March 20, 2023
Editor's note: Flyfisherman.com will periodically be posting articles written and published before the Internet, from the Fly Fisherman magazine print archives. The wit and wisdom from legendary fly-fishing writers like Ernest Schwiebert, Gary LaFontaine, Lefty Kreh, Robert Traver, Gary Borger, Joan & Lee Wulff, Dave Whitlock, Vince Marinaro, Rene Harrop, Doug Swisher & Carl Richards, Nick Lyons, and many more deserve a second life. These articles are reprinted here exactly as published in their day and may contain information, philosophies, or language that reveals a different time and age. This should be used for historical purposes only.
This article originally appeared in the March 1978 issue of Fly Fisherman magazine. Click here for a PDF of the print version of "Reading Water."
A fly-fishing guide is an arbiter, mediating between his client's expectations and the realities of the river on a particular day. He must inspire a certain confidence among the people in his boat. He also has to find the fish.
Finding the fish means reading the water accurately, quickly–and from a rapidly moving boat, without hesitation. Fly-fishing the big rivers of the West from a boat is really a course in the speed-reading of productive water. As soon as I kick the raft away from shore and start down the river, I begin a spiel of instruction: "We're looking for 'edges,' because they're the feeding spots for trout. Deep water against shallow water, fast water against slow water, cold water against warm water; places where a fish can sidle in and out of the food line from a comfortable hold. We're going to keep a fly on those feeding edges."
The people I guide are often experienced fly fishermen. Many of them are truly expert at fishing on foot, but when fishing from a moving boat, they find that the good-looking water passes by quickly and that they've only a moment to make a decision and pop in the fly. If they false-cast too often or miss on the first attempt, the chance for a fish is lost.
The perspective is different from a boat, causing frustration for a fisherman who is used to wading and stalking at his leisure. Many grumble about how it's tougher than it looks, but then the caster starts hitting the right pockets and catching a trout or two as he learns to read the water more quickly from a different vantage point. He not only becomes better at float fishing, he also becomes a better trout fisherman, because he can recognize fresh possibilities in familiar water types.
Let's examine some of those water types from the perspective of a float fisherman. I'll discuss some of their characteristics, how they might be fished, and give examples of what flies are especially appropriate for fishing them. Keep in mind that even if you always wade and fish, lessons learned in reading the water apply everywhere.
There are areas where a river spreads out, becoming an expanse of calm water that's often no more than a few feet deep. Especially during the bright sun of midday, these flats appear barren. If there isn't a shady bank along one of the sides, or an incoming stream–edges with a promise of fish–guides usually row their clients to the next riffle. These flats, however, can hold a surprising number of trout.
On one occasion, as my boat shot out of a rapid onto a calm stretch, another raft caught up from behind. The oarsman in the second boat nodded toward some riseforms in the flat and said, "Those are all whitefish."
After he rowed straight through, I began crossing the boat from one side of the river to the other. studied the bottom until I found a slightly deeper slot, a trench that funneled a slightly heavier current flow all the way down the flat. We pulled about 25 feet away from this natural ditch and cast our flies over it, picking up some whitefish, but taking five trout as well.
In such smooth water, fish often become selective, refusing the bulkier dry-fly types that are commonly used in the West. A double-wing Sidewinder dry fly not only matches a mayfly silhouette, it is more buoyant and visible than a regular No-Hackle. It's also more durable, a real asset for the rough, pop-pop casting of float fishing.
Too many boat fishermen slip into a casting rut, trying to drop the fly within inches of the bank every time. Against a productive shoreline, the fly needs to be close, but not every bank hides trout. The area next to the bank is often only a gradual sloping of the gravel beneath the water; the water nearest the dry shoreline rocks may only be a few inches deep and devoid of cover.
Even if there are no holding pockets close to the bank, there is usually a shelf 3-10 feet out, at which point the shallows drop more rapidly into deeper water. A dry fly floating directly above this edge, or a wet fly twitching into it from the bank side, often teases fish out of a sanctuary.
An effective wet-fly pattern for these situations is the Gimp, originated by Lacey Gee. In slow water, such as near a shallow bank, this little wet fly creates its own pulsing motion and, as it kicks into the main flow, it struggles against the current.
The current of a river hits the land hardest at an outside bend, digging against the earth or rocks. The steady erosion at the bend carves deep, cutting under tree roots and gouging pockets in the steep bank, and creating a shoreline to which the angler should cast within inches.
The Beaverhead River in Montana, where willow tangles line the banks, is no place for a timid caster. "What do you want to do," I remind worried clients, "save flies or catch fish?"
I keep the boat in fairly close to shore and follow the curve of the land, urging them to drop their wet flies or nymphs in between the tree branches. "Let it sink," I tell them when a fly drops in tight. "Just let it get down to the undercut so the fish can see it." This technique of drift-and-twitch loses lots of patterns, but I make promises: "I've never taken a trip down this river when we haven't caught at least one 20-inch fish."
The Girdle Bug is a standard pattern for the Beaverhead River, but it's effective elsewhere. Because of its rubber legs, it is a great choice for a drift-and-twitch presentation against the deep bends of a stream.
At certain places along a stream, farmers and highway builders can't allow the digging action of the current against a deep bank, so to prevent further erosion, large rocks are dropped along the bank. In my part of Montana, the stones used for this rip-rapping are large sharp-cornered chunks of volcanic basalt.
There are some good holding niches for trout among the blocks of stone where the current spins drifting food into the crevices. The trout hang back in the safety and cover of the dark water, ready to take food that drops from the land or passes in the main flow.
Actually, these rock holds are easy to fish from a boat, because there isn't any overhanging cover to interfere when casting toward shore. The caster, having an open, easy target, slaps a fly back into the niches, letting the swirling water give motion to the fly. Even a misplaced toss looks natural as it rattles off the stones. The Muddler is a universal fly because it looks right no matter what it is doing in the water. For the float fisherman, who may have to fish the same fly as a dry, nymph and streamer, it is indispensable.
Meadow rivers are great anywhere the water runs flush against a grass bank. Even if the streambed is shallow, the current often digs out a slot under the overhanging grass, creating a hidden, deep channel beside only inches of water.
For two days on the North Platte River in Wyoming, we had concentrated on the deep grassy banks only, but Doug Rame and I caught most of our dry-fly trout against the left side. The third morning we fished the left side all the way and took fish from both shallow and deep undercuts.
"It's shady on that bank," Doug speculated. "They can probably see our flies floating because the fish are hanging farther out into the current."
If we used nymphs that sunk down to the level of the fish, we drew them out from either side, but with dry flies it was consistently better to fish the shore where the shade extended out a few inches–even that minor "edge" proved critical.
Maybe the Spider, as it stands up on its hackle tips and flutters wildly, resembles a butterfly, as Hewitt claimed. Or maybe it's just attractive to a fish that's been taking terrestrials at random. For whatever reason, a Spider twitched against a grassy bank often draws frantic action; trout may bounce all over the hopping fly.
The first time down a section of river, I stop and check the water temperature at every spring and tributary. In most of these smaller feeders, which usually have less seasonal variation in temperature and flow than the main stream, the water will be warmer both early and late in the year and cooler in the summer. Other brooks, fed exclusively by either cold or hot springs, will remain at a near-constant temperature all year.
This information is valuable when the river isn't at prime temperature (64-70 degrees Fahrenheit, depending on species) for active feeding by trout. The feeder streams will carry fish all day, even when the main stream is dead, so concentrating on fishing these specific locations pays off.
Fish often hang next to the temperature "edge," holding in the tributary's more comfortable current, but moving across the temperature break to forage in the main stream where food is more abundant. A fly that's worked along the lane where the tributary and river currents run side by side will cover both positions at once.
It's frequently necessary to check the insect activity in a tributary, since it may differ from the main river. If trout are feeding selectively on a particular hatch at a creek mouth, it's often worth a change in fly pattern. Wright's Royal, a variation of the Royal Wulff developed by Phil Wright, has a downwing that presents a good silhouette of a terrestrial. It works well as a general pattern.
A rock jutting up through the water poses an obvious target, a spot that the guide points to and says, "Put a cast near that stone!" But a cast "near" the rock isn't exact enough. There will be a chance for only a presentation or two at most, so the fly had better cover the best holding water on the first shot.
First, every rock is different, deflecting the currents in unique directions. For example, it's not worth casting to the front of every obstruction. Not every stone pushes a buffer current back upstream as the water hits it; most of them have a rounded face that lets the flow sweep right up the front. The obstructions that form viable fish-holding buffers in front of them have a blunt or flat face that clearly doubles the current backward.
Usually, as the water splits around the rock, one side holds more fish than the other. There is a V of current that spreads out past the obstruction and the wing of the V that is deflected at the greater angle indicates the better side, the side of the stone along which the fisherman wants to brush his fly.
Fly fishermen often try to pop a fly in the dead wash directly in back of the obstruction, but it's typically a wasted cast. Fish that hold right behind the rock gather their food underwater from the eddies that filter from the bottom of the main flow. These trout seldom see a fly that dallies for only a moment.
The prime spot behind a mid-current rock is the edge between the still water and the main flow that trails for 3-10 feet downstream from the obstruction. The trout hang in the quieter current, easily picking up drift-items that spin off from the eddying tongues of the faster water.
Trout feed as selectively below the surface as they do on top, which is why the float fisherman needs other flies in addition to the usual Western attractor patterns. It's frustrating to cast over obviously working fish without results, especially when the boat is irretrievably slipping past the action.
Caddisflies are predominant on many Western rivers. Trout are apt to pay particular attention to the pupal stage during the early morning and late evening, and an exactly matching pattern catches many more trout than a general nymph.
The river hisses and spits where the channel pinches it tightly. This is heavy water where few people fish. Wading anglers can't stand up in such current; boats usually either hug a safe bank or shoot right through the flume. But in the places where boulders break the white spray, a guide can work his fishermen over big trout that seldom see a fly.
I let the back of the raft pass close to a large rock, and then row hard upstream to pull the boat into the quiet water behind the obstruction. In this way, I can "walk" the raft from stone to stone slowly down the swiftest run.
J. Marshall Edmonds, fishing a streamer through a slack patch in this kind of water on the Clarks Fork, took the largest trout ever caught by one of my clients, a brown trout of more than 12 pounds, which he played out after we beached in a quiet flat below.
A Marabou Muddler is prime for this fishing. Slap it, twitch it, drag it across the current so that it passes in front of a trout's nose. Repeat this at least twice for every patch of calm water in the run. The sinuous Marabou is often attractive enough to pull a fish out of a resting hold.
Below a heavy run the force of the current often breaks out into a natural basin. The flow dissipates over this large hole, and drifting nymphs and other food items are dropped as the velocity of the water decreases. The pool under this current forms a perfect holding spot for trophy-size trout.
I use a specific technique for these basins, a method called a "Matuka rip." This technique moves the fly very fast into the center area of the pool, actually exciting a "kill instinct" in the trout by simulating a minnow fleeing from the shallows. No other trick so consistently triggers big fish into an attack through 3-8 feet of deep water.
As soon as the raft slides through the white-water chute, I pull to one side of the main current. The client then throws a long line straight across the pool, and the faster band of flow in the center forms a drag belly in the line. The angler actually increases this belly by mending line downstream, starting the fly swimming in a wild and rapid arc toward the middle of the basin. This swing ends in the faster current, and the fly dies with a feeble twitch directly over the trout.
The method takes its name from the Matuka streamer, a New Zealand type popularized in this country by Doug Swisher and Carl Richards. The wing on the Matuka is lashed to the body, which prevents it from twisting under the hook during the fast retrieve. This style of streamer, either in a regular or Muddler type, seems to be becoming increasingly common on Western trout rivers.
Where the current is funneled into a dominant lane along a riffle, a foam line is often created. It presents an easy target, but casters in a boat usually fail to cover it properly. It's different from wading, in which case a fly fisherman can put the fly down repeatedly, because when fishing from a boat, the water passed is gone. It's important to get the fly out at an angle in front of the raft, with enough slack flopped in the line and leader to keep the pattern drifting naturally.
Sometimes it's very difficult to convince clients not to throw the "perfect" cast. A precise, tight loop that unfurls to stretch the line out straight on the water is the most useless cast for fishing the fly.
A dry fly or nymph should most often drift freely, but with a straight-line cast, drag takes over immediately. The client then typically false-casts a few times and lays down another straight line that only floats a few feet. Too much time is wasted with the fly in the air.
Charles Rowland was the best dry-fly fisherman I ever guided. As we covered the lower Missouri, he tossed a snaky length of line, but he put the fly accurately on the prime lane every time. For drifts of 200 yards or more, he never picked his pattern off the water, mending his line instead to prevent drag.
Nevin Stephenson, who has lived in the Butte, Montana, area all his life, is one of the top guides on the Big Hole River. He has studied the giant salmon fly (stonefly) extensively, both there and on other Western rivers, and has created a dry-fly pattern that simulates the frantic, flopping struggle of a giant stonefly on the water.
During this year's salmon fly hatch, his Fluttering Stone out-fished the other common stonefly imitations on the Big Hole, not only taking many trout, but usually the biggest. It should also be an effective pattern type when tied to match smaller stonefly species.
End of an Island
A few feet downstream from the end of an island, two currents collide; their force as they join dissipates into a surface slick. There is a pocket scoured out at the point of the V where the two currents mix, and the place to put the fly is just in front of the juncture. The prime spot is no more than one-foot square, a patch of calm water that collects debris from both sides.
For a distance of up to 25 feet downstream, depending on the amount of water, there is also a cozy buffer formed between the parallel flows. Fish that hold in this twin run are opportunistic feeders, snatching at any morsels that tarry in the slick.
On this rougher water, the fly fisherman needs a compromise fly: one that floats well, is generally imitative and can be easily seen. Frank Johnson's Water Walker fills the bill.
A shelving riffle is the result of a split current. The river never divides enough to form an island, but there is still a gravel-bottomed shallow between two lanes of current. When the stream-bottom configuration pushes the higher band of flow over into the lower one, a stepdown is created as the water bubbles through the rocks. The numerous current tongues, splashing and mixing as they flow, hold a lot of trout, but this is one spot that can't be fished properly from a moving raft. Rather than let the boat run over the good water, potentially spooking the fish, I beach upstream from the riffle and let the clients wade into position.
On an especially productive structure such as a shelving riffle, it's often best to stop and fish anyway. Not only do the clients get the chance to cover the river more thoroughly, it's also an opportunity for the guide to teach casting, collect insects and study the water.
Because of the extreme aeration produced by the mixing currents, there are usually concentrations of stonefly nymphs in shelving riffles. A dead-drift presentation straight up into the head of the riffle is the best way to explore the varying channels in the flow. One pattern that's been effective in this instance is the Black NaturalDrift Stone.
No spot on a river looks more promising than an eddy where the water curls foam and debris back against the direction of the main current. These flow-traps against a bank provide a fish with everything–easier current, cover and drifting food–and, especially during the high water of early summer, they're apt to hide lunker trout.
But eddies never produced well for my clients, even though I urged people to put a cast into them. "Something has to come ripping up through that foam," I kept saying.
Last summer I worked a trip down the Smith River for Streamside Anglers of Missoula, Montana. I told the Streamside guides–Frank Johnson, Fred Tedesco and Grove Hull–"I just don't catch trout out of eddies." They all knew the secret. "Wait until the boat drifts past the place," Fred explained, "and then cast back up into it."
"Of course!" I exclaimed.
This technique solved the problem of realistic presentation. Instead of a cast straight into the eddy, in which case the fly was ripped out as soon as drag grabbed the line, a cast back upstream allowed the fly to be pulled back against the reverse flow of the eddy. The fly struggling naturally against eddying current drew strikes.
Ever since Andre Puyans raved about the "killing" power of the Goddard Caddis, I've used it on Montana streams and lakes. It works best with an erratic retrieve, during which it forms a wake on the surface like a skittering adult caddis.
Art Aylesworth and Arnie Armstrong split the rowing time when they took me on my first float trip. But before we slid the boat into the Clarks Fork above Milltown Dam, I stood on the bank and popped a few casts on the riffle. When I quickly hooked a rainbow, it seemed a good omen for the day of fishing. It was, but not for me–they both tallied some nice trout during the morning. In spite of the instructions and advice they gave me, I never got a strike out of all the beautiful water we passed.
I wasn't at all sure then that I liked floating, just off my dismal angling performance. I was a little embarrassed and frustrated, but I just couldn't plan my casting as I could when I was wading. It wasn't until the afternoon drift that I began catching a few trout.
I marvel now at the patience Art and Arnie displayed during that float, giving me the best seat in the boat and first chance at the prime spots. It must have been exasperating for them, watching me botch water from which they could have taken trout.
During that day and others over the years, through the explanations and directions of experts like Art and Arnie, I learned more about the intricacies and demands of floating. One of the greatest dividends paid by that learning process has been an increased ability to pinpoint productive water quickly–with or without a boat.
Gary LaFontaine has spent the past several summers working as a guide on most of Montana's blue-ribbon trout waters, and is the author of the book, Challenge of the Trout.