I have an interest to declare this week, though I actually intend to argue against it. I come from a family that can trace its links with flour-milling back 600 years and has therefore benefited more than most from harnessing the power of water. My father ran one of the most famous mills depicted in western art: the one at Dedham, in Essex, once painted by John Constable. By my father's time it ran on electricity, but raising the floodgates after rain was still part of the job - the valley flooded if he didn't manage it in time. Built into the floodgates are eel traps, now disused because the eels have mysteriously disappeared. Gone, too, are the sea trout my father used to catch in the estuary, victims of a 1960s tidal barrage.  I thought I knew about all the compromises involved in taming a wild river. I was unalarmed by the prospect that we might need to compromise again in small ways in our quest for green, carbon-free energy by exploiting hydropower. But then I saw a shocking thing that made me think again. It was a picture of seven plump, healthy perch sliced in half by turbine blades on a river in Holland. It was on the cover of a report by the angling Trust into what is called run-of-river hydropower - small dams, weirs and in-river turbines, as opposed to power generated from big reservoirs, mainly in Wales or Scotland.  The circumstances in which the perch got chopped up are unclear, but the fact that they got so brutally sliced confounds two widely held beliefs; first, that small-scale hydro schemes are environmentally friendly and, second, that coarse fish, such as perch, pike and even minnows, do not migrate. In fact, anglers know that most coarse fish migrate to spawn, feed or move out of water that is either too warm or too cold. Shoals of minnows can be seen hurling themselves at weirs, and in France pike are regarded as migratory fish.  These erroneous beliefs used not to matter, but now they do. For the number of hydropower schemes submitted to the Environment agency (Ea) has risen from fewer than 20 to 100 a year and could exceed that because of the generous subsidised price offered by the feed-in tariff. and the Ea's guidelines for hydropower seem to have been drafted with little understanding of what one of its own advisers has warned could be a "looming tragedy" for fish populations everywhere - to the dismay of the Ea's own fisheries staff.  While it seems largely to be a myth that wind turbines kill birds - at least it seems they don't if the blades are large and slow - it is accepted that hydro turbines kill fish. a hydropower scheme on the Trent near Nottingham, licensed by the Ea, is actually allowed to kill 110 fish every 24 hours - up to 100 coarse fish and eels and up to 10 salmon and sea trout, of which the Trent has vanishingly few. The limits are unenforceable, unless Ea staff hang around trying to reassemble bits of minced-up fish.  The effect of hydro schemes on salmon, sea trout and other fish universally acknowledged as migratory is even worse than on coarse fish, for they have to run the gauntlet of more obstructions. On the west coast of america, the Fish Passage Center, a scientific body funded by the power companies, has found that 100 salmon smolts migrating downstream through a single turbine tend to be reduced to 15 by the time they reach the river mouth, not just as a result of being chopped up or bruised by the turbine blades, but because dams alter the flow and ecology o