A trout can feed in one of three ways when it begins taking pupae from just under the film: it can hold at the bottom and when it spots an ascending pupa it can follow the insect (seldom trying to grab it while it is swimming) and suck the pupa in as the insect begins struggling to shed the shuck; or the trout can hold at the bottom and watch for emergents already in the surface film and rise for the pupae as the insects pass over; or the trout can hold just under the surface and sip emergents drifting toward it. Which feeding mechanism a trout uses is determined by the quickness of the insects and the speed of the current. A trout only feeds the third way, holding under the surface, in the gentler flows. It feeds in either the first or second way in faster currents.

The peak feeding, with fish rolling and jumping all over the stream, should be a time when fly fishermen master the fish, but it can be a time of total failure and frustration. Too often anglers fail to realize that the insects are hesitating and concentrating under the film, not on it, and as a result they mistakenly assume that it is an occasion for dry flies.

Again, an effective tactic for fishing this stage of the hatch is one not usually associated with caddisflies. Especially when trout are holding just under the surface or rising to pupae already in the film, an emergent imitation designed to ride semidry generally outperforms any sunken fly fished with an active presentation. Used either dead drift or with a twitch and swim, the "damp" imitation stays in the surface film for the entire time, not just for a moment at the end of a swing or lift.

A technique that kicks the fly into a pulsing swim toward the top, the Leisenring lift, is generally effective when trout are rising all the way from the bottom to take emergents. A sinking pattern is cast upstream and allowed to drift near the bottom, but when it reaches a likely holding spot it is teased to life, hopefully in front of a trout. It is a method that demands skill in reading water if the angler is not casting to visible fish.

The feeding that occurs during this second concentration of pupae takes place in plain view, and a fly fisherman can figure out what to do when he sees the riseforms, but some anticipation of a particular hatch is still valuable. Possibly no other insect type inspires such strong selectivity in trout as a caddisfly pupa, mainly because the bright bubbles of air inside the pupal skin form such a visible characteristic. In this situation the proper method of presentation is not the overwhelming part of success that it is at other times. A fly fisherman cannot just pull out a general imitation and depend on his skill at mimicking the action of the natural. He needs both the proper presentation and the precise fly to consistently fool fish, and only a knowledge of local hatches can give him the chance to acquire or tie patterns beforehand that match a specific caddisfly.

The Third Area of Concentration

When the peak hatch is over and the surface of the river is blank, most anglers quit fishing, or at least stop trying to match caddisflies but there is still an hour of so of very exciting action left. There is one more concentration of insects that pulls fish, often the largest, into specific areas of the stream.

During any hatch there is a low percentage of natural cripples. Seldom, if ever, are these cripples numerous enough to cause selective feeding during the main emergence period. They would have to make up a much higher proportion of the hatch to create such a response. When fish feed selectively they feed with a regular rhythm, and they cannot feed regularly on a component of the hatch that appears irregularly.

The purpose of selective feeding, a set of patterned responses to the stimulae of the insect, is efficient use of energy. By moving with rhythmical repetition a trout ingests as many calories as possible with the least expenditure of energy. The stronger the selectivity to a food item the more regular the feeding motion. There is no way a trout can determine if an insect partially out of its pupal skin, struggling to complete emergence, is in that condition temporarily or permanently. Certainly the cripples are eaten during the hatch along with the healthy insects, but considering them important at that time contradicts some basic principles that are controlling the trout.

Cripples are usually not important during the main hatch because they are not concentrated. During a heavy hatch, over a prolonged period of time, there are nevertheless quite a few of them, and if they could be gathered into a small area they would become a significant food source. Fortunately, for both trout and trout fishermen, that is exactly what happens after the hatch. Even after they drown, these crippled insects remain quite buoyant and are swept along with the current. Eventually many of them join with bits of flotsam or foam and collect in a backwater of the stream to become a great, swirling cafeteria for the fish.

It is after the main hatch that the leftover cripples become the center of attention. The trout abandon the regular feeding lanes and gather under the foam. In the gentle eddies they sip the inert insects, their snouts and backs breaking or bulging the white covering and leaving a momentary spot of clear water in the foam.

The best patterns for mimicking cripples are adult imitations that not only look like, but also act like, bedraggled, half-emerged caddisflies. They sprawl flat in the surface film, wings fanning out along the sides, and quiver with the slightest twitch. They recreate the last, feeble struggle of the natural because they are tied with soft materials.

In backwaters that have a mixture of open and covered water an angler can see his fly floating if it is in a clear area and the casting and striking is quite normal. But in the foam, he has to slap it down hard enough to force it through the upper layer. The angler has to fish blind, striking when a bulge or gap occurs where the fly broke through.

Does the post-hatch feeding attract big fish? Graham Marsh rose an eight-pound rainbow three times on the lower Clark Fork River from a deep backwater. He watched the fish come slowly up out of the hole, but each time the drag on the line snatched the floating flyaway just before it got there. Even with his considerable skill, Graham could not stretch out an extra half-second of free drift with the standard techniques he was using, but it was a glorious sight anyway watching that huge trout rise to a dry fly.

AMERICAN FLY FISHERMEN have not begun to tap the potential of the caddisfly hatches popping out all around them, and they never will with general fly-fishing skills. They will need to understand the habits of the caddis flies in their home trout streams and the unique tactics for simulating the actions of those insects before they can work the full period of the hatch. To benefit from the current revolution in caddisfly knowledge, fly fishermen will have to collect and identify the important groups, at least to generic level, that they encounter regularly. Then they will have to study the biological information in angling entomologies for these insects so that they can anticipate and properly fish the hatches.