How industry shaped Severn salmon runs

The Severn is – and always has been - a hard river from which to catch a salmon on rod and line. This was true even when the river was at its peak as an early spring fishery. The need to spend many days on the bank to land a fish was the reality for the majority of anglers, most of the time during the heydays from the late ‘60s to the late ’80’s. Tales from the riverbank to the contrary are just that: angler’s stories. They don’t represent the true picture.

The nature of the Severn makes it a problematic river to fish. The impact of industry and geography means that there is perhaps little more than a mile in the bottom 55 miles of the river that could be described as salmon fishing water. Below Gloucester the huge tidal range – the second biggest in the world – stirs up vast amounts of silt and makes rod and line fishing very difficult. Between Gloucester and Stourport the impact of the industrial revolution and in particular the work to make the river navigable for large boat traffic in the mid nineteenth century had a massive detrimental impact on the Severn’s migratory fish species. Creating the modern navigable river  led to the construction of seven weirs, some of which (Diglis and Lincomb especially) are very considerable barriers to upstream migration. The dredging of the boat channel removed the pools and fords of the natural river and created what on low water is to all intents and purposes nothing more than a very slow flowing canal. Within a few years the Shad that had once spawned above the Welsh border could no longer get more than a few miles past Worcester. 

To get a sense of the environmental devastation wrought by the building of the Severn navigation weirs, imagine the Wye from Tintern to Hereford turned in to a waterway more like the Manchester Ship canal.

Diglis weir circa 1900

Above the navigation weirs the prospects for salmon fishing are very much dependent on water. Plenty of it and at the right time.

The result is that angling effort is heavily concentrated on a few key hot spots provided by the man made barriers to migration.  Writing in the seventies, at the peak of the Severn spring fishing, Ken Cope remarked in his Angling Times Book of the Severn that: ‘away from the weir pools Severn salmon fishing is a very chancy business’. In the old days when there were a lot more rods on the river, anglers would be fishing the weir pools shoulder to shoulder. 

This is not to say that salmon in the lower river only hold in the water below the weir pools. There are many locations within the 42 river miles from Stourport to Gloucester where salmon hold up (especially on low water). The problem is that they are difficult to catch mainly because of the ban on certain methods normally used on very slow canal like stretches, such as float fished shrimp.

The pattern of fish reaching the lower river weir pools then dropping back some way downstream was one recognised in the 19^th century by J. W. Willis Bund chair of the board of Severn Conservators who wrote an interesting book ‘salmon problems’ published in 1885. What he said about the pattern of salmon runs on the Severn is very important because he had access to an enormous amount of information from the net fishery.

‘A fish, this year (1885), was seen in a net at Powick. It got out of the net as it got over, one of the fishermen hit it with the boat-hook, and marked it. A fish so marked was caught two or three days afterwards, some way lower down the river, although there were several deep pools between where it might have stopped.


 If ascending fish reach a mill after waiting a day or two in the pool below the wheel, if not gaffed, or otherwise killed, they nearly always drop back down the river. A poacher who tries to gaff a fish, always, if he misses it, looks first in the pool below for the fish, never, if he knows his work, in the pool above. All this evidence goes to show that if a clean fish is interrupted in his journey up stream in fresh water he drops back. ‘

‘Does this apply to tidal waters? If fish were in the estuary, and met with any serious obstruction, would they drop back seawards.? The question is important in two ways. If the fish drop back and remain in the lower waters, then the nets and fixed engines get a double turn at them, and the chances of the fish escaping and ascending up the river are greatly decreased ; and if the effect of an obstruction like a weir is not merely to delay the fish in the pool at the foot, but to cause them to drop back, then the existing Fishery Acts require important amendments to prevent the fish that are ascending the river being caught.

 It was, by the Act of 1873, made illegal to fish 50 yards above, and 100 yards below, any weir, the idea being that the fish would rest in that space ; but if the fish drop down beyond that distance, then the existing law gives no protection whatever to the fish that are obstructed by the weir, and Boards of Conservators and the Legislature, who have considered the fish amply protected, are really imagining a vain thing.’

Today during conditions of low summer water most salmon stay out in the estuary. The same was true in the 19^th century. The combination of water temperature, low flows and poor water quality see to that. In a summer drought you can expect the river to be running at over 20c at midnight. This is outside the salmon’s thermal comfort zone and is not going to attract many fish to run the river. In addition to this low flows magnify the problems caused by domestic and industrial discharge. Despite undoubted improvements over the last 100 years the fact that over 6 million people live and work within the river catchment does impact on water quality and dissolved oxygen levels. This deters the salmon from entering the river. The tendency for summer and early autumn salmon to stay out in the estuary was another pattern that Willis Bund recognised very clearly:

‘There is an old fisherman's saying that "Salmon only run well on a spring

tide," and no doubt it is on a spring tide that the largest run of fish occur; but there is also no doubt that every tide the fish run up the estuary a certain  way, and fall back with the ebb. For the greater part of the year, — at some times much more than others, — the fish congregate in the estuary, and remain there. Some few work up each tide, but the majority wait for a fresh before they leave the estuary. ‘

How many salmon rivers have signposts for barges?

‘It is this that makes the estuary (net) fisheries so valuable ; the fish remain there tide after tide, and the fishermen have ample time to take them, while, when the fish have started inland, they are moving upwards, and, unless caught when moving, have passed, and

are not met with again.’

‘The fish seem to congregate most in the estuary in the summer and early autumn. I say " seem," for the reason that the water-bailiffs find more fish stranded on the sands after each tide during those months than during any others. It may be that the fish equally congregate in the winter and spring, but that the higher land-water coming down at those times

makes them take the rivers more readily than during the summer and early autumn, when, the rivers being low, the fish hang about, and appear to congregate more in the estuary.’

‘It may also well be that in low water they do not care to face the pollutions through which they have to pass, and wait for a rise of the water before they try it. Whatever is the reason, there can be no doubt the fish hang about the estuaries for some time in the summer and autumn.’

 Nowadays we know that a cooling of the water temperature and a small rise will tempt fish in to the lower river, especially on the biggest tides, and some fish will always push in despite the low water. However these fish (with the exception of the smaller grilse) do not usually make it past Diglis weir at Worcester. Diglis is the key barrier, because the weirs lower down are not obstacles on a big spring tide. Tewkesbury disappears on anything much over 10m at Sharpness, while an 8m + tide drowns Llanthony or Maisemore. Willis Bund observed exactly the same pattern:

‘Throughout the season, fish are moving up from the tideway. Of course, the number of these fish depends on the stock in the river ; but all the year round, in large rivers, clean Salmon are moving up; in the autumn very few, but more about December and January, and so on through the year. It is to the fish that return singly that weirs do so much harm ; the fish push up even in low water, and would make their way up the river little by little, but in low water they are stopped by the first weir they arrive at, and either drop back or remain in the pools below, and it is doubtful, if they once remain, how long they will stay. They seem to be content with their quarters, and to lose their migratory impulse.’

‘ A large run of fish never comes but with a rise in the water ; but these stragglers, until they make up their minds to stay in some pool, are always moving on up. One instance will show

this: the Worcester net fishermen always go out fishing on a Monday morning, however low the river, because they say fish will always work up so far. They do not fish the greater part of the week in dry weather; but the effect of the weekly close-time in stopping netting for two days generally gives them fish on the Monday. '

Bund’s analysis of how the fish stay out in the estuary in summer and autumn provides support for the idea that without sustained rain throughout the summer then the majority of the run will not enter the river, or fail to penetrate far, until after the season has finished. Those of us who keep an eye on some of the lower river weir pools in late autumn and early winter see this most years as a huge run of mainly coloured fish push upstream with the first water in November. 

This is also confirmed by looking at those years when we have had exceptional summer rainfall. In 2007 and 1988 there was a much higher proportion of grilse in the rod catch than usual, the peak months were in the later part of the season from July onwards and the middle and upper river accounted for a much higher percentage of the rod catch than the normal  figure for a low water year of just 25%.