Share this page
Alignment with SDGs
Albert Zimmermann & Thomas Nemecek Agroscope, Switerland
Food and feed production, together with its preliminary stages – from the provision of the means of production, through to the growing and processing stages – is associated with significant environmental impacts. Opportunities for reducing these impacts exist at both the production and consumption stages, e.g. through changes in production techniques and diet, respectively. As part of the ‘Green Economy’ strategy, the Swiss Federal Office for Agriculture (FOAG) deals with the challenges and potentials of resource-conserving food and feed production. Here, attention must be paid to the sustainable use of arable land and permanent grassland for food production
On behalf of the FOAG, Agroscope investigated how a diet for the Swiss population associated with the lowest possible environmental impacts might look like. Prerequisites were compliance with production-related process conditions whilst maintaining a productive domestic agricultural sector, coupled with a needs-based diet. Furthermore, certain additional framework conditions were stipulated in various scenarios.
The following issues were investigated:
- What would a needs-based diet for the Swiss population coupled with a reduction in environmental impacts look like?
- How would this alter agricultural production in Switzerland?
- What effects would this change have on imports and degree of self-sufficiency?
- To what extent could the environmental impacts be reduced?
The issue was investigated with the DSS-ESSA model system, used by the Federal Office for National Economic Supply to simulate food crises in Switzerland. This model system looks simultaneously at Switzerland’s agricultural production, imports and exports of food and feed, the processing of the products in question, and the diet of the Swiss population. The model was expanded within the context of the study as follows:
Milk production and grassland use were differentiated according to various levels of intensity.
- Imbalances in the model calculation between the manure generated by the animals and the fertiliser requirements of the crops were evened out by a reduced or increased demand for mineral fertiliser.
- Food loss at the consumption level was included and estimated on the basis of two studies.
- Nutritional requirements were significantly expanded and adapted to the latest findings. Several additional foods such as legumes, tofu or peanuts were included in the model.
- The SALCA life-cycle assessment method was used to determine environmental impacts for all activities depicted in the model, such as production processes and processing operations, or imported products. For this, 512 eco-inventories were used, most of which had to be adapted to or newly created for the DSS-ESSA model. The ReCiPe, Impact World+ and Environmental Impact Points (ecological scarcity method) endpoint methods were used to quantify environmental impacts. In addition, various indicators at midpoint level (e.g. greenhouse potential, eutrophication, ecotoxicity) or individual emissions and resources (e.g. ammonia, methane, phosphorus requirement) were included in the analysis.
The expanded Green DSS-ESSA model therefore investigates a food-supply situation that is optimised in terms of environmental impacts, and that takes into account on the one hand all production and nutritional requirements, and on the other the stipulated production specifications. Coffee and tobacco are not included in the model, as they do not contribute to the nutrient supply. Moreover, environmental impacts associated with the retail trade and consumer food preparation (Figure 1) have to date not been taken into consideration either.
The Reference scenario serves as a comparative scenario in the study. It describes the current situation by using the target function to minimise the deviations of the model solution from current production and dietary habits. The Min ReCiPe scenario is optimised in terms of environmental impacts. Depending on the life-cycle assessment method used, the environmental impacts can occur in Switzerland, or – where associated with imports – abroad. Based on the target function of Min ReCiPe, there were additional framework conditions to be complied with in three further scenarios.
- Reference Current situation
- Min ReCiPe Minimisation of environ-
mental impact of ReCiPe
- FP Composition of ration according to food pyramid
- FP/Cal Composition of ration and energy intake according to food pyramid
- FoodWaste Complete reduction in avoidable food waste during consumption.
Apart from the general model interrelation-ships, the following additional conditions were to be met:
- Average total calorie intake per person and day remained at the current level, so that the effect of a change in diet combined with unchanging energy supply could be analysed. Only in the FP/Cal scenario did reduced energy intake apply, in line with the relevant dietary recommendation.
- For each food, the current process yields and percentage losses along the food chain were assumed. Only in the FoodWaste scenario was the model permitted to completely reduce avoidable losses during consumption, thereby reducing the quantity of food required with the same intake before deduction of these losses.
- None of the scenarios permitted a further increase in current deviations of the average ration from the dietary recommendations (shopping-cart shares, nutrient supply).
- In order that currently consumed products do not disappear completely from the population’s diet, the consumption of all individual foods should not fall by more than 90% of the current quantity in each case.
- Food exports were kept constant in the current configuration; otherwise, the model solution would have led to an export, and hence to a non-imputation of environmentally damaging foods in particular.
- The use of the entire agricultural area of Switzerland was assumed. This condition serves two purposes: on the one hand, products from these areas contribute to supply security; on the other, an open landscape is maintained.
The effect of altered model assumptions was investigated by means of sensitivity analyses, to enable the robustness of the results to be assessed.
Overall, it appears that the environmental impacts of diet can be reduced by over 50% based on the assumptions made (Figure 2). Here, major improvements are possible with practically all environmental impacts. As regards deforestation, largely dispensing with certain imported products such as soya feedstuffs and cacao even allows us to achieve an 80% reduction. Major reductions are also possible for the individual emissions (greenhouse gases and ammonia, -50%; nitrate and phosphorus, -35%). Owing to the higher proportions of milk and vegetables and the lower sugar consumption, compliance with the food pyramid recommendations (FP scenario) leads to a smaller reduction in environmental impacts. By contrast, avoidance of food waste in the household yields a more significant reduction.
The composition of the average diet changes significantly (Figure 3). Key features of a resource-conserving diet of this sort are a significant drop in the proportion of meat (-70%) and a larger proportion of grains, potatoes or legumes (+35%) as well as oils or nuts (+50%), whilst milk consumption remains at the same level. This result can be explained by the major differences in environmental impacts between animal and plant foods, with milk nevertheless performing significantly more favourably than meat. By contrast, differences among plant foods are frequently very slight. Thus, the replacement of potatoes by grains, or nuts by vegetable oils and grains has very little influence on total environmental impact. The more-resource-conserving diet diverges less strongly from nutritional recommendations than the current diet, especially owing to the former’s lower meat and alcohol consumption, and its partial replacement of animal fats with vegetable oils and fats.
In line with the decrease in the proportion of meat in the diet, livestock populations – especially pig, fattening poultry, suckler cow and fattening-cattle numbers – also fall sharply in the model results. Grassland is used for dairy farming, and the proportion of higher-yielding dairy cows increases. Overall, animal populations – measured in livestock units – fall by almost half. Livestock feed rations also change: Cows are fed fresh and ensiled grass, hay, and those with a higher milk yield are also given maize kernels and barley. Thus, protein is increasingly provided via grass, whilst soybean meal, which is associated with high environmental impacts, disappears from the diet. A large proportion of the permanent grasslands is farmed extensively. The low nutrient grass from these lands is fed to the rearing cattle, sheep and goats. The sharp reduction in livestock populations means that feed imports can be largely eliminated. Arable land also continues to be used for forage cultivation, but to a significantly lower extent. Whereas a part of this land is used as temporary leys, which are important for a balanced crop rotation, significantly more grains for the human diet (+70%) are grown on arable land. There is also an increase in the area devoted to potatoes (+140%), vegetables (+100%; in the FP scenario, even +350%) and oilseed rape (+20%).
Food imports decrease (-28% in calories), whilst only small amounts of feedstuffs are now imported (-85%). As a result, the percentage of domestically produced products, and hence the degree of self-sufficiency, increases significantly, from 61% to around 80%. The total environmental impacts of imported foods fall by around 70%, and those of foods produced in Switzerland – despite the higher amount of calories produced – by 20% (ReCiPe indicator).
The choice of method for calculating environmental impact had only a slight influence on the result. A similar scale of percentage reduction to that of the ReCiPe indicator (-55%) was also achieved in the case of the minimisation of Impact World+ (-52%), Environmental Impact Points (-60%) and greenhouse potential (-61%). Here, dietary composition trends evolved in the same direction, but with differences for individual products.
A systematic focus on protection of the environment and resource conservation enabled the environmental impacts of the Swiss population’s diet to be more than halved whilst maintaining the use of all of Switzerland’s agricultural land, with unchanged exports and without an increase in existing deviations from dietary recommendations.
In order to achieve this, the average composition of the diet had to change substantially, involving on the one hand a significant increase in the consumption of (a) grains or potatoes, (b) nuts, and (c) fruit or vegetables, as well as the maintenance of dairy consumption in a predominantly unprocessed form; and on the other, a sharp reduction in meat and alcohol consumption, as well as a decrease in the consumption of edible oils, durum wheat products, rice, and processed dairy products. Sugar consumption would remain the same or fall on the basis of the nutritional recommendations.
At the same time, production processes would need to be optimised, especially in terms of the feeding of cattle, who would essentially exploit the grassland yields. Concentrates would almost cease to be imported and would only be cultivated domestically on a small scale.
An additional significant reduction in environmental impacts would be possible if we managed to avoid all avoidable food loss. Whilst losses at the production and processing stages are frequently inevitable, there is greater potential for avoiding them at the consumption stage.
An environmentally optimised diet would be accompanied by synergy effects: at the same time, it would largely meet current dietary recommendations. Furthermore, lower import levels would increase our degree of self-sufficiency, thereby reducing Swiss dependence on foreign sources.
All in all, the analysis shows that the current situation is far from the ideal of an eco-friendly, resource-conserving food and feed production system, and that a great potential for improvement therefore exists. In order to derive concrete measures for improvement, detailed investigations requiring a further expansion of the models and data sources used would have to be conducted. Moreover, it might also be necessary to consider economic aspects, depending on the issue examined. Such a far-reaching change in diet, however, would doubtless require the cooperation and willingness of both the population and the economic and policy-making sectors.
Zimmermann, A., Nemecek, T., Waldvogel, T.2017. Umwelt-und ressourcen-schonende Ernährung: Detaillierte Analyse für die Schweiz. Umwel, Agroscope Science. N 55.