Caddis Larvae--Part I

By Jeff Morgan

Gary LaFontaine's classic book Caddisflies greatly stimulated fly anglers' interest in all stages of caddis : larva, pupa, and adult. Numerous patterns were created to imitate an insect that many anglers had overlooked.

The great popularity of Elk Hair Caddis , CDC Caddis , Goddard Caddis , and Sparkle Pupa --flies nearly as common as the Parachute Adams and Prince Nymph--attest to how far caddis have risen in angler esteem. However, the focus has primarily been on caddis adult and pupae imitations; the larval stage has caught the interest of very few fly anglers.

One reason larval patterns get short shrift is the larvae's lack of visibility: it is hard to ignore choking swarms of adults on a summer's evening, but little critters slowly crawling around the bottom are rarely noticed by anglers.

This two-part article takes a fresh look at caddis larvae and suggestions some new approaches.

Imitating the Wrong Critters?

The two larval species most imitated by anglers are Rhyacophila (green caddis or green rockworms) and Dicosmoecus (October caddis ). However, the effort may be misplaced because these are not the caddis larvae most consumed by trout.

Rhyacophila are common in extremely fast riffles on medium-sized streams, and while trout sometimes feed heavily on the larvae, it's usually during the low light conditions near dawn and dusk. Most anglers cast imitatations in water that is too slow or during the wrong times of the day.

Dicosmoecus, on the other hand, are consumed by trout in slower water, and anglers sometimes imitate this species by drifting a Cased Caddis . Though LaFontaine praises these larvae as being very important, recent scientific research has disproved the importance of these larvae to trout. The fact is, trout seldom eat this species when it is drifting. Instead, trout pluck the larvae off the bottom. Trout may take drifting Cased Caddis imitations, but it's more likely that the fly is just an attractor pattern.

Another Approach to Big Larvae

So if drifting Cased Caddis are taken as attractors, why not go all the way? The Koudelka Killer is similar to a pattern created by Deschutes guide and fly tier Jim Koudelka. It is heavily weighted and the body is nearly all Ice-Dub, which makes it as flashy as an imitation can get.

I use this fly as a "tool" fly: a heavy pattern that is easy to tie (unlike a big stonefly nymph) that can sink a team of small flies quickly.

While originally designed for trout, I have found it very productive for Umpqua River smallmouth bass. Dicosmoecus larvae are a very important part of the smallmouth diet on this river. They are extremely abundant on any section of shallow bedrock in the upper river. Plus, the relatively flat bedrock allows these patterns to be slowly crawled without snagging. Crawling a fly 10 feet ahead of a cruising smallmouth can be as challenging and rewarding as fishing a crab pattern to a bonefish.

Uncased Glossossoma

On streams with lots of sunlight and algae growth, blue-winged olives mayflies and Glossossoma (saddle-case caddis ) may predominate. While the Baetis is lauded everywhere in the fly fishing world, the tiny Glossossoma remains relatively unknown to many anglers.

These caddis are exceedingly common in trout stomach samples during all of their life stages. The larvae normally are well protected in igloo-like pebble cases that are firmly attached to rocks, though like Dicosmoecus they may be plucked off and eaten, case and all.

During periods of stress, however, they will readily leave their case and drift naked for great distances. The little pinkish-tan grubs are much more visible than other drifting prey, and trout will feed on them at a higher rate than they appear in the drift.

The Uncased Glossosoma pattern is simple to construct and easily added to a dropper off a larger fly. The nice thing about Glossossoma is that their emergence period extends from February-November (depending on species), so there are larger larvae (size 16-20) actively drifting throughout the angling season. As a bonus, they are rarely imitated, making these imitations effective on high-pressure waters.

Dubbed Cased Caddis

This fly can be used to imitate a wide variety of cased caddis, and has quickly replaced the elaborate and time-consuming rocky cased caddis larvae in my boxes. Most trout don't eat stone-cased caddis in the drift, largely because the larvae don't drift very far. However, caddis that construct cases from bark and fir needles can and do get lost in the drift. Many species even drift near the surface. During a big storm on a lake, the churning water will dislodge some cased caddis and trap oxygen in their cases. They may float helpless in the surface film until the gases leave their cases.

The traditional fly for these situations was the Strawman, a pattern with a spun and clipped deer hair body. Unlike the monochrome and relatively neat Strawman, the Dubbed Cased Caddis suggests the hodge-podge of materials that go into the natural's case. It has an underbody of superglue, so you don't have to rib the pattern; it retains its shaggy look without sacrificing durability. Plus, this fly will float nearly as well as the Strawman (when treated), and is significantly easier to tie.


Brachycentrus caddis, the American Grannom , is the most common cased caddis larva found in trout stomach samples. That's because the larvae live in precarious positions and have a light, easily digestible case.

The Grannom Larva , a Brachycentrus larval imitation, is easy to construct. Simply use pheasant tail for the body, a twist of bright green dubbing for the thorax, starling for legs, and a black bead for the head. When fishing slow water, omit the black bead and use a black dubbed head. Some anglers have created complicated underbodies to match the 4-sided "box" construction of the natural's case, but I have found this far more time consuming than it is worth.

The best places to fish these patterns are below spring inlets, hatchery or irrigation returns, or tributaries that return a lot of sediment to the stream. The increased suspended sediments cause Brachycentrus larvae to congregate in these areas even when they are not found in large numbers in the rest of the river.

It is important to check on the size of the naturals: many Grannom larvae will be a size 12 in late spring, but by late summer or fall the next generation may only be a size 22 or 24. This pattern tends to only fish well prior to the early summer emergence of the natural insect.


New patterns in this article are:

Koudelka Killer
Uncased Glossosoma
Dubbed Cased Caddis
Grannom Larva

This article continues in Caddis Larvae--Part II

Jeff Morgan has written many articles for Westfly, mostly on entomology and fly tying. He is the author of An Angler's Guide to the Oregon Cascades and Small Stream Fly Fishing.