What we can do to let the UK’s tamed rivers flow wild and free again

WHEN I visited the river Derwent in Yorkshire in March, the landscape wasn’t as I remembered it from my youth. Normally, the floodplains would be inundated with water and the whole river system feel like a miniature recreation of the vast glacial lake that covered this part of northern England after the ice sheets last retreated. But an unusually dry February meant that floodwater was lacking. On the upside, it allowed me to approach the river across an area of grassland called Wheldrake Ings, which is a haven for rare plants.

The word “ings” derives from the Old Norse word for a water meadow. This was once Danelaw, an area administered by Vikings from Jorvik (now York), and their language is still etched on the landscape. In fact, this stretch of the Derwent looks much as it would have when Danish berserkers invaded in the late 9th century. That is a far cry from most UK rivers. Many have been heavily modified to suit our needs: blocked by dams and other barriers, corralled, straightened, widened or narrowed, disconnected from floodplains, sucked dry for drinking water and encroached on by housing estates and business premises. A river with no such obstructions or canalisation is called “free flowing”. Just 3 per cent of rivers in the UK fit that description.

This is an environmental disaster. Free-flowing rivers are a valuable part of healthy landscapes and ecosystems. They provide various ecological benefits, from carrying sediment and purifying water to allowing fish to migrate. How can we reverse the damage we have done and help our rivers run free again?

Humanity has been trying to tame rivers for millennia and, today, the results are clear to see. In 2019, when a team at McGill University in Montreal, Canada, and their colleagues produced the first global map of free-flowing rivers, they found that only about a quarter of very long rivers – those in excess of 1000 kilometres – are unobstructed along their entire length. These are almost all in remote regions of the Arctic and the Amazon and Congo basins. More encouragingly, 80 per cent of medium rivers (between 100 and 500 kilometres) and 97 per cent of short rivers (between 10 and 100 kilometres) were classified as free flowing. However, the researchers warned that those numbers are probably vastly inflated because their analysis excluded small barriers. This is especially true of highly developed regions in Europe and North America; the overwhelming majority of truly free-flowing rivers are in remote areas.

Shipping channels

More recently, a European Union project called Amber (Adaptive Management of Barriers in European Rivers) concluded that 3 per cent of UK rivers flow unobstructed. That is very low by global standards, but perhaps not surprising given the long history of river modification in this small, crowded country. One of the earliest approaches was what is now called channelisation, which entails raising riverbanks, deepening the channel, consolidating separate streams known as braids and sometimes straightening out meanders. Historically, this had two main purposes, according to Susanne Muhar at the University of Natural Resources and Life Sciences in Vienna, Austria. The first, and foremost, was to improve navigability for shipping, which is why it is also called canalisation. The second was to control flooding by segregating a river from its floodplains, with the knock-on benefit of draining floodplains, which could then be used for agriculture and settlement.

Until the industrial revolution, channelisation in the UK was quite benign, but the advent of steam power upped the ante considerably, says Muhar. River navigation came under intense competition from the railways and responded by fighting fire with fire: steam-powered watercraft became larger and there were extensive “improvements” to rivers. This process continued throughout the 20th century, compounded by the fact that channelisation was also employed to speed up river flow to help flush away industrial effluents.

Similar pressures were changing the course of rivers in other industrialising countries. Much of the early research on the detrimental ecological effects of channelisation was done in the US. It wasn’t until 1983 that geographers made the first such survey of rivers in England and Wales. They found that, in the previous half century, 8500 of around 35,000 kilometres of the main rivers had been subjected to major channelisation. (This was a much higher proportion than in the US, which had a “channelisation density” of 0.003 kilometres per square kilometre, compared with 0.06 in England and Wales.) Most of the rest was managed to some extent, including vegetation cutting and the removal of submerged ridges and banks to smooth out the flow.

Soon after, the Royal Geographical Society in London published a series of articles on problems associated with river channelisation. It concluded that the main impacts are twofold: on the river system itself and on the wildlife around it.

Unsurprisingly, the physical characteristics of a channelised river change dramatically; that is the whole point. The increased flow rate can cause a river to gouge ever deeper, further estranging it from its floodplains. This has knock-on effects downstream. If and when the river re-enters a non-channelised section, the flow rate decreases and masses of sediment are deposited, often requiring the next section to be channelised in a domino effect. This is what happened to the Danube in Austria, which is now completely channelised, according to Muhar. If a river has been channelised to reduce flooding, this often merely displaces the problem downstream, again inviting further channelisation. Then there is “retrograde erosion”, where the pull of the channelised section increases flow rates upstream, altering the course and depth of a river’s upper reaches and gouging out yet more sediment.

The negative effects on biodiversity result principally from homogenisation of the riverine habitat. Flow rates become more uniform, depriving aquatic species of their preferred water velocity. “You change the habitat structure,” says Ulrika Åberg at the International Union for Conservation of Nature in Gland, Switzerland. “You often get a very monotonous system that doesn’t have a large variety of riffles [shallow, turbulent zones] and pools and other morphological features that are important for the ecosystem.” Meanwhile, the removal of bankside vegetation can alter the water temperature and reduce the amount of nutrients falling in.

The value of a non-channelised river can be seen with the Derwent. The river itself supports many rare ecosystems and species, such as river lampreys and mats of vegetation that provide food and shelter for fish. And the floodplains are so important ecologically that much of Wheldrake Ings is a national nature reserve. Annual floods deposit masses of nutrient-rich sediment onto the ings, maintaining a rare, protected habitat called MG4 grassland. When I was there, it looked like an unremarkable field of turf. But this becomes a riotous flower meadow and, in late summer, will be a valuable hay crop, prized by specialist livestock farmers. “It’s a certain mix of plants and grasses,” says Craig Ralston at Natural England, which manages the reserve. “This area contains a significant proportion of [MG4 grassland], probably a third of the global resource, because it is globally confined to England and Wales.” The flooded ings also support an internationally significant influx of migrating waterfowl, with around 40,000 overwintering shovelers, teals, wigeons, Bewick’s swans, ruffs and golden plovers.

Nevertheless, not all is free flowing on the Derwent. Although largely unchannelised, except in its lowest reaches, it does have barriers that alter its flow and prevent the natural movement of sediment and animals. There are four old weirs upstream of Wheldrake, and downstream there is a large, modern lock-and-sluice system used to control water levels to enable the extraction of drinking water. There is also a tidal barrier at the river’s confluence with the Ouse at Barmby on the Marsh, which prevents brackish water entering and tainting the drinking water. “Human-induced changes in hydraulic conditions” is one of the major threats to the Derwent, according to the EU’s Natura 2000 programme, which has designated part of it a special protection area.

Obstacle course

Again though, the Derwent has far fewer barriers than many other UK rivers. In 2020, a team led by Carlos Garcia de Leaniz at Swansea University in the UK published an inventory of these obstacles across Europe: dams, weirs, sluices, culverts, fords, ramps and others, including small ones that had previously been overlooked. In the UK, they recorded 23,719 barriers on more than 68,000 kilometres of river. That suggests the UK has nearly 50,000 barriers in total, or about 0.7 per kilometre of river. The average for western Europe is 2.7, but that is heavily skewed by the Netherlands, which has 19.4 per kilometre.

One major problem with barriers is that they impede fish, especially migratory ones such as eels and lampreys, which swim far up and down rivers to complete their life cycles. On the Derwent, river lampreys are only found downstream of the lock, suggesting they can’t get through it. There are fish passes but they are “probably not very successful”, says Ralston. Critically endangered European eels also get stuck behind or in front of barriers. They migrate from the Sargasso Sea in the Atlantic deep up rivers across Europe and North Africa to mature, then return to the sea to breed many years later. River barriers are one factor in their precipitous 90 per cent decline since the 1980s, says eel expert Jack Wootton at the University of Hull, UK.

Barriers stop the natural movement of sediments too. “They get trapped, which means that the river downstream gets starved of sediment,” says Åberg, “When the river doesn’t get replenished by new sediment, the force of the water starts digging down into the bottom, so you can get a very strong incision.”

The obvious solution is to remove barriers where possible. “Often no one really knows what a weir was for,” says Wootton. “That isn’t the case for all weirs, dams and sluice gates, but for a significant amount of them, nobody has any want or need any more.” A key obstacle to getting rid of them is economic. “Money is an issue. You’re talking about hundreds of thousands of pounds, potentially, [for a stretch of river],” he says. Another issue is that removing seemingly redundant structures may have unintended consequences. Take the Derwent again. In the 1950s, engineers built up some of the riverbanks around Wheldrake to reduce the frequency of flooding. It would seem logical to raze them now, says Ralston. But climate change means that more water flows into the Derwent than before, so their removal might threaten the ings with excessive flooding. “That’s a really difficult one to solve,” he says.

Immovable barriers

Other barriers simply can’t be removed. These include hydropower dams and constructions designated as industrial heritage. Even where historic barriers aren’t protected, it may be impossible to get rid of them because of their settings. For example, says Ralston, a few years ago, Natural England was involved in a project to see whether it was feasible to remove a weir built in the 1800s on the Derwent at Howsham to serve a now-demolished watermill. As a trial, the water level of the millpond was lowered. This caused the water table to drop, leading to subsidence damage to the nearby ruins of Kirkham Priory, a grade I-listed medieval building, and the plan had to be shelved.

Where barriers can’t be removed, fish passes may at least help rescue biodiversity. They have been installed at dams for more than a century and are usually one of three types: a slippery slope, a ladder or an elevator. Slopes are water slides for fish; ladders are a series of small pools of increasing elevation to enable fish to jump a barrier in bite-sized chunks; elevators are holding tanks that accumulate fish and periodically hoist them over. “How successful they are is open to debate,” says Ralston. One problem is that they can be quite selective in the species they help. “A lot of them are very specific, usually focused on salmonids,” says Wootton.

Fish-pass technology is quite rudimentary and has barely advanced in a century, but now things are improving. A Seattle company called Whooshh Innovations is designing high-tech fish passages such as the Lampway, a water-filled Archimedes screw that transports lampreys over otherwise-insurmountable barriers. Some of these systems automatically record how many fish pass through so biologists can monitor populations. Other innovative solutions aren’t so high-tech. The Lough Neagh Fisherman’s Co-operative Society in Northern Ireland has perfected a pass called an eel rope. “It’s essentially a long, sausage-shaped bundle of straw that you drape over the top of a river barrier and it creates a climbing structure that eels can work their way up and over,” says Wootton. Each sausage costs peanuts. “I truly believe in 10 years, we will be looking at very, very different fish passes,” he says.

This is all well and good, but Ralston points out that fish passes don’t address the wider issue of water quality due to sedimentation around barriers. Only removing the obstacles will do that. Even then, remediation isn’t always possible. “If a dam is very large and has been in place for a very long time, it creates irreversible impacts,” says Åberg. Reversing the worst effects of channelisation may also be impossible. In many cases, the river has gouged a bed several metres lower than where it started, making it hard to reconnect to its floodplain. Besides, those floodplains are now often covered in houses and businesses that need flood protection.

One key lesson from all this is to value our remaining free-flowing rivers, because undoing the damage isn’t going to be cheap or easy. But the good news is that, if we embrace the challenge, we will have a surprising ally. What has been done can sometimes be undone simply by letting nature take its course, says Åberg. “Rivers can restore themselves if we just remove embankments and barriers and everything that constrains them.”

Graham Lawton is a features writer at New Scientist