Obviously, there is no great cut-off point, where you suddenly decide that in the next field you have moved from, say, the upper river to the middle river, and therefore a change of approach is needed. The transition from one area to another, and the change of approach that is required, is much more subtle than that. But a little about the Severn before we discuss the magnificent barbel that inhabit the river.

The Severn, Britain's longest river from source to tidal waters (about 180 miles) is a moody river, exceptionally prone to sudden rises in level and changes in colour. It can thank the Welsh hills for that, for the Severn - Hafren in Welsh - begins its run to the sea down the Plynlimon peak known as Bryn Cras in North Wales. For about 15 miles it tumbles through the valleys until it reaches Llanidloes where it is joined by the rivers Clywedog and Dulas, the first of many tributaries that feed and swell this great river.

The upper Severn is of little interest to the coarse angler until it reaches Newtown, and it is from this area to Atcham, below Shrewsbury, that is recognised as the upper Severn. From Atcham to Worcester, or just below, we think of it as the middle Severn. And from below Worcester to Gloucester we have the lower Severn.

Middle Severn barbel are well established and provide probably the best and most prolific barbel fishing in the country. Instigated by Angling Times, barbel were stocked in 1956 by the old Severn River Authority. Some 509 of them up to about 9lb were netted from Berkshire’s river Kennet and planted into the Severn at nine different points between Shrewsbury and Bewdley. It took about 10 years for them to make their presence widely known, but by the end of the 1960’s they were being caught quite consistently.

Upper Severn barbel are still in the process of establishing themselves, still creeping up the river to colonise new stretches. At present they are pretty well established as far up as Abermule, but it must be said that only very odd fish are reported from this area. From just above Welshpool, however, they have become the prime target of many anglers who now fish for them expecting to catch one, rather than merely hoping.

The Lower Severn was colonised over the same period as the upper Severn, but this wide, slower, deeper section of the river is home to the Severn’s biggest barbel.

I can’t say much about fishing the lower Severn for barbel for it has not been one of my regular haunts. But I do know the river is very canalised, wide and deep – as much as 20ft - and the usual methods used for the middle reaches work just as well on the lower reaches. It is more a matter of location than anything else which, due to the river being largely featureless, means that much trial and error is necessary to find the fish.

The greatest difference between the upper Severn and the middle Severn as far as barbel are concerned is that the middle Severn has a much greater population of both barbel and anglers. The barbel, therefore, are subject to a hell of a lot more pressure. Even today, with the upper Severn being more popular than it was a few years ago, it is still possible to fish many of its stretches week after week and not see another angler. This lack of pressure makes a great deal of difference to the approach to catching these fish.

Middle Severn barbel are subject to a constant bombardment from swimfeeders filled with hemp and caster or hemp and maggot. Feeders with 2 1/2oz leads are necessary in many swims to hold bottom in the strong flow. These must weigh in the region of 4oz when filled, and they crash over the heads of barbel with enough disturbance to rock a small boat. Yet these barbel continue to feed. Here is a case that proves as much as anything that fish can be trained to accept practically anything if they are subjected to it over a long enough period. Indeed, middle Severn barbel, and I have no doubt it applies to heavily pressured barbel in well-stocked rivers everywhere, actually wait for the crash of the feeder and at times grab it on the way from surface to bottom in their haste to feed on its contents. It is a case of the barbel coming to accept that maggot, caster and hemp are part of their staple diet, and that it is delivered in plastic containers that crash through the surface with a lot of noise. That there is an inherent danger of them being hooked and pulled from the water has become an unwanted but acceptable risk.

Upper Severn barbel, at least those well above the ‘boundary’ area between middle and upper, have not been subject to such a bombardment and are, therefore, more prone to shy away from such disturbances. Of course, swimfeeders tactics will catch upper Severn barbel, but it is usually a case of working up a swim with the feeder and then waiting a spell for the disturbance to die away before the barbel venture into the baited area and then feed with any confidence. Only rarely do they pounce on feeders immediately they dive through the surface.

Better techniques for upper Severn barbel are more akin to the stealth and cunning of the sniper at close range rather than the heavy bombardment from a distance. Namely the light lead and big bait approach in pre-baited swims, ie, walk the length and introduce a helping of hemp and hookbait samples into a few likely swims and then fish each in turn.

Significant also is the fact that upper Severn barbel are less inclined than middle reach barbel to feed during the hours of daylight. Again the difference is due to angling pressure, including the fact that the middle Severn is heavily match-fished in many areas, whereas the upper Severn is very rarely match-fished. Fish that are subject to heavy and continual feeding of hemp, maggot and caster through the daylight hours will obviously be more inclined to feed, to the point where daylight feeding will become a forced habit rather than the natural nocturnal habit of less pressured fish. As a result of this it is far easier to catch barbel in daylight from the middle Severn than it is to catch them in daylight from the upper Severn. Therefore the upper Severn barbel angler will generally time his visit to take in more of the hours of darkness than of daylight. If not actually fishing though the night, he will begin fishing an hour or two before dusk and continue to fish into darkness for several hours. Of course, coloured water following rain can persuade the barbel to feed during the day.

Very often my own approach to the upper Severn is to fish into darkness but not to use any feed whatsoever. I do not prebait any swims, but use a big bait in known swims. That bait is either a very smelly one, such as flavoured luncheon meat or designer paste, or a large lobworm. The choice is decided by the clarity of the water in that hour or two before darkness falls, trying lobworm in coloured water and then switching from one to the other after dark, but usually it is the smelly bait that does the trick most often. My reason for the no-feed approach is that if you fish a known swim for barbel that are not caught too often (if at all in some instances on the upper Severn) then that big bait should be enough to do the trick. When there is little or no competition from other barbel then it shouldn’t be a problem for the few barbel that are around to find a good, quality bait. And generating the confidence factor with a helping of hemp or other prebait material, should be unnecessary when that lone barbel does not have to compete with other fish. As often as not the competition for upper Severn barbel comes in the shape of chub.

I’ve left the most important aspect till last for, clearly, locating upper Severn barbel is more than half the battle, for there are still many areas of the upper river that are practically, if not completely, devoid of the species. Unlike many of the chalk streams of the south it is not so easy to walk the upper Severn in order to spot the fish. Most of the time the water is somewhat less than clear, if not downright murky, and spotting any fish, let alone identifying the species, is a thankless task. That leaves just a couple of options, the first, and easiest, being to walk the stretches and look for barbel anglers rather than barbel. Perhaps not the most ethical tactic to locate barbel but probably the most effective. Most barbel anglers are recognised by the designer seats they use, two rods in many cases, and the fact that there are no bait boxes in sight. Okay, I’m being cynical, but what I say is true in many instances. The other way, and the best in that you’ll get the most satisfaction from it when it pays off, is to fish on a trial and error basis, fishing those stretches and swims that look the likeliest, based on past experience.

So, when you fish for barbel on the Severn, the best way is to base your approach according to whether it is the upper, middle or lower reaches. Each area has either dense, less dense, or sparse populations of barbel and proportionately varying degrees of angling pressure. And that is the key to it all.