Abstract
The need for a livestock transition is widely acknowledged, entailing both production- and consumption-side adaptations. Yet the direction of this transition remains uncertain as a diversity of narratives and visions coexist. In this paper, we test the outcomes of dominant narratives and recurring elements of debates surrounding the livestock transition. We adopt a quantitative foresight scenario approach and apply it to the case of the Belgian livestock sector, testing four possible scenarios towards 2050 (Business as usual; Land sparing; Land sharing; Radical). Paying particular attention to the diversity of livestock farming systems and the inclusion of a multidimensional set of sustainability indicators, the results reveal the trade-offs that dominating visions entail. As no perfect scenario stands out, the exercise opens a space of options rather than singling out a unique vision. Beyond providing quantitative scenario results, we also take the opportunity to reflect on methodological choices and the usefulness of foresight approaches to advance sustainability transitions.
Similar content being viewed by others
Introduction
There is no question that current animal production and livestock sectors face important tensions and need to undergo a transition. Within each sustainability dimension, livestock may contribute to both positive and negative impacts1,2,3,4,5,6,7. At present, it has become clear that costs generally outweigh benefits and that the current extent of animal production has become unsustainable8. To relieve those tensions, two main approaches can be pursued, seeking to resolve two distinct questions9: (1) a quantitative question asks how many livestock should be produced and consumed to answer nutritional needs, meet societal expectations and remain within planetary boundaries; (2) a qualitative question asks which livestock should be produced, i.e. through which practices and farming systems can we produce animal products in a sustainable way.
Trying to solve these questions is no easy feat as food system transitions have to deal with the coexistence of several visions and narratives, and a multitude of potentially conflictual answers emerge10,11,12,13. In the case of the livestock sector, Jaisli and Brunori14 identify three main narratives: an efficiency path, pursuing sustainability through increased efficiency of production (sustainable intensification); a consistency path, aiming for circularity of livestock production and grassland preservation; and a sufficiency path, aiming to pursue a protein transition (i.e. the rebalancing of animal versus plant based protein in both production and consumption). Additional narrative debates can be found in the case of agroecology vs. sustainable intensification15, or in the case of land sparing vs. land sharing, i.e. either separating high-productivity farming from nature conservation, or pursuing both productive and conservation objectives on the same land16,17. A variation of the land sparing/sharing framework advocates for a three-compartment model, adding intermediate extensively managed farmland between the productive and conservation areas18,19. The sourcing of animal feed is another recurring element of discussion, trying to resolve a feed/food competition by deliberating on the use of agricultural land to produce human food or animal feed (or for other purposes). In this effort, alternative visions focus on decreasing the reliance on high-impacting feed ingredients such as soybean meal20,21, or developing low-cost livestock that valorises coproducts rather than primary agricultural products22,23,24. These coexisting visions are often linked to different disciplines and communities of practice, which although they agree on the general unsustainability of our food system, disagree on its causes and what food and farming systems should be prioritised as a result15,25. From a transition perspective, the compatibility and complementary of this diversity of visions remain an open question26,27,28. Thus, a third question emerges besides aiming to relieve the tensions faced by the livestock sector, asking whether we can progress towards a shared vision on the first two questions.
Two theoretical frameworks can be useful to ground these three questions and help to operationalise them. First, from a transdisciplinary perspective, research can contribute to societal problems through three types of knowledge26,29: systems knowledge (asking why and how processes occur and where change is needed), target knowledge (asking what better practices we should aim for) and transformation knowledge (asking how we can get from the current to the desired situation). Second, rooted in transition governance and transition management, several transition frameworks have been developed for the transition of agriculture and food systems30,31,32,33,34. Despite some variations in terms of specific steps or targets within the food system (farm-level, territorial-level, policy-level), all these frameworks share a step-by-step, often iterative, and goal-seeking approach, which could be summarised by the following sequence: assess (current/new situation)—envision (possible and desired futures)—implement (transition pathways). These steps largely align with the three types of transdisciplinary knowledge: systems knowledge is created during the assessment step, target knowledge is generated during the envisioning step and transformational knowledge fosters the implementation step (Fig. 1).
Three main stages of a transition cycle based on existing transition management frameworks, and three main types of transdisciplinary knowledge.
Foresight scenarios are commonly used tools in transition governance frameworks35. Through the use of scenarios, it is possible to discuss future developments and define long-term goals for which short-term targets can be set as well36. While quantitative scenarios provide quantified insights on the outcomes of alternative futures, qualitative scenarios develop narratives and storylines providing an opportunity to better capture contextual insights, uncertainties and challenges which cannot always be modelled37. In the case of quantitative scenarios, a diversity of approaches and methods can be found and a few attention points stand out38. With regards to the scope of scenario exercises, adopting food system perspectives ranging from production to consumption is important to gain a system-wide overview of challenges. Sectoral approaches targeting specific commodities and sectors (e.g. dairy, vegetables, cereals, etc.) are promising as they are more operational and closer to food system actors, thus facilitating possible implementation39. On the downside, by overemphasising the challenges of one particular sector, they risk overlooking sectoral interactions (e.g. crop-livestock) and system-wide perspectives. The spatial scale of analysis adds further complexity to the question of implementation, as large-scale exercises are broader in scope but harder to translate locally, while more localised exercises are easier to operationalise but risk missing interactions and constraints at higher levels38. These considerations apply both to the spatial extent (i.e. considered territory) and the spatial resolution (unit of analysis)40, with implications either for spill-over effects (trade-offs with other territories), or in terms of governance (defining the relevant governance levels). With regards to the diversity of visions and systems, scenario studies adopt different approaches depending on the purpose of the exercise, either assessing the feasibility of pre-identified visions (normative scenarios) or exploring a range of possibilities (exploratory scenarios). The diversity of systems and practices (e.g. farming systems) is not always considered in scenario studies, particularly in normative approaches. From a transitions perspective, accounting for diversity is important for two main reasons. First, as farming practices and systems are highly diverse within one sector, so are their sustainability outcomes. Highlighting diversity thus enables a better understanding of challenges and possible solutions (e.g. by identifying “promising” systems). Second, it allows overcoming polarising oppositions and better tailor the proposed solutions across different levels of action, thus bridging between practices and policies11,41. With regards to sustainability, while accounting for its multidimensional nature is largely promoted, few assessments effectively cover all sustainability dimensions38,42. Finally, the impact and contribution of scenario studies to long-term transition processes remain uncertain and understudied38,43.
In this paper, we seek to contribute to the three questions surrounding the livestock transition through the use of foresight approaches, specifically focussing on the case of the Belgian livestock sector. Belgium is no stranger to the debates and tensions surrounding the livestock sector. Animal production occupies an important place in the Belgian food and farming system. The country is a net exporter of livestock products, with self-sufficiency ratios above 100% for the five main productions (dairy, beef, pork, poultry and eggs), and over 200% for pork, poultry or skimmed milk powder, reaching respectively 214%, 216% and 203% in 202344,45. While close to half of farms are specialised in livestock productions in the country’s two main regions (Flanders in the North and Wallonia in the South)46,47, there are important differences in the sectors that developed in both regions and the resulting challenges. In Flanders, a higher population density (and thus increased pressure on land) and easy access to imported animal feed have led to adopting more intensive systems, in particular of monogastric animals. Flanders concentrates the vast majority of Belgian pigs and poultry (respectively 94% and 85% in 202348). This poses questions with regard to the dependence on imported feed, in particular soy, and its associated impacts in terms of land use change20,21. As other European countries and regions relying on intensive animal farming and high feed imports49,50,51, the region faces important challenges in terms of nitrogen pollution due to manure surpluses. This leads to pollution by air through the deposition of ammonia (NH3) and nitrogen oxides (NOx) causing eutrophication of natural areas, and to pollution of groundwater and surface water bodies through the leaching of nitrates. Aiming to tackle this issue and to comply with European legislation, the Flemish government has introduced two main policies, with recent updates in 2024. A manure decree is designed to tackle nitrogen and phosphate leaching into surface and groundwater bodies, while a nitrogen decree seeks to decrease ammonia and nitrogen oxide emissions in Flanders. The latter aims to reduce the region’s NH3 emissions by 40% by 2030 compared to 2015 levels, with sector-specific targets for ammonia emisisons in stables (−60% for pigs and poultry, −15% for dairy and beef). It also includes a 30% reduction target for the pig population by 2030. Wallonia on the other side is characterised by more available land and, in particular, an ample supply of permanent grasslands (42% of the region’s agricultural land in 2023, compared to 26% in Flanders or 30% at EU level48). It is therefore more suited for extensive systems, particularly of cattle production52,53. Yet it is not exempt of challenges either. As observed in other European regions, grassland areas have been decreasing over the past years and grass-based cattle systems are being replaced by maize-based systems which tend to be more input-intensive (pesticides and fertiliser) and linked to negative environmental outcomes (e.g. nutrient leaching and erosion)54,55,56. In parallel, cattle farmers face dire economic situations, with beef farmers in particular facing structurally low income levels and a high dependence on income support subsidies52,57. On the consumption side, current dietary patterns in Belgium have been shown to cause significant environmental impacts, particularly in terms of climate change and land use associated to ultra-processed food and animal products58. The resulting call for a protein transition in Belgium is being echoed at policy level, for example, through a Green Deal Protein shift in Flanders. Thus, the Belgian livestock sector is under significant pressure on both the production and the consumption side. It faces several sustainability challenges and needs to embark on alternative development pathways to meet its future challenges. For these reasons, it makes a particularly well-suited case study for this research.
In summary, this paper pursues a double objective: first, we seek to contribute contextual and objectivation elements regarding the Belgian livestock sector and possible transition scenarios. Four scenarios are tested. They simulate prevailing narratives or recurring elements of discussion in the debates surrounding the future of the livestock sector (Table 1). By adopting a territorial (all agricultural production and food consumption activities on the Belgian territory) and sectoral (five main livestock sectors: dairy, beef, pork, poultry and eggs) approach, we aim to conduct our research at a level of analysis which, from a research perspective, is believed to be actionable and operational for food system actors to facilitate potential implementation30,39. We strive to display the existing diversity of visions (by considering multiple scenarios) and of systems (by considering multiple farming systems and consumer diets) to remain open to multiple possible transition pathways (exploratory approach) and allow actors to situate themselves in the possible future scenarios. We include a variety of sustainability indicators in the assessment to pay attention to the multidimensional nature of the issue and reveal potential trade-offs. Second, by reflecting on the adopted process, we hope to improve our collective understanding on the complexity of such transition processes and the usefulness of foresight approaches. To do so, this research adopts an iterative and incremental approach building on a previous foresight exercise that was carried out on the Belgian livestock sector in a transdisciplinary research context in collaboration with civil society and sectoral actors9,59. By discussing methodological choices and looking at the perception and evolution of foresight exercises over time, we seek to reflect on the usefulness of such approaches for transition processes. So doing, we strive to contribute to all three types of knowledge: systems knowledge (current status of the Belgian livestock sector), target knowledge (possible futures and their outcomes) and target knowledge (progress towards a more sustainable livestock sector).
Results
Livestock numbers
Livestock numbers decrease in all scenarios, with variations in affected sectors (Fig. 2c). The business as usual scenario (BAU) scenario shows little change. In ruminant sectors, the expected decrease in the beef sector (suckler cows) is compensated by a continued increase in the dairy sector (dairy cows). In monogastric populations, the planned decrease in the pork sector (–30% as per the nitrogen decree) is counterbalanced by an increasing trend in broiler populations (Table S9). In the transition scenarios, livestock reductions are increasingly bigger from T1 through T3. While the land sparing scenario strongly reduces ruminant populations (for climate purposes), monogastric sectors remain less affected compared to scenarios T2 and T3. In the latter, monogastric populations are strongly reduced due to available feed options (EU-origin protein sources in T2 and EU-origin coproducts in T3), while ruminant populations are slightly less affected given the choice to preserve grasslands (Table 2).
a evolution of crop farming systems (% of agricultural area); b evolution of animal farming systems (% of animals); c evolution of livestock populations (livestock units, LSU).
Farming systems
All scenarios see a progression of organic and extensive systems (Fig. 2a, b). Organic livestock production remains limited in the BAU scenario (6% of animals) and increases incrementally in the transition scenarios, reaching 100% in scenario T3. Within non-organic systems, scenarios BAU and T1 remain dominated by productive systems (either as a result of current trends in scenario BAU or for productivity reasons in scenario T1), while scenario T2 engages in an extensification process (70% of animals in farming systems considered as extensive).
Agricultural land
Agricultural land (Fig. 2a) is left untouched in the BAU scenario. In the transition scenarios, a share of land is set aside for biodiversity conservation. This share is higher in scenario T1 (20%) following a three-compartment land sparing approach advocated for Belgium60, which maintains productive farming systems and separates them from biodiversity conservation land. Scenarios T2 and T3 also set aside some land for biodiversity (10%) but this share is more limited compared to T1.
Animal feed
Demand for concentrate animal feed is strongly reduced in the scenarios, ranging between –14% in scenario BAU and –79% in scenario T3 (Table S9). With regards to protein sources, scenarios T2 and T3 are soybean-free. The soybean meal demand represents over 300,000 ha in the current situation and in scenario BAU (i.e. around 25% of Belgium’s total agricultural area) and 200,000 ha in scenario T1 (Table S9). In scenario T3, the exclusive use of (EU-origin) coproducts to feed animals implies that no cereals are destined to animal feed. In terms of feed self-sufficiency (i.e. capacity of sourcing animal feed at Belgian or EU level), scenarios BAU and T1 entail little shifts compared to the current situation (except an increase in the self-sufficiency for protein-rich coproducts in scenario T1). Scenario T2 and T3 offer more perspectives in that regard. Both scenarios are self-sufficient in terms of protein-rich coproducts (i.e soybean meal alternatives) at EU level and cereal feed at the Belgian level (scenario T3 does not require any). Additionally, scenario T3 is also self-sufficient in protein-rich coproducts at EU level (given the objective to rely exclusively on coproducts) while in scenario T2 available energy-rich coproducts cover nearly 50% of the demand for energy-rich feed (Table 2 and Fig. 3).
All plots show the share of organic farming (% animals) on the x-axis. First row (top): biodiversity impact (Damage score) and nitrogen balance (% N from manure production/organic N demand for crop fertilisation). Second row: greenhouse gas emissions (kt CO2e/year) and coproduct feed self-sufficiency (% coproduct feed offer/coproduct feed demand). Third row: animal protein production (g animal protein/cap/day) and animal protein self-sufficiency (% animal protein production/animal protein demand). Fourth row (bottom): meat consumption (g meat/cap/day) and meat self-sufficiency (% meat production/meat consumption).
Livestock production
The production of meat decreases in all scenarios compared to the current situation. Reductions in total meat production vary between −10% in scenario BAU and −84% in scenario T3. Milk production only decreases in scenario T1 (–25%) while it increases in scenario BAU (+18% as per current trends) and in scenarios T2 and T3 (+10 and +18% as a result of the shift to dual-purpose cattle systems). Overall, total production of animal protein decreases are comprised between –10% in scenario BAU and –60% in scenario T3 (Table 2 and Fig. 3).
Human food
In terms of human consumption, the production potentials of scenarios have varying capacities to cover the demands for meat and animal protein, which vary according to the considered diets (Table 2 and Fig. 3). While scenario BAU maintains the current diet, scenarios T1, T2 and T3 are assessed under increasingly ambitious diets in terms of animal product consumption (respectively the TYFA diet, the Belgian FBDG and the EAT-Lancet diet). In scenarios BAU and T1, the production is largely sufficient to meet and even exceed meat demand (according to the current diet and the TYFA diet, respectively). In scenarios T2 and T3, the lower production potentials would require greater efforts in terms of meat consumption. Scenario T2 falls just short with regards to the Belgian FBDG while scenario T3 is nearly aligned with the EAT-Lancet diet. Looking at total animal protein, all scenarios exceed the theoretical demand for animal protein intake, assuming that 50% of total protein intake (estimated at 0.8 g protein/kg bodyweight) should come from animal products in the perspective of rebalancing animal and plant-based protein sources. As the scenarios shift to less animal-based and more plant-based diets, they lead to a growing demand for protein-rich crops for human consumption. While the demand for legumes in the current diet (current situation and scenario BAU) represents about 5000 ha (less than 1% of Belgium’s agricultural area), this demand increases to around 50,000 ha in the TYFA diet (scenario T1), 90,000 ha in Belgian FBDG diet (scenario T2) and 200,000 ha in the EAT-Lancet diet (scenario T3), representing respectively 4%, 7% and 16% of Belgium’s agricultural area (Table 2).
Export capacity
Assessing the self-sufficiency of meat and other animal products has implications for the potential exporting capacity of Belgium (Table 2 and Fig. 3). Scenarios BAU and T1 maintain significant exporting capacities (meat self-sufficiency levels of 155% and 189%, respectively, compared to 192% in the current situation). In scenarios T2 and T3, as production decreases are more important, the capacity to export becomes limited (with self-sufficiency levels of 97% and 95%, respectively). Dietary shifts offer a certain flexibility to adjust the exporting capacity as ambitious reductions in animal product consumption increase the self-sufficiency levels and thus the export potential (e.g. adopting the EAT-Lancet diet in scenario T2 would raise the self-sufficiency for meat to 170%, thus leaving a surplus of production which could be exported; Fig. 3).
Environmental impacts
The scenarios present contrasted situations with regards to environmental impacts (Table 2 and Fig. 3). All scenarios lead to reductions in absolute greenhouse gas (GHG) emissions, which are limited to –11% in the case of scenario BAU, while scenarios T1, T2 and T3 lead to increasing reductions of –38%, –46% and –54%, respectively (Fig. 3). While overall reductions for these three scenarios are close, the strategies are different (Table S9). Scenarios T2 and T3 show significant reductions in animal sector emissions (given the important decreases in animal populations). Scenario T1 exhibits smaller reductions for the animal sectors, and comparatively greater reductions in the plant-based sectors, as well as important negative emissions due to the conversion of agricultural land to biodiversity-rich land (considered to be forests). Reductions in this scenario reach −59% when accounting for the average sequestration of forests over the full scenario period (2022−205) rather than for the final annual emissions in 2050 (see Table S9). Decreases in relative GHG emission levels (per kg animal protein) are similar across scenarios, either resulting from the shift to more intensive and efficient systems in scenario T1, or from the decreased reliance on soybean meal in scenarios T2 and T3 (Fig. S1). With regards to the nitrogen cycle, the decrease in animal populations allows to reduce the production of animal manure, N emissions and NH3 emissions, thus decreasing the existing overproduction of animal manure with regards to cropping fertilisation needs and the resulting pressure on water and soil resources. Only scenario T2 seems to present a well-balanced situation when comparing manure production to organic nitrogen needs. Scenarios BAU (166%) and T1 (140%) remain in an excess situation, although improved compared with the current situation (193%), while scenario T3 results in a deficit of animal manure (73%). Finally, with regards to biodiversity impact, setting aside agricultural land for biodiversity conservation and shifting to organic systems decreases the potential impact on biodiversity. The greater reduction is achieved in scenario T1 (–20%), followed by scenario T3 (–15%) and scenario T2 (–10%), while scenario BAU shows little improvement compared to the current situation (–1%).
Socio-economic impacts
With regards to employment, the total number of FTEs remains similar in scenario BAU (−3%) but decreases by about 25% in the transition scenarios, mainly as a result of the reduced animal herd and agricultural area. Relative employment rates (per ha or per animal) show little variation for the plant-based sectors but show more variation for the animal sectors, with significant increases in scenarios T2 and T3 (Table 2). Gross value of agricultural production is increasingly reduced through the scenarios, reaching –45% in scenario T3, mainly due to the reduced sizes of the animal sectors. Relative value of production (per ha or per animal) shows signs of decrease in the case of the plant-based sector in scenario T3, but as in the case of employment, shows more variation for the animal sectors, with significant increases in scenarios T2 and T3 (Table 2).
Discussion
Along with a baseline scenario representing the current situation (2018–2022), the scenario exercise comprises four scenarios for the Belgian livestock sector in 2050. A BAU is complemented with three transition scenarios representing possible and increasingly greater shifts away from the current trends. Below, we discuss what we can learn from the results and from the methodological approach. Points of discussion include the attention for diversity of systems and visions, the multidimensional nature of the issue and the involved trade-offs, the implications of adopting sectoral and territorial approaches, and the general usefulness of foresight approaches to advance the livestock transition.
An objective of the study was to test a diversity of visions, i.e. currently prevailing narratives surrounding the future of livestock. The land sparing-land sharing continuum was tested by integrating both a three-compartment land sparing vision18,19,60 and a land sharing approach (scenarios T1 and T2, respectively). The three dominating livestock narratives identified by Jaisli and Brunori14 are reflected in our scenarios. Scenarios BAU and T1 are closer to an efficiency path, while scenarios T2 and T3 show similarities with the consistency and sufficiency visions. The scenario exercise also tackles a number of feed-related issues, such as the possibility of halting the use of soybean meal20, tested in scenarios T2 and T3, or the implications of a low-cost livestock sector relying exclusively on coproduct feed and thus resolving the feed/food competition22,23,24, imagined in scenario T3.
Reflecting such a diversity of visions is facilitated by the integration of a diversity of farming systems (in this case, livestock systems). For instance, the land sparing vision tested in scenario T1 is simulated by favouring highly productive systems while keeping organic and extensive systems more limited (as well as operating greater reductions in ruminant sectors compared to monogastric ones). Inversely, extensive systems, particularly grass-based ruminant systems, dominate in the land sharing vision tested in scenario T2, while intensive monogastric systems are strongly reduced (Fig. 2). Scenario T3 embodies a more radical vision entailing a fully organic scenario tested in several studies61,62,63,64. Different visions thus favour different systems, confirming that systems and visions are to some extent connected11,65. Systems can be seen as practical translations and implementations of visions, and conversely, visions can serve as argumentations in favour of certain systems. Besides the differences between animal products (inter-sectoral shifts), integrating this diversity adds additional nuance to sector-specific evolutions (intra-sectoral shifts) within each scenario. Discerning between different types of farming systems allows linking the exercise to the level of practices, thus facilitating identification to the scenarios for food system actors (e.g. farmers). Nevertheless, the practical implementation of a system is not limited to affinities with a certain vision, but might be constrained by external factors such as social, political, cultural or environmental contexts66.
Trying to answer the first two question related to the livestock transition (how many livestock and which livestock) is not straightforward as no ideal scenario emerges from the results. Trade-offs appear when comparing scenarios in terms of production, nutrition, climate change, nitrogen cycle, biodiversity preservation and socio-economic outcomes (Fig. 3). Yet, a few conclusions can still be drawn from the exercise.
Regarding the ambition of the scenarios, given the unsustainability of the current livestock sector, the results shows that a business as usual pathway is not enough (Fig. 3). While maintaining high production levels, it shows little progress on all considered environmental aspects, leading to minor decreases of GHG impacts (–11%), insignificant improvements in terms of biodiversity impacts (–1%) and little improvements on the nitrogen balance (surplus of 166%), despite planned measures to reduce part of the livestock herd (30% reduction of the pig population), or expected technology and efficiency gains. Specifically for Flanders, the scenario fails to meet the NH3 reduction objectives for all sectors except the beef sector (Table S10). This confirms that a significant turn and a radical shift away from current trends must be undertaken if the Belgian livestock sector is to reduce its impacts locally and contribute to global efforts, as pointed out in previous examples67,68.
Turning to the transition scenarios, no silver-bullet scenario stands out, but the results allow to draw certain conclusions. For instance, setting aside land for biodiversity brings significant benefits in terms of biodiversity conservation and climate goals. Yet, compartmentalisation with highly productive monogastric systems as in the land sparing approach does not solve nitrogen cycle balance issues or dependence on imported and high-impact animal feed such as soybean meal and would require greater reductions in livestock populations. On the contrary, a full shift to organic systems within a low-cost livestock scenario (T3) leads to a shortage in the nitrogen balance (74% self-sufficiency; Table 2), thus limiting the production potential of plant-based sectors, compromising the biophysical feasibility of such a scenario and confirming that nitrogen availability is a limiting factor for organic agriculture63. An avenue to solve this issue and meet the nitrogen deficit would consist in raising the share of nitrogen-fixing legumes in the Belgian agricultural landscape and crop rotations, which was not considered by the current modelling but would make agronomic sense given the higher share of such crops in organic systems and the shift towards plant-based diets, as considered recently for Wallonia69. Going further, scenario results confirm that both production-side and consumption-side measures are necessary. Lower-impact production requires shifts in dietary patterns, as previously shown in numerous scenario exercises5,67,70. The extent of the dietary shift is dependent on the ambition of the scenario. A fully organic and low-cost livestock scenario (T3) where animals feed exclusively on coproducts requires greater efforts on the consumption side. While it shows potential compatibility with the EAT-Lancet diet, as also found by van Selm et al.23, it does not with other diets such as the Belgian FBDG or the TYFA diet. The possibility to implement such changes, especially on the consumption side should be considered carefully71. Demand-side changes have been shown to be slow and challenging67, even with a specific policy framework targeting the protein transition in Flanders72.
In the light of such results, trying to single out one specific scenario or vision seems counterproductive given the trade-offs involved, as highlighted previously in the literature73,74,75. Rather, the learnings of this analysis might point to considering a space of solutions rather than a single point (Fig. 4). This option space is defined on the one hand by the considered scenarios (scenario option space) and on the other hand by established sustainability targets (sustainability option space). The crossing of both provides a new space in which scenario outcomes are compatible with sustainability objectives, thus delineating a possible sustainable space of operation. The extent of this space is flexible and can evolve depending on the scenarios considered and the acceptable sustainability boundaries. In the case of Fig. 4 (left), aiming for strict self-sufficiency of meat and nitrogen or allowing for import-export flows and complementation with synthetic nitrogen provides flexibility to expand the sustainability option space. Moving to ammonia emissions (here of the pork sector in Flanders; right of Fig. 4) shows another option space defined by specific policy objectives. Thus, the purpose of accounting for a diversity of visions is not to identify precisely which future is the most desirable but to shed light on part of the available option space. This asks the question of whether the coexistence of visions could be a viable project of transition41, which is further discussed below. In any case, going back to the quantitative and qualitative questions of livestock transitions, the results confirm that both downsizing and improving livestock systems are needed76.
Top: a scenario option space is defined by the outcomes of the tested scenarios while a sustainability option space represents sustainability targets. The extent of both option spaces can be adapted (dotted lines and arrows), e.g. by designing new scenarios or considering scenario variants, or by adapting the sustainability boundaries. Bottom: The intersection of the scenario and sustainability option spaces provides a new space where scenario outcomes are compatible with sustainability objectives.
The boundaries of the exercise are defined by three main methodological choices: the adoption of a territorial, sectoral and multidimensional approach. The scenario findings are constrained by their national (Belgian) scope and a theoretical assumption of self-sufficiency. In reality, the Belgian livestock sector, its actors and value chains largely operate in international markets with significant import-export flows and the flexibility to resort to a global food system. While scenario hypotheses might lead to hidden displaced effects due to disrupted trade flows (e.g. export flows in scenarios T2 and T3), efforts still need to occur at the national level. Certain issues are inherently played out at the local territorial scales (e.g. nitrogen balance, biodiversity conservation). Other challenges are global in scope, e.g. reducing GHG emissions or shifting towards lower animal product consumption5, but require local contributions. The adopted spatial scale also provides the flexibility to target more restrained territories (e.g. Flanders or Wallonia) and tailor the approaches to their specificities, challenges and levels of governance. The challenge resides in understanding livestock systems at regional scales to ensure local relevance while being broad enough to allow policy action66.
The scenario exercise focuses on Belgium’s five main animal sectors (dairy, beef, pork, poultry and eggs), also including plant-based sectors (although not the focus of the exercise), and considers both production-side and consumption-side activities. Such an approach allows to zoom in and out on parts of the system. On one hand, the aggregated zoomed-out system-wide approach adopted here ensures that interactions between sectors are not omitted (e.g. crop-livestock interactions). Although the exercise estimates the necessary area of legumes to meet the demand of alternative diets, and while the freed forage land resulting from lower animal feed demand is sufficient to cover this increase (in scenario T3), how and where shifts between crops and plant-based sectors would be implemented requires greater attention. On the other hand, while the approach makes the link between production and consumption, the role of mid-chain actors remains an important blind spot, as in many scenario exercises38,71. The capacity to progress on both the depicted production-side and consumption-side changes is greatly dependent on the adaptation of the system as a whole, including mid-chain actors which play a great role in steering the organisation of the livestock sector77,78. In this sense, zooming in on a specific sector (e.g. dairy or beef) might be useful to gather food system actors, including mid-chain actors, and facilitate practical exchanges on disaggregated challenges at the sectoral level39.
The set of output indicators provides a toolkit to characterise the outcomes of the Belgian farming and food system, spanning over multiple dimensions of sustainability. Additional indicators could be added to assess issues such as pesticide use, animal welfare or antibiotic use. While pesticide use can be reduced through both extensive practices favouring diversification and ecosystem services, or intensive systems relying on genetic and technological solutions79,80, results show that more extensive systems tend to favour higher animal welfare and lower antibiotic use, generally at the expense of higher GHG emissions per unit of product81,82, although not systematically83. Interestingly, while scenarios T2 and T3 operate an important shift to extensive systems, the average GHG emissions per kg of protein (Table 2) are not necessarily higher than those in scenarios BAU and T1, dominated by intensive high-productivity systems (due to a greater reduction in the GHG emissions related to soybean meal; see Fig. S1). On the socio-economic dimension, while the included indicators provide some information, a greater number of indicators could be integrated52,84. In this research, they remain limited to the agricultural production stage. The developed modelling tool fails to account for dynamic evolutions, better represented in economic models (e.g. partial and general equilibrium) generally centred on trade85. As a market dominated by organic and extensive productions would not look like the current one, and given the current economic precarity and subsidy-dependence faced by certain livestock farmers7,52, further research is needed on the socio-economic impacts and the necessary changes in economic paradigms required by such transitions (e.g. through the instauration of taxation measures, healthy food access schemes, etc.). Such a two-step approach where socio-economic implications are researched after environmental framing has been applied in other foresight exercises, such as the TYFA modelling exercise86,87. Importantly, while the study assesses the theoretical feasibility of certain scenarios and their expected outcomes, it does not demonstrate the political feasibility of the outlined scenarios.
Rooted in the three steps of transition cycles (assess – envision – implement), this paper seeks to contribute to the three types of knowledge on the topic of the livestock transition through quantitative scenarios, used as a tool to draw and objectify possible futures. Upon completion of the analysis, we take the opportunity to reflect on this double theoretical framework.
This study particularly takes on the assessment and envisioning steps, thus contributing systems and target knowledge. Regarding the former, while the study seeks to adopt a system-wide approach (from production to consumption, integrating all agricultural activities in Belgium and a comprehensive set of sustainability indicators), full system knowledge can in practice rarely be attained26 as certain elements are out of scope (e.g. mid-chain actors, market-related consequences, spill-over effects on other territories, etc.) or would require more nuanced qualitative approaches37. Regarding target knowledge, the study aimed to consider and compare a diversity of narratives surrounding the livestock transition. In this sense, the adopted approach can be considered hybrid between exploratory and normative as it explores a range of preidentified visions38. While it does not seek to demonstrate the feasibility of a particular vision, the specific (normative) choices of scenarios and scenario hypotheses risk favouring certain futures over others despite the initial exploratory purpose28. Still, the scenario outcomes point to the trade-offs between narratives rather than to a single future. In this sense, the idea to reach a consensual and shared vision might be both optimistic (in terms of reconciling different worldviews) and unfeasible (from a biophysical perspective), while overlooking issues such as power relationships, diversity and eventually delaying action26. Rather, research should pursue and foster a shared sense of responsibility and environmental virtues26, while striving for greater openness to alternative and radical futures and careful actor participation37,88.
Before moving to an implementation step, the objective of this paper is also to reflect on the usefulness of the approach, and how it can contribute transformation knowledge on the third question of livestock transitions: is it possible to progress towards a shared vision. Considering the results presented above with those of a previous scenario exercise performed on the Belgian livestock sector59, including an analysis of the reactions of food system actors (in particular farmer unions) to this exercise9, provides an opportunity for medium- to long-term evaluation. The short-term reactions to the first scenario exercise showed imported tensions and little openness to the approach. In particular, paying attention to a diversity of visions and systems was largely overlooked by food system actors9. These challenges resonate with a critique of transition management approaches as being too managerial, creating a sense of controllability and putting too much emphasis on identifying solutions and generating transformation knowledge while there may be more potential in fostering empowerment, justice and agency26,89. Nonetheless, the study continued to live on in subsequent years, generating a ‘snowball effect’. Study results were mobilised in civil society circles for advocacy purposes or in policymaking contexts, e.g. during a parliamentary session in June 2022. Among other factors, it contributed to establishing more interactions between civil society (environmental NGOs) and the farming sector (farmer unions), e.g. during negotiation rounds for the local implementation of the EU Common Agricultural Policy. The scenario exercise presented in this paper allowed to overcome a series (not all) of limitations and criticisms made: the set of indicators was expanded to include some socio-economic aspects, and plant-based sectors were integrated to account for crop-livestock interactions. It remains to be seen how this new set of scenarios will be taken up and used by different food system actors. Yet, we do believe that such research can contribute to the cyclic and iterative processes that are transition cycles. The primary purpose, at least in exploratory approaches, resides in opening up debates and triggering new ways of thinking rather than pin-pointing specific pathways38,90. The paradox for foresight approaches lies in a strong commitment to drive change, as per the objectives of transition governance frameworks30,31,32,33,34,36, which conflicts with the difficulty of assessing its capacity to achieve such change43. Rather than willing to measure a potential impact of foresight approaches, we believe there is value in acknowledging a slow nature of foresight. Going further, the process must eventually feed into an implementation step to contribute to concrete transition pathways. As discussed above, the adopted territorial, sectoral and diversity approach may prove useful in this regard if certain attention points are taken into account (e.g. spill-over effects, sectoral interactions, etc.). Further building on the option space approach proposed here, actor participation and backcasting approaches are put forward to foster concrete implementation of pathways88,91.
This paper set out to apply the use of foresight scenarios to the case of the Belgian livestock sector. The objective was to study both the object and the methodological process. While the presented foresight approach allows to draw important conclusions for the future of the Belgian livestock sector, limitations must be taken into account with regards to certain methodological choices (e.g. the boundaries of the exercise). The results confirm that designing foresight scenarios on the livestock sector is no easy task given the multiplicity of existing visions and the trade-offs involved. While measuring the impact of the approach is complicated, the long-term research process provides arguments in favour of a slow foresight approach as a valuable tool in the management of transition cycles.
Methods
For the purpose of this research, a modelling tool was developed. It simulates the functioning of the Belgian food and farming system. Its general functioning and construction are detailed below.
Type of model
The modelling tool is a biophysical process-based model. It follows a bookkeeping approach using agricultural land and animal populations as initial modelling units to assess a series of related biophysical flows (production of biomass, nutrient flows, gaseous emissions, etc.). As other biophysical models61,69,92, the developed modelling tool does not account for market effects resulting from production and price changes, which are better apprehened by specialised economic models. As the main purpose here lies on assessing the biophysical outcomes of certains scenarios, the inclusion of such economic effects would likely come at the expense of other modelling choices prioritised here (e.g. focus on a diversity of farming systems within each sector, inclusion of a diversity of environmental indicators, etc.).
Scope and boundaries of the model
The scope of the model is the Belgian agriculture and food system. It accounts for all agricultural activities within both Belgian regions: Wallonia and Flanders. The model adopts a ‘territorial LCA’ approach in the sense that it accounts for all agricultural production and food consumption activities that occur on the Belgian territory93, although the production side remains the main entry point. All agricultural production occurring in Belgium is included in the model and assessed through a series of indicators. This includes impacts occurring outside of Belgium, but which are necessary to allow production in Belgium (e.g. the production and imports of feed ingredients such as soybean meal to feed Belgian animals). On the consumption side, the demand for food is compared with local agricultural production, but impacts related to imported foods (e.g. tropical fruit) are not assessed. In short, the approach focuses on all agricultural productions taking place in Belgium (whether they are exported or not) and all the impacts related to those production activities (whether these are direct impacts occurring in Belgium or indirect impacts occurring abroad).
Input variables and sequence of calculation
The model relies on three types of input variables: production factors, farming systems and control variables (Fig. 5).
General structure of the model.
Two main categories of production factors are considered to estimate agricultural productions and outcomes: agricultural land (expressed in hectares) and animal populations (expressed in animal numbers). Human population can also be considered a production factor which is necessary to estimate food demand. The model encapsulates 56 different agricultural products, grouped into 16 sectors and two categories of agricultural production (Table S1). The model goes beyond animal sectors and also includes plant-based sectors to account for interactions between both. Yet, for the purpose of this study, the primary focus lies on the animal sectors (dairy, beef, pork, poultry and eggs). Production factor data comes from the Belgian statistical office. Other production factors such as farms and capital are not covered given the type and purpose of the model (biophysicial process-based model).
A specificity of the approach resides in the consideration of a diversity of farming systems within agricultural sectors to represent the diversity of sustainability outcomes and better tailor the proposed scenarios to different types of actors. For each sector, a typology comprising two to six farming systems is proposed (e.g. enriched cage, indoor, free-range and organic in the case of the egg sector). Sector-specific farming systems are reclassified into generic categories (intensive, intermediate, extensive and organic) to facilitate comparison across sectors. Typologies are mainly built based on differences in practices. As no national statistical data is collected on farming systems, sources for farming system typologies include pre-existing results and expert consultations through an informed participatory process (Table S2).
Control variables are those that make the link between production factors, farming systems and the model outputs. They are generally expressed per production factor (per hectare; per animal; per capita). Control variables include elements such as agricultural yields, nitrogen and dry matter contents of crops, final uses of crops (food, feed, bioenergy, etc.), animal feeding practices (feed consumption and composition), input uses (e.g. N fertiliser and pesticides), human diets, etc. (Tables S3–S7). Data sources are diverse and have been compiled in ‘sectoral fiches’, which group all the information that is necessary to run the model. These fiches were sent to sectoral experts for validation and refinement of the typologies and their associated parameters. Depending on data availability, control variables are detailed at the farming system level or at higher levels (e.g. sectoral level).
Model outcomes and output variables
The combination of production factors, farming systems and control variables allows the calculation of a series of output variables and indicators (Fig. 5 and Table 3). The choice of the indicators was guided by the three criteria of indicator selection84: parsimony (non-redundancy of indicators), consistency (necessary indicators for interpretations, i.e. getting a comprehensive picture of the main challenges related to the Belgian livestock sector and comparing different scenarios) and sufficiency (indicators cover the three dimensions of sustainability). Included indicators are grouped in three main dimensions: production and consumption indicators; environmental indicators; socio-economic indicators. Additional indicators (e.g. antibiotic use, animal welfare, trade, health aspects)38 and sustainability dimensions (e.g. time, permanence and people)94 would be relevant to consider but were not included either due to a lack of available data (e.g antibiotic use, health aspects) or because not easily covered by the type of model (e.g. time, trade).
Production and consumption indicators allow to assess production, demand and self-sufficiency levels for both human food (with a focus on animal products) and animal feed. Results are expressed in absolute terms (kt/year) and relative terms (g/cap/day). Human food is analysed in terms of animal products and protein contents. Food losses and waste along the chain are taken into account, based on product-specific estimations95. Departing from the total demand of legumes for human food consumption (considered in the different diets), the land use represented by this demand is calculated (based on average yields for legumes in each scenario). Similarly, the land use related to the demand of soybean meal for animal feed is calculated.
Environmental indicators cover three main themes. Climate change is assessed through the GHG emissions associated to each farming system. Calculations include main emission sources from the IPCC guidelines (on-farm energy use, enteric fermentation, manure management, soil management, land use and land use change) and three additional sources (production of synthetic nitrogen, pesticides and imported feed, in particular soybean meal; see Table S4 for more detail). The nitrogen cycle is analysed by assessing animal nitrogen emissions through the production of animal manure and the self-sufficiency of organic nitrogen (animal manure) for crop fertilisation purposes. Nitrogen emissions are assessed based on animal feeding practices, nitrogen content of feed ingredients and nitrogen use efficiencies, following the method and data from Hou et al.96. Organic nitrogen needs of crops are based on grey literature (see Table S5 for more detail). Ammonia emissions from manure management in stables are calculated based on nitrogen emissions of animals (see above) and animal-, stable type- and region-specific emission factors which allow to estimate the fraction which is emitted as ammonia (see Table S5 for more detail). Finally, impact on biodiversity is assessed through damage score factors provided in Chaudhary and Brooks97, illustrating potentially disappeared species of a number of land use types (arable land, grasslands, forests, etc.) and land use intensities (intense use, light use, minimal use), including spatially-explicit data at country level (specific characterisation factors for Belgium; see Table S6 for more detail).
Finally, two socio-economic indicators are included in the assessment: employment level (expressed in full-time equivalents; FTE) and gross value of agricultural production (expressed in euros; €). Data is collected through an agricultural census carried out every 5 years by the national statistics office.
Calculation and analysis of output variables
All output indicators are calculated via an R script. A few exceptions are calculated separately, in Microsoft Excel spreadsheets. This is the case of socio-economic indicators, biodiversity indicators, food demand indicators and food self-sufficiency indicators. All output variables are compiled and analysed in a dedicated Microsoft Excel spreadsheet.
Validation of results
In order to verify the robustness of the model and ensure the validity of the scenario results, model outcomes for the current situation (reference year 2019) have been compared against external sources for certain output variables (Table S8). In certain cases, when differences appeared to be too significant, correction factors have been integrated in the model to ensure better calibration with external data. While some internal feedback loops are considered in the analysis (e.g. in terms of forage availability or organic nitrogen self-sufficiency), these are checked for scenario consistency but do not constrain the model (e.g. nitrogen deficit in scenario T3). Additional feedback loops and spillover effects expanding beyond the scope of the model are not included (e.g. on trade or production and impacts abroad).
Building scenarios and scenario variables
Scenarios are built by setting hypotheses on certain input variables, thus changing their default values and assessing how these changes affect the output variables. Five input variables are used to design scenarios (Table 4): evolution of livestock populations, shares of farming systems, reconfiguration of agricultural areas, optimisation of technical parameters, evolution of diets (Table S7). The last variable is exogenous to the model while the others are endogenous.
Process of identifying scenario narratives
Scenarios were designed with the objective of simulating prevailing narratives and recurring elements of discussion in the debates surrounding the future of the livestock sector (Table 1). Scenario building relied on an iterative and mixed expert-based and participatory approach. A first set of scenarios for the Belgian livestock sector was built following a transdisciplinary research approach in collaboration with a civil society organisation and during which actors from the livestock sector had the opportunity to provide feedback on hypotheses and suggest additional scenarios9,59. The set of scenarios examined in this paper builds on the first scenario exercise. It was proposed by the authors with the aim of reflecting dominant visions for the future of the livestock sector while refining the first scenarios, taking certain comments into account.
Setting scenario hypotheses
Elements of scenario narratives were proposed by sectoral actors (e.g. technological gains; no feed/food competition), based on Belgium-focused literature (e.g. three compartment vision60; no soybean meal21) or based on choices by authors (e.g. land sharing). Scenario narratives were translated into changes in scenario variables for the modelling exercise based on own expertise and the literature (Table 5). While the scenario exercise aims to adopt an exploratory approach, it inevitably involves some sort of normativity in the choice of tested narratives or in setting values for scenario variables. Alternative scenarios and narratives could have been investigated, e.g. focussing on ‘agroecology’ vs. ‘sustainable intensifcation’, testing a ‘no-livestock/vegan’ scenario, putting greater emphasis on technological gains, etc. All scenarios were tested with a 2050 time horizon to allow for sufficiently significant changes and reconfigurations of the system. A reference scenario representing the current situation (2018–2022) constitutes a baseline and is used as a point of comparison.
Data availability
All data generated or analysed during this study are included in this published article and its supplementary material. Full datasets are available from the corresponding author on reasonable request.
References
Steinfeld, H., Gerber, P. J., Wassenaar, T., Castel, V. & De Haan, C. Livestock’s Long Shadow: Environmental Issues and Options. Food Agric. Organ. U. N. FAO https://www.fao.org/3/a0701e/a0701e.pdf (2006).
Gerber, P. J. et al. Tackling Climate Change through Livestock: A Global Assessment of Emissions and Mitigation Opportunities. Food Agric. Organ. U. N. FAO https://www.fao.org/3/i3437e/i3437e00.htm (2013).
Erb, K.-H. et al. Exploring the biophysical option space for feeding the world without deforestation. Nat. Commun. 7, 11382 (2016).
Notarnicola, B., Tassielli, G., Renzulli, P. A., Castellani, V. & Sala, S. Environmental impacts of food consumption in Europe. J. Clean. Prod. 140, 753–765 (2017).
Willett, W. et al. Food in the Anthropocene: the EAT–Lancet Commission on healthy diets from sustainable food systems. Lancet 393, 447–492 (2019).
Jordon, M. W. et al. A restatement of the natural science evidence base concerning grassland management, grazing livestock and soil carbon storage. Proc. R. Soc. B Biol. Sci. 291, 20232669 (2024).
Kortleve, A. J., Mogollón, J. M., Harwatt, H. & Behrens, P. Over 80% of the European Union’s Common Agricultural Policy supports emissions-intensive animal products. Nat. Food 1–5 https://doi.org/10.1038/s43016-024-00949-4 (2024).
Torpman, O. & Röös, E. Are animals needed for food supply, efficient resource use, and sustainable cropping systems? An argumentation analysis regarding livestock farming. Food Ethics 9, 15 (2024).
Antier, C., Riera, A. & Baret, P. A participatory prospective approach fails to ignite a debate on the future of the livestock sector in Belgium. In Proc. 14th European IFSA Symposium 576–589 (2020).
Gasselin, P., Lardon, S., Cerdan, C., Loudiyi, S. & Sautier, D. Coexistence et confrontation des modèles agricoles et alimentaires. (éditions Quae, 2021). https://doi.org/10.35690/978-2-7592-3243-7.
Penvern, S. et al. Addressing the diversity of visions of ecologization in research and in support to agroecological transitions. Agroecol. Sustain. Food Syst. 47, 1403–1427 (2023).
Freibauer, A. et al. Sustainable food consumption and production in a resource-constrained World. Summary Findings of the EU SCAR Third Foresight Exercise. EuroChoices 10, 38–43 (2011).
Béné, C. et al. When food systems meet sustainability – Current narratives and implications for actions. World Dev. 113, 116–130 (2019).
Jaisli, I. & Brunori, G. Is there a future for livestock in a sustainable food system? Efficiency, sufficiency, and consistency strategies in the food-resource nexus. J. Agric. Food Res. 18, 101496 (2024).
Fischer, K., Vico, G., Röcklinsberg, H., Liljenström, H. & Bommarco, R. Progress towards sustainable agriculture hampered by siloed scientific discourses. Nat. Sustain. https://doi.org/10.1038/s41893-024-01474-9 (2024).
Phalan, B. What have we learned from the land sparing-sharing model? Sustainability 10, 1760 (2018).
Green, R. E., Cornell, S. J., Scharlemann, J. P. W. & Balmford, A. Farming and the fate of wild nature. Science 307, 550–555 (2005).
Finch, T. et al. Bird conservation and the land sharing-sparing continuum in farmland-dominated landscapes of lowland England. Conserv. Biol. 33, 1045–1055 (2019).
Finch, T., Green, R. E., Massimino, D., Peach, W. J. & Balmford, A. Optimising nature conservation outcomes for a given region-wide level of food production. J. Appl. Ecol. 57, 985–994 (2020).
Karlsson, J. O., Parodi, A., Van Zanten, H. H. E., Hansson, P.-A. & Röös, E. Halting European Union soybean feed imports favours ruminants over pigs and poultry. Nat. Food 2, 38–46 (2020).
Jennings, S. & Schweizer, L. Risky Business: The Risk of Corruption and Forrest Loss in Belgium’s Imports of Commodities. (3keel for 'WWF Belgium, 2019).
Van Zanten, H. et al. Defining a land boundary for sustainable livestock consumption. Global Change Biol. 24, 4185-4194 (2018).
van Selm, B. et al. Circularity in animal production requires a change in the EAT-Lancet diet in Europe. Nat. Food 3, 66–73 (2022).
Frehner, A. et al. The compatibility of circularity and national dietary recommendations for animal products in five European countries: a modelling analysis on nutritional feasibility, climate impact, and land use. Lancet Planet. Health 6, e475–e483 (2022).
Brunori, G. et al. Has transformation of food systems reached an impasse? Considerations on the role of agri-food research. Agric. Food Econ. 12, s40100-024-00308–8 (2024).
Kueffer, C., Schneider, F. & Wiesmann, U. Addressing sustainability challenges with a broader concept of systems, target, and transformation knowledge. GAIA - Ecol. Perspect. Sci. Soc. 28, 386–388 (2019).
Levidow, L., Birch, K. & Papaioannou, T. Divergent paradigms of European agro-food innovation: the knowledge-based bio-economy (KBBE) as an R&D agenda. Sci. Technol. Hum. Values 38, 94–125 (2012).
Levidow, L. & Papaioannou, T. Policy-driven, narrative-based evidence gathering: UK priorities for decarbonisation through biomass. Sci. Public Policy 43, 46–61 (2016).
Hirsch Hadorn, G., Bradley, D., Pohl, C., Rist, S. & Wiesmann, U. Implications of transdisciplinarity for sustainability research. Ecol. Econ. 60, 119–128 (2006).
Duru, M., Therond, O. & Fares, M. Designing agroecological transitions: a review. Agron. Sustain. Dev. 35, 1237–1257 (2015).
Dendoncker, N. et al. How can integrated valuation of ecosystem services help understanding and steering agroecological transitions? Ecol. Soc. 23, art12 (2018).
Gaupp, F. et al. Food system development pathways for healthy, nature-positive and inclusive food systems. Nat. Food 2, 928–934 (2021).
Prost, L. et al. Key research challenges to supporting farm transitions to agroecology in advanced economies. A review. Agron. Sustain. Dev. 43, 11 (2023).
Meynard, J.-M. et al. Unravelling the step-by-step process for farming system design to support agroecological transition. Eur. J. Agron. 150, 126948 (2023).
Loorbach, D., Frantzeskaki, N. & Avelino, F. Sustainability transitions research: transforming science and practice for societal change. Annu. Rev. Environ. Resour. 42, 599–626 (2017).
Loorbach, D. Transition management for sustainable development: a prescriptive, complexity-based governance framework. Governance 23, 161–183 (2010).
Gupta, B., Zurek, M., Woodhill, J. & Ingram, J. S. I. Advancing the next-generation of global food system scenarios: a critical review of existing narratives. Front. Sustain. Food Syst. 9, 1620374 (2025).
Riera, A., Duluins, O., Antier, C. & Baret, P. V. Which types of quantitative foresight scenarios to frame the future of food systems? A review. Agric. Syst. 225, 104270 (2025).
Borman, G. D. et al. Putting food systems thinking into practice: Integrating agricultural sectors into a multi-level analytical framework. Glob. Food Secur. 32, 100591 (2022).
Gibson, C. C., Ostrom, E. & Ahn, T. K. The concept of scale and the human dimensions of global change: a survey. Ecol. Econ. 32, 217-239 (2000).
Gasselin, P., Lardon, S., Cerdan, C., Loudiyi, S. & Sautier, D. Ambivalences des modèles et défi de la gouvernance territoriale de la coexistence. in Coexistence et confrontation des modèles agricoles et alimentaires. Un nouveau paradigme du développement territorial? 225–238 (Editions Quae, Versailles, 2021).
de Olde, E., Konefal, J. & Hatanaka, M. The role of farming by metrics in transforming food systems sustainably. Npj Sustain. Agric. 3 1-5 (2025).
Dernat, S., Etienne, R., Hostiou, N., Pailleux, J.-Y. & Rigolot, C. Ex-post consequences of participatory foresight processes in agriculture. How to help dairy farmers to face outcomes of collective decisions planning? Front. Sustain. Food Syst. 6, 776959 (2022).
Statistics Belgium. Meat Supply Balance in Belgium (2010-2023). https://statbel.fgov.be/en/themes/agriculture-fishery/supply-balance-sheets-meat (2024).
BCZ-CBL. Annual Report - Reporting Year 2023 (BCZ-CBL, 2024).
Agentschap Landbouw en Zeevisserij. Landbouwcijfers. www.vlaanderen.be/landbouwcijfers (2024).
Direction de l’analyse économique agricole. Etat de l’Agriculture Wallonne. https://etat-agriculture.wallonie.be/ (2024).
Statistics Belgium. Agricultural Figures 2023. https://statbel.fgov.be/en/themes/agriculture-fishery/farm-and-horticultural-holdings (2024).
Papangelou, A. & Mathijs, E. Assessing agro-food system circularity using nutrient flows and budgets. J. Environ. Manag. 288, 112383 (2021).
Leip, A. et al. Halving nitrogen waste in the European Union food systems requires both dietary shifts and farm level actions. Glob. Food Secur. 35, 100648 (2022).
De Pue, D. & Buysse, J. Safeguarding Natura 2000 habitats from nitrogen deposition by tackling ammonia emissions from livestock facilities. Environ. Sci. Policy 111, 74–82 (2020).
Riera, A., Duluins, O., Schuster, M. & Baret, P. V. Accounting for diversity while assessing sustainability: insights from the Walloon bovine sectors. Agron. Sustain. Dev. 43, 30 (2023).
Battheu-Noirfalise, C. et al. Classifying and explaining Walloon dairy farms in terms of sustainable food security using a multiple criteria decision making method. Agric. Syst. 221, 104112 (2024).
Peeters, A. Importance, evolution, environmental impact and future challenges of grasslands and grassland-based systems in Europe. Grassl. Sci. 55, 113–125 (2009).
Lebacq, T., Baret, P. V. & Stilmant, D. Role of input self-sufficiency in the economic and environmental sustainability of specialised dairy farms. Animal 9, 544–552 (2015).
Reinsch, T. et al. Toward specialized or integrated systems in Northwest Europe: on-farm eco-efficiency of dairy farming in Germany. Front. Sustain. Food Syst. 5, 614348 (2021).
Duluins, O., Riera, A., Schuster, M., Baret, P. V. & Van den Broeck, G. Economic implications of a protein transition: evidence from Walloon beef and dairy farms. Front. Sustain. Food Syst. 6, 803872 (2022).
Dénos, C. et al. Contribution of ultra-processed food and animal-plant protein intake ratio to the environmental impact of Belgian diets. Sustain. Prod. Consum. 51, 584–598 (2024).
Riera, A., Antier, C. & Baret, P. Study on Livestock scenarios for Belgium in 2050. UCLouvain - Sytra https://sytra.be/publication/scenarios-livestock-belgium/ (2019).
Honnay, O., Avermaete, T., Govers, G. & Keulemans, W. Een slimme zonering van landgebruik verzoent landbouwproductie, biodiversiteit en klimaat. Natuurfocus 20, 157–165 (2021).
Muller, A. et al. Strategies for feeding the world more sustainably with organic agriculture. Nat. Commun. 8, 1290 (2017).
Smith, L. G., Kirk, G. J. D., Jones, P. J. & Williams, A. G. The greenhouse gas impacts of converting food production in England and Wales to organic methods. Nat. Commun. 10, 4641 (2019).
Barbieri, P. et al. Global option space for organic agriculture is delimited by nitrogen availability. Nat. Food 2, 363–372 (2021).
Borghino, N., Wissinger, L., Erb, K.-H., Le Mouël, C. & Nesme, T. Organic farming expansion and food security: a review of foresight modeling studies. Glob. Food Secur. 41, 100765 (2024).
Darnhofer, I., Gibbon, D. & Dedieu, B. Farming Systems Research: an approach to inquiry. in Farming Systems Research into the 21st Century: The New Dynamic 3–31 (Springer, Dordrecht, 2012).
Kazanski, C. E. et al. Context is key to understand and improve livestock production systems. Glob. Food Secur. 45, 100840 (2025).
Röös, E. et al. Greedy or needy? Land use and climate impacts of food in 2050 under different livestock futures. Glob. Environ. Change 47, 1–12 (2017).
Röös, E. et al. Agroecological practices in combination with healthy diets can help meet EU food system policy targets. Sci. Total Environ. 847, 157612 (2022).
Desmarez, T., Bindelle, J. & Dumont, B. Towards sustainable diets and farming systems through land use optimisation. npj Sustain. Agric. 3, 49 (2025).
Billen, G. et al. Reshaping the European agro-food system and closing its nitrogen cycle: the potential of combining dietary change, agroecology, and circularity. One Earth 4, 839–850 (2021).
Saujot, M. & Waisman, H. Mieux représenter les modes de vie dans les prospectives énergie-climat. IDDRI Étude 2, 42 (2020).
Rubens, K. et al. EI-MEET: Monitoring Eiwitinname En -Aankopen in Vlaanderen 2024. Green Deal Eiwitshift Op Ons Bord. (Departement Omgeving, Vlaamse overheid, 2025).
Auclair, O., Eustachio Colombo, P., Milner, J. & Burgos, S. A. Partial substitutions of animal with plant protein foods in Canadian diets have synergies and trade-offs among nutrition, health and climate outcomes. Nat. Food 5, 148–157 (2024).
Mahon, N., Crute, I., Di Bonito, M., Simmons, E. A. & Islam, M. M. Towards a broad-based and holistic framework of Sustainable Intensification indicators. Land Use Policy 77, 576–597 (2018).
Mahon, Crute, N., Simmons, I., Islam, E. & Md, M. Sustainable intensification – “oxymoron” or “third-way”? A systematic review. Ecol. Indic. 74, 73–97 (2017).
Herzon, I. et al. Both downsizing and improvements to livestock systems are needed to stay within planetary boundaries. Nat. Food 5, 642–645 (2024).
Guthman, J., Butler, M., Martin, S. J., Mather, C. & Biltekoff, C. In the name of protein. Nat. Food 3, 391–393 (2022).
Baudish, I. et al. Power & protein—closing the ‘justice gap’ for food system transformation. Environ. Res. Lett. 19, 084058 (2024).
Zimmermann, B. et al. Mineral-ecological cropping systems—a new approach to improve ecosystem services by farming without chemical synthetic plant protection. Agronomy 11, 1710 (2021).
Mora, O., Berne, J.-A., Drouet, J.-L., Le Mouël, C. & Meunier, C. Foresight: European Chemical Pesticide-Free Agriculture in 2050 (INRAE, 2023).
Bartlett, H. et al. Trade-offs in the externalities of pig production are not inevitable. Nat. Food 5, 312–322 (2024).
Karlsson, L., Keeling, L. & Röös, E. What is a better chicken? Exploring trade-offs between animal welfare and greenhouse gas emissions in higher-welfare broiler systems. Sustain. Prod. Consum. 55, 203–216 (2025).
Hashemi, F., Mogensen, L., Van Der Werf, H. M. G., Cederberg, C. & Knudsen, M. T. Organic food has lower environmental impacts per area unit and similar climate impacts per mass unit compared to conventional. Commun. Earth Environ. 5, 1-18 (2024).
Lebacq, T., Baret, P. V. & Stilmant, D. Sustainability indicators for livestock farming. A review. Agron. Sustain. Dev. 33, 311–327 (2013).
Teeuwen, A. S., Meyer, M. A., Dou, Y. & Nelson, A. A systematic review of the impact of food security governance measures as simulated in modelling studies. Nat. Food 3, 619–630 (2022).
Poux, X. & Aubert, P.-M. An Agroecological Europe in 2050: Multifunctional Agriculture for Healthy Eating. Findings from the Ten Years For Agroecology (TYFA) Modelling Exercice (2018).
Schiavo, M., Le Mouël, C., Poux, X. & Aubert, P.-M. An Agroecological Europe by 2050: What Impact on Land Use, Trade and Global Food Security? (IDDRI, 2021).
Moallemi, E. et al. Models of future food systems should address transformation complexity and uncertainty. Nat. Food https://doi.org/10.31223/X5WB19 (2025).
Kaljonen, M. et al. Deliberating justice in food systems transformation pathways: a transdisciplinary approach applied in Finland. Environ. Res. Food Syst. https://doi.org/10.1088/2976-601X/ae1dfa (2025).
Newell, R., Dring, C., King, L. & Hemphill, M. Food system scenarios in uncertain futures: a case study on long-term local food system planning in Revelstoke Canada. Agriculture 15, 611 (2025).
Duygan, M., Binder, C. R., Meylan, G. & Fischer, M. Carving the transformation pathways to sustainable futures: a novel analytical framework for backcasting in sustainability transitions. Sustain. Futur. 9, 100553 (2025).
Karlsson, J. O. et al. CIBUSmod 25.09: a spatially disaggregated biophysical agri-food systems model for studying national-level demand- and production-side intervention scenarios. Geosci. Model Dev. 18, 8589–8611 (2025).
Loiseau, E. et al. Territorial Life Cycle Assessment (LCA): What exactly is it about? A proposal towards using a common terminology and a research agenda. J. Clean. Prod. 176, 474–485 (2018).
Seghezzo, L. The five dimensions of sustainability. Environ. Polit. 18, 539–556 (2009).
Income consulting - AK2C. Pertes et Gaspillages Alimentaires: L’état Des Lieux et Leur Gestion Par Étapes de La Chaîne Alimentaire (Income Consulting - AK2C' for 'ADEME, 2016).
Hou, Y. et al. Feed use and nitrogen excretion of livestock in EU-27. Agric. Ecosyst. Environ. 218, 232–244 (2016).
Chaudhary, A. & Brooks, T. M. Land use intensity-specific global characterization factors to assess product biodiversity footprints. Environ. Sci. Technol. 52, 5094–5104 (2018).
Conseil Supérieur de la Santé. Recommandations Alimentaires Pour La Population Belge Adulte (Conseil Supérieur de la Santé', 2019).
CELINE-IRCEL. et al. Belgium’s Greenhouse Gas Inventory (1990-2023) - National Inventory Document Submitted under the United Nations Framework Convention on Climate Change. https://unfccc.int/sites/default/files/resource/NID_2024_101224.pdf (2025).
CELINE-IRCEL. et al. Belgium’s Greenhouse Gas Inventory (1990-2021) - National Inventory Report Submitted under the United Nations Framework Convention on Climate Change. https://unfccc.int/sites/default/files/resource/NID_2024_101224.pdf (2023).
Vrints, G. & Deuninck, J. Technische En Economische Resultaten van de Varkenshouderij Op Basis van Het Landbouwmonitoringsnetwerk. Boekjaren 2011-2013. (Departement Landbouw en Visserij, 2014).
Vrints, G. & Deuninck, J. Rentabiliteits- En Kostprijsanalyse Vleesvee (Departement Landbouw en Visserij, 2015).
Van der Straeten, B. Rentabiliteits- En Kostprijsanalyse Melkvee Op Basis van Het Landbouwmonitoringsnetwerk. Boekjaren 2007-2012 (Departement Landbouw en Visserij, 2015).
Petel, T., Antier, C. & Baret, P. Etat des lieux et scénarios à horizon 2050 de la filière lait en Région wallonne. UCLouvain - Sytra https://sytra.be/fr/publication/filiere-lait-wallonie/ (2018)
IPCC. Contribution of Working Group III to Fifth Assessment Report of the Intergovernmental Panel on Climate Change - Chapter 11: Agriculture, Forestry and Other Land Use (AFOLU) (Intergovernmental Panel on Climate Change (IPCC), 2014).
Acknowledgements
This research was produced under the Horizon Europe project “FOOODCoST—redefining the value of food”, funded by the European Union (grant 101060481). Inceptive works on the topic were supported by Greenpeace Belgium.
Author information
Authors and Affiliations
Contributions
A.R.: conceptualisation; data curation; formal analysis; investigation; methodology; software; writing – original draft; writing – review & editing. N.V.: data curation; methodology; software; writing – review & editing. A.S.: data curation; methodology; software. Q.V.: data curation; methodology; software. P.B.: conceptualisation; funding acquisition; writing – review & editing.
Corresponding author
Ethics declarations
Competing interests
The authors declare no competing interests.
Additional information
Publisher’s note Springer Nature remains neutral with regard to jurisdictional claims in published maps and institutional affiliations.
Supplementary information
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article’s Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article’s Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Riera, A., Vandevoorde, N., Squilbin, A. et al. Narratives, trade-offs and scenarios to explore the livestock transition in Belgium. npj Sustain. Agric. 4, 16 (2026). https://doi.org/10.1038/s44264-025-00122-9
Received:
Accepted:
Published:
Version of record:
DOI: https://doi.org/10.1038/s44264-025-00122-9
