Select a fly from feeding evidence rather than habit

In this guide

1. Watch one feeding area
2. Describe before naming
3. Build a short choice list
4. Run a useful fly change
5. Keep a practical box
6. Learn from refusals and blank sessions

Choose a fly by asking what the fish appears to be eating, where it is eating and how that food moves. A familiar pattern is a reasonable starting point, but it is not evidence. Slow down long enough to separate an insect in the air from the food a fish is actually taking.

Watch one feeding area

Choose a position where you can observe without disturbing the water or compromising your footing. Watch a repeatable feeding lane rather than every ripple across the pool. Note the apparent feeding level, timing and whether visible food disappears. Do not treat a splash shape as certain identification of an insect or fish.

Tom Rosenbauer’s Orvis hatch guidance explains that fish may select a particular life stage and may ignore a conspicuous insect while taking a less obvious food item. That is why matching the largest insect you can see may miss the real feeding situation. Presentation faults can also produce refusals that look like a fly-selection problem.

Describe before naming

Start with plain observations: approximate size, body outline, colour and position in the water. Orvis’ selection guidance recommends matching the general size and shape of available food and looking for clues around the bank and shallows. A photograph can help you compare later without carrying specimens away or disturbing habitat unnecessarily.

Learning local insect groups is useful, but uncertain Latin names should not become invented certainty in your notes. Write “small pale winged insects, fish feeding just below surface” if that is all you can establish. A local club or knowledgeable instructor can help refine the identification.

Build a short choice list

Evidence	Candidate approach	Check before changing again
Fish repeatedly take visible surface food	A dry presentation with a similar general profile.	Is the fly drifting without obvious unwanted pull?
Activity appears just beneath the film	An emerger or shallow subsurface approach.	Is the imitation actually at that level?
No surface feeding, but subsurface food is plausible	A controlled nymph presentation.	Are depth and drift suitable?
Fish respond to moving prey or follow a streamer	A streamer presentation matched to the observed profile.	Can a change in path or retrieve explain the response?

These are hypotheses for testing, not rules that identify a diet from one surface sign. Keep alternatives open when the fish’s behaviour is unclear.

Run a useful fly change

1. Choose a candidate that broadly matches the observed food and feeding level.
2. Check leader condition, hook point and the presentation before casting.
3. Make a few controlled presentations through the selected lane, allowing the fish room and time.
4. If the fly behaves badly, correct drift or delivery before judging the pattern.
5. If the presentation is controlled but the response remains poor, change one visible feature, such as size or profile.
6. Record the original and replacement, the presentation and any response. Do not credit a fly change automatically if the current, position or fish also changed.

Keep a practical box

Organise flies by the job you understand them to do: surface, near-surface, drifting subsurface or actively moving. Within each group, make it easy to compare profiles and sizes. This is an editorial organisation method, not a demand to buy more flies. An identifiable small set is easier to use deliberately than a crowded box of patterns whose purposes you cannot explain.

Keep a note when a fly needs floatant, a different leader or a different line system according to its design and the intended technique. A pattern cannot be evaluated fairly if the surrounding setup prevents it from fishing as intended. Replace a damaged or fouled fly before drawing a conclusion about fish preference.

Learn from refusals and blank sessions

A refusal can suggest something is wrong, but does not identify the cause. Record whether the fish approached, changed direction, ignored the fly or stopped feeding entirely. Include visibility and your confidence in the observation. If you could not see the response clearly, write that rather than reconstructing it later.

After the session, choose one lesson to test on a comparable visit. “Try a smaller profile with the same drift” is actionable. “This colour always works” is a claim your notes are unlikely to support. Build useful local knowledge through repeated observations, while keeping room for the next session to contradict the previous one.