It’s brilliant that the Environment Agency has, through the national framework, identified the flow deficits that exist on our chalk streams. 

It’s great that Environment Agency has signalled to the national framework groups that chalk streams should become a priority in terms of addressing those deficits.

Thus far, however, the Environment Agency has inclined to stand back – at least publicly – from guiding the decision-making that will be needed to apportion those abstraction reductions strategically and cost-effectively over time. This is being left to regional groups and partnerships, but it is not yet clear how these decisions will actually be made or if they will be consistent and logical.

Our CaBA chalk stream strategy called for the collaborative development of a prioritisation process and while everyone agrees the need, it still hasn’t quite happened.

In all of the meetings I have ever attended in which abstraction reduction is discussed the idea is aired that we will need evidence to justify and ensure wise decision-making, including evidence that ecological gains will follow mooted abstraction reductions.

On the face of it, the call for evidence seems only prudent and sensible – after all public money is at stake. 

But the idea that such evidence could ever exist is a chimera.

Clear cause-and-effect evidence according to a robust, before / after / control / impact method of proof, is – I argue – impossible to acquire. And it is so, precisely because of the infinitely complex web of cause and effect that is leant upon to justify the call for evidence.

Water companies will argue, and rightly, that reducing abstraction is expensive and, therefore, that there’s no point doing it if no benefit follows. Or if the potential benefit is neutered by some other factor such as a heavily modified river channel, or pollution from farms.  In certain settings we run a real danger of spending millions reducing abstraction, when other factors – like the fact that the river is navigable and impounded by locks and weirs – are as big or even a bigger brake on the ecological health of the system from which there is no possible relief. 

There are also settings where the cost of reducing (some of) the abstraction could be more cost-effectively spent (in terms of ecological gain per buck) improving the physical habitat. We have transformed canal-like channels in Norfolk into vibrant, wild and free-flowing streams for modest amounts of money, all things considered. £200,000 per km is dwarfed by the £4 million cost of replacing 1Ml/d of water. For that you could rebuild 20 km of knackered chalk stream.

I’m all for the intelligent and undogmatic trade-offs and counter-intuitive thinking that will be needed if we really are to balance the needs of society and nature.

But the call for evidence is self-deluding at best and a delaying tactic at worst.

Why? 

In the insanely busy and pressured landscapes we are talking about it is virtually impossible to strip out the variables: the physical condition of the channel; the micro and macro stressors of water quality which are highly complex and some of which we barely understand; shifts in the global agricultural markets which might generate or ease an agricultural pressure beyond one’s control or easy quantification; road run-off which might be terrible in a year when a local farmer is rearing pigs, or not too bad when the farmer gets rid of the pigs or in a mild, dry winter; the weather; the climate; the impact of invasive species like signal crayfish, or predation from cormorants when a cold winter forces them off the reservoirs: etc. etc. etc.

I defy anyone to design an experiment into the teeth of those variables, that could possibly isolate the beneficial or non-beneficial impacts over time of one single action. 

The only way you could construct such an experiment would be to select a stream where abstraction is pretty much the only pressure and a significant one, gather baseline data for at least five years, ideally a decade and then COMPLETELY TURN OFF THE ABSTRACTION in that stream and all nearby streams (because you need the signal to be significant to rise above the variables you can’t eliminate no matter how hard you try) and study for another five to ten years. The study periods would have to either equally include or exclude periods of drought and very wet years too. 

When helping to write Defra’s now scandalously abandoned chalk stream recovery pack, I was looking for exemplar case studies of where abstraction reduction had made a significant and demonstrable beneficial impact to ecology. I struggled to find a slam-dunk example, mostly because the reductions that have been made – though significant – have been made from very high totals and are actually quite small against the volumes still abstracted. For example, on the River Ver, while abstraction once exceeded 50% of recharge, it is still 30% of recharge.

I cited the River Piddle in the end, even though the changes made there since the dark days of the late 20thC when the river dried up regularly, include flow augmentation as much as abstraction reduction. The Piddle, however, is indeed much better now than in 1989 – 93. I know because I’m lucky enough to co-own the bit that used to dry up and it is now an exemplar of chalk stream health. It is an example.

But even so, water companies and others will often say abstraction reductions made thus far haven’t yielded the hoped-for results. Either in terms of flow or clear ecological gains.

In terms of flows, this is not true. John Lawson’s analysis of the flow-recovery following abstraction reductions shows unarguably that flows do recover in proportion to abstraction reduction. But when the abstraction is really high and you only reduce it a bit … hmmm.

John’s report also shows that the reductions made, though significant and expensive in water resource terms, have been far too small relative to the size of the overall catchment abstraction and far too small to rise above the “noise” made by all the other variables (and many of these variables haven’t been attended to properly, either. We still have a lot to learn about high-quality and cost-effective process-based habitat restoration).

In the only really good long-term BACI type flow scenario that exists, flows on the River Ver reduced and then recovered exactly in sync with the abstraction increase and then reduction. Of course they did: where else would the water have gone to?

So, if slam-dunk cause and effect evidence that reducing abstraction Y will lead to X ecological recovery doesn’t exist and can’t be found, how do we approach the problem? 

Without knocking the idea that data and evidence are useful tools to guide our decision-making, we should not abdicate our own common sense. In the same way that we don’t need science to tell us that it’s warmer in summer (though we need science it to tell us why) we don’t need science to tell us that abstraction adversely impacts the ecology of a river (though we do need it to tell us why).

Many fine minds have spent a lot of time discussing and agreeing that sustainable abstraction in chalk streams should generally cause less than a 10% reduction in natural low flows, (which also, give or take, amounts to 10% or less of the average aquifer recharge). That is why we have the Environment Agency’s Environmental Flow Indicator, which is based on the UK Technical Advisory Groups deliberations on exactly this flow / ecology balancing act.

What we need beyond this work (that has already been done!) is not so much more impossible-to-find evidence but rather a screening process that aids and brings logic, common sense and consistency to the thorny issue of how to spend public money most cost-effectively in our collective goal of achieving sustainable abstraction on chalk streams. 

To give a really obvious example: we need a screening process that stops us spending billions of pounds reducing abstraction in a river that is navigable and therefore doesn’t really have a flow-dependent ecology, but compels us to crack on with spending millions of pounds reducing abstraction in iconic chalk streams which can also be physically restored for 200k per km!

That really shouldn’t be too difficult.

I have made a start below … comments welcome.