Educating Trout

Bob Wyatt’s thought-provoking new book, What Trout Want: The Educated Trout and Other Myths, made me consider my experiences with ‘selective trout’ in New Zealand. The debate on selective feeding and learned hook-avoidance seems to have been fuelled as much by selective angler experience, and perhaps the lack of a clear definition, as by the vagaries of trout behaviour.

SELECTIVE FEEDING

Most anglers consider selective fish to be those that refuse their flies. They persist in this belief even if the real reason for the refusals is poor presentation. For a biologist, selective fish are those that choose specific prey in preference to other available items. In their stomachs, the preferred prey will occur in greater proportion than it occurs among all the potential food items in the water.

One reason trout feed selectively is the adaptive advantage of maximising energy intake, and optimising the trade-off between the amount of food eaten and the energy spent catching it. For example, during mayfly hatches Mataura River brown trout feed selectively on emergers, the dominant and most vulnerable prey. Once the hatch begins to peter out, that strategy loses its advantage. No longer locked exclusively on that prey image, the fish invariably begin paying more attention to natural duns and the indicator dry flies some of us use when fishing emergers. This does not necessitate conscious decisions by trout. According to optimal-foraging theory, it is a hard-wired genetic response with an evolutionary advantage.

But there is another explanation for selective feeding. The notion of intelligence and learning by trout is contentious. However, nine years of fishing backcountry bush streams has convinced me that brown trout do become more wary of fly patterns representing certain prey groups.

The season opens when trout are mostly feeding on nymphs, so getting flies down to them with a decent presentation is generally all I have to do. By December the fish are ‘looking up’ more often. They frequently rise to medium-sized terrestrials and will rise a little further to intercept a nymph, so dry-fly/nymph combinations are highly effective. Come mid-January the trout have become frustratingly ‘smart’ to the extent they are immune to Royal Wulffs, Humpies and even the most realistic blowfly and bee imitations I can devise.

Then chorus cicadas hit the water in late January or early February, and suddenly the recalcitrant fish of mid-January confidently attack anything resembling a #8–#6 cicada (see FlyLife #50).

Chorus cicadas usually begin hatching on the coast two months before they appear in the cooler inland bush valleys. Some years, my first eagerly- anticipated ‘cicada trip’ has preceded the hatch by about a week. I persevered with a cicada pattern on several of these occasions, but although a few fish had a casual look I did not have a single take. On occasions when the hatch was just beginning and a few cicadas were about, I managed to catch some fish on an imitation, but also had a good many refusals. But when my first trip has coincided with a valley filled with chorusing cicada Romeos, big brown trout have charged down my Sparkle Wing Cicada from all directions, moving surprising distances, vertically and horizontally, to confidently swallow the fly. This strongly suggests that trout have to learn each season that cicadas represent good food.

Unfortunately, just as backcountry browns learn to avoid medium-sized dry flies in mid-summer, so too do trout become wary of large cicada imitations. On my first trip up a particular section of a local stream I usually expect to hook 15–20 fish, provided I am the first to fish that waterway once the cicada hatch has fully developed. Typically the weather packs in soon after this trip. It is two or three weeks before the flows recede to manageable levels and I have that all-important sunny day for my next visit. On the second trip I am likely to hook only two to six fish on the same stretch of water, generally not any of those caught previously. When I move on to water not fished during my first visit, I’m usually back to confident head and shoulder rises, and fingers bloodied from removing Sparkle Wing Cicadas from the throats of big brown trout.

I have rarely caught the same trout twice in one season on a SWC. My observations on local bush streams are consistent with the results of a series of scientific angling experiments conducted by Roger Young and John Hayes (Cawthron Institute) on four backcountry rivers, with the trout becoming harder to catch and requiring more fly changes as angling pressure increased.

SELECTIVE EXPERIENCE

Much of my understanding of learn-ed hook-avoidance in brown trout is based on my experience on lightly-pressured North Island backcountry streams. Had I worked mainly lowland rivers or more heavily fished backcountry streams, my conclusions may well have been different.

My experience of the Mataura River, for instance, has been in autumn when there were plenty of other anglers about and the sections I fished had been pounded throughout the season, yet I still caught trout, both during and between hatches. Densities in lowland rivers are much higher than those in backcountry streams, with the result that even if some fish have become ‘educated’ for a time, there are plenty more uneducated fish to sustain satisfactory catch rates using impressionistic patterns. Lowland rivers also have abundant food supplies which often lead to selective feeding for optimal foraging. When hatches are so prolific, the sheer number of naturals might quickly re-educate previously hooked fish to the fact that the organism is safe to eat. As long as imitations are close enough, one could expect multiple recaptures within a season using the same pattern.

Research on the Owen River — a heavily-fished backcountry stream — revealed that catch rates were lower than in the little-frequented Ugly River, despite similar trout densities, because less of the population was visible at any one time, and sighted fish were also more difficult to catch (Young & Hayes 2004). However, the participating anglers did not experience a drop in catch rates on consecutive days fishing a given stretch of water, as they did on the Ugly River. It seems that as fish were caught and went into hiding for a period on the Owen, other fish replaced them and became vulnerable to being caught.

THEORETICAL MODEL

Observations which should be ex-plained by any model of trout feeding behaviour may be summarised as follows:

1. Trout feed selectively on mayfly emergers during prolific mayfly hatches.

2. Trout feeding on mayflies during hatches on heavily-fished rivers (in the USA) may sometimes prefer one pattern imitating the selected organism over others imitating the same organism.

3. Trout on lightly-fished back-country streams will accept a wide variety of flies.

4. With increasing angling pressure, trout on lightly-fished backcountry streams become fussier about flies and are harder to catch.

5. Trout on backcountry streams will make a big effort to take large cicada imitations when the naturals are about.

6. Trout caught on a large cicada imitation are unlikely to be caught again on a cicada during the same season.

If I had to come up with an explanation of feeding behaviour that accounted for all of the observations listed, it would have to include learning. It would look something like this:

Trout are opportunistic predators that feed on a wide range of prey types, but will feed selectively during prolific hatches to optimise energy intake. This form of selective feeding is more likely to be hard-wired than learned, and is probably the result of trout selecting prey using only one search image during abundant hatches. Although trout are opportunistic predators, they do need to learn through repeated exposure that new hatches of unusual prey forms are edible. Trout caught and released may learn to avoid flies imitating unusual prey types once they have been caught on them.

This model suggests that apart from the angler’s presentation, there is more than one mechanism determining whether a trout will eat a fly, and that there is some involvement of learning.

While active avoidance is fairly straightforward, passive avoidance — the result of a trout not recognising a fly as food — is a bit more complex. Rather than specifically avoiding a particular fly, educated trout may simply have learned to feed more selectively, focusing on smaller more abundant prey (in the absence of prolific hatches).

But what could all this have to do with catching more trout?

PRACTICAL IMPLICATIONS

Anyone who has spent time fishing New Zealand’s crystal-clear backcountry streams will appreciate just how important presentation is when targeting spooky browns. However, assuming presentation is up to scratch, accounting for the ability of trout to learn to adjust their feeding behaviour to avoid recapture, is guaranteed to produce more fish.

If you’re planning to visit a lightly-fished backcountry waterway with the specific aim of enjoying high catch rates, try beginning your trip just after a period when the river is likely to have experienced no angling pressure (e.g. just after inclement weather or high flow rates), or visit it at the start of a new terrestrial hatch (e.g. during the cicada season before fish have learned to avoid cicada imitations). Employing these two strategies, it is possible to experience world class fishing on many low-profile streams with typically low trout densities.

The timing of trips to large lowland rivers with decent trout populations is less critical to success. Having said that, you are likely to have the most fun when mayfly hatches are at their best, such as autumn on the Mataura. Just remember to carry some emergers in your box.

When fishing popular backcountry rivers it is good policy to carry nymphs imitating a range of common local prey, just in case fish are feeding selectively upon one or two items only. This often means smaller foods such as Deleatidium mayfly nymphs of 8–10 mm (#18–16). However, there are occasions when trout are keyed in to large green stoneflies (Stenoperla) or spiny-gilled mayflies (Coloburiscus). Timing trips to popular destinations to coincide with the start of the cicada hatch will also ensure good fishing on heavily-angled streams.

Since selectively feeding fish probably pay more attention to those triggers used to identify selected prey, incorporating key triggers into patterns is a good idea. One would expect anatomical detail, such as accurate profile and appendages (e.g. wings), to be more important for bigger, more visible prey (e.g. cicadas), than for smaller ones. Although this may be less important for small prey items, my experiments with stretch cord over the last five years (FL#53) suggest that translucency is a key trigger for nymphs #16 or smaller. Most small aquatic nymphs are strongly translucent, especially when viewed from below.

Exaggerated trigger features, recommended by some authors, will ensure patterns stand out. But while they may increase strike rates from naive fish, they could also result in the pattern being so different to the natural that selectively feeding fish may not choose them.

Assuming trout are capable of learning to avoid prey groups, and are capable of discerning key triggers on selected items, it is also possible that they learn to avoid prominent unnatural features on flies, such as gold beads or excessive flash. Gold bead-head nymphs are generally ineffective later in the season on backcountry streams, especially on bright days. I have even seen fish spook after inspecting them, in much the same fashion as some of the fish approaching my SWC for the second time in a season. I tend to avoid excessive flash on flies I tie for bush streams.

There will be readers who believe I may have spent too many days ‘up the creek’ on my own. But the next time you are casting Royal Wulffs and Hare and Coppers to trout that do not appreciate your near-faultless presentations, you may wish you had given more thought to the flies in your box, or the timing of your trip.

And if you want to experience some of the most exhilarating fly fishing imaginable, visit a backcountry bush stream near the beginning of the cicada season. Hang around for long enough and the trout may even teach you something about how they learn to disregard artificial flies.