Recent research on biodiversity impacts of food consumed in Finland conducted by Kyttä et al. (2023) focusing on land use and land-use change showed that chicken had a larger biodiversity footprint than beef due to the imports of soybean for its feed. The results probably surprised many, as previous research systematically showed that beef production generally results in the largest land use amounts (Crenna et al., 2019; e.g. Mazac et al., 2022; Poore & Nemecek, 2018). However, the researchers were able to explain their results due to the large biodiversity impacts caused by deforestation in favour of soybean production. Soybeans are one of the many ingredients of chicken feed and are grown mostly in Brazil and the United Stated of America (U.S.). A recent article on chicken feed showed that as much as 17% of Finnish chicken feed consists of soybeans ingredients (Usva et al., 2023). However, current preliminary results produced within the BIODIFUL project show that the average beef consumed withing Finland — that is a combination of Finnish produced beef and imported beef — still typically results in higher overall global biodiversity loss than chicken production, as presented in Figure 1, and included more biodiversity loss drivers to calculate biodiversity impacts.
Biodiversity impacts can be calculated using environmental life cycle assessment —shortly referred to as LCA. When estimating the biodiversity impact of a product, LCA practitioners commonly do this by focussing solely on the species richness – the number of different species that are potentially lost due to human activities. This is officially referred within LCA as the ‘potential fraction of species disappeared’ globally or shortly PDF. Loss of biodiversity occurs for many different reasons, but its main drivers were identified by the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES) as land use (i.e. loss of habitat), direct exploitation, climate change, pollution, and invasive alien species (see Figure 2).
Within LCA we can estimate the PDF of many of these different biodiversity loss drivers and combine these into one biodiversity loss estimate. For example, biodiversity loss caused through land use is estimated by comparing the species richness of natural land areas to the number of different species found within the human-managed landscape. Many life cycle impact assessments (LCIA) use a limited amount of land categories to estimate these differences in species richness: annual cropland, perennial cropland, urban environments, extensive and intensive forestry, and pasture. A good method makes a distinction between the different levels of biodiversity found in different places in the world — species richness in tropical areas in many times higher than that of Finland, for example. The LC-IMPACT method is one method that has a regionalized approach (see Figure 3). The method thus reflects that using 1m2 of land in Finland leads to smaller loss of biodiversity in comparison to using 1m2 in Ecuador. Additionally, different farming methods leads to a variety in biodiversity loss intensity. New developments in LCIA methods, such as seen in the GLAM method, make it possible to distinguish between three different intensities levels, thereby increasing the accuracy of biodiversity impact assessment of less intense farming practices, such as organic farming.
Deforestation (e.g. to produce soybeans or for the establishment of grassland) falls under loss of habitat and are considered as part of land stress calculations within LCA. Land stress not only considers the removal of the natural ecosystem but also includes the continuous occupation of land over time. This land occupation is considered a major threat for biodiversity as many species are unable to survive within the human-created landscapes. The biodiversity impact of land use (or occupation) and land-use change (or transformation) together are referred to as land-stress within LCA.
The sole focus on land stress in many biodiversity studies is logical because this is the main driver of biodiversity loss for terrestrial ecosystems. However, looking at Figure 2 shows that land use only explains approximately 30% of all biodiversity loss and only for terrestrial ecosystems. LCA methods, unfortunately, do not yet give estimates for freshwater and sea use, although their impacts are significant. Nonetheless, current LCA methods do give estimates for quite a number of the other drivers, including climate change, eutrophication (the disruption to aquatic ecosystems due to an excess amount of nutrients – for example caused by fertilizer use), water use, and ecotoxicity (through pesticide application, for example).
Estimates of species lost due to climate change, for example, is based on the proportion of species that is not able to keep up with the shift of ecosystems distribution due to the predicted increase in global temperatures. One well-known example is the bleaching of coral reefs, where the increasing temperature of the seas and oceans stresses corals and causes them to expel the algae upon which they depend. Although, global warming is listed as the third or fourth (depending on the ecosystem) most important driver for biodiversity loss, it is expected that increasing global temperatures will also increase its contribution as driver for biodiversity loss. For example, IPBES projected that if nothing changes, climate change will be responsible for 40% of biodiversity loss within the Americas by 2050 (IPBES, 2022). Including climate change within biodiversity impacts assessments in therefore crucial to better understand the biodiversity impacts of food production.
New estimates on biodiversity impact of meat consumed in Finland
The preliminary results presented in this blog for meat options are part of a large study of BIOFIDUL research together with researchers from China Agricultural University Beijing in China, Institute of Ecological Economics in Vienna, Austria, Institute of Environmental Sciences (CML), Leiden University, The Netherlands, and Royal Institute of Technology, Sweden, that will be submitted this autumn for a peer-review process of a scientific journal. However, due to the large interest in the results of different meat options, we thought it important to share our preliminary results already here. In total the biodiversity footprint of over 600 products have been analysed and will be used to establish an optimal biodiversity-respectful diet for the future.
To calculate the biodiversity footprint of these products, we have used several data sources. Food products were selected from the national FINDIET survey reporting the average daily food consumption of Finns. This data was combined with the Food and Agriculture Biomass Input–Output model, shortly called FABIO (Bruckner et al., 2019). FABIO is a database that contains the information on complex flows on agricultural products within the global economy utilizes data from FAOSTAT as well the related environmental flows, such as land use. The database contains 191 countries, 123 products, and nowadays extents to the year 2020. Combining the FINDIET data with the FABIO data makes it possible to establish where Finnish consumed food items are generally coming from and what the related environmental flows are. When the environmental flows per food product were identified they could be multiplied with so-called characterization factors that translate, for example, land use to PDF (i.e. biodiversity loss). The most up to date characterization factors for biodiversity impacts assessment produced by the GLAM project were used in the assessment considering different farming intensity levels globally and landscape fragmentation (Scherer et al., 2023). Combining this information with the GLAM and LC-IMPACT method enables the calculation of average biodiversity footprints for products consumed in Finland. For example, when calculating the biodiversity impact of 1 kg of beef, the impact is the representation of the shares of domestically produced as well as imported beef.
Figure 1 shows biodiversity impacts for average-meat options consumed in Finland, as well as tofu as a meat-replacement. Impacts were divided into the Earth’s three different ecosystems: terrestrial, freshwater, and marine. Each of these ecosystems has different drivers included. For example, terrestrial ecosystems include biodiversity loss caused by land use (or land occupation), land-use change (or land transformation) and climate change. Unfortunately, marine ecosystems currently only include eutrophication; the disturbance of aquatic ecosystems caused by the artificially increase amounts of nutrients through fertilizers application, for example. This is especially problematic when trying to estimate the impacts of fisheries.
Comparison with previous studies
The main difference between the previous results presented by Kyttä et al. (2023) is the inclusion of other biodiversity drivers. By expanding on the drivers leading to biodiversity impact, we can show with our results that land stress is not the only important contributor to the biodiversity impact of meat options. Climate change plays a significant contribution both for terrestrial and aquatic ecosystems. Especially when it comes to ruminant’s meat production, methane emissions caused by the burping of the cow (caused by enteric fermentation) are a significant source of greenhouse gases (GHG) and thus contribute substantially to biodiversity loss. Our results showed that most of the terrestrial species’ losses caused by beef were related to its emissions of GHGs. For chicken, loss were mostly caused by land stress for feed production. As climate change will continue to rise as a contribution to biodiversity loss (IPBES, 2022), including its impact is critical for understanding the biodiversity footprint of different food items and making informed decisions when stirring towards biodiversity-respect diets and food production.
A second difference appears when we focus only on the land stress results in Figure 1. While they differ from the results presented previously by Kyttä et al. (2023), they are in line with previous results showing that beef has a high biodiversity impacts cause by the substantially higher land requirements than poultry production and its contribution to climate change. One possible explanation for the difference in results is that both studies used different databases and methods. While the BIODIFUL study relies on data from the FABIO input-output database in combination with FAO data, the study by Kyttä et al. (2023) relied solely on FAO data, potentially overestimating the total land-use change per product. Additionally, the biodiversity impact for beef seemed to be much smaller in the Kyttä et al. (2023) study than for our average beef impacts presented here. Most biodiversity loss of beef, and especially land-stress related species loss, was linked to imported beef. When looking at purely Finnish-produced beef, our results for land use related biodiversity loss of Finnish beef were close to the values presented by Kyttä et al. (2023). When looking into Finnish meat production including only land-stress, the impacts of Finnish poultry production appear to be somewhat larger than Finnish beef. However, adding the impacts of greenhouse gas emissions flips the balance and results in Finnish beef causing 3 times more losses than Finnish poultry.
Regardless, the main conclusion should be that we need a reduction in meat consumption, not a shift from one meat to another or beef production in high biodiverse places to lower biodiverse places. Meat production — and especially beef production — requires significant amounts land to grow their feed needs and emit a substantial amount of greenhouse gasses. The often intense way meat currently is produced leads to large losses of biodiversity, especially when livestock or its feed is produced in species rich areas. Consumption needs to decrease to mitigate climate change and further degradation of our life-supporting ecosystems. In addition, we need to change our way we produce livestock-based products in a manner that is more respectful to the livestock, their farmers, and the environmental upon which both are dependent.
There are more differences between the studies though. For their calculation the researcher from the BIODIFUL project used more recent soybean content data provided by Atria and combined this with previous data to estimate a soybean content of average Finnish broiler production. Atria uses as little as 9% of soybean meal compared to the 17% assumed in the LUKE research. We assumed a conservative estimated national average of about 14% (assuming other poultry producers still use about 17%).
Another difference between the two studies is the selection of life cycle impact methods. In this study we were able to use the latest so-called characterization factors produced by the GLAM team that are used to estimate biodiversity impacts. As explained earlier, the GLAM factors distinguish between three different management intensities whereby the biodiversity impact increase as the management level increases thereby increasing the accuracy of land-stress related biodiversity impacts (Scherer et al., 2023). As land stress is the largest contributor within terrestrial ecosystems and biodiversity losses can vary substantially between different management intensities, biodiversity assessment can really benefit from this increased resolution.
Lastly, the results presented by the BIODIFUL researchers here are also somewhat different from a study conducted by Peura et al. (2023) who calculated the biodiversity impacts of the S-group. Although the results by Peura et al. are based on economic data, therefore complicating the comparison with weight-based data, the results showed that pork has the smallest biodiversity impact of the three different meat options. The calculations were based on the same LC-IMPACT method but did not use the updated characterization factors provided by the GLAM project. Within the Peura et al. (2023) calculations the EXIOBASE database was used. This allows for including environmental impacts beyond the farm level (as did the study by Kyttä et al. 2023) but has highly aggregated food categories. The more disaggregated the results, the higher the precision of the results. For example, the food category ‘fruits’ contain both bananas and apples, the former being produced in tropical areas with high biodiversity losses and the latter in temperate climates with lower biodiversity losses. A clear explanation for the low biodiversity impacts of pork (that also relies on soybean inputs for feed to some extent (https://yle.fi/a/3-10941259)) of this study performed by Jyväskylä researchers could not be found.
Although our studies showed differences with those of previously published studies, each study contributes to our overall understanding of this complex impact assessments. Each study allows researchers to learn for the others and build upon this or find explanations why differences occur. Having conversations about these differences, such as was done between LUT and LUKE researchers, is very valuable as is transparent reporting on choices, methods, and assumptions made during research.
Reducing biodiversity impacts through changes in consumer and producer behaviour
While there was noticeable difference between the 3 studies conducted on the biodiversity impacts of Finnish food consumption, there is one aspect in common: all studies focussed on products grown by current average production processes. These processes are often intense, relying greatly on external inputs such as fossil fuels, pesticides, and fertilizers for example. These are not only bad for the biodiversity but are also causing harm to farmers and consumers. The commonly used pesticide glyphosate, for example, has been linked to prevalence of Parkinson’s disease as well as to increased mortality rates of bees. Additionally, current agricultural methods rely on an estimated 50% of all habitable land, cause one third of all globally emitted anthropogenic GHG emissions, and redraw about 70% of freshwater.
We can therefore conclude that there are two ways to improve the biodiversity impacts of our diets. The first way is by making smarter choices as consumers as addressed by the research above. We can choose products that have a lower biodiversity footprint. The goal of the biodiversity footprint calculation of the 600+ products is to guide consumers to make biodiversity-respectful dietary choices. But another way to decrease the current decline in biodiversity is by improving the production systems. Changing agricultural practices is a key to reduce biodiversity loss and the consequential loss of ecosystem services that are in turn necessary for healthy food systems and sustaining yield. Alternatives do exist and have been gaining attention. Good examples are, fortunately, endless and include practices based on agroecology, regenerative agriculture, organic farming, agroforestry, permaculture, and carbon-smart farming. One example of producing food while preserving biodiversity are Finnish rural biotopes – among the most endangered biotopes – that can be protected by extensive grazing management of livestock. Irene Kuhmonen wrote in a previous blogpost on the potential of alternative production systems and the need to support our farmers who are trying to make this change (https://biodiful.fi/blogi/ovatko-viljelijat-ruokajarjestelman-kestavyyssiirtyman-muutosvoima/)
Unfortunately, current most LCA methods to calculate biodiversity are mostly based on conventional agricultural practices and therefore distinguish poorly between agricultural management practices when it comes to impacts of land stress. In a new project funded by the Kone foundation – that connects to the BIODIFUL project – LUT researchers in collaboration with HY researchers will try to turn this around by measuring the biodiversity levels of regenerative agricultural practices by using environmental DNA method and improve estimates of land-related biodiversity impacts of regenerative farming. This allows a better estimate on the biodiversity savings of biodiversity-respectful farming methods.
References
Bruckner, M., Wood, R., Moran, D., Kuschnig, N., Wieland, H., Maus, V., & Börner, J. (2019). FABIO – The Construction of the Food and Agriculture Biomass Input-Output Model. Environmental Science and Technology, 53(19), 11302–11312. https://doi.org/10.1021/ACS.EST.9B03554/ASSET/IMAGES/LARGE/ES9B03554_0003.JPEG
Crenna, E., Sinkko, T., & Sala, S. (2019). Biodiversity impacts due to food consumption in Europe. Journal of Cleaner Production, 227, 378–391. https://doi.org/10.1016/j.jclepro.2019.04.054
IPBES. (2022). Global Assessment Report on Biodiversity and Ecosystem Services – Implications for the Global Biodiversity Framework – Science Biodiversity Workshop COP 15, Montreal 6 /12/22. https://files.ipbes.net/ipbes-web-prod-public-files/2022-12/Global Assessment_ CBD COP 15_Science Day_6 December.pdf
Kyttä, V., Hyvönen, T., & Saarinen, M. (2023). Land-use-driven biodiversity impacts of diets—a comparison of two assessment methods in a Finnish case study. International Journal of Life Cycle Assessment, 28(9), 1104–1116. https://doi.org/10.1007/S11367-023-02201-W/FIGURES/5
Mazac, R., Meinilä, J., Korkalo, L., Järviö, N., Jalava, M., & Tuomisto, H. L. (2022). Incorporation of novel foods in European diets can reduce global warming potential, water use and land use by over 80%. Nature Food 2022 3:4, 3(4), 286–293. https://doi.org/10.1038/s43016-022-00489-9
Peura, M., El Geneidy, S., Pokkinen, K., Vainio, V., & Kotiaho, J. S. (2023). Väliraportti: S-ryhmän luontojalanjälki. University of Jyväskylä
Poore, J., & Nemecek, T. (2018). Reducing food’s environmental impacts through producers and consumers. Science, 360(6392), 987–992. https://doi.org/10.1126/science.aaq0216
Scherer, L., Rosa, F., Sun, Z., Michelsen, O., Laurentiis, V. De, Marques, A., Pfister, S., Verones, F., & Kuipers, K. J. J. (2023). Biodiversity Impact Assessment Considering Land Use Intensities and Fragmentation. Environmental Science & Technology, 57. https://doi.org/10.1021/ACS.EST.3C04191
Usva, K., Hietala, S., Nousiainen, J., Vorne, V., Vieraankivi, M. L., Jallinoja, M., & Leinonen, I. (2023). Environmental life cycle assessment of Finnish broiler chicken production – Focus on climate change and water scarcity impacts. Journal of Cleaner Production, 410, 137097. https://doi.org/10.1016/J.JCLEPRO.2023.137097
Verones, F., Hellweg, S., Antón, A., Azevedo, L. B., Chaudhary, A., Cosme, N., Cucurachi, S., de Baan, L., Dong, Y., Fantke, P., Golsteijn, L., Hauschild, M., Heijungs, R., Jolliet, O., Juraske, R., Larsen, H., Laurent, A., Mutel, C. L., Margni, M., … Huijbregts, M. A. J. (2020). LC-IMPACT: A regionalized life cycle damage assessment method. Journal of Industrial Ecology, 24(6), 1201–1219. https://doi.org/10.1111/JIEC.13018
