CNUM3Legume fallow

What it actually is

A fallow with a seed mix and a species requirement attached to it. GOV.UK's stated aim is that there is an established legume fallow with plants that grow and flower from late spring and during the summer months, and that provide green cover over the winter months. The page sets that in the context of managing nutrient efficiency and improved soil health, providing food for farmland wildlife such as pollen and nectar for pollinators and farmland birds, and supporting an integrated pest management approach.

On the established fallow, grazing with livestock is not permitted, the area must not be used for regular vehicular access, turning or storage, and cutting is permitted only to prevent blackgrass setting seed or to control other annual grass weeds.

Source: GOV.UK, CNUM3 action page, checked 02/08/2026.

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 8 hectares is an illustration.

Is it worth it on your ground?

Step one: what the seed costs, and how many times you buy it. The timing requirements mean the seed goes in twice, not once. You establish the fallow by the autumn within 12 months of the start date, maintain it at the same location for a period of time that can reasonably be expected to achieve the action's aim, which GOV.UK says will usually be until the end of the second summer after sowing, and then re-establish it by the following autumn.

Source: Boston Seeds' own listed price and stated sowing rate for the CSS6 Legume Fallow NUM3 mix, looked up 23/08/2026. The site displays prices with an inclusive and exclusive of VAT toggle and £125.00 was read in the inclusive state. The timing that produces two sowings is from the GOV.UK CNUM3 action page, checked 23/08/2026. Everything else in the table is arithmetic on those two figures.

Step two: what the action is worth after seed. Taking the annualised seed cost off the payment leaves between £448.67 and £469.50 a hectare a year, before cultivation and drilling.

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 payment rate at the top of this page less the seed costs in step one, 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.

The gap above leaves out what the fallow does for the crop that follows it. GOV.UK sets the action in the context of nutrient efficiency and improved soil health rather than putting a figure on either, and we are not going to put a figure on them either, because what they are worth depends on your soil, your rotation, your nitrogen prices and how you would have managed the ground otherwise. We have not found a published figure we would stand behind on a page about money.

Where that leaves it. On ground that performs like the rotation averages, the crop is ahead by something in the region of £295 to £387 a hectare a year, and the fallow has to make that up out of the following crop and out of whatever the block was doing badly in the first place. If the block was going to be fallowed anyway, or if it is the field where you already know the nitrogen goes furthest, the picture changes a good deal. On your best land it does not.

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.

The fit needs setting out before the findings, and on this action it matters more than usual. Conservation Evidence assess Plant nectar flower mixture or wildflower strips, which is a sown flowering mix left standing rather than cropped. That is what CNUM3 is, and a good deal of the work below is on legume-rich mixes specifically, which is closer still. Two differences run through it. The assessed practice is a strip, where this action is a block or a whole fallow. And a fair part of the work is on grass-free or grass-light mixes, where CNUM3 names cocksfoot and timothy among the options you may include.

The larger gap is what the evidence does not touch. GOV.UK leads this action's aim with nutrient efficiency and improved soil health, and puts pollen, nectar, birds and pest management after it. Conservation Evidence is a conservation database. Nothing below speaks to the nutrient or soil half of the aim, which is the half the action is named for.

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 comes out at Beneficial, their top category, on 104 studies, with 100 per cent for effectiveness, 75 per cent for certainty, and harms not assessed. Those two are not the same measure. Effectiveness scores only the size and direction of the reported effects, 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 score. Certainty is the separate score for how many studies there are, how well built they are and how much ground they cover.

Source: Conservation Evidence, action 442, Plant nectar flower mixture/wildflower strips, Farmland Conservation synopsis, checked 22/08/2026. 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 individual studies investigated the effects of flowering strips on biodiversity, and 64 of them showed some benefit to one or more groups.

Sixty-five looked at invertebrates. Fifty reported positive effects, including 41 studies across eight European countries finding higher invertebrate numbers, species richness or diversity. Fifteen reported mixed or negative effects and six found no significant effect. Twenty-one looked at plants: 16 found higher plant cover, flower numbers, diversity or species richness, four found negative or no effect, and five described establishment or management methods rather than outcomes. Seven looked at birds: four replicated controlled studies from Switzerland and the UK and one European review found a positive influence, with skylarks foraging and nesting in or near sown patches and less likely to abandon territories containing them, against one Swiss trial where barn owls avoided sown wildflower areas and two UK winter recording periods that found no more bird species or individuals than on control margins. All five studies on small mammals found benefits, in abundance, density or species richness.

On particular species, 19 studies reported positive effects from sowing named plants including borage and red clover, and three replicated studies found negative or no effects from sowing 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".

Source: Conservation Evidence, action 442, key messages, checked 22/08/2026.

What we have not found. Nothing on the nutrient efficiency and soil health half of this action's stated aim. Nothing measuring CNUM3 itself. Nothing comparing a block or a whole fallow against a strip of the same area. And no published figure for what the fallow is worth to the crop that follows it, which is the number the money section further up this page also declines to invent.

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 what happens to a legume mix over time, because it is the finding that lands hardest on this action. Gardiner and others, on arable margins sown with a legume and grass seed mix and followed from 2001 to 2004, recorded almost complete cover of alsike clover and red clover a year after establishment, and more bumblebee forage plant species than on naturally regenerated margins. Then it went backwards. Alsike clover fell from a peak of about 33 per cent cover in 2002 to 2.5 per cent in 2004. Red clover peaked at around 85 per cent in 2003 and fell to 20 per cent in 2004. The clover-sown plots were invaded by perennial grasses including false oat grass in the third and fourth years, as the clover cover dropped away.

Edwards, working across 41 farms in England and Scotland from 2005 to 2008, found more worker bumblebees on margins sown with a legume-rich mix than on grassy margins or track edges, and recorded a decline in foraging workers on the legume-sown margins once they had been established for more than three years. That came from five farms and no formal statistical analysis was done on it.

Then what to put in the mix. Greiler, comparing up to 65 arable sites in Germany over three years, found invertebrate numbers from suction samples lowest on set-asides sown with phacelia at around 500 individuals per 5 square metres, intermediate on naturally developed set-asides and cereal fields at around 1,000, and highest on set-asides sown with clover and grass mixes at around 1,500. Malaise trap catches were also highest in rye fields and clover-grass mixes. The same study found plant species richness running the other way, lowest in sown set-asides at 10 to 15 species per 49 square metres and highest on naturally developed set-asides at 37 to 45.

Phacelia comes out poorly more than once. Gathmann and others found newly sown phacelia set-aside attracted many honey bees but no cavity-nesting solitary bees made nests in reed stems placed there, while 12 bee species nested in stems placed in two-year-old naturally regenerated set-aside in the same study.

Then cutting. Feber and others, across four years in Oxfordshire, found more adult meadow brown butterflies on margins left uncut or cut in spring or autumn than on margins cut in summer, at 14 to 52 a 50 metre plot without a summer cut against 4 to 22 with one. In a companion paper they found summer cutting reduced butterfly diversity and density while spring and autumn cutting did not, and that a summer cut caused an immediate drop in flowering plants, though by September the cut margins carried more flowers than the uncut ones. Marshall and Nowakowski found a single cut in June significantly reduced both sown and unsown plant diversity, at two sown species a quadrat against four.

That same study is the one that mentions blackgrass. Unsprayed plots were dominated by it, and it was eliminated by a herbicide application in December.

One finding bears on the pest management part of the aim. Hausammann, working in a winter rape field in Switzerland, found lower numbers of pollen beetles and cabbage weevils near a sown weed strip than near an ordinary field boundary, at least early in the season, with no difference in the abundance of predators and parasitoids between the two.

Source: Conservation Evidence, action 442, supporting evidence from individual studies, checked 22/08/2026.

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 requirement on this page that looks like an unnecessary expense turns out to be the one the research supports, which is not something we have been able to say much on these pages. GOV.UK makes you re-establish the fallow before the third year, and that second sowing is the largest single cost in the money section above. Two independent pieces of work found a legume mix collapsing on exactly that timetable: alsike clover down from 33 per cent to 2.5 per cent cover in two years, red clover from 85 to 20, perennial grasses moving in during years three and four, and bumblebee use falling away once the margins had been in more than three years. Read that way, the second seed bill is not the scheme being awkward. It is the scheme rebuilding the thing before it fails.

Alsike clover is worth a second look when you choose the mix. GOV.UK names it among the legume options, and it is the species that fell hardest in the study above. That is one study on one set of margins rather than a rule, and we would not tell you to leave it out, but it is the sort of thing worth knowing before you read a bag label.

The grass question is one this action leaves to you and the evidence answers in two directions. GOV.UK permits cocksfoot and timothy in the mix. Greiler found clover and grass mixes holding the most invertebrates of anything compared, more than natural regeneration and far more than phacelia. Set against that, the same study found sown ground carrying the fewest plant species. Our reading is that grass in a legume fallow is a choice about what you want the block to do rather than a mistake either way, and this is one of the few actions where you get to make it.

The cutting rules and the cutting research sit awkwardly, and it is worth knowing before you commit ground. This action permits cutting for one purpose, to stop blackgrass setting seed or to control other annual grass weeds. The research is consistent that summer cutting is the expensive one, costing butterflies and flowering plants where spring and autumn cutting did not. Blackgrass sets seed in early summer. So the one cut you are allowed is the one the evidence says costs the most, and the study that found blackgrass dominating an unsprayed sown margin cleared it with a December herbicide rather than a cut, which is not an option this action gives you.

The honest limit on all of it is the same one the money section runs into. This action is named for nutrients and soil, and everything above is about insects, birds and plants. Nothing here tells you what a legume fallow does for the nitrogen in that block or for the crop that follows it. That figure would settle the action, and we have not found it.

What you actually have to do

Establish a legume fallow using a seed mix containing at least six flowering species, including legumes. GOV.UK names alsike clover, bird's foot trefoil, black medick, common vetch and lucerne among the legume options, alongside non-legume flowering species and grasses such as cocksfoot or timothy. Once it is established, you must maintain it, and GOV.UK puts that as managing it in a way, and for a period of time, that can reasonably be expected to achieve this action's aim.

Establish the legume fallow by the autumn, within 12 months of the action's start date. Maintain it at the same location for a period of time that can reasonably be expected to achieve this action's aim, which GOV.UK says will usually be until the end of the second summer after sowing. Re-establish it by the following autumn and maintain it to the end date.

Source: GOV.UK, CNUM3 action page, checked 02/08/2026.

What you cannot do on it

Once the fallow is established: no grazing with livestock. No using the area for regular vehicular access, turning or storage. No fertilisers and no manures.

Cutting is permitted only to prevent blackgrass setting seed or to control other annual grass weeds. Pesticides are not permitted at all, except herbicides to weed wipe or spot treat to control injurious weeds, invasive non-native species, nettles or bracken.

Source: GOV.UK, CNUM3 action page, checked 02/08/2026.

Where it can go

Agricultural land below the moorland line, on arable land, temporary grassland, fallow land or permanent horticultural crops, with compatible land covers and land use codes. It can be rotational or static, so you can move it in years two and three or keep it in the same place.

SSSIs are eligible but you must give notice to Natural England and get SSSI consent before the Rural Payments Agency offers you an SFI26 agreement. Land with historic or archaeological features needs an SFI HEFER before you do the action, and GOV.UK says not to sow deep-rooted legumes such as lucerne on the areas it identifies.

Source: GOV.UK, CNUM3 action page, checked 02/08/2026.

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.

Two sowings, not one. The timing requirements put an establishment at the front and a re-establishment before the third year. Anyone budgeting this off a single seed invoice is out by the price of a second one, and on a mix at this price that is not a rounding error.

Six flowering species is a countable requirement, and the bag is the evidence. It is not "legume-rich" as a general description. The number is on the page, so the composition on the invoice is the document that answers it, and a mix sold under the code still has to be read.

Cutting is not a management tool here. It is permitted to prevent blackgrass setting seed or to control other annual grass weeds, and that is the extent of it. Topping a fallow because it looks untidy is not one of the permitted reasons on the page.

No regular vehicular access, turning or storage. A block that is out of cropping is a convenient place to turn, park a trailer or drop a heap, and the prohibition rules that out. Worth knowing before you pick which block goes into it.

Deep-rooted legumes and historic features do not mix. If your HEFER identifies anything, lucerne is specifically the species GOV.UK tells you to keep off it, and that is a decision made when you choose the mix rather than when you drill.

The merchant prices above were read inclusive of VAT. Merchant sites can display either basis, and a quote you are given may be on the other one. Worth settling before the figures go into a budget.

Compatible on the page is not the same as compatible on your parcels. The GOV.UK action page sets out its compatibility lists separately for actions on the same area and actions on the rest of the parcel, and a legume fallow occupies the ground for the whole period, so on those hectares it will not sit alongside an action that requires a crop to be grown.

Keeping the evidence

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

The action page asks you to keep evidence of what you have done, giving field operations at parcel level with their invoices, and photographs or other documentation, as the examples.

For this action the two seed invoices are the backbone, because they are the only documents that show both what went on and that it went on twice, and because the species breakdown on them is what speaks to the six-species requirement. Photographs are worth taking more than once across the year, since a legume fallow in flower and the same block in November look like two different fields, and a single picture speaks only to the day it was taken. Cutting is worth recording with its reason attached, because the permitted reasons are narrow and a cut with no note against it says nothing about why it happened.

The part only your own ground can answer

The gap above is worked on national averages because that is all a website has. What it cannot know is what this block returns in your rotation, or what the fallow is worth to the crop that follows it on your soil. Both of those live in your own records, and Tilth is where we keep them, parcel by parcel and year by year. Request an invite.