Friday 22 February 2013

Fast Food & Broad Beans

In the last post I looked at how to present the fly at the right depth to prompt a take.  We now turn to the horizontal plane, which comprises first getting the fly to the right place, and making it behave in a way that may stimulate a salmon to take.  However, I am going to take those two elements in reverse order, because the desired behaviour of the fly drives its delivery.

We do not know exactly why a non-eating salmon takes a fly, but let us assume that there is some connection with embedded reflexive responses.  It takes an object into its mouth because something in the nature or movement of the object triggered a feeding reflex.  No one teaches a salmon to eat, but their life is characterised by the unrelenting consumption of protein leading to spectacular growth.

Once the alevin sheds the yolk sac and becomes a fry, it automatically starts to feed on small invertebrates.  A parr eats progressively larger invertebrates, mostly ephemera nymphs and stonefly larvae.  As this blog is about salmon rather than trout, for simplicity I use the abbreviation ‘nymph’ to include all the different species and types of waterborne larvae (and hope that the experts forgive me).  The time spent as a parr and the amount it eats embed high-speed food recognition and instinctive response.  Salmon are short sighted as a consequence of their very wide, indeed near-spherical field of view.  Survival is not cost free.  Parr feeding in flowing water therefore do not have much time to decide: the slow-witted starve.  At this stage of life it needs around 8% of its body weight daily in protein, which is a lot of nymphs - thousands per month.  Consequently the size, shape, tone and movement of nymphs become printed in the forefront of the young salmon’s brain in order to achieve the necessary speed of decision and action.  Those features are:

Size – small, in the range 3-20 mm
Profile – roughly cylindrical with legs sticking out
Tone – dull, but in the final ascent towards hatching, gas bubbles on the body may reflect ambient light brightly.

Movement – very slow, in the range 1-5 cm/sec (< 2”/sec): try timing your finger along a ruler.  Most travel is horizontal and near the bottom, but the ascent to hatch (when their food value is highest) is upward and thus clearly visible in Windows 2 & 3.

Once the smolt reaches the sea it has a cornucopia of high-protein prey species.  Sand eels and sprats congregate in huge numbers.  For their part the smolts form shoals for protection from predators; improved observation and prey detection; and increased hunting efficiency to counter the prey species’ defences.

The sand eel has a tough life at the top of every fish’s menu.  It is bite-sized, easily swallowed and high in protein.  Once the sand eel is deprived of its shallow water defence of burrowing into the sand, its 2 metre dash is no defence against smolts or salmon.  An adult salmon carving at speed through a shoal of sand eels may swallow several in each pass, an exercise conducted at such speed that only reflex can deliver the required coordination of tail and mouth.  Although the sand eel’s range is limited to shallower waters, its importance lies in being the first and most plentiful food that the salmon encounters during its early development in the North Sea and Shetland Banks.

The sprat, which is more widely distributed across the salmon’s range (the Faroese call them brisling), shares top of the menu with the sand eel.  Better still, it’s even easier to catch and has exceptional protein density.  It’s shorter and fatter than the sand eel, lacks its dash speed, and forms very dense shoals.  A mature salmon might be taking over 1lb/0.5Kg of sprats per day to maintain its growth rate.

The key characteristics of the 2 species are:
Size – in the range 3-15 cm.
Profile – oval, with a large length to diameter ratio.
Tone – grey green with silver sheen that increases with ambient light.  In a grey-green ocean in winter light it’s lateral movement that gives them away.
Movement – from slow at cruising speed to fast in survival dash (about 1.5-2m/sec, which is very fast for such a small fish).  Movement tends to be in straight lines and broadly horizontal unless panicked.  They wriggle continuously.

The two listings of characteristics frame the images we probably need to create in the salmon’s brain in order to trigger a response.  From that we can derive how best to fish the 2 generic types of fly:


Blue Charm 14
Note legs & gas bubbles in the dressing
The movement is slow, ideally upwards and interspersed with short pauses.  Therefore:
  • There’s not much logic in casting small flies into fast water where they will tumble and move chaotically like the rest of the debris coming downstream, and be similarly disregarded.
  • Remember that it would take a live nymph 2-3 seconds to traverse the palm of your hand, so slow the fly down to match.
  • The fly can only move upwards if it first went down: use a fluorocarbon tippet and/or an intermediate polyleader, but always keep the fly above the salmon’s sight-line.
  • At all costs stop a belly forming in the line which will otherwise accelerate the fly to sand eel warp velocity.
  • Towards the dangle, slowly retrieve a few inches; pause to allow your fly to sink back; and then repeat.


Sunray Shadow 12 cm
Off-centre dressing increases wiggle
The movement is steady but brisk, horizontal and animated.   Therefore:
  • Obey Newton’s Law: you will only get lateral movement and wiggle by applying tension on the line. 
  • If the leader does not lay out straight at the completion of your cast, take up the slack immediately, otherwise the fly will just hang in the water pointing skywards for much longer than you think.  It’s better to fish a smaller area effectively than waste the first third of the swing.
  • Eliminate line belly: a fly moving downstream achieves nothing; and a fly at warp speed even less.

Of course none of that guarantees success.  But we can help to improve the odds by eliminating things that are handicaps and hindrances to securing the fluke, even if a fish with a brain the size of a broad bean will remain unimpressed for 99.9% of the time.

You may wonder about the Broad Beans in the post's title.  This is a sad little story of a failed experiment and humiliation.  I got myself into position nearly 20 feet above a nice clear pool crammed with plainly visible salmon milling about soaking up the oxygen before launching themselves at a testing ascent of the falls behind me.  They were all awake, alert and active.  I attached an exceptionally realistic imitation plastic sand eel, cast beyond the fish and stripped it back through their midst at the appropriate speed.  There was not even a flicker of interest, despite the eel’s seductive wiggling only inches from their noses.  Ten minutes of repetition achieved nothing: their bean-brains were obviously wholly focused on the task in hand, and any way, the eel was at or below their sight line.  I gave up trying. Twenty feet below me at the water’s edge Patrick rolled a small Stoat into their midst, stripped it back heading downstream and promptly hooked a nice hen fish.  You can generalise about broad beans, but there’s no two that are identical.  So no matter how confident my assertions may be, remember that it only takes the intellect of a broad bean to prove me wrong!

Monday 18 February 2013

Deep Thinking

Carrying on from last week’s ‘Windows on the World’ post, the sketch represents the assumed ‘taking zone’ for an alert active hen salmon in a non-residential lie.  Viewed from above it comprises a fan shape, probably extending out to some 60 degrees left and right.  However, its actual size in 3 dimensions will be determined by the nature of the lie, the degree of shelter that it provides and the strength of the current beyond its immediate confines.   In most cases an energy-conserving hen fish is unlikely to expend much effort in moving to inspect a fly, no matter how attractive, if it’s outside this zone.  Her embedded survival will probably prevail over a passing urge, and that is the basis of my working assumption of a 6’/2 metre radius of action.
Obviously there are no firm rules: lies come in every shape and size from a shallow depression in the bottom to submerged boulder fields, while their viability and nature changes with the water level.  In every case the key to success is to present the fly in a realistic fashion in the taking zone within and around those spaces.
That is not straightforward because we have to achieve it in 3 dimensions, in flowing water.  It is not a simple matter of just advancing downstream one pace per cast in the belief that we will thereby cover the whole volume of water and all the taking zones within it.  That steady advance addresses part of the longitudinal (up & down the river) and lateral (left/right) dimensions, but by no means all, because the swing of our fly does not describe a smooth regular arc across the river.  Differing flow speeds in the breadth of the river affect the movement of the line and can have disproportionate impacts on the fly’s direction and speed of travel.  Unless you do something to correct this problem, your fly will miss some lies completely.  To make matters worse, even if it enters the zone it may not be moving properly and the fish will totally disregard it.  I’ll be looking at these problems and their solutions over the next couple of posts.
However, as its title implies, this week’s post is concerned with the third dimension of depth.  This is fundamental but simple.  If you do not present the fly at the right depth you will miss every taking zone in a pool completely, no matter how well you cast, mend the line or control the horizontal movement of the fly.  Fortunately you can solve the depth problem with a mixture of technology and observation to establish the depth of the water in which the salmon are lying.
If you combine the hen’s 6’/2m range with the 45 degree angle of attack you arrive at an answer of 4’ 6”/1.4m.  This gives you a rough estimate of the maximum depth at which you can reasonably hope to catch a hen salmon with a fly fished near the surface in slow or medium speed water.  The good news is that many good fishing pools on Scottish rivers are not very deep; and give or take the boulders their bottoms are quite level leading to uniform depths in the head and middle sections. 

Of course rivers have pools that are narrow, on bends or astride some geological fault which makes them significantly deeper, but at Tomatin 4 of the best pools are relatively shallow with an average low-water depth around 3’ 6”/1.1m.  This means that you can catch fish with a floating line and simple leader until the river has risen by 10”/25cm.

Tomatin House Pool at +12", perfect for floating line, with most of the water within casting range up to 4' deep.  The lies in this shot are not adjacent to the obvious rock in the middle, but rather 10' short and left at 8 o'clock from it; and directly in front of your feet.

However, at +10" and higher the fish are very awake, alert and vulnerable to the attractions of a fly.  It probably offers your best chance to catch salmon, but you won’t catch many if the fly is above their taking zone.  You have to get deeper, which requires a sink tip or a weighted fly.  I prefer to add weight to the leader before adding weight to the fly because it makes casting and turnover easier.

This simple chart shows the depth (in feet) you can expect to fish an unweighted fly with each grade of sink tip, using the colour codes of the popular Airflo Polyleader range, and the Rio T-14 depth-charge device (which needs a skagit line to cast it).  However, there are other things to bear in mind when making your choice for reaching the necessary depth:

  • The ‘lift’ on the line and tip increases with the water speed, so in faster water add an extra unit of weight – leader or fly.
  • The depth to which the tip sinks is time-dependent.  If the lies you wish to get at are in the early part of the swing; the surface of the water is especially rough; or you are fishing a narrow river or section, go heavier still to get down quickly (but don’t dawdle at the ‘dangle’ or else you will snag the bottom).
  • Be ready to change either the tip or the fly as the depth and flow speed reduce further down the pool.
  • Large doubles sink faster than small.
  • Look at the fluff to metal ratio – fluffy flies generate lift and ride higher in the water.
  • With smaller and lighter tubes the size of hook you use will have an effect on their fishing depth.
  • There’s a bewildering range of tube flies in the tackle shops and in many anglers’ boxes.  To avoid indecision I ration myself to a little box of 9 tubes, subdivided into light (plastic), medium (copper) and heavy (brass).
  • I like coneheads: people say that the extra weight at the front balances the hook at the back, but whatever the physics, they work for me.
  • When fishing near the surface I always use a fluorocarbon leader or tippet because it is heavier than water, nylon and copolymer.  Good reliable fluorocarbon is expensive – don’t be tempted by cheap tippets when you’re spending a lot of money to hook a salmon.  In my experience, Riverge is strong and reliable, while Seaguar, optimised for salt water work, is especially abrasion resistant, which is handy in rocky rivers.
  • Depending on the colour of the water, you don't need much tippet on the end of a polyleader: 3'/1m suits most occasions.

Tomatin Garden Pool at +2' and 6' deep on the far side, flowing medium to fast.  Floating line, green slow sink polyleader,  and 1" Conehead Cascade, with an 11 lbs hen fish firmly on.

The longer 10’/3m sink tips obviously go deeper, but their extra weight loads the rod more heavily and demands more effort to extract from the water at the start of the next cast.  Conversely that extra weight will turn over a heavier tube better than a short tip, provided that the sum of the two is within the capacity of your rod and line combination.  With a big hefty 15’ #10/11 this shouldn’t be a problem, but at 14’ #8/9 you will feel the difference (future posts will look at rod and line choices).
You can also get lines with interchangeable tips: I have one of the original  Loop Multi-Tip lines, which casts well, especially with the medium sink tip.  On the other hand, in full floating configuration its handles a 5’ sinking polyleader with ease.  However, if you are buying a new line today, I wouldn’t recommend an interchangeable tip.  They are a great deal more expensive than the ‘parent’ line plus a selection of polyleaders, so you are paying a premium for an optimised balance that exceeds most people’s casting capabilities.
Finally, there are full sinking lines and their shooting head cousins.  The Upper Findhorn is not a big river or very deep, so in 10 years I don’t recall a situation where a full sinking line would have been the answer to my prayers.  On a big river, especially early in the season, a full sinking shooting head would give you the combination of distance, balance and depth to meet the range and depth challenges better than a floating line – polyleader combination, but those conditions are beyond the remit of this blog.  Anyway, I worry about loading myself with yet more reasons for indecision!
Having addressed the critical issue of depth, in the next post I shall look at the factors that influence the management and presentation of the fly in the horizontal plane.

Tuesday 12 February 2013

Windows on the World

The picture shows the salmon’s view of its world divided into 3 windows.  When viewed from above these form concentric circles, but the boundaries between them are not fixed or sharp.  They vary with the light and river conditions, and the transition between them is progressive.  Throughout we must remember that the salmon's eyesight is broadly monochromatic (but not black and white) and works much better than ours at low light levels.  However, that extra sensitivity, its very wide field of view and the lack of eyelids makes the salmon much less tolerant of excessive brightness.

Window 1 is the murky world of the straight ahead view through green- tinted, burgundy-coloured Scottish water, which is strewn with rocks and cluttered with debris and bubbles.  In September, if the river is high, this window is dull and lacks contrast, which declines with depth.  Maximum sunlight penetration generally occurs between 10 am and 3 pm.  Small objects may not stand out and be visible at short range when moving laterally.  Stimulation probably requires something bigger and brighter (Flying C perhaps?). 

In this world of dull backgrounds and low contrast, the probability of a salmon seeing, reacting to and taking a fast-moving fly is correspondingly reduced.  A brief explanation: the salmon is lying in comparatively slow water to conserve energy, but the speed of your fly is determined by the interaction between the faster current on the surface and your floating line, even when you are using sink tips and tubes.  If you are presenting the fly deep in Window 1 then controlling the lateral speed of the fly is absolutely vital, otherwise it just flashes past her nose.

Window 2 is dominated by the silver reflective underside of the surface.  Unlike the smooth mirror reflecting the gravel bed of a calm trout stream, the turbulence and waves of the fast water of a Scottish river make the surface that the salmon sees jumbled and rapidly changing.  Viewed from below it comprises a kaleidoscope  of randomly shaped tiles, each transmitting light of differing intensity, and some none at all.  This is caused by the roughness of the surface, as every wave-slope reflects or refracts the light differently.  The proportion of light to dark tiles generally increases with the elevation of the sun, so you can expect the brightness and visibility in Window 2 to increase throughout the morning to a midday peak, albeit in September this is only about 45% of the midsummer value.  (Photo by Mark Conlin).

Beyond the brightness and elevation of the sun, the other factors affecting the salmon’s image of a fly in Window 2 are the clarity or turbidity of the water (which is related to the river’s height); the depth at which the fish is lying; and the angle of its view of the fly (the steeper the angle the clearer the image).  A fly in Window 2 is above the salmon’s horizontal sight line and so is viewed against the surface kaleidoscope, which is lighter and brighter than the Window 1 background.  This allows a salmon to detect small objects displaying lateral movement at considerable distances, even in peat coloured water.  If you can catch grey sprats in the dark winter waters of the Barents Sea, spotting a #6 Ally Shrimp in the Findhorn is relatively easy.  Conversely, intense oblique sunlight on a cloudless day may make the Window 2 surface generate bright flashes of light that might degrade the salmon's eye function. 
Watching from on top of a high bank I have seen salmon react to flies in Window 2 at distances up to 6 metres in calm water that provided optimum sub-surface reflection and image quality.  However, whilst cock fish may attack a fly at that distance (see the ‘Morning Glory’ post), the more circumspect and energy saving hens appear to restrict their taking radius to a couple of metres at most.  In all cases the fly’s lateral movement appeared to play a key part in the triggering process.  The issue of fly size hinged on visibility: in clear calm conditions I saw  a #12 Ally Shrimp secure a response by an alert active fish at  5 metres, which underlines the acuity of the salmon’s vision in this zone.  But the simple fact is that we need the fish not only to react but also to take, and for hens that means putting the fly within comfortable reach in both depth and range.

Window 3 is an inverted cone around and above the salmon, described by an angle of about 45 degrees (everything beyond 48 degrees is reflected into Window 2).  For a fish lying at a depth of 4 feet/1.2 metres, the radius of Window 3 is about 5 feet/1.5 metres at the surface.  This is the brightest and clearest window in which the salmon can easily detect all objects and motion, and may therefore be most directly stimulated.  The salmon will see land-based objects in Window 3 if it is either lying close to the bank or there tall trees or other high features nearby, which may complicate its view of a fly.

There are, however, three features of Window 3 that the angler must understand.
  • First, for a salmon looking upwards, the brightness of Window 3 causes relative darkening of the underside of Window 2.  An object moving from 2 to 3 will be immediately visible in the salmon's peripheral zone and potentially stimulating.
  • Second, within the brightness of Window 3 there is a natural limit at which the salmon finds the intensity of the light uncomfortable owing to the wide focal aperture of its eyes and the lack of any shielding mechanism.  The threshold is set by the elevation and intensity of the sun; the roughness and turbidity of the water; and the depth of the lie.  When the light crosses the comfort threshold the salmon has 2 options: to switch off; to move into deeper water; or both.
  • In Window 3 the salmon sees everything in the water above in most part as a dark silhouette against the sky.  The underside of the fly is mostly in shadow and its colour therefore suppressed.  This raises the question – in Window 3 does the fly pattern matter, or is it just a matter of the size and lateral movement of a dark object, and the contrast between its component parts?

My observations indicate that salmon may have particular optical acuity around the boundary layer between Windows 2 and 3, i.e. at an elevation around 35-55 degrees.  This angle broadly equates to the salmon’s preferred angle of attack on prey species in the sea and possibly with its parr feeding experience.  In my next post I shall combine this angle with the hen’s’ 6 feet/2 metre taking radius as the basis for a rule of thumb as to where we need to present the fly to a salmon lying at a given depth in order to optimise stimulation and the prospects of a successful take.

Friday 1 February 2013

Good Morning Ladies

In my last post I discussed the daft behaviour of over-sexed September cock salmon in the early morning.  This week it’s the turn of the more leisurely hens.

The graph shows that hen fish activity is more evenly spread over the day. There is an obvious bulge in the mid to late morning period, with lesser peaks in the afternoon and early evening. Actually, it’s not quite as simple as it appears.

 The expert statisticians would say that the shape of the graph and its peaks have more to do with my pattern of behaviour than the salmon's. I fish frantically from first light until last call for breakfast at 9; eat the hearty full English, shave and re-emerge onto the water at 10; and then fish through to beat changeover and a late lunch. I get back onto the water between 2.30 and 3, and then carry on until the call of bath and gin overcomes my fanaticism. You can see their logic: it could be correct, although it doesn’t explain the voids between 12 and 3 pm, and 4 and 6 pm. However, what blows the sceptics’ argument away is the evidence of the Tomatin House fishing records for this week. The catches of normal people show an even more pronounced mid-morning bulge.

The next thing we have to do is go through the data to discard the abnormalities. We have to be slightly cautious here because arguably all of these catches reflect abnormal salmon behaviour (“it’s all a fluke” – Falkus). However, this sample does contain some very flukey outliers towards the end of the day. For example, the 2 fish I took in near darkness in 2005 in low water, both small and stale (but very welcome as they broke my duck after 2 blank years). The big resident hen fish taken at 6 pm in freezing cold water in 2012 described in my ‘Brass Monkeys’ post was another bit of outrageous luck against the odds. Finally we have to look for exceptional circumstances, which in this sample is the 2011 season (sadly I missed the extraordinary week in the 2004). In 2011 the fishing conditions were good for 5 of the 6 days. There were lots of fish in the river that were running strongly but resting frequently in the face of the heavy water. As a result, and exceptionally, I fished for all but one day of the week with sink tips and tubes, catching 9 fish. (Whereas the great majority of the other fish in the 10 year sample were caught on floating lines, plain leaders and un-weighted flies). I suggest that the exceptional factors of 2011 tended to spread my catch more evenly across the day – for example the 2 lunchtime fish were unique to 2011. However, in 2011 the other rods’ catches still showed a strong mid-morning bias, possibly because most of them continued to fish normal floating tackle. Having taken my odd behaviour, the abnormal data and exceptional circumstances out of the equation, we are left with 2 marked peaks of hen fish activity. The larger one is centred at 11 am and the second, about half the size, at 3.15 pm. Is there any explanation? The regularity of the Tomatin September morning bulge rules out transient phenomena such as river level, barometric pressure, water and air temperature, oxygen levels, cloud cover and wind direction as the acute variable, because none of them vary regularly with the time of day. Obviously the conditions had to be right to catch fish (whatever the unknowns may be that ‘right’ entails – possibly all of the above). But all other factors being favourable and equal, there’s something that is time of day related that causes the 11 o’clock peak, and the only thing that regular is the sun.

If the sun is the factor, how might its light trigger the behaviour? The graph shows that the light intensity is about the same at 11 am and 3 pm in this week. However, I’m not proposing a theory that an intensity of X lumens stimulates a take. Instead I suggest that given otherwise favourable water and weather conditions and a fish that is inclined to take, this light level gives it the most stimulating view of a fly fished on a floating line and leader. Exactly what ‘stimulating’ comprises, I don’t know, but I’ll explore it in my next post.