Circularity in Europe strengthens the sustainability of the global food system

The global food system faces multiple pressures, including pandemics, geopolitical conflicts, resource depletion, biodiversity loss and climate change, while population growth continues. This creates a paradox: food production is essential for life but current practices threaten future food production capacity. The central question addressed is how to provide healthy diets for all while safeguarding planetary health. Circular food systems, which minimize waste and recycle unavoidable residues, have emerged as a promising redesign strategy and are a key EU priority. Prior work has highlighted the role of animals as recyclers of by-products, but a holistic assessment of circular food systems is lacking. This study assesses how redesigning the EU27 + UK food system using circularity principles can secure food availability and minimize environmental impacts by evaluating three scenarios of increasing circularity that include changes in consumption, crop and animal production, and fertilizer use within a self-sufficient European system.

Existing studies demonstrate environmental benefits when livestock are fed non-food-competing feedstuffs, such as by-products and grass, and have explored EAT-Lancet-aligned diets and reductions in animal-derived foods. However, many modeling studies remain constrained by linear food system assumptions and provide only incremental improvements. Evidence shows that healthier diets can reduce GHG emissions and land use substantially (for example, a WHO-guideline diet reducing GHG by ~29% and cropland by ~7%; partial replacement of animal foods with plant-based foods reducing GHG by 25–40% and cropland by ~23%). Despite these insights, a comprehensive circular food system assessment integrating supply- and demand-side measures, feed-food competition, by-product recycling, and nutrient adequacy across the EU27 + UK has been missing. The role of animals as recyclers remains important, but must be balanced with nutrient adequacy (for example, vitamin B12) and system constraints.

The authors developed the Circular Food System (CiFoS) model, a biophysical, data-driven optimization model coded in GAMS, to explore circular redesigns of the EU27 + UK food system. CiFoS minimizes agricultural land use or maximizes population fed (depending on scenario) while meeting dietary constraints for nutrients, protein, energy, and EAT-Lancet food group bounds. Emissions are calculated ex post and constrained to avoid unrealistic transport. Scenarios: (1) AgriBase (baseline), calibrated to empirical FAOSTAT production, land use, exports/imports, and minimizing deviation from current food supply; (2) CirAgri, applying supply-side circularity while meeting current protein supply at the food-group level and minimizing land; (3) CirHealth, combining supply-side circularity with healthy diet constraints (EAT-Lancet ranges and EFSA nutrient requirements) and minimizing land; (4) CirPop+, combining supply-side circularity with healthy diets and maximizing the number of people nourished using EU land. Circularity principles include reducing waste and recycling unavoidable residual streams (food losses and waste, by-products, human excreta) as fertilizer or feed, prioritizing non-food-competing feed for livestock, and closing biomass and nutrient cycles. Data and system components: - Human nutrition: 42 nutrients tracked (macros, micro, vitamins, essential amino acids) using USDA SR Legacy (2018) composition; EFSA requirements applied in CirHealth and CirPop+; EAT-Lancet bounds govern food group intakes. - Cropping system: 43 food crops and 8 fodder crops; crop choice constrained by climate-soil zones (850 zones from GAUL, GAEZ v4, IPCC soil classes), observed yields/areas from SPAM 2010 v2.0 with disaggregation using EARTHSTAT for vegetables and tree nuts; crop rotations implemented as average shares with frequency constraints (for example, sunflower max once per 6 years), static equilibrium; fertilizer requirements for N and P derived from crop harvest and residues with fixed N:P in organic fertilizers; AgriBase uses current artificial fertilizer (12.3 Mt N+P) and sludge; circular scenarios allow organic sources (manure, compost, human excreta) complemented by artificial fertilizer as needed. Total agricultural land is 171.7 Mha (FAOSTAT), split into cropland, temporary grassland, permanent pasture, and rangeland with harmonized maps (HYDE, Yu et al. 2020). - Livestock and aquaculture: dairy, beef, pigs, broilers, layers, and farmed fish (Atlantic salmon, Nile tilapia) with productivity levels and full herd/stock accounting; rations formulated from co-products, food waste (monogastrics/fish only), grass resources, and limited human-edible biomass; ruminants excluded from food waste; feed-food competition minimized via prioritization of non-food-competing feeds. - Fisheries: capture fisheries constrained to maximum sustainable yield (MSY) for key Northeast Atlantic stocks aligned with EU quotas; edible fractions and by-products accounted for feed-food competition. - Residual streams: by-products from processing using technical conversion factors; food losses and waste quantified by stage and product group, with 35% of consumption waste assumed usable as wet feed for monogastrics/fish; manure partitioned between grazing deposition and manure management systems; 36% of sewage sludge assumed available as fertilizer with specified N and P contents. - Transportation: processed products transported by truck between EU countries; food waste and grass used in country of origin. - GHG accounting: IPCC tier 2 for livestock (enteric CH4, manure CH4/N2O, grassland N2O) and aquaculture N2O; IPCC tier 1 for cropping N2O and fertilizer production emissions; composting CH4/N2O; transport CO2; peatland N2O from drained organic soils; global warming potentials: CH4=28, N2O=265 (100-year). Emissions are calculated per scenario; a per-country transport emission cap of 1,500 kg CO2e per capita per year prevents unrealistic flows.

• Baseline (AgriBase): 644 MtCO2e per year (1.17 t CO2e per person per year, 2020 population), agricultural land use 172 Mha, consistent with FAOSTAT.
• CirAgri (supply-side circularity, meet current protein supply): GHG −22% to 515 MtCO2e (0.91 t CO2e per person per year); agricultural land −71% to 50 Mha. Cropland use −53% to 50 Mha; grassland area nearly eliminated. Reduced cereals and fodder crops; increased diversity with more pulses, sorghum, vegetables, and tropical fruit. N and P fertilizer use reduced versus AgriBase; food waste largely used as animal feed with manure used as fertilizer.
• CirHealth (supply + healthy diet): GHG −29% to 458 MtCO2e (0.83 t CO2e per person per year); agricultural land −71% to 50 Mha (cropland 49 Mha; grassland nearly eliminated). Larger crop diversity than CirAgri; N and P fertilizer use reduced. Protein intake 64 g/person/day meeting EFSA and EAT-Lancet constraints; all macro/micronutrients within recommended bounds.
• CirPop+ (healthy diet, maximize population): EU can nourish EU27+UK plus an additional 767 million people (+149% relative to EU27+UK population). Per capita GHG −38% to 0.72 t CO2e/person/year, but total GHG +55% to 998 MtCO2e due to higher population served. Total agricultural land ~167 Mha (−3% vs current); cropland 102 Mha (−3% vs current cropland), grassland 66 Mha (−2%). Artificial fertilizer use (N and P) increases relative to other circular scenarios due to limited organic fertilizer availability.
• Diets and protein: Current actual protein intake estimated at 69 g/person/day (after subtracting losses); current supply corresponds to 103 g/person/day. CirAgri provides 83 g/person/day; CirHealth and CirPop+ provide 64 g/person/day. Animal:plant protein ratio shifts from ~60:40 (current/CirAgri) to 37:63 (CirHealth) and 34:66 (CirPop+). Animal protein intake decreases ~51% (to 24 g/person/day) in CirHealth and ~55% (to 22 g/person/day) in CirPop+, driven mainly by reduced red and chicken meat.
• Livestock sector: Under healthy circular diets, large reductions vs AgriBase in beef (−91% CirHealth; −99% CirPop+), pigs (−78%; −100%), broilers (−79%; −73%), layers (−33%; −93%); dairy and fish show small changes in CirHealth and large increases in CirPop+. Feed-food competition declines across circular scenarios, yet fodder crops remain important, especially in CirAgri.
• Fertilizers: AgriBase uses 12.3 Mt artificial N+P. CirAgri and CirHealth reduce artificial N and P use; CirPop+ requires relatively high artificial fertilizer use due to limited organic sources when livestock numbers are low and population served is high.

Redesigning the EU27 + UK food system around circularity can deliver nutritionally adequate diets while substantially reducing agricultural land use and per capita GHG emissions. Despite reductions, per capita emissions remain above the proposed planetary boundary of ~0.5 t CO2e per person per year, indicating a need for additional measures. Key system insights include: (1) crop rotations constrain circularity potential and can drive fodder crop production when food crop options are limited, underscoring the need for greater crop diversification; (2) animals remain important recyclers to meet nutritional adequacy (for example, vitamin B12), though herd sizes decrease markedly under healthy circular diets; alternative nutrient requirements or supplementation could further shift animal-plant protein ratios; (3) organic fertilizers alone are insufficient, especially with low livestock numbers; increased use of green manures and safe nutrient recovery from human excreta could help close nutrient cycles; (4) reduced grassland use in circular scenarios enables rewilding opportunities and biodiversity gains or low-density multifunctional grazing; (5) shifting consumption towards healthy diets with more fruits, vegetables, whole grains, legumes and plant oils is essential, complemented by processing choices that improve nutrient bioavailability and enabling food environments; (6) self-sufficiency and circularity likely reduce indirect land-use change emissions associated with imports, though these indirect effects were not quantified here. Achieving circularity requires coordinated, systemic transitions across production, processing, distribution and consumption, with social acceptance and economic transformation as critical enablers.

The CiFoS-based assessment shows that transitioning the EU27 + UK food system to circular principles can substantially reduce agricultural land use and per capita GHG emissions while meeting healthy dietary requirements and, under a global shortage scenario, significantly increasing the number of people nourished using European land. Circular redesigns depend primarily on systemic reorganization of existing practices rather than novel technologies, but require societal acceptance, policy alignment and economic restructuring. The study provides a comprehensive blueprint of dietary shifts, crop diversification, livestock system redesign, and nutrient cycling needed to safeguard human and planetary health. Future research should refine nutrient recovery from human excreta, expand crop portfolios and rotations compatible with diverse diets, evaluate trade-offs across feed-food-fiber-fuel systems, and quantify indirect global effects of EU self-sufficiency, including land use change and trade dynamics.

The analysis focuses on circularity within the food system and does not fully integrate feed-food-fiber-fuel competition. Emissions from indirect land-use change (linked to trade) and broader global market feedbacks are not quantified. The model is static (average rotation shares) and omits crop sequence constraints. There are notable data uncertainties and inconsistencies (for example, FAOSTAT production/consumption, fertilizer use statistics vs yields, land area maps across years, harmonization of dietary guidelines). Assumptions include 36% availability of sewage sludge for agriculture, fixed N:P ratios for organic fertilizers, full availability of organic fertilizer nutrients at steady state, and restrictions on ruminants consuming food waste. Nutrient adequacy relies on EFSA recommendations; alternative requirements or supplementation (for example, vitamin B12) could change outcomes. Results are calculated for GHG emissions and not optimized, and per-country transport emissions are capped to avoid unrealistic flows.