In a seminar on May 12 2022, researchers got together to discuss the role of animal livestock in a circular production system. Professor Imke de Boer from Wageningen University & Research (WUR) started with a presentation of a recent paper on circularity in animal production in Europe and how that relates to the diet proposed by the EAT-Lancet Commission on Food, Planet, Health in 2019 (see links to publications below).
Feeding 10 billion people within planetary boundaries
Numerous studies have proposed healthy and sustainable diets. In the EAT-Lancet diet, the aim was to provide a future population of 10 billion people in 2050 a healthy diet operating within planetary boundaries. The results were that we drastically need to reduce our consumption of (red) meat and eggs and starchy vegetables such as potatoes, and increase our consumption of fish, fruits, vegetables and whole grains.
Limited amount of animal-source food from circular systems
– In our study, we wanted to assess whether feeding livestock solely leftovers from plant-source food and grass resources is compatible with the EAT-Lancet diet, said Imke de Boer. PhD student Benjamin van Selm, who was the lead author of the publication, has modelled three scenarios of circular food production systems.
It was shown to be difficult to produce the quantities of animal-source food from specific species that was recommended in the EAT-Lancet diet. In the EAT-Lancet diet, poultry meat was preferred over beef and pork from a health perspective, but from a circularity perspective pigs and ruminants are better suited to upcycle food leftovers and grass resources into animal-source food. All three scenarios had lower greenhouse gas emissions and lower land use compared to the EAT-Lancet reference scenario.
Our diet, however, influences the quantity and quality of food leftovers.
– By including co-products from refined grains, we increased the room for poultry, but also increased the land use, said Imke. Thus there is a trade-off between consuming healthy wholegrains or producing healthy poultry meat and eggs.
Reducing our dependence on soybean import
Dr Johan Karlsson from the Department of Energy and Technology at SLU talked about how halting European Union soybean feed imports would favor ruminants over pigs and poultry. Today, EU is dependent on soybean import, mainly from Brazil, Argentina and USA. In total, soybean constitute 28% of the protein given to livestock (excluding protein in grass and other roughages). EU’s Farm to Fork Strategy includes reducing our soybean dependence.
Life cycle analysis studies have shown that replacing soy with local protein crops (such as broad beans), by-products from food and energy industry (for example rapeseed meal and ethanol distillation residue) can reduce environmental impacts. But considering the large dependence of soybean imports such solutions might not be able to scale.
–We have modelled different scenarios of EU livestock production without soy. Livestock numbers were optimized for supplying human edible protein from available feeds. Milk and meat from ruminants that are grazing or eating low quality feeds were favored because these ruminants convert low-quality feed to high-quality food. Laying hens were favored because egg production is highly efficient. In contrast, pig and poultry meat decreased when no soybean meal was available. To maintain current levels of protein and fat in EU diets it would be necessary to increase consumption of plant protein and fat, said Johan.
What is an appropriate scale for evaluating circular systems?
After the two presentations, a discussion led by professor Lotta Rydhmer, from the Department of Animal Breeding and Genetics at SLU, ensued. When modelling circular systems, how do you take local or national conditions into account?
– The appropriate scale depends on whom you are talking to - who are going to make use of the scenarios? When working on a large scale, you of course loose some of the details and it is harder to consider local particularities. As for example grasslands are defined very differently in different countries, said Johan.
– When you model larger scales, it gets more complicated. For me it is simpler to think on a regional level said Sophie Julie Krizsan, researcher at the Department of Agricultural Research for Northern Sweden at SLU. Modelling studies should be framed in accordance with stated aims concerning diet and environmental impact.
Models are thought experiments
Imke agreed with Sophie about having aims as starting points and said that her research group more and more focus on country-based models. However, if you model per country, you have to handle import and export of feed and food, which is tricky to optimize across Europe.
– Animals have more functions than food production, such as ecosystem services, and that is hard to model. Remember that models are thought experiments and it is hard to include all aspects in a model, said Imke. The conclusions from our models should be tested in field studies showing how circular systems function in practice.
– If you look at land use and greenhouse gases, you often talk about marginal land. If you increase the production of for example rapeseed, what happens in other countries? The Swedish narrative has been that we always can import food. In the Netherlands, people instead think that they should export food. However, with geopolitical and climate changes this is rapidly changing, added Christel Cederberg, professor at Chalmers.
Imke also pointed out that people need to be more connected to their food, which suggests that we should have a more local approach to production and consumption.
Main gaps and constraints in the models
– What are the main constraints or gaps today that are not tackled in present models? asked Lotta.
– We need to add competences within One Health in our research teams, to ensure food safety. Product quality is also important in the retail chain. For example, retailers want farmers to deliver uniform products. That is a gap we should consider, said Mikaela Lindberg, senior lecturer at the Department of Animal Nutrition and Management at SLU. An additional gap highlighted in the chat, is that most models do not handle the production systems’ resilience to disturbances.
– Our European scenarios were not set in a specific year, said Johan, when talking about gaps. We are now working towards accounting for the time dimension as well, starting with the current situation and then look at different pathways into the future. This would also allow us to study for example soil carbon changes and model greenhouse gas emissions over time.
Anna-Karin Modin-Edman from Arla Foods asked about inclusion of carbon dynamics and sequestration in the models. It is important to keep carbon in the soil and you need nutrient models to really grasp the carbon dynamics. Pasture is positive for carbon sequestration, but more ruminants on pasture result in more greenhouse gases.
– Models can always be refined, but the scenarios we have modelled demonstrate the key direction that we should take already today. There is a lot of knowledge out there that is not used for different reasons, said Imke. You can argue about the timescale or the type of grassland or fuel used in specific countries in the model, but the main results are clear: We need to reduce food-feed competition and go for a diet with less animal-source food, Imke stated.
Including sustainability and ecosystem benefits
How do we take into account all aspects of sustainability and how do you incorporate different community and ecosystem benefits, asked Lotta.
– In our models, we are now starting to include land-use. We need to become better at explicitly address trade-offs in nutrient cycling. EU policy is going to influence agriculture in Europe, said Anna-Karin Modin-Edman from Arla Foods.
– Models are abstract. It is important to do experiments on the farm level already now and see what is possible or not, said Imke.
– One thing that is often missing in the models is the impact of market and politics and a deeper understanding of the whole food system. I get the impression that to implement the conclusions from these studies we need a society with a planned economy, said the journalist Ann-Helen Meyer von Bremen. Reducing land use for production and giving agricultural land ‘back to nature’ is not going to happen in a market economy.
– The results from our studies can be used to guide decision makers by showing them what is biophysically possible, said Johan.
– Political governance is important for a more sustainable production and consumption, the politicians need to initiate transformations and to get people to accept necessary changes, added Lotta.
–A lot of things need to change, for example the model for economic growth, before we can expect things to change in the direction towards a more healthy diet within the planetary boundaries, Imke remarked.
Main messages from the webinar
It is impossible to include all aspects in a model, but models can indeed show directions and highlight the need for change.
- An aspect that is difficult to include in today´s models are biodiversity assessment. The reason is that it is a very complex aspect. It is difficult or almost impossible to include all aspects in a model. Another aspect not yet considered are e.g. food safety.
- Production systems differ between countries and models should consider local aspects, without neglecting the impact of import and export.
- By feeding animals with leftovers from plant-source food and grass resources and reducing animal-source food in our diets, we can achieve a more sustainable food consumption with lower climate impact and lower land-use.
- A circular food system based on food leftovers and grass resources, as well as a strong reduction of soy in the feed, favor ruminants over pigs and poultry.