Why we cut the meadow exactly once

People assume a wildflower meadow is what happens when you stop mowing. It is very nearly the opposite. Stop mowing a field in this country and in fifteen years you have scrub, in forty you have woodland, and at no point on the way do you have a meadow. A hay meadow is a crop. It exists because somebody takes one cut off it a year, at the same time each year, and then puts animals on it.

We began ours in 1994 with green hay — freshly cut, still-seeding hay carted from a surviving meadow eight miles away and spread on ground we had ploughed that autumn. It is slow, it is unreliable, and it is still the best method anybody has. Thirty-one years on, the 1994 block carries around thirty-four species per square metre. The block we did in 2016 carries nineteen. The difference is time and nothing else.

The date does the work

The single most consequential decision on ninety-eight hectares is which week we cut. Cut in June and the sward looks immaculate, the hay is superb, and the yellow rattle has not set seed — so next year there is less rattle, so the grasses are less suppressed, so there is less of everything else, and in six years you have a nice green field with no flowers in it.

Rattle is the hinge. It is a hemiparasite: it taps the roots of vigorous grasses and takes about a third of their vigour away, which is what lets the slower, smaller species get light. Everything we do is arranged around letting it seed. That means late July, and it means accepting hay that is coarser and lower in feed value than any dairy farmer would choose.

Cut a week early and the hay is better and the meadow is worse. That trade is the whole job.

Then the animals

Lifting the hay is only half of it. What actually returns seed to the ground is cattle, on from the middle of August at about one to the hectare, treading and dunging and opening up small patches of bare soil that seed can get into. Sheep will not do it — they graze too tight and too selectively, and they leave you a lawn.

They come off when the ground turns, usually late October. After that nothing happens at all until the following July. No feeding, no topping, no rolling, no fertiliser of any description. The temptation to intervene over the winter is considerable and it is always wrong.

What we got wrong

Two things. The first is that we spread green hay on three blocks in 1998 without ploughing properly first, because it was a wet autumn and we were behind. Those blocks are still the worst on the estate and the only fix is to start them again, which we will do in 2028.

The second is that for the first decade we cut the whole ninety-eight hectares in the same week. It is efficient and it is bad practice: it removes every source of nectar on the estate on a single day in late July, which for invertebrates is a cliff edge. Since 2009 we have left roughly one hectare in eight standing until September, rotating which. It costs us about four tonnes of hay a year and it is the cheapest thing we have ever bought.

The meadow will keep improving for another twenty years or so on current trajectory, and then it will hold. Nobody working here now will see the 2016 block reach what the 1994 block is at today. That is normal, on this estate, and it is not a complaint.