Climate change and biodiversity loss are critical challenges - now and into the future - and matter to us. Pretty much all human activities have an off-target impact - something that happens as a direct or indirect consequence of what we do, as well as the thing we were aiming for. Even interventions aimed at helping fight global warming or habitat loss can have off-target impacts too. Farming is no exception to this and much is made of the environmental effects it has. Unfortunately, at times this information, on both sides of the debate, is inaccurate or misrepresented. 

Context is everything - the carbon footprint of a steak from a cow raised in an American feedlot system, fed grain, will be different than one from cattle reared on grassland torn out of virgin rainforest in "slash and burn" systems. Both of these are a long way from an extensively raised animal almost or completely grass-fed during its lifetime, making use of pastures that could not be used for other types of food production. Plant-based foods have their own impacts too - out of season varieties flown in from abroad, greater reliance on so-called plant protection products (read herbicides and pesticides), the carbon cost of building the machinery involved and the biodiversity effects of monoculture system, to name just a few.

Alongside this there is much still to learn about the real impacts of shifts in natural carbon cycles. After all, there have been ruminant mammals around on earth for a very long time, and methane (CH4), whilst being a more potent warming gas than carbon dioxide, (CO2) does not last for anywhere near the time that CO2 does. If that methane is the by-product of eating grass that has grown that season, using carbon drawn from the atmosphere during photosynthesis by the plant, then all we are doing is moving the same carbon around the cycle, not adding more.

This isn't to say that farming and, more specifically, meat production is somehow blameless or has no role to play in mitigating off-target impacts - quite the opposite. But, in turn, that starts with an awareness and an honest reflection on what those impacts may be and what can be done to avoid, reduce or balance them. Nor is it to suggest that only one side in the plant v meat debate is "right" - again, quite the opposite.

To inform ourselves and identify ways we can avoid, reduce and balance, we have read, attended workshops, and completed our own carbon audit of the farm. The process isn't perfect - trying to account for every little thing is incredibly difficult to do - and the models have their own limitations - a better way of accurately representing the impacts of methane is needed. Nor is the information secret - we share our findings, and the assumptions and estimates that underpin them, below in the hope it increases wider understanding of the issue and encourages more of us to have the curiosity to find out and the courage to try and make things better.

What follows below is an honest explanation of how we went about working out what to input into the calculator and the assumptions and estimates we have needed to make along the way. We intend to apply the same methodology going forwards so that we can compare like-with-like in future, unless we discover that there are fundamental problems with something and correction would be better all round.

The calculator is divided into a number of sections depending on the type of impact in question - for example fuel use versus livestock. We'll tackle them individually for clarity. We have used a Scope 3 approach as far as possible - this includes not just our direct emissions, but any that arise from our electricity use, as well as emissions that occurred to produce the items we have bought or used.

• Red diesel and petrol totted up from the receipts of each purchase (we don't use a lot!). If a receipt was missing then the payment amount was used to estimate the volume using the price per litre from other receipts. We assumed that all of the fuel was used on the farm and ignored that some of the petrol, for example, might be used in a domestic lawnmower.

• Any items delivered were assumed to generate 10 delivery miles at 5 miles per litre of road diesel.

• Electricity was harder as the farm doesn't have a separate meter from the household (and we have ignored the household as everyone needs a place to live - it isn't contributing to the farm's emissions itself any more than you would be contributing to your employer's emissions because you also have a house). We have solar panels and assumed, for the sake of simplicity, that they would cover the electricity needed to run our electric fences and recharge batteries, e.g. for power tools. This probably grossly underestimates the net gain here as the panels generate far more than we use in this way. We have allowed for running LED lights in our winter poultry shed (2x 36W on for 12h per day for 180 days). As this is longer than they are actually used for, we have ignored the intermittent use of lights elsewhere in the building, e.g. when getting tools or feed. We have not included the running of freezers to keep produce frozen, but may aim to do this in future. There is no mains electricity elsewhere on the farm.

• The farm does not use any solid fuels, but we do have an odd bonfire of brash if we have trimmed the hedgerows to manage them. We have estimated 50kg burned this year.

• Miles travelled in our cars to collect/deliver produce, obtain supplies, or other farm tasks, have been estimated.

• We have been able to include accurate measurements of contractor operations involved in making our haylage and have included an estimate of the diesel used by groundworks contractors this year based on the operating times of a digger and dumper. Our own operations are already accounted for in the red diesel use above.

• The vast majority of the carbon impact of materials has come from groundworks undertaken to install a concrete pad and hardstanding area. Although there is a cost to this they were also done in order to reduce the risk of diffuse pollution caused by soil erosion or slurry run off. Even concrete can have benefits!

• The actual quantities of stone and concrete used have been included, based on invoices received. Rock salt for the livestock was added as the same weight of general stone to reflect the emissions from extracting and crushing it.

• LDPE bale wrap was added to reflect the number of bales produced from our meadow plus any bought in.

• Wood has been included using the closest fence post dimensions in the pick list or by calculating the volume, e.g. for fence rails.

• Metalwork, such as nails and screws, have been included under fencing staples, by weight.

• Fence fittings were added by estimated weight of galvanised steel.

• The farm disinfectant was listed as an option, so included. Iodine was added as the same volume of this disinfectant, as it is iodine based.

• Geotextile membrane, although not made of polypropylene, was included as such, as no better etimate was apparent.

• MDPE fittings have been estimated at an average of 100g each and included by weight. Electric fence insulators were estimated as 15g each.

• A bucket and new water trough (estimated as 15kg) were added.

• We use very little mains water for the farm - limited to washing bowls or cleaning dosing guns - so have ignored this. The overwhelming majority of water used on the farm is collected from springs or harvested from rainfall. We do not have any mains sewerage.

• Items bought in 2025 were included and the weight of metal estimated by looking at similar models online.

• This does exclude the carbon impacts of items we have bought previously, e.g. machinery, hurdles, hay racks, trailers and so on, but the calculator depreciates the whole item in the year it arrives, even though it will likely last decades. As such we have decided to take each year "as it comes".

• Any herbicides acquired were added as specific items - none bought in 2025 as we use very little.

• This is undoubtedly the most complex part of the tool and where we had to develop our approach to working out the data to include. This was made more difficult because we are not a typical farm raising the usual breeds of stock in the standard way. Our animals are generally smaller breeds and they are fed and housed differently to the "average" farm. Mature Hebridean ewes are about 35-40kg in weight, whereas a more common North of England mule would typically be closer to 70kg. Because a ewe not producing milk will consume roughly 1.5% of her bodyweight in dry matter per day you can see that the actual weight of the animal has a big impact on how much food it eats and therefore how much CO2 and CH4 it produces. Housing affects the way manures emit greenhouse gases, and the standard model assumes that pregnant ewes are housed for 6 weeks each year at lambing time. Our sheep perhaps have 2-3 days inside after they lamb, so we have had to gather more detailed data to model more accurately. In many commercial systems lambs are feed creep feed, a concentrate based on grains, to speed their growth. We do not use any creep feed (in 8 years we have bought one 25kg sack to try and support an orphan lamb that was struggling, but it largely refused to eat it and most of the sack went out of date!). Similar differences exist for our Dexter cattle.

• Livestock feed bought in, including extra haylage above the bales we make here, was included by totting up the quantities from invoices. We assumed that everything bought in 2025 was used in 2025 but did not include anything already in stock at the start of the year (and will apply the same logic going forwards). Some feed had to be added as the nearest equivalent - for example pig feed is not listed, so we used 14% and 16% sheep feed as it is likely the closest to the 14% and 16% pig feed. Similarly 16% goat mix was recorded as 16% lamb finisher. 

• Lick buckets were added as the same weight of molasses (from sugar beet).

• Heston bales of straw were included alongside weight-based estimates of chopped straw and wood shavings used for chicken bedding.

• In terms of the livestock numbers, weights, manure management, and feeding regimes, we needed to apply more thought!

• Pigs:

  • We assumed a linear growth rate from 15kg on arrival as weaners to 95kg at slaughter after 26 weeks. This gave an average weight of 55kg during this period.

  • All manure recorded as in field.

• Chickens:

  • We estimated hen numbers at various points during the year, which gave an average of 32 hens year-round, and left the weight to calculate based on the industry average. This will overestimate the impact of our poultry as they are predominantly smaller types than commercial hybrid layers.

  • Manure recorded as 50% in field, when they are free-ranging, and 50% FYM, when they are housed in the winter enclosure.

• Goats:

  • We used our weight records, that we gather regularly through the year, to calculate an average weight per goat across the year. We did this separately for adults (66.3kg) and kids (11.2kg) and calculated the number of goats present for each period.

  • Manure was recorded as described for chickens above to reflect that they largely use a small barn for winter accommodation (goats hate the rain!).

• Sheep:

  • The flock was divided into a number of groups, based on how their energy needs would be calculated, and their time on farm. We again used our weight records to calculate an average weight within the groups:

    • Ewes that lambed present the whole year (37.5kg) and ewes that lambed present part year (37.5kg).

    • Ewes that didn't lamb present the whole year (25.8kg) and ewes that didn't lamb present part year (25.8kg).

    • Rams, including wethers, present the whole year (59kg), and those present part year. (36.3kg)

    • Lambs before weaning (14.9kg) and after weaning (24kg). All sheep born in 2025 were counted as lambs and all those born before 2025 were counted as adults.

  • Where sheep were on farm for less than a year, the number of individuals were multiplied by the number of weeks for each case. These were added up and the total divided by the number of animals to give an average time per animal.

  • Ewes that lambed had manure management recorded as 99% in field and 1% FYM to reflect that they may spend a few days inside after lambing. All other sheep were recorded as 100% in field as they are not housed.

  • Then came the complexity of feeding! The default method was inappropriate for us as we do not use the quantities of concentrate feed it describes. To arrive at a better estimate we had to apply the ADHB guides on percentage of bodyweight in dry matter intake per day to the weight of that group during the year. For the non-lambing ewes and males this was a straight 1.5% throughout. Lambs were recorded as 0% for the first 4 months until assumed weaning, and then at 4% for the remainder of the year. Breeding ewes were 1.5% outside of lactation and 2% for 5 months of lactation. Rather than apply the 1.7% for ewes with singles and 2.1% for those with twins, we assumed most have twins and used a straight 2% for all. This overcounts the impact of lactation (as we elsewhere assumed that lambs were weaned at 4 months, not 5) which offsets the underestimation of the higher percentage for lambs (4% after weaning) in the first 4 months of the year (we included the previous year's lambs as adults, i.e. 1.5%, in January-April).

  • These percentages were multiplied by the average weights above to give us dry matter intake (DMI) figures per head per day. For the breeding ewes we then calculated a weighted average across the year in a similar way to calculating average head count for animals that were only present part year. This has led us to use daily DMI/head figures of 0.64kg for breeding ewes, 0.39kg for ewes that didn't lamb, 0.89kg for our older wethers, 0.54kg for the younger tups, and 0.94kg for post-weaning lambs.

  • Next we had to describe the make up of their diet. Post-weaning lambs were recorded as 100% grazed on moderate quality grass forage. We may look to refine this in future as grass quality does vary across the farm with some fields probably being low quality here in the uplands. Tups were recorded as 80% grazed and 20% eating moderate quality hay. There was no haylage option in the calculator and since our haylage is generally very dry it is far closer to hay than silage. The ewes were trickier as they do get some concentrate feed in the 6 weeks pre-lambing. This really was a bit of a back of the envelope, albeit informed, guesstimate, but we went for 80% grazed, 15% haylage, and 5% sheep nuts across the year (based on assuming the sheep nuts make up 40% of their diet in those 6 weeks).

• Cattle:

  • We used a simlar approach to the sheep but had to make estimates of weight across their age groups because we do not have a way to weigh the cattle regularly at present. Using information online we began from the suggestion that a Dexter calf weighs around 20kg newborn, around 130kg after 6 months, and rises to 350kg by 2 years for cows and 450kg for bulls. In between these times we assumed linear growth rates, so could work out an average weight within each group:

    • Last year's calves whilst under 1 year of age (178kg).

    • Last year's calves once over 1 year of age (226kg).

    • This year's calves (74.3kg).

    • Mature cows (350kg).

    • Mature bulls (450kg).

  • As above, we applied these weight figures to the ADHB data on percentage DMI (1.5% year round for bulls, 2.5% for last year's calves, 3% for this year's calves, and a range for cows starting from 1.5% when dry, 2.5% in early to mid-lactation, and 2% in late lactation).

  • This gave daily DMI/head figures of 4.45kg for last year's calves under 1 year, and 5.65kg once over a year, 2.23kg for this year's calves, 11.25kg for the bull, and 6.71kg for the mature cows.

  • We worked out their manure management (100% in field as they are not housed - although this does ignore the small amount of FYM we create in their winter shelter - we may try and estimate this better in future).

  • Again, similarly to the sheep, we then considered their diet. This year's calves, last year's calves over 1 year and the bull are 100% grazed as we are not feeding haylage at those times of the year (we tend to open a bale in late December and feed these through to late April). Last year's calves under 1 year were calculated to be 36% grazed and 64% haylage reflecting that the months included here are January to May, when we assume that 80% of the diet is haylage. The mature cows worked out to be 73% grazed, 27% haylage across the year based on 100% grazing most of the year, but 80% haylage January to April. 

• We estimated 150kg of plastic to include bale wrap (approx 2kg per bale) and the materials used to package our meat produce, and assumed that it all ends up in landfill.

• We estimated 100kg of cardboard to include the meat boxes and miscellaneous other packaging.

• In future years we may need to keep records of other items diposed of, e.g. old fencing materials removed, but did not feel there were significant quantities in 2025.

• Road miles have already been included in the Fuel section.

• Refrigeration estimates required calculations based on the amount of coolant gases used. As we do not use gases - we have domestic freezers - we have assumed this to be zero, and instead will account for freezers on purchase in future.

• Until now everything was adding up the emissions so it was good to get to a point where we could include the carbon captured on farm.

• We used our RPA field mapping details to record the amount of woodland (with a small estimated rounding to allow for the riparian woodland in one field). Although this woodland has been around for over 200 years, many of the trees are much younger than this, having been replanted in a millennium project and subsequently. We therefore assumed an average age of 60 years across the woodland.

• We used Google Earth to tot up the approximate numbers of in-field trees across the remainder of the farm (80) and estimated an average canopy area for them (20m2).

• We added the measurements of the different sections of hedgerow along the track and for the new shelterbelt planted on our western boundary.

• We used the Countryside Stewardship GS3 (low input grassland in SDA) for the majority of the grassland as this reflects our management approach, with Broad Acre recorded separately under HK15 (the HLS option applied there).

• Cardboard and plastic for meat boxes is already included under Waste disposal.

• There did not seem to be a way to account for the carbon emissions from the abattoir and butcher.

• No further items to record.

• We could not find a way to include the following:

  • 2 spools of electric fence rope

  • Medicines, vaccines, and mineral drench

  • 5 litres of antifreeze

  • 50 fibreglass fence posts

  • The carbon emissions associated with insurance, accountant services, farm software, the SIM cards in our monitoring cameras, and processing and refrigeration at the abattoir and butcher

• We feel that, overall, they are unlikely to make a major difference to the overall outcome.

• So, what did we find? Click on to the next page to discover the outcome of our farm carbon audit.