more important.

Furthermore, accurate colour matching between different materials (insect/dubbing) is likely to be utterly impossible for the following reason:
All natural colouring matter has a complex spectrographic character. A particular blue pigment, for example, reflects not only that narrow band of "blue wavelengths" which mark its apparent position in the spectrum, but actually reflects a mixture of separate wave bands which our eye (and the trout's) can resolve only as if it were a single wave band. (Very much unlike our ear which with training can identify every note - or wavelength - in a chord.) Different blue pigments are likely to have very different wave band mixtures. . The result of this is a phenomenon known as "metamerism" whereby two surfaces which are coloured one with, say, a reddish blue pigment and the other with a greenish blue toned with a touch of red might look alike in one condition of lighting but will appear different when the light changes as it does when filtered through water. They will also appear similar only to people with exactly the same balance of 3-cone response. Since the trout has response which is quite different to ours at least in the red zone, it can be more or less guaranteed that any attempt at accurate colour matching will fail.

Finally, with reference to the aforementioned rod/cone changeover at dawn and dusk:
the change occurs over several hours so there is no sudden change of response. Nevertheless, at night a trout has no colour response whatever and in the evening it is fading with the light. Contrast therefore becomes much more important so that black and white is likely to be the most effective combination. Silver tinsel could have value if conditions are such that moonlight is significant.

The author hopes that this description of the objective aspects of colour vision, together with the last section which examines some of the ramifications, will assist in realistic deduction as to what to offer according to the conditions of the day