Book Excerpt: Understanding Vertical and Horizontal Stratification in a River

Here’s a fascinating excerpt from Tom Rosenbauer’s latest book, The Orvis Guide to Finding Trout, which came out in June. It’s an excellent resource that’s all about how to find trout in all types of water, in all seasons. Tom also discusses what methods and flies may work best in each water type.

You probably know that the current near the bottom of a river is slower than at the surface, which is why trout hold there. When water comes in contact with a solid object, molecules of water that touch the solid object stick to it, and the effective velocity is zero. If the bottom of a river was completely smooth, above a certain velocity turbulence would still form because the viscosity of the water forms an internal resistance and this spreads to adjacent layers, producing mixing and turbulence. But the bottom of a river is not a smooth pipe, so irregularities on the bottom form turbulence at any speed as water bouncing off an object raised above the bottom bumps into water flowing above it, and cells of turbulence form that carry up through the water column, although as it extends farther from the bottom, energy is dissipated so turbulence is lower. The faster the water, the more extensive the turbulence, with bigger cells and a more widespread effect.

In the more turbulent layer, energy is lost through friction with the bottom and downstream velocity is lowered, because instead of moving in a steady downstream direction, some of the energy is direction upward and sideways and backward. This loss of energy creates heat, especially from friction along the bottom, which causes the cells of turbulence to slowly migrate upward, as warm water is less dense than cold water. The heat produced is negligible because organisms living in it because of the thermal resistance of water, any temperature increase is insignificant. You don’t need to worry about fishing water with heavy turbulence in warm weather because it will never warm the water enough to even be measured with your stream thermometer so it won’t hurt the trout—and anyway turbulent water offers more gas exchange with the air, so oxygen content will increase in areas of heavy turbulence.

This is an approximation of vertical stratification in a trout stream. Trout rest very close to the bottom where velocity is at a minimum, and tip up into the current above to feed. They prefer to capture food that is below the boundary layer, because above that layer velocity increases quickly and they have to expend too much energy swimming above it because it’s much harder for them to regain their initial position.

The difference in velocity in a stream channel, from bottom to top, is not a straight-line relationship. It’s more of a logarithmic relationship; in other words, the velocity increases slowly in the boundary layer as you rise from the bottom to a certain point and then it rises quickly to a maximum. The rougher the bottom, the higher into the water column this slower region extends. At the surface is a little variation from a true curve because friction between the surface of the water and the atmosphere slows the velocity a small amount. This slightly slower velocity extends up to 20 percent of the depth, so in 4 feet of water it extends down to almost 10 inches below the surface (although, being a curve, it’s most pronounced right at the surface). When fishing dry flies and nymphs, I’ve always favored a leader that floats (I grease it up with a little dry-fly paste), as I somehow feel this gives me a better presentation, and I think that a sinking leader or tippet extends into the faster water just below the surface and pulls on my dry fly or nymph at a speed faster than the current at the surface, producing drag.

This difference also has important implications in fishing subsurface flies. In fast water, trout hug the bottom where the current is 2 feet per second or less, where fish can maintain their position without expending too much energy. If the current in the middle of the water column is much faster, somewhere around 5 or 6 feet per second, trout will tip up into the layer of slower water where the velocity curve is not as steep, but will avoid going above the inflection point where the curve gets much steeper. This is why, particularly in fast water, you need to get your flies close to the bottom, in that calmer layer, because even if trout see your nymphs or streamers in the middle of the water column, they’ll be reluctant to brave the flow. Not only do they use more energy getting to the faster layer, but once they get there, the current pushes them backward, forcing them to swim upstream to regain their position.

The composition of the bottom, assuming the same current velocity, will determine the size of the boundary layer and how many trout it may support. On a sand or other smooth bottom like bedrock, the only place trout will hold is where you find a depression in the bottom. Rough gravel gives a slightly larger boundary layer and more places for trout to live. A cobbled or bouldered bottom offers the most places for trout to live and an easier place to feed, so it should hold more trout per square foot.

The size of this slower layer is related to the size of objects on the bottom. And unless the layer is bigger than the height of a trout’s body, it’s difficult for a trout to find protection from the current. In water with an abundance of large rocks, the calm layer extends farther into the flow, but over a smooth bedrock or sand bottom it is narrower. Over a very smooth bottom like a layer of sand or fine gravel, you may not find any trout in the main current unless there is a slight divot on the bottom, at least as long as a trout. Fish lying in these depressions on a smooth bottom are reluctant to move very far into the water column. This makes fast but smooth-bottomed waters more difficult to fish with subsurface flies.

Vermont’s Battenkill is one example of this conundrum. The surface velocity in the Battenkill is deceptively swift, and because the surrounding land is composed mostly of glacial till, smaller gravel, there are few large rocks on the river bottom. Getting a nymph down to these fish is a challenge, and over the years I have found that to successfully fish nymphs, I need to find heads of pools, where the fish lie in the pocket just below a riffle, where the water deepens and slows quickly, because it’s much easier to get my fly to them here than in the fast, smooth water in the middle or tail of a pool.

In contrast to the Battenkill, the Madison in Montana is carpeted with rocks and boulders of all sizes, and I’ve found that as long as I stay out of the swift main current in the middle of the river, I don’t need to set my indicator as far up on the leader, or can fish a dry-dropper with a shorter dropper. That calm layer extends much farther up into the water column, sometimes right to the surface, so trout here are more likely to venture farther from the bottom when eating subsurface food and are also more likely to take a dry fly, as it does not consume as much energy as it does in a river like the Battenkill.

The smooth, flat water of much of the Battenkill is deceptively swift, as you can see by the wake behind Tyler Atkins. It’s a tough place to fish a nymph because the fast surface current, unbroken by much turbulence, whisks away the line and leader before the fly has a chance to sink. It also does not hold many trout except for places where a shelf on the bottom or an obstruction like a rock or log slow the current.

It’s important to note that the shear stress that causes turbulence is proportional to the square of the velocity. So the faster the current, the wider the boundary layer, and it doesn’t take much of an increase in current to widen the slower boundary layer below. We worry about trout getting washed away during a severe flood until we realize that the increase in current speed gives the trout a wider boundary layer of slower current to escape the ravages of the torrent.

Water in contact with the banks also forms turbulence and slows the flow of current. Just as a rough bottom of cobbled rocks gives the trout a wider comfortable layer, a rough bank lined with rocks, logs, or dimples and bays along the shore will offer more comfortable places for trout, especially if the current along the bank is fast. Even a shallow, sloping bank produces some turbulence and slower current, from gentle undulations of the bank and the bottom. The comfort zone of calmer water will extend from the bottom of the river into the shallower water along the bank, and trout will slide into these places to feed. They prefer to feed here because when the calmer zone is mostly uniform, all the way to the surface, a trout can feed in the whole water column without the need to dart into any faster, more difficult current. They can slide sideways using little energy, never venturing into faster current.

Water in contact with both the banks and the bottom creates isobars of current. Here is a hypothetical and greatly simplified view of the differing currents in feet per second, in a cross-section of a river. Trout are comfortable holding in current less than 2 feet per second, so you can see the places where they’ll prefer to hold and feed—close to the banks and close to the bottom. Objects in the streambed or along the banks will change these velocities, but this shows what a non-obstructed section of river would exhibit.

One cold, dreary day I was fishing the West Branch of the Delaware and unusual for that river, I could not find a single surface-feeding trout. I must have covered a mile of water looking for noses poking into the surface film when I came to a large pool I had never fished before. The right bank was much deeper and faster, not a high-percentage spot for finding rising trout, but since nothing was happening I decided to wade up that bank anyway, more to explore it than expecting to find any fish. About halfway up the bank was a slight projection, just a little finger poking into the faster current, and I thought I saw the wink of a rise there. As I got closer I saw the fish again, with a slow, broad rise that pushed a wide swath of foam aside as it fed. All I could see on the water were a few size 22 olive mayflies, and as much as I avoid fishing a small fly to large fish, I tied one on. After a few casts the fish took the fly with barely a ripple and I actually landed it, a brown trout of 22 inches measured against a reference point on my rod. It was the only trout I caught that day but the biggest fish of the week.

Tom Rosenbauer’s The Orvis Guide to Finding Trout is available online and in bookstores everywhere.

Leave a Reply

Your email address will not be published. Required fields are marked *