- Details (segmentation, curve, legs, antennae, eyes, etc.)
According to all I've read and experienced, this is the most important trigger. If the size of your fly isn't close to the size of the food they are fixating on then you ain't catching.
Fish have to make a fast decision on whether to eat something, especially in faster water. Once the size is determined to be correct, the fish has to make sure the outline looks like something it is familiar with as a food item. The fish has to decide in a short moment if that thing drifting by is food or junk, and the outline of the item is one method it used to sort the item into the "eat" or "ignore" pile. Also, fish in many cases don't have the luxury of seeing detail. The lighting and the water clarity are much worse than what we see above the surface, so silhouetting its prey against a contrasting background is essential to identifying it.
3. Motion or action.
I think this is a really underestimated element in fly tying. Most tiers of "realistic flies" tie up flies that are anatomically correct but that have little movement to them. When compared to a living, breathing insect this isn't very realistic at all, is it? I suspect that's why most realistic flies have failed to make a difference.
Motion is the missing element. I believe, that once a fly is the right size, with roughly the right silhouette, if it moves like it's alive then the fish will take it. Think about how a non-moving fish in a stream is very hard for us to spot, but as soon as it moves, it's easier for us to see it. Well, I think that applies to how action on a fly makes it stand out more to the fish as well.
To that end, I think fly action is best achieved by using long, soft hackles in various places on the fly, as well as using soft materials like long, buggy dubbing and very soft feathers like marabou and feather aftershaft fluff. You can go even further and build in articulation for larger flies. All in all, I think you make a fly more lifelike by imparting movement to it rather than by making its body hyper-realistic in the details. This explains a lot of the success of more impressionistic flies.
People love their colors but I think the fish are less impressed.
Color perception varies widely among fish species, so I'll stick to Trout and Salmon in rivers, where the water is usually less than 10 feet deep where we are fly fishing. From my reading, based on the amount of cones in their eyes, it appears that these fish don't see colors as well as humans do, nor does it appear that they need to. Also, according to this article, human vision is about 14 times better at resolving images than salmonids'.
Based on the above, as well as the fact that visibility is generally much worse underwater than in air, it appears that color would be much harder for fish to perceive than we do in our fly tying vise. Boiled down to the essentials, I think one could be very successful by just deciding where in the range of light cream colored, light brown, or dark brown the fly needs to be (hare's ear vs. pheasant tail anyone?), depending on what's being imitated. In pressured, clear, and slow water I think a closer color match becomes more significant but still to a lesser degree than size, silhouette, and motion.
Considering color from an evolutionary perspective, what does a fish need to see to survive? Well, for people, color is very useful to identify food. However, in the much more muted colors of the subsurface world, color is less important. Fish deal in shades of gray with some brown, green, and blue. Maybe red can be significant since it indicates gills or blood, although water quickly absorbs red spectrum. However, for the most part, if you were a fish seeking food or avoiding danger, what would you look for? I'd say: size, silhouette, and motion. Color is way down the list of useful things to a fish. Everything looks sort of light or dark brown underwater. (Blue flies are an exception to be discussed in another posting).
That said, color is useful to the fisherman in one significant way: contrast. If you can provide the proper contrast of your fly to the water color, or contrast within the fly body itself, I believe you can make your fly both visible and interesting to the fish. Water is less clear than air, and it varies widely in its color and clarity. If the water has color to it, there are colors you can incorporate into your fly to make it stand out. I think this is one of the reasons purple is so popular for steelhead flies. Purple looks essentially black under water but in green water, it will stand out a bit better than black. Chartreuse is known to work in off-color water as well. Likewise for the contrast of pink on black in an Egg Sucking Leech fly or Ralph Cutter's Goblin streamer. These flies have contrast within themselves, so when they move, the contrast makes them stand out more than a monochromatic fly, especially in off-color water. I believe that contrast put into any fly, for example, a dark head or thorax and light body on a nymph, will be noticed by more fish than a fly of a single color. However, I think this is way down the priority list.
Finally, Ralph Cutter, Gary LaFontaine, and others have documented how many nymphs appear to shine or glow underwater because they carry an air bubble with them. This provides some shine to the nymph and I do believe that trout key in on this shine, especially on sunny days. I think a small amount of flash is more important than color for subsurface flies.
5. Body details. These are the items that realistic fly tiers focus on. While they are works of art, I don't believe they catch many fish. This includes details like eyes, legs, antennae, egg sacs, and who knows what else. Maybe in super selective trout situations on clear, slow spring creeks these details make a difference but I still bet they take a back seat to the above items.
A word on eyes. Many people swear by realistic looking eyes on streamers. I don't think they are useful but that's purely my opinion rather than based on any evidence. Ralph Cutter has some good reasons why eyes are counterproductive in streamers. Among them is that in nature, one of tropical fishes' evolved natural defense is to have patterns that look like eyes on their tail to deter predators. That evolution alone is enough to convince me.
In the body details, here are what I think are the most important priorities.
- Segmentation: This is a cool one, albeit unscientific. If you think about it, segmentation is one of the things that visually distinguishes insects (and some other animals) from the detritus that's floating around in a stream. Think about when you're walking around a stream looking for bugs, what do you think catches your eye? Well, for me it's the organization of the insect's body. The body organization is not random, and the segmentation and symmetry make it stand out from the background junk. My brain recognizes this on an instinctual level, much like I bet a trout or other predator does, and it immediately recognizes this thing as a living animal. When I see an orderly segmentation or pattern, that to me signals a living thing.
- Body curve. Most fly tiers take care to make their flies straight. However, if you look at subsurface nymphs floating in the current, they are wriggling, and often their body curves as they move. I tie all my nymphs on curved caddis and scud hooks for 2 reasons. First, to mimic this curve. Second, to hide more of the hook shank in the curve of the insect's body. On a curved hook, much less of the straight point sticks out, making the silhouette that much more convincing. The scud/caddis hooks have a continuous bend all the way to the point section so the body shape follows a plausible curve.
So this is my list. What it boils down to is that I tie a lot of soft-hackle type flies, some with skinny bodies, some slightly fatter. The fatter ones are usually made that way using marabou or after-shaft feather fluff, palmered hackle, or less often, buggy dubbing. I also tend to tie a lot of small flies. Rick Hafele has made the point, which I have corroborated on the river, that in most cases, the natural nymphs/flies we are imitating are smaller than they look, so we should use smaller flies than we think. This means that I tie lots of 16s - 20s with skinny bodies and sparse hackles. For color, I tie some dark brown flies and a lot of light cream colored flies. The advantage of the light flies is that I can quickly darken them with magic marker on the river if needed. The one detail I add to most of my flies is that I rib them with silver or copper wire. I do this to reinforce their body material but mainly to provide segmentation and a small amount of flash.
Hope this is useful to someone. Figuring all this out has helped me simplify my fly tying, reduce the number of patterns I carry, and most importantly catch more fish.