Checking for ground nesting birds before harrowing.
Changeable Furlongs was mown for silage and hay was made on the Neyotts meadows and parts of Long Mead by Luke and George (once students at Bartholomew School), who use it to feed their cattle and sheep on the farm that they have taken on from their ‘Gramp’ in Wootton. (Changeable Furlongs are the fields beyond the Wharf Stream are leased from Smiths of Bletchingdon).
Bales of hay on Long Mead. The remaining standing patches of meadow were harvested as 'green' hay.
The next task was to take the first step in creating a new wildflower meadow in Changeable Furlongs for our wildlife and for us all. You may have joined in the hedge-planting two winters ago. You may have joined the dock-pulling in the eastern field the following spring. You may even have been part of the recent team of NRNers who walked Changeable Furlongs with a rope between them to check for skylarks’ nests before the big machines arrived: those of you walking along the Wharf Stream Way would have noticed the recently harrowed field.
Disc harrow on Changeable Furlongs.
The first step towards creating a new meadow seems deceptively simple: harrow a field and sprinkle it with wildflower seed, then stand back and watch it grow. In practice, most attempts to make wildflower meadows fail – research from the Open University in 2020, that looked at 167 floodplain meadow creations undertaken during the last 30 years, found that the failure rate was a disappointing 75%. This is unfortunate, because the seed is not only rare and expensive, but the actual work required to do the initial seeding takes considerable planning, cost, effort, and skill in execution. With every failure our rare habitat continues to decline, as do the species that depend on them – some to the verge of extinction.
The fact is that while the general principles of meadow-creation are known, there are no detailed guidelines for creating a new floodplain wildflower meadow, and many contractors are not familiar or practiced with the techniques or hazards, so most attempts are carried out in a state of optimistic ignorance.
At Long Mead we are in the fortunate position of having an ancient floodplain wildflower meadow that contains one of the rarest plant communities in the UK. Only a total about 4 square miles remains in the UK., mostly in small fragments. The plant community is classified as the Alopecurus pratensis-Sanguisorba officinalis community and is usually located on a floodplain. It is a plagioclimax community, which means that the plant community is stable, because its natural succession - to scrub and then woodland - has been arrested by human activities, in this case its use for agriculture. Over the centuries this activity has produced in one of the most species-rich areas in the UK and this biodiversity is the inspiration for the Thames Valley Wildflower Meadow Restoration Project (TVWMRP), which aims to create a continuous riverside corridor of species-rich floodplain meadows along this stretch of the Thames.
The Changeable Furlong field to be restored. Wharf stream on the left.
Through baseline and follow-up surveys of a number of restorations over the past years we have been able to accumulate useful data to guide the project ad optimise the methods used. The target for restoration this year is a 5.6 hectare field that is one of the fields known as 'Changeable Furlongs'. It once was a hay meadow, but was ploughed up for arable crops after the 2nd World War. The meadow flora was destroyed for crops, but the arable field has been uncultivated in recent years. Fortunately, the analyses of our soil samples showed the fertility to be relatively low, (i.e. its phosphate and nitrate levels are low enough for it to be suitable for creating a wildflower meadow). High levels of phosphate, in particular, enable grass species to flourish, making it hard for the wildflowers to establish from seed. Our baseline botanical surveys of the field indicated that it was relatively species-poor and was dominated by Meadow-foxtail grass. This condition was the result of it producing a monoculture crop of oil-seed rape for many years. Significantly, the field floods, which makes it problematic for arable crops, but creates the ideal conditions for the rare plant community that thrives just upstream on Long Mead.
Planting the hedge along the fence of the field to be made into a meadow.
In January 2024, NRNers together with OCV planted 240m of hedge along the new fences that Smiths put in after they bought the farm. This young hedge is flourishing and, as it matures, it will form an important habitat for insects, reptiles, birds, and small mammals. Since we took on the lease, we have put the two fields that form Changeable Furlongs back into meadow management. That means taking a hay cut in July and then grazing the aftermath with cattle or sheep in the Autumn. This is the cycle that has been repeated on Long Mead for over 1000 years. It ensures that fertility is kept low, since the only fertiliser that goes onto the meadow is silt, which is trapped by perennial plants during Winter floods, and the dung and urine deposited by the livestock in the Autumn.
In order to gather objective data to guide future floodplain meadow creation by ourselves and others, we have adopted an experimental approach; we are leaving the east riverside field of Changeable Furlongs to ‘restore itself’ by putting it back into the traditional hay-meadow regime of management. Making hay and grazing the aftermath should keep the grasses and more vigorous plants in check, giving flowering plants the best chance of re-establishing. We will compare this with the more interventionist method of meadow creation (harrowing and seeding with green hay from another site), which we are adopting on for the west riverside field of Changeable Furlongs.
Even here, we are trialing different methods of preparation. One of the key challenges in creating new meadows is dealing with competitive species such as docks and thistles. The conventional method is to spray these off with a herbicide like glyphosate before you start. However, the harms of herbicides, particularly in our floodplain environment, where fertiliser, pesticides and herbicides wash off into the river, are now all too well-known. Although many environmental organisations still use herbicides to prepare the ground for meadow creation, we use a traditional method of dealing with invasive species, taught to us by Graham Podbery (whose grandparents used to own the farm where the Eynsham Post Office and Evans Road now stand). Before the invention of modern herbicides, farmers used to run over their fields with a plough, turning up the roots of perennial weeds to burn them off with the sun. You would do this several times through the season – ‘keeping the field moving’ or ‘fallowing it’ as Graham describes it.
To study the consequences of fallowing vs. not fallowing, we disc-cultivated half of the field twice at 2-week intervals, with the heatwave doing its job in scorching the exposed roots.Then, just before spreading green hay, the entire field was disc-cultivated to create a fine tilth. In more solid clay soils, we would use a power-harrow after the disc-cultivator to generate a finer tilth. As luck would have it, rain intervened on the day we had planned the green hay harvest and spreading, and we had to delay the procedure to the next fine day.
Close-up and Professional: Tim and Will foragiing in perfect formation.
There are two basic methods to obtain the seeds for spreading. One is to use a brush-harvester to gather the ripe seeds, which are then distributed with some sort of spreader. This requires specialist equipment and may not capture the smallest seeds. How much seed is lost by the mechanical disturbance of the brush-harvester is also unknown.The other is the ‘green hay’ method, where the standing plants of a species rich donor meadow are harvested with a forage harvester, which chops the seed heads, leaves and stalks into small pieces and blows them into a muck spreader that is then used to distribute the green hay over the prepared recipient meadow. This method uses standard farm equipment and captures all sizes of seed because the whole above-ground plant is harvested, so it is the method we use. One disadvantage is that as soon as the plants are harvested, they start to compost and heat up, so the operation is time-critical. As with the brush-harvester, however, the seed content is a snap-shot of what is ripe on the day of harvesting, so neither method can capture the full array of seeds from the 120 plant species on the ancient donor meadows like Long Mead. This is the reason we hand-propagate the rarer, slower-growing species using seed from Long Mead and plant them out in the restoring meadows.
Modern indirect-cut forage harvester.
Early versions of trailed forage harvesters are ‘direct cut’, i.e. they mow the standing hay, shred it, and blow it out in one operation. Modern trailed forage harvesters, however, do not mow. Instead, in a separate operation the meadow has to be mown and then the forage harvester has to pick up the mown hay from the ground and shred it. This indirect method inevitably means that some of the seed falls on the ground during the mowing and is effecively lost from the green hay.
Direct- cut forage harvester trailed by a classic Massey Ferguson 188 tractor.
After years of searching, Long Mead has now managed to source a veteran direct-cut forage harvester (a ‘Wilder Twin-chop’ made in Wallingford), giving us the opportunity to compare the extent of seed loss with modern indirect harvesting. For our experimental design, half of Changeable Furlongs received green hay from the direct-cut forage harvester and the other half received green hay from the indirect cut harvester; each covered equal areas of fallowed and un-fallowed field, thus giving 4 areas that received different treatments. (Watch this space to see how these different treatments influence the outcome).
The muck-spreading process itself can be tricky: if you spread it too thick, the green hay will turn to compost, if you spread it too thin and there will be insufficient seed. From experience, we find that the green hay should cover the ground sufficiently to give it a lace-like appearance, with the dark tilth still visible though a mesh of green. This typically requires several passes of the muck-spreader, with someone on the ground to monitor the spread and guide the tractor driver.
At Long Mead we selected areas that had the most diversity of ripe seed and included a generous number of yellow rattle plants –one of the rare annual plants called ‘the meadow maker’ because it is hemiparasitic and reduces the vigour of grasses. These separate areas were cut in elegant curves to maximise capture the yellow rattle, although this created real problems for the later stages, where straight paths are best for the efficient harvesting of green hay by the two machines operating in close proximity.
The remainder of the meadow was baled for hay. The ratio of areas of the donor meadow to the recipient field is an inexact science and it varies according to the prevailing conditions. Previously we have had success in the range of 1 part donor area to 2-4 parts recipient area. This year, due to the lack of rain, the sward of Long Mead was low, which meant the ratio of seed to stalk and leaves was higher than in most previous years, but we opted to use 1.85 hectares worth of Long Mead green hay to cover the 5.64 hectares of Changeable Furlongs. (In case you are wondering about the second decimal, it comes about because the NRN/Long Mead team deploy a special high resolution GPS survey instrument to measure the areas and monitor our rare plants – thanks to a generous grant from our County Councillor, Dan Levy).
Old and young, veteran and contemporary, operating in tandem to harvest the green hay.
The decision to use two different forage harvesters added considerably to the complexity and choreography of the operation. Our team included Matt and Will, two of Chris Strainge’s farmers from Peashell Farm, who took a break from combining to drive the mower, indirect forage-harvester and the muck spreader. Tim Cox was 'volunteered' to operate the veteran direct-cut forage harvester. We were uncertain as to how the veteran harvester would perform and we knew it could not be operated with the huge modern computer-controlled tractors that were used for the indirect harvester. Fortunately, Tim, whose tractors carry the Carnival Queen at Eynsham’s Carnival, agreed to bring his own classic – a Massey Ferguson 188 - which, by coincidence, had been owned by the late Norman Butler-Miles who had used it to make hay on Long Mead in the early 2000s and in the Neyotts in the 1950s and 1960s. After some coughs and splutters and t.l.c. the Wilder Twin-chop was nursed into action by spanner-wielding Tim and Matt and soon Will’s muck-spreader was collecting the stream of green hay emitted from the spout.
Spreading the green hay: first pass.
When the first muck-spreader was filled to the brim, Will chugged off to Changeable Furlongs to spread its load under the watchful eye of Long Mead’s Kevan. After two loads had been thinly spread on about 2.5 hectares, we decided we had to reduce the turn-around time by packing more into the muck spreader. This was achieved by 3 large men jumping on the load to compact it. A further two loads produced the desired lace-like appearance.
Matt mowing the hay for the indirect forage harvester to pick up.
By then it was already late afternoon, but Matt and Will were keen to complete the job, so now it was the turn of the indirect forage harvester, which quickly picked up rows of hay mown earlier by Matt. The mowing was closer to the ground than the Wilder Twin-chop, so produced more material, but since all the loads were compacted by jumping, the remaining half took 3 loads. We were interested to see whether the yellow rattle pods still contained seeds after the mowing. Most pods had shed their seeds, but a few stalks still rattled, so we could be sure that some yellow rattle seeds were still picked up by the indirect-cut forage harvester.
Matt and Kevan checking whether any yellow rattle pods still have their seeds after mowing (some did).
In the gathering gloom, it was not possible for anyone on the ground to assess the coverage, so Will did his best in the headlights of the tractor. By 23:00 the last load was spread and we all headed home.
Rolling to get the seed in contact with the soil.
All that remained to be done was to roll the field to ensure good contact between the seed and the soil. We hoped to do this the next morning, but again rain intervened. Will rolled it a few days later, which gave a chance for our stalwart Monday and Wednesday Meadow Restoration Groups to scatter yellow rattle seed by hand in the gaps inadvertently left by Will when he spread the green hay in the Friday evening darkness.
Landscape-scale meadow creation with all its big machines, precise timing, and host of improvisations on the day, is not for the faint-hearted or inexperienced. Gratifyingly, the Thames Valley Wildflower Restoration Project is bucking the dismal statistic of failure in 75% of attampts. Baseline and follow-up botanical surveys are critical if we are to optimise the methods. Along with the botanical surveys we are fortunate in having a team of entomologist studying the insect communities in the meadows under restoration and the ancient meadows to understand the dynamics of the fauna as well as the flora.
Many many questions remain to be answered, but the tabula rasa of Changeable Furlongs has given us an opportunity to create something rare and special, something that enhances the biodiversity our local Nature, that connects and protects our floodplain meadows, and that involves the community in something that gives us joy, and last, but not least, something that advances our understanding of the key factors that influence the successful restoration of the rare plant community that floodplain meadows support.
Hand-broadcasting yellow rattle seed mixed with kiln-dried sand to give heft and visibility.
Kevan AC Martin