SHOP BOARD — FLOWS POSTED WHEN THE GAUGES REPORTFREE U.S. SHIPPING OVER $49 · SHOP STATUS: OPEN — CLOSED WHEN THE WATER'S RIGHT
Where trout sit in a riffle
reading-water-trout-behavior

Where trout sit in a riffle

Drew··7 min read

Trout hold on the gravel in dips in the riffle bed — most often on the upstream lip where the bottom starts to fall away, with their noses within a couple of inches of the rocks.

Most of a riffle is too expensive for a decent fish to sit in, so there are only a handful of those spots. Because drifting food builds as it travels, the bottom third of a long riffle out-fishes the head.

You get to the water with an hour of light. The riffle at the top of your home run is knee-deep gravel, sixty yards of it, and it looks the same end to end — so you start at the head, because that's where the water comes in and that's what you were taught. You fish the seams. You catch six-inch fish. Every one of them, all evening, and by the time the light goes you've decided that's just what the riffle holds and next trip you'll walk past it to the pool.

That conclusion costs you the best water in the run. Two things put you there. You spent the hour on the first ten yards of sixty, and your nymph rode six inches off the bottom the whole time — which felt right, because six inches up is where you could see the food going past.

Nothing about that looks like a mistake while you're doing it. Which is why it pays to be exact about where a trout in a riffle actually sits.

Where do trout hold in a riffle

Most of a riffle is water a trout can't afford to sit in. Confine decent-sized young cutthroat to riffle habitat and they don't just grow slowly — the larger juveniles lost weight in riffles while the same age class gained in pools. The current bills more than the food pays back. Small fish clear that bar easily, which is why your riffle is full of six-inchers and the better fish is in the run below it.

The limit is speed, and it's tighter than it looks. At around a foot per second a trout catches most of what it goes for. Take it to two feet per second and capture success drops to about 10% — it can't pick items out in time, and getting back to its station costs more than it can hold. The fastest chute isn't lying to you about the food. It just isn't a place to eat from.

So you're not looking for food. You're looking for the few spots where the bill drops far enough for a decent fish to turn a profit, and those spots are dips in the bed. When holding positions were mapped fish by fish, every trout sat in a depression — and 40% were on the upstream slope, against 27% down at the deepest point. That upstream lip is the spot. The flow has already lifted off the bed as it comes over the crest, the fish is out from under the worst of it, and it's looking uphill at everything the riffle is sending down.

They're also lower than you picture them. 85% of noses within 5 cm of the bed, bellies averaging 2.3 cm off the rocks and most often flat against them. Not hovering mid-column. On the gravel, looking up a couple of inches at what comes past.

One more thing separates a lip a fish uses from one it ignores, and it isn't depth. It's how much the water over it churns — and that's the part most of us have backwards.

What does that look like from the bank?

The thing the drawing looks wrong about, the first time you see it, is which of those tongues is the rough one.

Isn't the fast water the rough water?

Fast looks rough. Stand in a riffle and the loud broken water is what you brace against, while the smooth tongue sliding down the middle looks like the hardest place a fish could possibly sit. It's the other way round. Turbulence goes up as speed comes down. That smooth fast tongue is the most orderly water in the riffle. The ragged slow margins — the churn off the back of a gravel bar, the boil where flow trips over cobble and loses its shape — are the water that shoves a fish off balance and makes it work just to stay upright.

Trout choose on it. In the same mapping of daytime positions, occupied sites had lower turbulence than comparable unoccupied ones, and turbulence intensity ran inversely to current speed. Two spots that look identical from the bank are not identical to the fish, and it takes the calmer one.

So slow doesn't mean restful. It depends whether it's smooth slow or ragged slow, and you can see which from the bank once you're looking. Smooth slow water has a surface that slides. Ragged slow water has a surface that keeps rearranging itself.

This also puts a ceiling on the sit-in-the-calm-and-dart-out plan. Leaving a velocity refuge to grab something raises cost per attack about 65% and cuts the odds of getting it. A fish would rather the food arrived where it already is, which is exactly what the upstream lip of a dip gives it: slow water at nose height, little churn, and a clear view uphill. The old surface rule of thumb, holding water running about 1–2 ft/s, is describing the same place from above.

That's which foot of a riffle a fish can live in. It says nothing about which end of sixty yards deserves your hour, and the food decides that one.

Get letters like this in your inbox

This Week's Path — one skill, one water-finding trick, one first-fish story. Tuesdays.

Should I fish the top of the riffle or the bottom?

A riffle makes food and then hands it downstream. Per patch of bottom it produces about 3.5x the drift of pools, and the caddis larvae and mayfly nymphs coming off it run roughly three times the size of pool bugs.

It builds as it travels. Total drift rises with riffle length, so water arriving at the tail of a long riffle carries what the whole riffle has shed, not just the last ten feet of it. And it isn't recycled downstream: few invertebrates reach the next riffle, because the pool between them is a sink. The bottom third of a long riffle is the richest water in the piece. It's also the part you wade through on your way to the head.

I fished the head of mine all morning for a season's worth of six-inchers. Then one evening, half packed up, I threw a couple of lazy casts into the flat bottom end where it tips into the run. Same depth, no seam, no rock, nothing to look at. Hooked a fish twice the size of anything I'd touched up top. Next trip I checked my depth properly and found I'd been running the rig six inches off the bottom the whole season.

Both fixes fit in one trip. Start at the tail and work up, so you cover the richest water before you push silt and shadow through it. Add weight until you're ticking gravel, then stop. And know what you're up against in daylight: mid-teen fish only at low light in the riffle itself, the rest of the day spent in the deep run carved below it.

Fishing it in that order raises the same two questions every time: how you know you're really on the gravel, and whether an ankle-deep riffle is worth the walk.

Take it to your own riffle

Next time out, give one riffle a full pass in this order and see what changes.

  • Walk to the tail first. Fish the flat twenty yards above where it tips into the run before you go anywhere near the head.
  • Add weight until the rig ticks gravel and hangs up now and then. That's the depth. Six inches up is a different piece of water.
  • Read the surface for dips: darker patches, slicked-over glides in the chop. Put the fly on the upstream edge of each one, not in the middle.
  • Compare smooth fast against ragged slow. Fish the smooth first.

A few more questions people ask about riffles

How do I know the fly is on the bottom? You touch it. Add weight until you're ticking gravel and hanging up every dozen casts or so, then leave it alone. A drift with no contact at all is a drift you can't judge.

Is an ankle-deep riffle worth fishing? The sheet of it, not much. The dips in it, yes. Adult fish want roughly a foot over their backs, and holding water inside a riffle tends to run 1–3 ft deep. A pothole a few inches to a couple of feet deeper than the bed around it is enough.

Is the riffle only worth fishing at dawn and dusk? Not for everything. Brown trout yearlings moved into riffles during daylight in summer and early fall, so the middle of the day gives you honest small and medium fish. The better ones use it on the edges of the light.

Upstream or downstream? Upstream, short casts, zig-zagging across the bed from the tail up. You keep your fly in front of fish that are facing uphill, and you're not walking through water you haven't fished yet.

The short version

A riffle can cost a trout more than it pays
Larger juvenile cutthroat confined to riffle habitat consistently lost weight, while the same age class gained in pools. Energetics alone explain why bigger fish avoid riffles.
They sit on the upstream lip, not the deep part
Of trout mapped in bed depressions, 40% held on the upstream slope where the bottom starts to drop, against 27% at the deepest point.
They're on the gravel, not mid-column
85% of trout noses were within 5 cm of the bed, with bellies averaging 2.3 cm off the bottom and most often flat against it.
There's a speed limit on eating
Prey capture success falls from about 65% to about 10% as current rises from 0.29 to 0.61 m/s. The fastest chute carries food a trout can't catch.
The tail of a long riffle is the richest water in it
Riffles produce about 3.5x the drift of pools with invertebrates about three times larger, and total drift increases with riffle length.
Slow water is the rough water
Turbulence intensity rises as current speed drops, and occupied holding sites had lower turbulence than comparable empty ones. The smooth fast tongue is the calmest place in the riffle.

Sources

  1. Growth of juvenile cutthroat trout confined to riffle and pool habitats[study]Larger juvenile cutthroat lost weight in riffles while gaining in pools; energetics alone explain riffle avoidance by bigger fish
  2. Turbulence and habitat selection: daytime holding positions of stream trout[study]85% of noses within 5 cm of the bed; belly-to-bottom mean 2.3 cm; all fish in bed depressions with 40% on the upstream slope and 27% at the deepest point; occupied sites had lower turbulence and turbulence intensity ran inversely to current speed
  3. Effects of water velocity on prey capture by drift-feeding trout[study]Capture probability falls from roughly 65% to roughly 10% as velocity rises from 0.29 to 0.61 m/s
  4. Invertebrate drift production in riffles and pools[study]Riffle drift production about 3.5x pools, riffle invertebrates about three times larger, and total drift flux increasing with riffle length
  5. Riffles as drift sources and pools as sinks[study]Few invertebrates leaving one riffle reach the next riffle downstream
  6. Energetic cost of feeding from a velocity refuge[study]Leaving a velocity refuge to feed raises cost per attack about 65% and cuts success
  7. Habitat use by brown trout of different ages in a stream[study]Young-of-year brown trout more numerous in riffles; yearlings moved into riffles during daylight in summer and early fall
  8. Tom Rosenbauer: how to find trout in a river[industry]Trout hold in water running roughly 1–2 feet per second
  9. Fly fishing the riffle[industry]Fish in the mid-teens use the riffle only at low light; the deep run below is where they spend most of their time
  10. Reading the water[industry]Adult trout want roughly a foot of water over their backs; riffle slicks mark deeper or slower pockets
  11. Where to fly fish for trout in rivers: the riffles[industry]Trout favour 1–3 ft depths within riffles and small depressions a few inches to a couple of feet deeper than the surrounding bed
Written by

Drew

Founder, Smolt Co.

I grew up on the Little Susitna and learned to fish by asking every question there is, mostly of people who were kind about it. Smolt is the shop I wanted then: no gatekeeping, no assumption that you have $1,100 to spend, and no naming of anyone's creek.

Reading waterBeginner fly fishingTrout behaviourConservation and fish handling
Where do trout hold in a riffle

This Week's Path

One skill, one water-finding trick, one first-fish story. Tuesdays, free, no gatekeeping.