Here's Looking at You

Following the ‘Windows on the World’, ‘Deep Thinking’ and ‘Fast Food’ posts I was asked “what do salmon actually see?”  The short answer is that I don’t know, but nor does anyone else, because salmon do not respond well to questions.  However, we can draw some inferences from the construction of the salmon’s eye and the effects of water on light, which together may be relevant to our choice of fly.

Clever Blue Eyes

Let down by my cones!
Deceived by a small fly in crystal clear water
on the only sunny day in August 2012

The salmon's vision is an example of smart evolution that contains some interesting features. 

• To correct the spherical aberrations that would otherwise arise from its 160+ degree hemispherical field of view from each eye, the fluid in salmon's lens has differential refractive indices. It does not have to move the head to get a better view of objects in the periphery.  Your fly is in focus from the moment it first appears within visible range.
• The eye and vision automatically stabilises to compensate for movement of the tail, which maintains a consistent view of its prey and your fly during manoeuvre.
• The retina contains a mixture of 'cones' (acuity & precision in good light - helpful for parr eating nymphs) and 'rods' (sensitivity & low light performance based on contrast - essential for hunting prey in dark oceans and running at night).  The balance between them is adjusted by pigmentation to suit the circumstances.  This change is slow compared to the human eye because the the salmon's eye has no iris to control the amount of light entering the eye.  It therefore finds abrupt increases in light levels disorientating and disturbing (see 'Morning Glory').
• The salmon's capability extends across the visible spectrum and possibly beyond. This may include some sensitivity to UV  and near-IR light (even sticklebacks have proven UV detection related to their mating colouration).  
• Salmon have the optical capacity to detect red, green and blue light, but their capability for integrating this into full colour vision is unknown (a computer printer uses only 3 colours of ink to generate full colour images in your brain).  Just because it's red to our eye doesn't mean it's red to the salmon's, and what looks good to us may not to her.  Nevertheless, fly colour may be significant when brighter water and light conditions permit differentiation.

Two observations flow from those characteristics.

• In bright and clear water conditions when the 'cones' are dominant, a small dull-coloured fly will be less easily identified as 'false' than a large one owing to its closer approximation of the prey nymph's appearance.
• In dull light and darker water when the 'rods' are dominant, contrast and movement rather than colour will give the strongest visual cues.

Water and Light

The phenomena of refraction and reflection determine how much of the ambient light enters the water, and the ways in which the fish sees its environment as I described in ‘Windows on the World’.  Once the light has entered the water its subsequent behaviour is influenced by its wavelength and the amount of things like salts, vegetable and other biological matter, mud and debris suspended in the water.  A fair amount of the red-orange portion of the spectrum is absorbed in the first 2.5m/8’ travelled.  That’s not the same as depth, because at mid-afternoon in September the light is entering the water at an angle of 30 degrees, so 2.5 metres travelled equates to a depth of only 1.2m/4’.  Suspended peaty matter increases the red absorption and scatter, which explains why the upper levels of Scottish rivers have that classic burgundy colouration.  Similarly, much of the ultra-violet component and some of the blue and yellow wavelengths are absorbed quite early.

This means that the appearance of your fly will change a great deal depending on the time of day (the elevation of the sun), the depth at which it is presented and its distance from the fish.  Consequently, the apparent colour of your fly may be very different.  The diagram shows a fly 2m down and the same distance from a fish.  In peaty water, only 40% of the core red-orange arrives at that depth to reflect from your fly, and by the time it reaches the salmon, that figure falls to 24%.  That’s a rather dim image, even before you add effects of clouds, rain and the low elevation of the sun at Tomatin’s latitude of 57 degrees North.

Using a picture of nice bright Cascade Conehead and the wonders of Adobe Photoshop we can reduce the relevant bits of the red-orange spectrum to replicate the effect.  The scenario is 3 pm at Tomatin in mid-September, with the water at +12” and rising, cloudy with sunny intervals: in other words, perfect fishing conditions.  The photographs have been size-corrected to allow for the salmon's viewing range and tinted to match the water's colour.  The background is based on Window 1 to avoid the complications of reflection in Window 2 or the shadowing in Window 3.



Our view in air, white light

Surface, 1m range

Surface, 2m range

Depth & range 1m

Depth & range 2m
sunny

Depth & range 2m
dull

Although the last 2 images look very faint to the human eye, they recreate the low-contrast conditions for which the salmon's 'rod' vision is optimised - the high-speed pursuit of grey sprats in the grey-green waters of the sub-Arctic. They do, however, suggest that at 2m depth the brightness and colour of your fly's dressing becomes less important than its size, definition, and above all, its movement. You will see that the black materials in the mid-body of the Cascade stand out more strongly than you might expect. This underlines the wisdom of the Tweed ghillies' preference for large black/yellow tubes for deep early season fishing, as evidenced in the pictures below.

Black & Yellow Tube
Black predominant

Yellow & Black Tube
50/50

You can see that the black portion of the tube is much more clearly defined and stands out better in the murk than the ostensibly brighter yellow.

Two metres is very deep fishing on most medium-sized rivers, not least because the salmon could be another metre of more below that level.  That is sinking line or T-14 territory, and there are only small parts of 3 pools on the Tomatin House water with that depth.

There are, however, 2 other conditions in which the evidence of these images may be relevant:

• When presenting a fly around 1m depth in coloured water with poor light conditions between 10am and 3pm, when a mixture of dark body definition and flash might be useful.
• When fishing at any depth in normal water in the rapidly declining underwater light levels that occur after 3pm.  By 4.30 pm your fly fished at 1m depth has the same dull appearance as it would at 2m at 2.30 pm because the underwater light level has almost halved over the intervening two hours.  Two out of the 3 late taking pre-dinner fish in my sample took black flies.

The water-light relationship suggests, if it's deep, dirty or dull, try a bigger darker fly.