FarmLCA: A novel approach to assess agroecological innovations in Life Cycle Assessment

Research Paper FarmLCA: A novel approach to assess agroecological innovations in Life Cycle Assessment Simon Moakes a,b,* , Philipp Oggiano a , Jan Landert a , Catherine Pfeifer a , Laura de Baan a,* a Research Institute of Organic Agriculture (FiBL), Department of Food System Sciences, Ackerstrasse 113, CH-5070 Frick, Switzerland b Institute of Biological, Environmental & Rural Sciences, Aberystwyth University, Gogerddan, Aberystwyth SY23 3EE, Ceredigion, UK HIGHLIGHTS GRAPHICAL ABSTRACT •FarmLCA assesses trade-offs and synergies of agroecological innovations. •It models crop-livestock interactions, on-farm and upstream environmental impacts. •Feed-no-food scenarios of a Scottish mixed beef farm were assessed with FarmLCA. •Circular farms avoiding food as feed produce more protein and reduce environmental impacts. ARTICLE INFO Editor Name: Paul Crosson ABSTRACT Context: Agroecological innovations are seen as solutions to reduce environmental impacts of agriculture but can potentially lead to trade-offs with food production. Appropriate tools are needed to better understand synergies and trade-offs among environmental issues, resource efficiency and food production. Objective: This study presents the FarmLCA tool, which models farms as interconnected crop-livestock systems and assesses environmental impacts from farms and farm-inputs. A mixed beef farm serves as case study to assess synergies and trade-offs of avoiding human edible feed in beef production. Methods: FarmLCA allows the calculation of cradle-to-farm gate life cycle assessments (LCA). Emissions of environmentally harmful substances from crops and livestock are modelled based on the farm management. Upstream impacts from imported inputs (including fertilizer or feed) are accounted for with life cycle inventory data. Yields and nutrient requirements are checked for plausibility, based on management handbooks, while manure availability and composition are calculated based on livestock production. Environmental impacts, nutrient use efficiency and food production for a typical mixed beef farm in Scotland were calculated (baseline) and compared to alternative farm management scenarios: a Feed-no-Food scenario, avoiding concentrate feeds resulting in a smaller herd size and a circular Feed-no-Food scenario, additionally optimizing productivity and synergies between crop and livestock (e.g. more legumes in crop rotation, reduced replacement rate and feed waste). This article is part of a Special issue entitled: ‘Agricultural LCA methods’ published in Agricultural Systems. * Corresponding authors. E-mail addresses: [email protected] (S. Moakes), [email protected] (L. de Baan). Contents lists available at ScienceDirect Agricultural Systems journal homepage: www.elsevier.com/locate/agsy https://doi.org/10.1016/j.agsy.2025.104560 Received 25 July 2025; Received in revised form 21 October 2025; Accepted 4 November 2025 Agricultural Systems 232 (2026) 104560 Available online 25 November 2025 0308-521X/© 2025 The Authors. Published by Elsevier Ltd. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). Results and conclusions: In the Feed-no-Food scenario, the beef production was reduced by 25 %, but more calories and protein were produced overall due to cereal and legumes now being available for direct human consumption. However, slower growth of livestock led to increased environmental impact of beef, whilst reduced livestock numbers required more mineral fertilizer for crop production to replace on-farm manure. In the circular Feed-noFood scenario, beef and overall calorie production were slightly reduced compared to the baseline, but 1.5 more high quality protein (expressed by the Digestible Indispensable Amino Acid Score, DIAAS), were produced. Environmental impacts of beef were reduced and nitrogen self-sufficiency improved due to increased legume share in the rotation. Significance: Existing LCA approaches often fail to capture the complex dynamics of integrated crop-livestock systems and agroecological practices. FarmLCA addresses this by modelling both on-farm processes and upstream inputs, enabling a consistent assessment of environmental impacts, nutrient use efficiency, and food production. It offers a more holistic and systemic view of the consequences of agroecological innovations and enables the identification of synergies and trade-offs between environmental protection, resource efficiency, and food production. Guest Editor: Paul Crosson. 1. Introduction Agriculture is a main contributor to humanity exceeding the planetary boundaries (Campbell et al., 2017), particularly for the global climate and biodiversity crises. Over the past decades, livestock production drastically increased around the globe and is responsible for a large share of these environmental impacts (FAO, 2006; Herrero et al., 2016). About half of all agricultural land is used for feed production and a third of the cereals produced globally are fed to livestock (Mottet et al., 2017). Importing feeds and fertilizer strongly contributes to a disruption of global and regional nitrogen and phosphorous cycles and to the environmental problems associated with this (Billen et al., 2021). Considering these challenges, there is growing recognition that incremental efficiency gains may not be sufficient to address the environmental impacts of livestock systems (Van Zanten et al., 2018). This has led to increasing interest in transforming agriculture through more circular and ecologically integrated approaches (Wezel et al., 2020). Agroecological innovations are seen as solutions to achieve more sustainable food systems and achieving the Sustainable Development Goals (Bicksler et al., 2023). Principles of agroecology include both elements within the food system and at the agroecosystem level. Within the food system this may include a dietary transition to more plant based diets and avoiding feed and food competition, e.g. Storyline 4 “Local-agroecological-food-systems” (R¨ o¨ os et al., 2022). For the latter, the main principles encompass recycling of nutrients and biomass, reduction of purchased inputs and increased self-sufficiency, enhancing soil, animal health and biodiversity, using synergies between elements of the agroecosystem (i.e. livestock, crops, trees, soil and water) as well as economic diversification (Wezel et al., 2020). Recent studies showed that dairy farms with lower concentrate inputs improved net protein contribution to the food system (Wild et al., 2025), whilst improved efficiency and use of grassland can reduce competition for crop land (Ineichen et al., 2023). However, agroecological farming systems can also create trade-offs because they are often more land demanding due to lower productivity which can sometimes result in higher environmental impacts per unit of food produced (Boschiero et al., 2023; Mathis et al., 2022; Seufert and Ramankutty, 2017). To support decision-makers such as farmers, advisors, policy makers, retailers or consumers in identifying sustainable solutions for food production, farm level holistic and dynamic tools are needed (Prost et al., 2023) to capture the complexity, synergies, trade-offs and feedback mechanisms within agroecosystems. Life Cycle Assessment (LCA) is a common methodology to assess environmental impacts of food production. Typically, the emissions of environmentally relevant substances to different environmental compartments (air, soil, water) occurring on-farm during crop and livestock production are quantified, as well as environmental impacts related to production and transport of inputs (e.g. fertilizer, chemicals, machinery, feed). However, LCA has been criticised as not being able to capture the benefits of agroecological systems, such as organic systems and favouring high-input and intensive systems (van der Werf et al., 2020). To calculate agricultural LCAs, specific tools are available with different scope, which often are well suitable to assess conventional agriculture or a single sector, but differ in the extent to which they account for the complexity and interconnectedness of more complex agroecological farms. While some tools were specifically developed to compile agricultural life cycle inventory (LCI) data and to calculate on-farm emissions (e.g. MEANS-InOut (Auberger et al., 2018)), others allow to additionally assess environmental impacts (e.g. Swiss Agricultural Life Cycle Assessment (SALCA) (Nemecek et al., 2024), including indicators for biodiversity and soil quality). To assess impacts of livestock farms, SIMS DAIRY (Del Prado et al., 2011) or Sustell™ (dsm-firmenich, 2025) can be applied, whilst others including Crop.LCA (Goglio et al., 2018b), focus solely on LCA of crop production. Tools like Agrecalc (2025) and Cool Farm Tool (Hillier et al., 2011) allow the calculation of farm carbon footprints, the latter also assesses water use and biodiversity (Crowther et al., 2024). The online platform HESTIA provides open-source models for calculating emission, gap-filling missing data and impact assessment and allows data sharing in a harmonized way (HESTIA, 2024). These tools and models are generally well-suited for assessing the environmental impacts of specialized and conventional agricultural systems in Europe, but may not adequately reflect the complex interactions and internal feedbacks of mixed and agroecological systems, where crop and livestock components are interlinked. For example, a change in crop fertilization may influence the availability of livestock feed, which in turn affects the quantity and composition of manure applied to fields and such interdependencies are central to agroecological farm design yet remain largely unaccounted for in existing assessments. To address this gap, a novel tool called FarmLCA was developed that combines a flexible farm system model with LCA. This Excel tool allows the calculation of “standard” agricultural LCAs, but additionally also allows assessment of a farm as a system, modelling crops and livestock as interconnected dependant units. The tool includes the compilation of agricultural life cycle inventories (LCI), linking background data from ecoinvent (Wernet et al., 2016) or Agribalyse (Colomb et al., 2015), together with emission modelling from fields (including changes in soil organic carbon) and livestock. It also comprises multiple allocation methodologies, calculating environmental impacts though Life Cycle Impact Assessment (LCIA), as well as displaying results using different functional units (e.g. mass, area, livestock cohort, farm enterprise), and S. Moakes et al. Agricultural Systems 232 (2026) 104560 2 contribution analysis as process groups (e.g. tillage, fertilization, harvesting, field emissions). Furthermore, allocation of impacts between coproducts is improved by allocation of both inputs and emissions to specific products from the same plot, such as pasture used for grazing, silage and hay production. In the first part of this article, we present the overall structure of the FarmLCA model, implemented emission models, procedures to fill data gaps, conduct plausibility tests and propose indicators to assess nutrient circularity of farms in addition to LCA impact categories. Then, we illustrate the usefulness and limitations of the model by applying it to a typical mixed conventional farm in Scotland, producing arable crops and beef. Two potential future scenarios are modelled for this farm showing the consequences of avoiding both external feed and on-farm produced human edible feeds (Feed-no-Food) as well as an additional optimization of the farm to further reduce mineral fertilizer inputs and increasing yields (circular Feed-no-Food). Finally, a sensitivity analysis is conducted to show the effect of uncertain parameters on the environmental performance of the scenarios. 2. FarmLCA model 2.1. Model structure The FarmLCA tool is fully programmed in MS Excel and consists of five main components (Fig. 1). First, data on crop and livestock production and on farm management can be specified into a user interface. The tool provides standard data to prevent data gaps as well as plausibility checks to avoid implausible data. Second, a farm system model quantifies the bio-physical flows of nutrients between the crop and livestock system as well as the agricultural products produced on the farm. Third, a set of sub-models quantify emissions from fields, livestock and manure management. Fourth, the life cycle inventory is compiled by collating external inputs and emissions emitted from fields or livestock to the outputs (plant or livestock products) and defining allocation procedures for co-products. Finally, a full cradle-to-farm gate life cycle impact assessment is calculated for all the products produced per farm for a range of impact categories (Fig. 2), and a contribution analysis can be displayed for different functional units, for example per kg, per hectare, or per farm. In addition, indicators for nutrient circularity and food production can be calculated at farm scale. Optionally, data on costs or farm operations and prices of sold products can be indicated, to calculate economic gross-margins (see Weiner et al., 2024). In the following, each step is explained in more detail. 2.2. Data entry, data gap filling and plausibility testing 2.2.1. Plant production In a first step, the different types of land use of a farm (e.g. different crops in a rotation, permanent crops or grassland) are entered in separate columns. General production parameters, such as country, climate, soil data, field sizes, production system (organic/conventional) as well as the main crop(s) produced need to be specified. A staggered dropdown system is implemented that allows selection of standard life cycle inventory datasets from ecoinvent (Wernet et al., 2016) or Agribalyse (Colomb et al., 2015), which can be specified for each plot or by each crop or forage type, best fitting the respective crop or grassland in terms of production intensity and region. After selecting an inventory for the main crop, standard values per hectare are displayed in each column for all crop management such as tillage, sowing/planting, fertilization, grazing, pesticide usage, irrigation, harvest, yields, off-farm material and energy inputs as well as off-farm transports, waste or use of natural resources. For all these parameters, the inventory values can be replaced with farm-specific values, as available. When primary farm data are not available, locally relevant management books for values such as crop yields should be used as these strongly affect the impacts per kg due to for example, climate and soil characteristics. As agroecological farmland is often multifunctional, the model allows the entry of two different land uses per land use type (such as the main crop production and a cover crop with for example livestock grazing or mechanical harvesting). Furthermore, for each land use type up to four different products can be specified that result from a field (such as wheat grains, straw and fodder from cover crops). For each plot, the share of crops harvested and sold needs to be indicated, to determine crop residues left on field, as well as crop output and available feed for on-farm use (e.g. by livestock). In addition, if feed or bedding is used onfarm, this must be specified for each plot. Finally, a potential change in land management in the past 20 years can be specified, which is relevant to soil carbon calculations. Fig. 1. Structure of the FarmLCA model. S. Moakes et al. Agricultural Systems 232 (2026) 104560 3 Management aspects, including tillage, inputs and residue management can be specified. For fertilization, the total amounts of macronutrients (N, P, K) applied per field from all sources, such as organic fertilizer from on-farm livestock as well as imported mineral and bio-based fertilizers and manure are calculated. These values are compared with yield adjusted crop nutrient requirements from management recommendations in GRUD („Grundlagen für die Düngung“, Richner et al., 2017) and ecoinvent (Wernet et al., 2016) and overor under-fertilization is visualized in colours as part of the plausibility checks. This allows crosschecking of data and correction where necessary. 2.2.2. Livestock production For farms with livestock production, FarmLCA allows the modelling of a wide range of livestock species including dairy and beef cows, beef finishing, pigs, sheep, horses, alpacas, laying hens and broilers. Each herd or flock is specified and split into different livestock categories to reflect their nutritional requirements (e.g., the lactation phase of dairy cows or the multiple production phases of pigs from piglet to finishing). For each livestock category, the breed type, number of livestock, mortality rate, age of first birth, liveweight, traded livestock, days of feeding, bedding, housing and manure management, as well as produced output (meat liveweight, milk, wool, eggs, etc.) needs to be indicated. Based on this data, the daily feed requirement regarding dry matter, metabolizable and gross energy, crude protein and other parameters are calculated for each livestock category. Equations and reference values are derived per livestock type from the following sources: dairy cows (Gruber et al., 2004); beef and heifers (Lfl, 2024), pigs (AHDB, 2025), sheep (AHDB, 2024), alpaca/horses (IPCC, 2019), laying hens (Leinonen et al., 2012b) and broilers (Leinonen et al., 2012a). If a farm purchases external concentrate feed, the tool allows to flexibly specify different concentrate feed mixtures. Based on a comprehensive list of feed ingredients compiled from available data from ecoinvent or Agribalyse, the respective inventories can be selected for each component of the concentrate feed mixture. The nutrient composition of the feed mixture is also calculated based on the typical nutrient composition of the different feed ingredients (based on data from Feedbase, 2025). Finally, for each livestock category, the rationing of each feed type is specified per animal per day. This includes the rationing from on-farm concentrate feed, roughage and grazing, but also from off-farm feed mixtures. The FarmLCA tool thereby calculates the amount of on-farm feed still available (or wasted), as well as a comparison between supplied feed and required feed for each livestock category. This supports in checking the plausibility of data provided by farmers or compiled from statistical data and correcting it to ensure a balanced supply and demand. This is particularly relevant when modelling impacts of potential management changes of farm systems to ensure plausibility. In addition to feeding, the amount and sources of bedding should be indicated. Bedding produced on the farm as well as purchased bedding can be added. The housing system is specified for each livestock category, as well as the time spent in housing, open yards or grazing. In addition, the type of solid and liquid manure storage system, the amount of manure directly applied, stored or used for biogas as well as the storage duration can be indicated. This is then used to calculate emissions from manure storage and housing as well as to quantify manure availability. 2.3. Farm system model: Crop–grassland–livestock interactions FarmLCA allows the modelling of interactions of crops, grassland and livestock. For each animal species, the amount and N and P content of Fig. 2. System boundaries of FarmLCA model, included processes and emission models. S. Moakes et al. Agricultural Systems 232 (2026) 104560 4 liquid and solid manure is calculated based on the actual feed livestock receive and the nutrient retention in livestock, using Tier 2 equations from IPCC (Gavrilova et al., 2019), thus manure nutrient content reacts to diet composition and animal productivity. In addition, the type of housing and the time spent grazing are used to calculate the amounts of manure collected in the housing system based on GRUD (Richner and Sinaj, 2017). The amount and type of manure applied can be indicated for each single field of crops or grassland. Thus, the model can capture changing nutrient availabilities from manure, if the feeding or herd size is changed. A change in the crop land use, on the other hand, affects the availability of feed and bedding for livestock. In both cases, users receive instant feedback on the respective changes. This allows ex-ante assessments of systemic changes in farm management and supports the quality checking of data balancing of supply and demand for fertilizer and feed. Grazing livestock can be modelled on temporary and permanent grassland, but also on crops or residues (such as winter cereal grazing) or in agroforestry systems, such as between fruit trees. The type of forage, including the legume proportion can be specified as well as the feed intake during grazing on different types of land, which is determined based on estimated total feed intake minus concentrate and conserved feeds. Plausibility tests for energy and protein intake are also conducted, since the feed intake through grazing is often uncertain. With reference to current trends in grazing management, it is possible to indicate the utilisation rate of forage, such that practices like mob grazing or deliberate underutilisation with the aim of increased residues can be modelled. Furthermore, feed quality such as crude protein level can be adapted, allowing for differentiation between for example, young versus mature grass. The nutrients excreted on this land whilst grazing are automatically calculated based on the time spent grazing on specific plots. 2.4. Modelling direct emissions of on-farm activities Direct emissions from both crop and livestock activities are estimated through the implemented emission models in the FarmLCA tool (Table 1 and Fig. 2). The main GHG emissions related to fertilization (mineral and organic, including pasture deposition), manure storage, housing and enteric fermentation are estimated using the most recent methods from IPCC (2019). Where possible, Tier 2 methods are used to improve accuracy, but Tier 1 (including disaggregated values e.g. for N 2 O) are used in case of limited data availability (e.g. unknown sitespecific values). Other agricultural emissions are estimated using specific models such as EMEP/EEA (2023) for ammonia or SALCA (Nemecek et al., 2024; Oberholzer et al., 2006) for phosphorus and heavy metals. For pesticides, no specific emission models have been implemented so far, but could be modelled separately with PestLCI (Birkved and Hauschild, 2006) and UseTox (Rosenbaum et al., 2008). Further details are provided in the Supporting Information (SI) section 1.1–1.2. For soil organic carbon (SOC) changes and its potential impact on climate change, no full consensus exists so far on how impacts should be assessed (Joensuu et al., 2021), but recommendations on how to model SOC in agricultural LCAs were made recently (Pelaracci et al., 2025). In FarmLCA, the IPCC (2019) Tier 2 steady state method based upon the Century model (Parton et al., 1988), is implemented and results are reported separately from other GHG emissions, as recommended by the European Commission-Joint Research Center (2010), considering both a 20 and 100 year time frame (see also SI section 1.1.4). 2.5. Life cycle inventory and impact assessment The FarmLCA tool generates the life cycle inventory (LCI) data for each plot or livestock category by combining the farm-specific input values with default inventory values from ecoinvent and Agribalyse on crop management (e.g. tillage, seeding, fertilizer, pesticides, irrigation) and livestock management (e.g. feed, housing; Fig. 2). For fields or animal species producing multiple co-products (such as wheat and straw or milk and meat), different options for impact allocation are implemented. For crops, the default is economic allocation, but the user can alternatively select mass allocation, based on either dry or fresh-matter. For livestock products, economic, biophysical (IDF, 2015 for dairy) or mass allocation can be selected. Results can finally be displayed for typical functional units, such as per ha, per livestock category, per kg product or for the full farm. In addition, results can also be assessed per 100 g protein, per kcal or 100 g Digestible Indispensable Amino Acid Score (DIAAS), which reflects the protein quality of produced products (McAuliffe et al., 2023b). For life cycle impact assessment, the most recent version of the FarmLCA tool (v4.1) calculates impacts with both Impact World+ methodology v1.3 (Bulle et al., 2019) and Ecological Scarcity 2021 method (BAFU, 2021). Impacts for on-farm emissions are estimated based on values modelled by the tool (see Section 2.4), whilst for other processes, (for example ploughing, fertilizer or external feed inputs), pre-calculated impacts are imported into the tool as well as the characterization factors for all relevant substances. Therefore, impacts per unit of input are according to the original inventory assumptions including soil organic carbon. The collated impacts are first calculated Table 1 Overview on implemented emission models for plant and livestock production. Emission Process Methods Plant production Carbon Dioxide (CO 2 ), to air Plant absorption ecoinvent 3.8 (Wernet et al., 2016) Lime IPCC (2006) Tier 1 Urea IPCC (2006) Tier 1 Dinitrogen oxide (N 2 O), to air Field emissions IPCC (2019) Tier 1 (disaggregated factors) Ammonia (NH 3 ), to air Mineral and organic fertilizers (including pasture deposited manure) EMEP/EEA (2023) Tier 2, IPCC (2019) Tier 1 Nitrate (NO 3 ), to groundwater Leaching: crops/ grassland SQCB (Faist Emmenegger et al., 2009) IPCC (2019) Tier 2, expanded with drainage factor of % of area drained to surface water Leaching: crops/ grassland Nitric oxide (NO x ) to air Field emissions EMEP/EEA (2023) Tier 1, IPCC (2019) Tier 1 Phosphate (PO 4 ), to groundwater Leaching: PO 4 SALCA-P (Nemecek et al., 2024) to surface water Runoff: PO 4 SALCA-P (Nemecek et al., 2024) Runoff: P SALCA-P (Nemecek et al., 2024) Trace metals (Cd, Cu, Cr, Hg, Ni, Pb, Zn) to ground water Leaching SALCA-SM (Nemecek et al., 2024) to surface water Erosion SALCA-SM (Nemecek et al., 2024) to soil Emissions SALCA-SM (Nemecek et al., 2024) Optional: Soil organic carbon (SOC), to air Tillage land (crops and temporary forages) IPCC (2019) Tier 2, steady state Permanent grassland IPCC (2019) Tier 2, steady state (Bolinder et al., 2007) Livestock production Methane (CH 4 ), to air Cattle, enteric IPCC (2019) Tier 2 Sheep or other ruminant, enteric Belanche et al. (2023), IPCC (2019) Tier 1 Pig, poultry Jørgensen et al. (2011), FEON (2023) Manure storage IPCC (2019) Tier 2 Dinitrogen oxide (N 2 O), to air Manure storage/ housing IPCC (2019) Tier 2 Nitric oxide (NO x ), to air Manure storage IPCC (2019) Tier 2 Ammonia (NH 3 ), to air Manure storage/ housing EMEP/EEA (2023) S. Moakes et al. Agricultural Systems 232 (2026) 104560 5 per field (ha) and livestock category. In a next step, LCIA results are calculated for multiple plant or livestock products, using different functional units and allocation methods (see above). 2.6. Indicators for nitrogen circularity and self-sufficiency To quantify and compare the degree of circularity and selfsufficiency of agroecological farms, FarmLCA calculates additional indicators for nitrogen. First, the nitrogen use efficiency (NUE, Eq. 1) for the whole farm is calculated based on the total N output of the farm in relation to the required N inputs (EU Nitrogen Expert Panel, 2016). NUE =(N_outputfarm/N_inputfarm)*100 (1) With: NUE: nitrogen use efficiency (%) N-outputfarm = {crops} + {straw} + {trees/branches (net) } + {livestock (net) } + {livestock products (milk,egg,wool) } (2) N-inputfarm = {mineral fertilizers} + {biological nitrogen fixation} + {atmospheric N deposition} + {compost and bio-based fertilizer} + {seed and planting material} + {manure (net) } + {irrigation water} + {feed and fodder (net) } + {bedding material} (3) In addition, two indicators on N-self-sufficiency (Eq. 4) of plant as well as livestock production were developed, which quantify the share of N inputs needed for plant or livestock production that is produced onfarm. Here, only inputs controllable by the farmer are accounted for, thus atmospheric N deposition and N in irrigation water are not considered as inputs. N-self-sufficiencyi=N-inputs(on-farm,i)/(N-inputs(on-farm,i) +N-inputs(off-farm,i))*100 (4) i: type of agricultural product (plants or livestock products). N-inputs,on-farm,plants = {biological nitrogen fixation} + {on-farm produced compost} + {on-farm produced manure} + {on-farm produced seed or planting material} (5) N-inputs,off-farm,plants = {mineral fertilizers} + {purchased bio-based fertilizer/compost} + {purchased manure} + {purchased seed and planting material} (6) N-inputs,on-farm,livestock = {on-farm produced feed and fodder} + {on-farm produced bedding material}(7) N-inputs,off-farm,livestock = {purchased feed and fodder} + {purchased bedding material}(8) 2.7. Indicators for food production To show the change in food produced on the farm after a management change, the total energy (kcal) and proteins (kg proteins and kg proteins corrected for their quality by DIAAS) can be calculated for all sold human edible outputs (livestock products and crops), enabling multiple products to be compared on the same basis, e.g. calories or protein. For the meat, a conversion from live weight to human edible meat was undertaken based on carcass yield (Coyne et al., 2019; Mosnier et al., 2021) and percent of human edible meat in carcass, including typical losses until retail (Caldeira et al., 2019). Typical losses from farm gate to retail were also considered for crops (Caldeira et al., 2019; Gatto et al., 2023). For both plant and livestock products, typical nutritional composition of foods were taken from Public Health England (2021). For the different products, average DIAAS (for persons older than 3 years) were calculated based on Adhikari et al. (2022), Herreman et al. (2020) and Ertl et al. (2016). A more detailed description of implementation for the case study can be found in SI Table S1–S3. 3. Case study description 3.1. Goal and scope A cradle-to-farm gate LCA was performed on a typical Scottish mixed beef farm, with the aim to show the environmental consequences of avoiding human edible feed in the cattle ration. The functional unit was 1 kg liveweight produced per herd (finished beef and slaughtered beef cows). Thus, no allocation was performed between the finished beef and the slaughtered cows. In addition, the environmental impacts of the main cereals produced on farm (wheat and oats) were calculated per kg crop, using economic allocation between grain and straw. To allow for a more holistic assessment that combines crop and livestock impacts, results are also presented as impacts per hectare functional unit. 3.2. Baseline For the baseline, a typical mixed beef farm in Scotland was assumed, with typical data for livestock and plant production derived from the Scottish Farm Management Handbook (FMHB, SAC Consulting, 2023). If data was missing, it was supplemented with data from mixed Scottish beef farms recorded within the MIXED project (Moakes, S. and Oggiano, P., 2025), with default data from ecoinvent, Agribalyse or based on expert knowledge. An overview of input data and sources is provided in the SI Tables S4–S6. For the livestock herd (180 animals in total), the following livestock categories were considered: beef cows (calving 6 times before being culled), spring-born calves (0–6 months), overwintering suckler beef (6–12 months), finishing beef (male: 12–18 months; female: 12–20 month), as well as replacements units of beef cows. Livestock were assumed to be fed with concentrate feed produced off-farm (97.5 % soybean meal, 2.5 % mineral supplement) and on-farm produced winter barley, grass silage (from temporary and permanent grassland), and direct grazing on temporary and permanent grassland (Fig. 3A). The farm had a total size of 512 ha, including 115 ha permanent grassland and 397 ha under crop rotation. The assumed 6-year crop rotation consisted of wheat-oat-wheat-barley-grass-grass, where the wheat (Triticum aestivum), oat (Avena sativa) and a proportion of the barley (Hordeum vulgare) were sold, while the grass (mainly Lolium perenne) and remaining barley were used as feed and the cereal straw used as bedding. The available liquid and solid manures were assumed to be applied on temporary grassland only, while fertilization via direct excretion during grazing was directly allocated to permanent and temporary grassland. In addition, both types of grassland received mineral fertilizer to fulfil crop nutrient requirements. The on-farm produced crops were fully fertilized with mineral fertilizer, while standard inventory data was used for the off-farm produced concentrate feeds. Typical feed waste of 19 % was assumed, similar to assumptions within Agribalyse (Koch and Salou, 2016). 3.3. Feed-no-food scenarios Two different feed-no-food scenarios were introduced to the baseline farm. In a first scenario “Feed-no-Food” (FnF); (see Fig. 3B), no external concentrate feed was imported and the on-farm produced barley was also not fed to the cattle but sold for human consumption. The farmland area, crop rotation and grassland management were assumed to be S. Moakes et al. Agricultural Systems 232 (2026) 104560 6 Fig. 3. Farm system model of (A) baseline; (B) feed-no-food scenario; (C) circular feed-no-food scenario. System elements in grey font were not analysed here. Elements marked with plus or minus sign were adapted compared to the baseline (changes in resulting emissions are not illustrated). S. Moakes et al. Agricultural Systems 232 (2026) 104560 7 unchanged, and no external fodder was bought. It was assumed that the lower overall feed quality and quantity resulted in a two month slower growth of finishing beef (Doyle et al., 2023; FMHB, SAC Consulting, 2023), and therefore a smaller overall herd size (total 138 animals), leading to reduced manure production and meat output. In a second step, a “circular Feed-no-Food scenario” (cFnF) was introduced (Fig. 3C), still assuming the same land area, but including additional management changes for plant and livestock production to optimize productivity and better use synergies within the farm (i.e. increase circularity). For plant production, the crop rotation, permanent grassland and fertilization were adapted. The crop rotation was adapted to wheat-oat-fava beans-wheat-grass clover-grass clover, as legumes including fava bean (Aschi et al., 2017; FAS, 2024) and grass clover leys (Berdeni et al., 2021) reduce dependency on mineral N and have been shown to induce positive soil changes that enhance cereal productivity within a rotation. In doing so, the barley was replaced by fava beans (Vicia faba for human consumption) which in combination with the grass clover (e.g. Lolium perenne and Trifolium pratense), reduced the required amounts of N-fertilizer (mineral and organic), due to N-fixation by legumes (average N-fertilizer use reduction of 45 %). Also, the permanent grassland was improved by inter-sowing grass clover on 25 % of the area which reduced required manure and mineral fertilizer inputs (average N-fertilizer use reduction of 58 %). Hence, more manure was available to be applied to the arable fields, which further reduced mineral fertilizer inputs (see SI Table S5). Adding legumes to temporary grassland has shown to also increase the total biomass yield, thus we assumed an average of 25 % more available biomass (Barneze et al., 2020; Lüscher et al., 2014). For the livestock production, the productive lifespan of beef cows was increased by one year and cows were culled after 7 calvings (instead of 6 in the baseline), reducing the required replacement units. Feed waste was also assumed to be reduced by 4 % through more careful management (from overall 19 % based on typical Agribalyse values, to 15 % feed wasted), as waste reduction is a typical approach farms take when improving efficiency. Due to the improved grassland management, more feed (25 % yield increase in temporary forage with legume inclusion, (Lüscher et al., 2014)) and better quality (crude protein assumed to be +20 % for temporary and +25 % for permanent grass with legumes (Feedbase, 2025)), was assumed to be available, supporting higher livestock numbers (total herd size of 170) and meat output on the farm compared to the simple feed-no-food scenario. To understand, how much the assessment of changes in soil organic carbon (SOC) would affect results of all three scenarios, SOC changes of the crop rotation (including temporary grasslands) and permanent grasslands were modelled separately, using different time horizons (20 and 100 years) as recommended by e.g. Goglio et al. (2015) and Joensuu et al. (2021). First, the SOC of the baseline system was calculated. This data was then used as a starting value to assess how a management change (FnF or circular FnF) could change the carbon stored in soils. For the indicator on food production, we assumed that under both feed-no-food scenarios, human edible concentrate feed (imported soybeans, on-farm produced wheat, oat, barley, fava beans) would still be produced but directly consumed by humans. 3.4. Life cycle impact assessment (LCIA) Impacts were assessed with Impact world+v1.3 midpoint indicators (Bulle et al., 2019). The following impact categories were selected: Climate change, short term (referred to as Climate change); P-Eutrophication (Freshwater eutrophication); N-Eutrophication (Marine eutrophication); Acidification (Terrestrial acidification); Fossil and nuclear energy use (Energy use); Mineral resources use (Material use); Water scarcity; and Land occupation, biodiversity (Land occupation), which accounts for both the area used as well as the potential biodiversity impacts of different types of land use (Bulle et al., 2019). Each scenario was compiled and assessed in a separate FarmLCA Excel file and results were further analysed and processed in R software (R Core Team, 2025). 3.5. Sensitivity analysis In the FarmLCA tool, assumptions on crop yields and livestock productivity are entered based on primary data, literature or expert opinion, which can have a strong impact on LCA results per kg crop or meat. Therefore, sensitivity analyses were conducted on the circular FnF scenario, to reflect the uncertainty in the assumed changes in productivity of grassland (10 % lower or higher yields) and animal growth (+/−1 month of required fattening period before slaughter). These values are expert-derived realistic ranges. The area of each crop was assumed to be constant and the herd size that can be sustained with the on-farm feed was adapted based on feed availability estimation within the model. 4. Case study results 4.1. Food production The baseline system produced the highest meat output (101 t LW y −1 ), the FnF and circular FnF showed a 24 % and 6 % lower meat production, respectively. In terms of on-farm crop production (fresh matter, FM), 11 % more crops were produced in the FnF scenario compared to the baseline and 3 % less in the circular FnF scenario. The quantity of wheat and oats produced for sale was the same in all scenarios (1060 t FM wheat, 497 t FM oat), while in the FnF scenario the quantity of barley sold was increased by 70 % to 497 t FM. In the circular FnF, the barley was replaced by fava beans (total yield of 238 t FM). In terms of total edible FM (i.e. assuming typical losses until retail and that the 74 t FM of soybean meal could additionally be available for human consumption since not used as animal feed anymore), the FnF scenario produced 14 % more edible FM while the circular FnF scenario produced the same amounts of edible FM than the baseline (Fig. 4). Regarding the nutritional value, the FnF scenario produced 15 % more calories and 20 % resp. 25 % more proteins and DIAAS adjusted proteins compared to the baseline. The circular FnF scenario produced 3 % less calories than the baseline, but 31 % and 52 % more protein and DIAAS adjusted protein respectively. Across scenarios, the beef delivered 4–7 % of all DIAAS adjusted protein, while legumes contributed to 0–41 % and cereal to 55–93 %. The produced calories resulted from cereals (85–99 %), 0–14 % from legumes and 1 % from beef. 4.2. N-circularity The nitrogen use efficiency (NUE) of the baseline was 48 %, 54 % for the FnF and 56 % for the circular FnF scenario (Table 2). The nitrogen self-sufficiency of the animal production was 90 % for the baseline and increased to 100 % in both FnF scenarios. For the crops, the nitrogen self-sufficiency was 8 % for the baseline, 6 % for the FnF and increased to 55 % for the circular FnF scenario. 4.3. Environmental impacts 4.3.1. Livestock The baseline system beef impact for climate change was 21.3 kg CO 2 eq. per kg LW. Compared to the baseline, the FnF scenario showed lower impacts per kg LW for climate change and land occupation (−1 % and −11 %), but between 6 and 15 % higher impacts for all other impact categories (Fig. 5). The circular FnF scenario showed 6–38 % lower impacts than the baseline for all impact categories except for terrestrial acidification (+5 %). The environmental impact was dominated by on-farm forage production (all impact categories), enteric fermentation (climate change), manure management (acidification) and stall infrastructure (material, S. Moakes et al. Agricultural Systems 232 (2026) 104560 8 energy and water use). For the baseline, external feed production was relevant for climate change, land occupation and freshwater eutrophication, on-farm concentrate for energy use, land occupation and water scarcity. 4.3.2. Effect of including soil organic carbon When including SOC into the GWP calculation of 1 kg live weight beef, impacts change marginally for all scenarios (Table 3). For the circular FnF scenario, accounting for SOC changes would reduce the climate impact by 1.3 % over a 20-year time horizon and by 0.5 % over a 100-year time frame. For the other scenarios, accounting for SOC would alter the GWP of beef between −0.2 % and +0.1 %. 4.3.3. Crops The environmental impacts of 1 kg wheat and 1 kg oats remained unchanged between the baseline and the FnF scenario (Fig. 6 and Figs. S1–S2 in SI). For the circular FnF scenario, the impacts of both crops were lower than the baseline for climate change, energy use, material use and water scarcity (between −13 % and −3 %). For land use, they were the same as the baseline. Higher impacts than the baseline were 0 500 1000 1500 2000 Baseline FnF circular FnF total t e d i b le FM 0e+00 2e+09 4e+09 6e+09 Baseline FnF circular FnF total kcal 0 100 200 Baseline FnF c ircular FnF total t protein 0 50 100 150 Baseline FnF c ircular FnF total t DIAAS−corr. protein Commodity Soybeans* Beef Fava beans Barley Oat Wheat Fig. 4. Total amount of food produced on the farm and on external fields (imported soybeans that were previously used for feed, marked with asterisk) in the three scenarios based on edible fresh matter (FM), calories (kcal), protein and DIAAS corrected protein. For all indicators, typical losses until retail are considered. Table 2 Nitrogen inputs and outputs (in kg N / farm/ yr), nitrogen use efficiency and nitrogen self-sufficiency (in %) of the three scenarios. Inputs and outputs calculated according to the EU Nitrogen Expert Panel (2016). Inputs and outputs not present in this case study are not displayed. Parameter Baseline FnF Circular FnF Inputs Mineral fertilizers 68,457 69,508 33,048 Feed and fodder (import) 5145 0 0 Biological nitrogen fixation 0 0 35,538 Atmospheric N deposition 6149 6149 6149 Seed and planting material 1149 1149 1654 Manure (net import) -5 −2−3 Outputs Crop products, incl. straw 36,392 39,712 40,458 Livestock (net) 2419 1837 2278 Nitrogen use efficiency (NUE) 48 % 54 % 56 % Nitrogen self-sufficiency livestock production 90 % 100 % 100 % Nitrogen self-sufficiency crop production 8 % 6 % 55 % S. Moakes et al. Agricultural Systems 232 (2026) 104560 9 Oberholzer, H.R., Weisskopf, P., Gaillard, G., Weiss, F., Freiermuth, R., 2006. Methode zur Beurteilung der Wirkungen landwirtschaftlicher Bewirtschaftung auf die Bodenqualit¨ at in ¨ Okobilanzen – SALCA-SQ, p. 98. OECD, 2025. Measuring Carbon Footprints of Agri-Food Products: Eight Building Blocks. OECD Publishing. https://doi.org/10.1787/8eb75706-en. Oggiano, P., Weiner, M., Pfeifer, 2025. Netto-Null für die Stadt Zürcher Landwirtschaft: Ein erreichbares Ziel? Agrarforschung Schweiz 16, 96–102. https://doi.org/ 10.34776/afs16-96. Parton, W.J., Stewart, J.W.B., Cole, C.V., 1988. Dynamics of C, N, P and S in grassland soils: a model. Biogeochemistry 5 (1), 109–131. https://doi.org/10.1007/ BF02180320. Paustian, K., Lehmann, J., Ogle, S., Reay, D., Robertson, G.P., Smith, P., 2016. Climatesmart soils. Nature 532 (7597), 49–57. https://doi.org/10.1038/nature17174. Pelaracci, S., Goglio, P., Moakes, S., Knudsen, M.T., Van Mierlo, K., Adams, N., Maxime, F., Maresca, A., Romero-Huelva, M., Waqas, M.A., Smith, L.G., Oudshoorn, F.W., Nemecek, T., de Camillis, C., Grossi, G., Smith, W., 2025. Harmonizing soil carbon simulation models, emission factors and direct measurements used in LCA of agricultural systems. Agr. Syst. 227, 104361. https:// doi.org/10.1016/j.agsy.2025.104361. Pfeifer, C., Lozano-Jaramillo, M., Leiber, F., Walkenhorst, M., subm.. Extending the productive lifespan in Swiss dairy herds: impacts on greenhouse gas emissions, land use and gross margin. J. Dairy Sci.. Prost, L., Martin, G., Ballot, R., Benoit, M., Bergez, J.-E., Bockstaller, C., Cerf, M., Deytieux, V., Hossard, L., Jeuffroy, M.-H., Lecl` ere, M., Le Bail, M., Le Gal, P.-Y., Loyce, C., Merot, A., Meynard, J.-M., Mignolet, C., Munier-Jolain, N., Novak, S., Van Der Werf, H., 2023. Key research challenges to supporting farm transitions to agroecology in advanced economies. A review. Agron. Sustain. Dev. 43 (1). https:// doi.org/10.1007/s13593-022-00855-8. Public Health England, 2021. Composition of foods integrated dataset (CoFID). McCance and Widdowson’s “composition of foods integrated dataset” on the nutrient content of the UK food supply. (Version 2021) [Dataset]. https://www.gov.uk/government/ publications/composition-of-foods-integrated-dataset-cofid. R Core Team, 2023. R: A Language and Environment for Statistical Computing [Software]. https://www.R-project.org/. Renard, D., Tilman, D., 2019. National food production stabilized by crop diversity. Nature 571 (7764), 257–260. https://doi.org/10.1038/s41586-019-1316-y. Richner, W., Sinaj, S., 2017. Grundlagen für die Düngung landwirtschaftlicher Kulturen in der Schweiz (GRUD 2017). Agrarforschung Schweiz 8 (6), 1–276. Richner, W., Flisch, R., Mayer, J., Schlegel, P., Z¨ ahner, M., Menzi, H., 2017. Eigenschaften und Anwendung von Düngern. In: Grundlagen für die Düngung landwirtschaftlicher Kulturen in der Schweiz (GRUD) (Bd. 4). Roberts, A.J., Petersen, M.K., Funston, R.N., 2015. Beef Species Symposium: can we build the cowherd by increasing longevity of females? J. Anim. Sci. 93 (9), 4235–4243. https://doi.org/10.2527/jas.2014-8811. Rodriguez, C., Dimitrova Mårtensson, L.-M., Zachrison, M., Carlsson, G., 2021. Sustainability of diversified organic cropping systems—challenges identified by farmer interviews and multi-criteria assessments. Front. Agron. 3, 698968. https:// doi.org/10.3389/fagro.2021.698968. R¨ o¨ os, E., Mayer, A., Muller, A., Kalt, G., Ferguson, S., Erb, K.-H., Hart, R., Matej, S., Kaufmann, L., Pfeifer, C., Frehner, A., Smith, P., Schwarz, G., 2022. Agroecological practices in combination with healthy diets can help meet EU food system policy targets. Sci. Total Environ. 847, 157612. https://doi.org/10.1016/j. scitotenv.2022.157612. Rosenbaum, R.K., Bachmann, T.M., Gold, L.S., Huijbregts, M.A.J., Jolliet, O., Juraske, R., Koehler, A., Larsen, H.F., MacLeod, M., Margni, M., McKone, T.E., Payet, J., Schuhmacher, M., van de Meent, D., Hauschild, M.Z., 2008. USEtox—the UNEPSETAC toxicity model: recommended characterisation factors for human toxicity and freshwater ecotoxicity in life cycle impact assessment. Int. J. Life Cycle Assess. 13 (7), 532–546. https://doi.org/10.1007/s11367-008-0038-4. SAC Consulting, 2023. The Farm Management Handbook 2023/24. Seufert, V., Ramankutty, N., 2017. Many shades of gray—the context-dependent performance of organic agriculture. Sci. Adv. 3 (3). https://doi.org/10.1126/ sciadv.1602638. Sykes, A.J., Topp, C.F.E., Rees, R.M., 2019. Understanding uncertainty in the carbon footprint of beef production. J. Clean. Prod. 234, 423–435. https://doi.org/10.1016/ j.jclepro.2019.06.171. van der Werf, H.M.G., Knudsen, M.T., Cederberg, C., 2020. Towards better representation of organic agriculture in life cycle assessment. Nat. Sustain. 3 (6), 419–425. https://doi.org/10.1038/s41893-020-0489-6. Van Zanten, H.H.E., Herrero, M., Van Hal, O., R¨ o¨ os, E., Muller, A., Garnett, T., Gerber, P. J., Schader, C., De Boer, I.J.M., 2018. Defining a land boundary for sustainable livestock consumption. Glob. Chang. Biol. 24 (9), 4185–4194. https://doi.org/ 10.1111/gcb.14321. Weiner, M., Moakes, S., Raya-Sereno, M.D., Cooper, J., 2024. Legume-based crop rotations as a strategy to mitigate fluctuations in fertilizer prices? A case study on bread wheat genotypes in northern Spain using life cycle and economic assessment. Eur. J. Agron. 159, 127267. https://doi.org/10.1016/j.eja.2024.127267. Wernet, G., Bauer, C., Steubing, B., Reinhard, J., Moreno-Ruiz, E., Weidema, B., 2016. The ecoinvent database version 3 (part I): overview and methodology. Int. J. Life Cycle Assess. 21 (9), 1218–1230. https://doi.org/10.1007/s11367-016-1087-8. Wezel, A., Gemmill Herren, B., Bezner Kerr, R., Barrios, E., Goncalves, A.L.R., Sinclair, F., 2020. Agroecological principles and elements and their implications for transitioning to sustainable food systems. A review. Agron. Sustain. Dev. 40 (40), 13. Wild, M., Komainda, M., Bettin, K., Jürgens, K., Isselstein, J., 2025. Feed the green for a sustainable and protein-efficient dairy production. Agr. Syst. 223, 104216. https:// doi.org/10.1016/j.agsy.2024.104216. Willett, W., Rockstr¨ om, J., Loken, B., Springmann, M., Lang, T., Vermeulen, S., Garnett, T., Tilman, D., DeClerck, F., Wood, A., Jonell, M., Clark, M., Gordon, L.J., Fanzo, J., Hawkes, C., Zurayk, R., Rivera, J.A., De Vries, W., Majele Sibanda, L., Murray, C.J.L., 2019. Food in the Anthropocene: the EAT–Lancet Commission on healthy diets from sustainable food systems. Lancet 393 (10170), 447–492. https:// doi.org/10.1016/S0140-6736(18)31788-4. Yang, X., Xiong, J., Du, T., Ju, X., Gan, Y., Li, S., Xia, L., Shen, Y., Pacenka, S., Steenhuis, T.S., Siddique, K.H.M., Kang, S., Butterbach-Bahl, K., 2024. Diversifying crop rotation increases food production, reduces net greenhouse gas emissions and improves soil health. Nat. Commun. 15 (1), 198. https://doi.org/10.1038/s41467023-44464-9. Zhen, H., Goglio, P., Hashemi, F., Cederberg, C., Fossey, M., Trydeman Knudsen, M., 2025. Toward better biodiversity impact assessment of agricultural land management through life cycle assessment: a systematic review. Environ. Sci. Technol. 59 (15), 7440–7451. https://doi.org/10.1021/acs.est.5c02000. S. Moakes et al. Agricultural Systems 232 (2026) 104560 16