CAHL1Pollen and nectar flower mix

What it actually is

You sow a flower mix in a block or a strip and keep it flowering from late spring right through the summer. GOV.UK sets that in the context of providing food for pollinators and for the insects that predate crop pests.

The mix has to be grass-free and hold at least six flower species, with no single species making up more than half of it by weight, and it must include at least two from common knapweed, musk mallow, oxeye daisy, wild carrot and yarrow.

It goes in between early spring and early autumn, within twelve months of the action's start date, and it is maintained until the end of the second summer after sowing and then re-established.

Source: GOV.UK, CAHL1 action page, checked 04/08/2026.

What it pays, and what that comes to

The rate is in the panel at the top of this page. Multiplied out:

Source: arithmetic on the payment rate shown at the top of this page, which comes from Tilth's copy of the GOV.UK action catalogue and carries its own last-checked date. The 4 hectares is an illustration.

Is it worth it on your ground?

Step one: what it costs to put in. Seed, and it is the biggest establishment cost of any action on this site. Three mixes sold for this action, with the merchant's own listed prices and stated sowing rates:

Source: Boston Seeds, own listed prices and stated sowing rates for mixes sold for CAHL1, looked up 04/08/2026. The per hectare column is arithmetic on those two figures. Price your own mix. Cultivation and drilling are your own cost and are not in the table.

Step two: what that comes to a year. A three-year mix sown once and spread across the agreement, against a one-year mix sown every year:

Source: seed costs as above, divided across three years where the mix is a three-year one. The final column is the payment rate at the top of this page less the seed cost, and is subtraction.

A caution on the three-year mixes. The action requires the mix to be maintained until the end of the second summer after sowing and then re-established. Where a re-establishment falls inside your three years, a three-year mix is bought twice rather than once, and the seed cost roughly doubles. On the standard mix that moves it from £29 to £38 a hectare a year up to £58 to £77, and the worth per hectare per year down to £662 to £681. Which it is depends on when your agreement starts against when you sow, so work it on your own dates rather than ours.

Source: the re-establishment requirement, GOV.UK CAHL1 action page, checked 04/08/2026. The rest is arithmetic on the seed prices above.

Step three: what the ground would otherwise return. Gross margins per hectare for England, non-organic, in the 2023/24 crop year. Variable costs have already been taken off these figures, so there is nothing further to deduct.

Source: Farm Business Survey, Crop Production in England 2023/24, Rural Business Research, England, non-organic, 2023/24 column, checked 02/08/2026. The winter wheat figures come from a sample of 572 farms.

The agreement runs three years, so the comparison has to run three years too, and that ground will not be in wheat for all three of them. Three rotations, worked from the crop figures above:

Source: arithmetic on the crop gross margins in step three, from the Farm Business Survey, Crop Production in England 2023/24, checked 02/08/2026. These three are illustrations. Work your own out from the same crop figures.

Step four: the gap.

Source: the after-seed figures from step two, set against the rotation averages in step three. The remaining figures are subtraction.

Three things to set against those figures. What follows is Tilth's own reading rather than scheme guidance.

The survey reports one winter wheat figure, not first and second wheat separately. A rotation with two wheats in it carries a second wheat at the same gross margin as the first, which is unlikely to be how it turns out on your ground. We have not found a published split to cite, so we have not applied a discount of our own invention. Read the rotation averages as the generous end for the crop.

It is one year, and not a strong one. The same report's 2022/23 column put winter wheat's gross margin at £1,570 a hectare against the £988 above for 2023/24. A three-year decision resting on a single crop year is a shaky thing, and this particular year was the lower of the two.

Establishment is a cost the table does not carry. The seed price is in it. The cultivation pass and the drilling are not, because they depend on your own kit and your own costings, and we are not going to put a number on them for you. They move the gap against the action rather than for it.

Where that leaves it. On ground performing at national rotation averages the crop is ahead, by something in the region of £54 to £187 a hectare a year depending on which mix you buy. That is the narrowest gap of any of the flower and grass actions on this site, and it is narrow enough that the mix you choose matters as much as the field you choose. Buy the cheapest three-year mix that meets the species rules and the gap is close to nothing on average ground and gone entirely on ground that returns below your rotation average. Buy the enhanced mix and sow the smaller area, and you are paying for a better stand out of your own margin.

Does it work, and what makes it work better

This section is what we have found in published sources about the practice this action pays for. It is not scheme guidance, it is not Defra's or the RPA's view, and it adds no requirement beyond what the GOV.UK action page sets out. We have named every source and given the date we checked it. What it means for your ground is your decision.

This is the best-covered action on this site. It is also the one where the published work and the scheme wording pull slightly apart, so that is worth saying before the numbers.

CAHL1 asks for a grass-free mix. What has been studied is nectar flower mixtures and wildflower strips, which Conservation Evidence describe as mixes that may include agricultural varieties of flowering plants such as clovers. The two strongest UK studies in the set both sowed their margins at 80 per cent grass to 20 per cent wildflower by weight, and a fair slice of the earlier continental work tests a strip of phacelia, or a commercial mix built around phacelia, rather than a diverse flower mix. The evidence reaches this practice, and reaches it better than it reaches anything else on this site, but the version of the practice it mostly measures has grass in it. Three studies looked directly at what taking the grass out costs, and they are below.

The graded assessment. Conservation Evidence, based at the Department of Zoology at the University of Cambridge, summarises the published studies behind individual conservation actions and has independent subject experts score each one. This practice sits in the top category, Beneficial, on 104 studies, with 100 per cent for effectiveness, 75 per cent for certainty, and harms not assessed. Effectiveness scores only the size and direction of the effects that were reported, and Conservation Evidence say themselves that it does not consider the quantity or quality of the studies, so a single poorly designed study could produce a high effectiveness score. Certainty is the separate score for how many studies there are, how well built they are and how much ground they cover. Those are the highest scores attached to any action we have written up. What sits underneath them is in the next paragraph, and it is not all one way.

Source: Conservation Evidence, action 442, Plant nectar flower mixture/wildflower strips, Farmland Conservation synopsis, checked 06/08/2026. Cite as Dicks et al. (2020), Farmland Conservation, pages 283 to 321, in What Works in Conservation 2020, Open Book Publishers.

What the studies found, both ways. Eighty studies looked at biodiversity, and 64 of those found benefits to one or more groups. Of 65 studies on invertebrates, 50 found benefits, 15 found mixed or negative effects and 6 found no significant effect. Of 21 studies on plants, 16 found benefits and 4 found negative effects or none, with the remaining 5 comparing establishment methods rather than outcomes. Of 7 studies on birds, four replicated controlled trials and a review found benefits, and skylarks foraged and nested in or near sown patches, but barn owls in a Swiss study significantly avoided sown wildflower areas, and both winter recording periods of one Warwickshire trial found no more bird species or individuals than on the cropped control. All 5 small mammal studies found benefits. Nineteen studies found benefits attached to particular sown species, phacelia, borage and red clover among them, while 3 replicated studies found negative effects or none from phacelia. Conservation Evidence say themselves that studies are not directly comparable or of equal value, and that study size, design, reported metrics and relevance to your own situation matter more than "simply counting the number of studies that support a particular interpretation". The individual studies run from 1989 to 2010, so every one of them predates this scheme and this action.

Source: Conservation Evidence, action 442, key messages and supporting evidence from individual studies, checked 06/08/2026.

What we have not found. Nothing in the set measures a grass-free mix built to CAHL1's species rules, so nothing here tells you what dropping the grass does to the whole picture rather than to the three groups measured below. We have found no study of the six-species minimum, the 50 per cent by weight limit or the five named species the action asks you to pick two of. And we have found nothing measuring CAHL1 itself.

What the sources say makes it work better. All of the following comes from the studies listed on that Conservation Evidence page, and it is a description of what was found rather than advice from us.

Start with the grass, because it is the part of the specification that differs. Marshall, in the UK, found wildflower-mix margins carried fewer grasshoppers and crickets, under four individuals of a single species a margin, than 2 metre grass-and-flower margins, which held ten individuals of four species; the wildflower margins had the lowest grass cover of any treatment, under 60 per cent against 100 per cent for the sown grass-and-flower. Woodcock and others found a mix of mainly flowering plants with no tussock grass in it carried fewer beetles and fewer beetle species than both a tussock-grass-and-forbs mix and a grass-only Countryside Stewardship mix. In a later paper the same group found more herbivorous beetles where wildflowers were in the mix, but more predatory beetle species and individuals in any margin containing tussock grass, wildflowers or not. Pywell and others, in the pollinator arm of the SAFFIE trial, found a flower mix designed for pollinating insects supported no more butterflies or bumblebees than a floristically enhanced tussocky grass mix, though both beat grass alone by a distance, at 35 to 47 bumblebees of four species and 18 to 20 butterflies of six species per 125 square metres against ten bumblebees of two species and twelve butterflies of five species on the grass-only margins.

Then what goes in the mix, which is where the largest effects in the whole set sit. Pywell and others, surveying 32 grid squares across England, counted 86 bees a transect on pollen and nectar margins and 43 on wildflower margins, against 6 to 8 on grassy margins and 0.2 on the cropped control. Heard and others recorded 26 bumblebees per 200 square metres on sown forage patches against 2 on controls. Carvell and others, over six sites and four years, found pollen and nectar margins supported significantly more foraging bumblebees than cropped, grassy or naturally regenerated margins. Pywell and others found the result depended on the key forage species being present, naming red clover, bird's-foot trefoil and borage; in a related study 70 per cent of the pollen collected by buff-tailed bumblebees came from borage and 76 per cent of that collected by common carder bees came from red clover, with short-tongued bees preferring the annual mix and long-tongued bees the perennial one.

The timing within the mix came out of the same work. Carvell and others found wildflower margins carried around 4,200 forage flowers a plot in May and June falling to 2,000 in July and August, while the pollen and nectar mix ran the other way round, few flowers early and around 9,000 in July and August. They also found native red clover and bird's-foot trefoil flowered earlier than the agricultural varieties of the same plants. Their review of the two mix types recommends a pollen and nectar mix where the aim is a rapid effect on numbers, and a wildflower mix where the aim is a wider range of species across the season.

Phacelia is the clearest case of a mix that lifts the count and narrows the range. Steffan-Dewenter and Tscharntke found around 10 wild bee species on phacelia set-aside against about 27 on two-year-old naturally developed set-aside, with fewer red-listed and specialist species; in a later study they found similar bee numbers but fewer species, 13 a field against 15 to 29, with the bees on phacelia mainly common bumblebees and furrow bees while the endangered and specialised bees were on the naturally regenerated ground. Gathmann and others found phacelia set-aside attracted many honey bees but no cavity-nesting solitary bees nested in trap nests placed there, against 12 species in naturally regenerated set-aside. Greiler found phacelia set-aside had the lowest plant richness and the lowest invertebrate numbers of the treatments compared. Against those, Carreck and Williams found two commercial mixes containing 25 and 40 per cent phacelia attracted 14 bee and wasp species including all six common UK bumblebee species, with phacelia taking 87 to 99 per cent of bee visits while the buckwheat in the same mixes took 1 per cent or less of bee visits but up to 36 per cent of hoverfly visits. Pywell and others recorded 32 to 85 bees a plot on phacelia and borage plots against 1 to 22 on the others, and in another trial found butterflies highest not on those but on lucerne or red clover.

Cutting is the part of the management the studies are clearest on, and the pattern is consistent. Feber and others found 14 to 52 meadow browns per 50 metres on sown margins left uncut in summer against 4 to 22 where a summer cut was taken; in a later study they found summer cutting reduced butterfly diversity and density while spring and autumn cutting did not, and that although a summer cut knocked flowering back immediately, by September the cut margins carried more flowers than the uncut. Haughton and others found cut plots had significantly lower numbers of all invertebrates, spiders, true bugs and leafhoppers than uncut, that two cuts a year beat one, and that spring-and-autumn cutting gave the higher numbers. Baines and others found sown margins held more spiders but not more spider species, and that cutting, summer cutting in particular, significantly reduced abundance. Bell and others counted 95 pseudoscorpions in unmanaged plots against 19 to 53 in cut plots, the worst being the spring-plus-summer cut. Marshall and Nowakowski found a single June cut significantly reduced both sown and unsown plant diversity. Smith and others found spring-and-autumn cutting raised the number of plant species while other treatments fell.

The other side of cutting is what it does for the stand over time. Smith and others, following Oxfordshire margins from 1987 to 2000, found plots cut twice a year retained 50 to 60 per cent of their sown species against under 40 per cent for plots cut once or not at all.

Which matters, because this action runs three years and the stands in these studies did not hold. The same Oxfordshire work recorded plant species per quadrat falling from 23 or 24 in 1988 to 9 to 12 by 2000, with the sharpest loss in the first two years as the annuals went. Carvell and others found flower numbers in the pollen and nectar mix declined in year three while the wildflower mix held steady. De Cauwer and others, in Belgium, found sown plots fell from 22 or 23 species a plot in 2002 to 13 to 16 by 2004, by which point they resembled the naturally regenerated plots. Gardiner and others watched alsike clover peak at about 33 per cent cover in 2002 and fall to 2.5 per cent by 2004, and red clover peak at about 85 per cent in 2003 and fall to 20 per cent in 2004, as perennial grasses came in during years three and four. Edwards, surveying 41 farms, reported a decline in foraging bumblebee workers on legume-rich margins established more than three years, without statistics attached.

On fertiliser and weeds, which the action also speaks to. May and Nowakowski found fertilised wildflower margins carried fewer plant species than unfertilised ones, around 30 against 35 a metre. Kleijn and others found sown wildflower margins held more plant species than naturally regenerated or rye grass margins, and far less creeping thistle and couch grass, at 0.1 and 6 grams a square metre against 33 and 28 on the naturally regenerated. Jacot and others found problem weeds established in sown margins but did not come to dominate, and that mixes containing more than 75 per cent grass tended to produce a grass-dominated margin. Pywell and others found no difference in bumblebee use between ten different seed mixes, but that a grass-specific herbicide raised both bumblebee numbers and bumblebee species, taking grass cover from 45 per cent down to 2 per cent and sown wildflower cover from 24 per cent up to 56 per cent, while also taking bare ground from 4 to 16 per cent and undesirable weeds from 4 to 14 per cent.

And on where the patch goes rather than what is in it, Heard and others found bumblebee density was higher where there was more arable land within a kilometre, and that the size of the patch, at a quarter, a half or a full hectare, made no difference to it.

Source: Conservation Evidence, action 442, supporting evidence from individual studies, checked 06/08/2026.

What follows is Tilth's own reading of how those findings sit against this action, and it adds no requirement beyond what the GOV.UK action page sets out.

The 100 per cent effectiveness score is real and it is earned on pollinator numbers. Bees went from near nothing on cropped ground to tens per transect on sown flowers, in study after study, in England, in Scotland and across Europe. If the question is whether sowing flowers puts more bees on a field than growing wheat there, the answer in this set is not in doubt.

The grass finding is the one that sits least comfortably against CAHL1. Four studies, from three research groups, found something measurable lost when the grass came out: grasshoppers and crickets, predatory beetles, and in the SAFFIE trial no pollinator gain at all over a tussocky grass mix with flowers in it. None of them is a test of a grass-free mix sown to this action's species rules, and none of them found the flowers themselves did harm. What they suggest is that a grass-free block is a sharper instrument than a mixed one, better at the thing it is aimed at and worse at some of the things it is not. The action is called a pollen and nectar flower mix, so that may be exactly the trade the scheme intends. It is still a trade.

The phacelia cluster is the same shape of finding in a different place, and it is the one we would keep in mind at the merchant's counter. More bees and fewer kinds of bee is a real result, reported by three research groups, and the bees that went missing were the specialists and the red-listed ones. The action's six-species minimum, its 50 per cent by weight limit and its list of five natives all push against exactly that failure. Reading the studies, those rules look less like paperwork and more like the thing that decides what the block ends up worth.

The decline evidence bears on this action in a way it would not on a longer one. Five separate studies watched sown stands thin out, with the clover-heavy ones collapsing fastest and the sharpest losses in the first years. CAHL1 runs three years and then asks for re-establishment inside the agreement. On this evidence that requirement is not administrative tidiness, it is the point at which several of these trials had already lost most of what they sowed, and the seed table further up this page is where that lands financially.

The cutting evidence is the most useful thing here for anyone who has already decided to do it. Eight studies point the same way: a summer cut is the expensive one, and spring or autumn cutting costs little or nothing. One found two cuts a year held on to 50 to 60 per cent of the sown species against under 40 per cent where the margin was cut once or left alone. The action does not prescribe a cutting regime, it asks you to maintain the block in a way that can reasonably be expected to achieve its aim. Cutting twice, away from the middle of summer, is what these studies did when it worked.

What none of this settles is the money question further up the page. The gap between this action and a rotation average is narrow, and nothing in 104 studies puts a figure on the other side of it.

What you actually have to do

Establish a grass-free seed mix holding at least six flower species, with no single species more than 50 per cent by weight, including at least two of common knapweed, musk mallow, oxeye daisy, wild carrot or yarrow.

Sow between early spring and early autumn, within twelve months of the action's start date. Maintain the block or strip in a way that can reasonably be expected to achieve the action's aim, until the end of the second summer after sowing, and then re-establish it.

Source: GOV.UK, CAHL1 action page, checked 04/08/2026.

What you cannot do on it

No cutting or grazing in a way that prevents the action's aim being achieved. No fertilisers or manures. No pesticides, except herbicides used to weed wipe or spot treat injurious weeds, invasive non-native species, nettles or bracken. That is the whole list, and the method is named as well as the target. No regular vehicular access, turning or storage.

GOV.UK also advises avoiding cultivation and fertiliser within 2 metres of the centre of a hedgerow, and checking for nesting birds before cutting.

Source: GOV.UK, CAHL1 action page, checked 04/08/2026.

Where it can go, and the cap

Arable land, temporary grassland, fallow land and permanent crops below the moorland line, registered with an eligible land cover and declared with a compatible land use code. You can put it in as a block or as a strip. It is rotational or static, so you can move it in years two and three or leave it where it is.

Source: GOV.UK, CAHL1 action page, checked 04/08/2026.

CAHL1 is one of the ten limited area actions under SFI26. The rule is that the total eligible area you enter into them must not be more than 25 per cent of the total agricultural area of your farm. The other nine are CIPM2, CAHL2, CAHL3, CIGL1, CIGL2, WBD3, AHW7, AHW9 and AHW11. How the 25 per cent limit works.

Source: GOV.UK, SFI26 scheme rules and guidance, section 4.5, checked 23/08/2026.

Things worth watching

This section is Tilth's own reading of where this action is easy to get wrong. It is not scheme guidance, it is not sourced from Defra or the RPA, and none of it adds a requirement beyond what the GOV.UK action page sets out.

Grass-free is the word that catches people. A great many stewardship mixes on the market carry a grass component, and some of them are sold under headings that mention this action's older codes. The requirement is a grass-free mix. Read the species list on the bag rather than the name on the invoice.

The 50 per cent rule is by weight, and vetch is heavy. No single species may be more than half the mix by weight. Some perennial mixes lean hard on common vetch and sainfoin because they establish well and are cheaper by the kilo, which is precisely why the rule exists. It is a specification the merchant should be able to confirm in writing, and that written confirmation is worth having on the file.

Two of the five named species is a minimum, not a target. Common knapweed, musk mallow, oxeye daisy, wild carrot, yarrow. At least two have to be in there. A mix carrying a token amount of one of them meets nothing.

The re-establishment is inside the agreement, not after it. Maintain until the end of the second summer after sowing, then re-establish. That is a second sowing to plan and pay for, and it is the difference between the top and the bottom line of the seed table above.

No fertiliser, and pesticide only by weed wipe or spot treat. The only permitted pesticide is a herbicide, applied by weed wipe or spot treatment, against injurious weeds, invasive non-native species, nettles or bracken. Blackgrass is not on that list. On a block that has come out of a field with a blackgrass history, that is three years of living with the problem rather than three years of managing it.

Where you turn matters, again. No regular vehicular access, turning or storage. A flower block in a field corner and a place to turn the sprayer want the same ground, so pick the location with the machinery in mind rather than the map.

The 25 per cent cap is shared, not per action. CAHL1 draws on the same allowance as CAHL2, CAHL3 and the rest of the ten. Whatever goes into one is not available to the others, so total them together before the form is filled in rather than deciding them separately.

Keeping the evidence

Also Tilth's own reading. GOV.UK sets the requirements; what follows is a practical suggestion about record keeping.

GOV.UK gives field operations at parcel level, invoices detailing the seed mix used, and photographs as its examples of what to keep.

The invoice is doing more work on this action than on any other. It is the only document that shows what was in the bag, which is where every one of the species rules lands. An invoice that says "wildflower mix, 60 kg" proves the weight and nothing else. What is worth asking the merchant for, at the point of ordering rather than a year later, is the species list with the percentages by weight against it.

Tilth keeps the sowing operation against the parcel with the invoice attached to it, so the two are in the same place three years later.

The part only your own ground can answer

The gap above is worked on national averages, and on this action it is small enough that your own ground decides it. A block on your rotation average leaves the crop ahead. A block that quietly returns below it does not, and the seed table is then the whole argument. Which block that is sits in your own yield records rather than in a national survey. Tilth keeps them field by field, alongside the invoice that shows what you sowed. Request an invite.