Blind spots of modelling biodiversity loss in food systems

Blind spots of modelling biodiversity loss in food systems | Research Square window.SnipcartSettings = { analytics: { enabled: false } }; (function() { var accessVector = localStorage.getItem('access_vector') || ''; window.dataLayer = window.dataLayer || []; if (accessVector) { window.dataLayer.push({ user: { profile: { profileInfo: { snid: accessVector } } } }); } })(); (function(w,d,s,l,i){w[l]=w[l]||[];w[l].push({'gtm.start':new Date().getTime(),event:'gtm.js'});var f=d.getElementsByTagName(s)[0],j=d.createElement(s),dl=l!='dataLayer'?'&l='+l:'';j.async=true;j.src='https://www.googletagmanager.com/gtm.js?id='+i+dl;f.parentNode.insertBefore(j,f);})(window,document,'script','dataLayer','GTM-K279D39R'); Browse Preprints In Review Journals COVID-19 Preprints AJE Video Bytes Research Tools Research Promotion AJE Professional Editing AJE Rubriq About Preprint Platform In Review Editorial Policies Our Team Advisory Board Help Center Sign In Submit a Preprint Cite Share Download PDF Article Blind spots of modelling biodiversity loss in food systems Felipe Cozim Melges, Wendy Margarida Nina Jenkins, Rob Alkemade, and 13 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6735266/v1 This work is licensed under a CC BY 4.0 License Status: Under Revision Version 1 posted 9 You are reading this latest preprint version Abstract Global biodiversity is in crisis, with food systems identified as a major driver of its decline. Various modeling approaches are currently used to provide a critical step guiding policy making. In this paper, we reviewed ten global models assessing food systems, including seven global biodiversity models (GBMs) and three life cycle impact assessment (LCIA) models to map their capacities to capture the diverse effects on biodiversity both within and beyond agricultural fields. LCIA models more comprehensively focused on pollution and climate change, while GBMs captured spatial impacts like fragmentation better with a more diverse range of indicators such as compositional intactness, species richness, extinction, and habitat suitability. Gaps found in both GBMs and LCIAs highlight the need for complementary aspects to be included in current models to more holistically assess biodiversity in food systems.Combining GBMs and LCIAs could address their individual limitations and integrate consumption with shifting spatial dynamics. Earth and environmental sciences/Ecology Earth and environmental sciences/Environmental sciences Earth and environmental sciences/Planetary science Agroecosystems LCIA Land-use GBM Biodiversity food systems models Figures Figure 1 Introduction Beyond its intrinsic value, biodiversity is essential for human health and well-being through the provision and maintenance of ecosystem services such as crop pollination 1 . Despite its importance global biodiversity is declining at unprecedented rates 2 . Food systems have been identified as the primary driver of global biodiversity loss 3,4 , caused by different activities, ranging from land expansion into natural areas 5,6,7,8 to the intensification of production methods 9,10,11,12 or over-harvesting of wild foods 13 . International policy agreements like the Kunming-Montreal Global Biodiversity Framework (GBF) 14 or the EU 2030 Biodiversity Strategy 15 , aim to reverse the trend of biodiversity loss. Effective nature conservation strategies combined with improvements to supply chains, from production to waste disposal, will be needed to bend the curve of biodiversity loss 16 . Capturing the effects of the whole supply chain on biodiversity however requires a food systems approach 17 . Global food systems models attempt to capture the effects of the whole supply chain on a variety of impacts. Many different approaches to modelling biodiversity have been put forward in global food systems models, each with their own set of assumptions and simplifications to assess biodiversity 18,19,20 , e.g. different systems boundaries. The different assumptions from different models can yield different results for recommendations to reduce biodiversity loss in the food system 21,22 . This review aims to create an overview of what is currently being done and map how current approaches model biodiversity loss. We compared the models by looking at biodiversity pressures, indicators and food system components in each model, as per the definitions in the text box ‘Terminology Used’ below. First, we looked at how models are accounting for the most significant biodiversity pressures as outlined by the Global Assessment of the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES) 2 , i.e., natural resource use and exploitation, pollution, climate change, land use, invasive species and occupation (table 2). Second, we explored which indicators were being used to assess biodiversity and clustering them based on the biodiversity impact they assess, as described by Pereira et al. 23 : species extinction, extent of suitable habitat, compositional intactness and species richness (table 3). Finally, we developed a comprehensive overview of how biodiversity is modelled across the components of the food system as defined by the FAO 24 as production, aggregation, distribution, processing, consumption and waste disposal. Multiple different venues to model biodiversity in food systems currently exist. Life Cycle Impact Assessments (LCIAs) are a step in the assessment method of the life cycle analysis, which is often employed when looking at the impact of a product 25 . LCIAs work by translating the effects of raw material use into environmental outcomes, such as climate change or water use which are then converted into biodiversity impacts 25 Click here to enter text.. Global Biodiversity Models (GBMs), using Integrated Assessment Models (IAM), are models which are used in tandem with designed scenarios to assess linkages between human activity and the biosphere, often focusing on land use activities, such as agricultural production 26 . Like LCIA, GBMs translate impacts such as climate change and land use and occupation into their impacts on biodiversity. Unlike LCIA, GBMs tend to make assessments clustering different midpoint outcomes into common impact coefficients factors, which are then used to estimate biodiversity impacts. Our results showed that LCIAs are able to encompass more specific impacts of the food system but lack the spatial dynamics and biodiversity indicator breadth of GBMs. We argue that using the models in tandem has potential for addressing some of the blind spots in both modelling approaches and increase the range of drivers covered, allowing for more robust and realistic considerations of biodiversity loss and potential interventions to bend the curve of biodiversity loss. This more comprehensive consideration is direly needed for the urgent action proposed in the Kunming-Montreal Global Biodiversity Framework mission 27 and the EU 2030 Biodiversity Strategy 28 . Box 1. Terminology used To better merge terminology from different types of models we distinguish our definitions of the concepts below: 1) Biodiversity: “the variability and interaction among living organisms from all sources including, inter alia, terrestrial, marine, and freshwater ecosystems and the ecological complexes of which they are part of; this includes diversity within species, between species and of ecosystems” 29 . 2) Food systems: “encompass the entire range of actors and their interlinked value-adding activities involved in the production, aggregation, processing, distribution, consumption and disposal of food products that originate from agriculture, forestry or fisheries, and parts of the broader economic, societal and natural environments in which they are embedded” 24 . 3) ‘Midpoints’ is a commonly used term in LCAs and are intermediate measures of impact of an activity known to, directly and/or indirectly, affect biodiversity loss e.g. the total amount of fertilizer used will contribute to the midpoints such as ‘GHG emissions’. We clustered ‘midpoints’ into broader categories of the ‘IPBES pressures’ (land use, natural resource use and exploitation, pollution and climate change). Clustering the midpoints from models into IPBES pressures allows us to see how each model is or is not addressing a biodiversity pressure. For example, for the IPBES pressure ‘land use’, some models may only account for the midpoint ‘land occupation’ while others may also account for ‘fragmentation’ or other land use related impacts. Each midpoint can be translated into a ‘biodiversity impact’. 4) ‘Biodiversity impact’ is the broad aspect of biodiversity encompassed by the indicator used in the model. ‘Biodiversity impact’ classifications were defined based on an adaptation of Pereira et al. 23 . To address which aspect of biodiversity is being assessed in models, ‘biodiversity impact' classifications can be used to report the component of biodiversity being assessed across indicators. For example, the biodiversity intactness index and mean species abundance are two different indicators for the compositional intactness of biodiversity but compare different baselines and under different assumptions/perspectives. This classification allowed us to assess how holistic/broad the biodiversity assessment of the combined models is 2 . Results To determine how biodiversity is accounted for in global food systems models, we conducted a literature review, retrieving a total of 126 papers, which after screening and filtering resulted in identifying ten distinct food-system models that assess global biodiversity (table 1, figure 1 methods). Two contrasting approaches emerged, Life Cycle Impact Assessment (LCIA), which captures product level impacts, and Global Biodiversity Models (GBM), which capture spatially dynamic impacts and tradeoffs assessing effects of the food system on biodiversity loss: Life Cycle Impact Assessments (LCIAs) (3 models: IMPACT world + ReCiPe 2016 and LC-impact) and Global Biodiversity Models (GBMs) (7 models: Map of Life, BILBI, PREDICTS, GLOBIOM, C-SAR 2018, inSIGHTS and GLOBIO – table 1). In the section below we will discuss the ‘IPBES pressures’, ‘midpoints’, ‘biodiversity indicators’, ‘biodiversity impacts’ and ‘food system activities’ that were found in the models. An overview of the blind spots and strengths of the models can be found in table 2. Broadly speaking, GBM models account for a more diverse range of indicators compared to the LCIAs, however in terms of ‘IPBES pressures’, the LCIAs more consistently included the pressures, particularly regarding pollution. The only IPBES pressure not included in any of the retrieved models was invasive species. In terms of food system activities production was the most represented with GBMs also explicitly including aggregation. It should be noted however that scenarios enable more diverse ‘food system activities’ to be included in model outputs. For LCIAs the food system activities are dependent on the scenario chosen and its associated system boundaries, however like GBMs they are limited in the effects across the supply chain they can assess. The details of these results will be discussed in the sections below. Biodiversity indicators and metrics Overall GBMs used six indicators for measuring biodiversity (PDF, SHI, BII, MSA, RLI and ESH) corresponding to all the ‘biodiversity impact’ categories outlined by Pereira et al. 23 . LCIAs varied much less than GBMs using only the indicator PDF, corresponding to the biodiversity impact ‘species extinction’. The GBM inSIGHTS was the only model to account for multiple biodiversity indicators in the same model (RLI and ESH) covering the ‘biodiversity impacts’ ‘species extinction’ and ‘extent of suitable habitat’ (table 3). C-SAR is the only GBM that uses the indicator PDF, and along with inSIGHTS is the only other GBM to account for the ‘biodiversity impact’ ‘species extinction’. The ‘biodiversity impact’ ‘extent of suitable habitat’ was used by Map of Life with the indicator SHI, which is unique in that it uses GIS data to calculate ecosystem quality and diversity relative to a baseline year. The ‘biodiversity impact’ ‘compositional intactness’ represented by the indicator BII was used by BILBI, PREDICTS and GLOBIOM. BII looks specifically at changes in native terrestrial species to estimate the compositional intactness. Similarly, GLOBIO uses the ‘biodiversity impact’ ‘compositional intactness’ linked to the indicator MSA, the difference being, that while both the indicator BII and MSA look at all species native/original in a given region, BII looks at the percentage of loss by using total species abundance and a compositional similarity index, while GLOBIO does so referring to the mean species abundance of a given area. Table 3 depicts the pressures assessed and indicators used by the models retrieved. Food systems’ activities and IPBES’ pressures In this section we present the results of the models retrieved in terms of the IPBES pressures. In table 4 we present the ways in which these pressures are represented in the models. We find that while all pressures are covered, they are mostly done with individual midpoints. 1) ‘Land use’ was the most included ‘IPBES pressure’ with all the GBMs and LCIAs including it. For LCIAs, only the ‘midpoint’ ‘land occupation and transformation’ was included referring to the land use type, if it has been converted from an alternate land use activity e.g. nature areas to pasture. For the GBMs, both Map of life and GLOBIO included the ‘midpoint’ ‘habitat fragmentation’ relating to the ‘IPBES pressure’ ‘land use’. ‘Habitat fragmentation’ causes a habitat to be split into a larger number of smaller fragments. This reduces the area and population size of each fragment, which reduces likelihood of persistence of the population in the fragment. ‘Agricultural area’ was used to assess ‘fragmentation’ and was dependent on location’, with land closer in proximity to natural areas and larger agricultural areas having higher impacts. The presence and type of road also impacted the severity of effects with larger roads having a bigger impact. GLOBIO was the only GBM to account for the ‘midpoint’ ‘infrastructure disturbance’ which looks at the effects of presence of infrastructure on neighboring species. ‘Land use activity’ and ‘settlement population size’ were both used by models to calculate the effects of the ‘midpoint’ ‘infrastructure disturbance’. In GLOBIO the impact zones related to ‘infrastructure disturbance’ are predicted to change as the result of increasing human population sizes with the severity of impact is dependent on the type of land cover. The impacts of ‘infrastructure disturbance’ differ from fragmentation because they arise from a species preference to avoid developed areas and roads as opposed to the effect of a physical barrier to movement. Similar to fragmentation, the ‘presence and type of roads’ also impacted the effects of infrastructure on biodiversity. Other factors mediating the effects of infrastructure were factors including distance to the coast and distance from protected areas. ‘Land occupation and transformation’ was calculated using ‘land use activities’ by all models. ‘Land use activities’ varied from model to model with ReCiPe 2016 having the most different types of land use activities amongst the LCIAs (used forest, pasture and meadow, annual crops, permanent crops, mosaic agriculture) and GLOBIOM including the most amongst the GBMS (cropland (18 crops globally), managed grassland, short rotation plantations, intensity (subsistence, low input rainfed, high input rainfed, high input irrigated)). 2) Natural resource use and exploitation was included in all LCIA models in the form of the ‘food system activity’ ‘water use’. The impacts of ‘water use’ were determined through assessing the reductions in availability for local plants and shifts in river discharge on terrestrial and marine species. GLOBIO was the only GBM to look at water use. GLOBIO calculates the effect of water limitations on ‘land use activities’, meaning that limited water would shift the ‘land use activity’ to one that requires less as it is based on optimization. GLOBIOM was the only GBM to include the ‘midpoint’ ‘water use’. The impacts of ‘water use’ were however differently accounted for than in the LCIAs as GLOBIOM assessed available water as a limitation for irrigation. In this way the effects on biodiversity loss were instead determined because of shifts in land use as available water limited what could be grown. The LCIAs instead accounted for the impacts of water shortage on terrestrial and marine species loss. GLOBIO additionally looked at the ‘midpoint’ ‘encroachment impacts’ which were linked to the effects of hunting on an area. ‘Distance to cropland’ was used to estimate the effects of the ‘midpoint’ ‘encroachment’ by determining an impact zone around agricultural/human land use areas under the assumption that hunting (disturbance), will take place at certain rates in this zone 3) Pollution was included in the greatest detail in the LCIAs. The IMPACT world + model had the highest number of ‘midpoints’ related to ‘pollution’ including seven different types of pollution categories. LC-impact and ReCiPe included the same number and type of ‘pollution’ ‘midpoints’ including four different types (i.e. toxicity, fresh water/marine eutrophication, ozone and terrestrial acidification). While the ‘midpoint’ ‘nitrogen deposition’ was not directly included in any LCIAs, the pollution ‘midpoints’ which are included were predominately linked to pesticide and fertilizer production and application, and hence might be considered to indirectly account for the effects of nitrogen. ‘Pollution’ was the least represented in all the GBM models, with only GLOBIO including it with the midpoint ‘nitrogen deposition’. ‘Nitrogen deposition’ refers to the effect of a critical load of nitrogen on a given area and in the case of GLOBIO, refers specifically to the impacts on different land use types such as forests and grasslands GLOBIO uses the IMAGE model to create scenarios for present and future atmospheric ‘nitrogen deposition’ based on ‘land use activity’ e.g. pasture or cropland, with some ‘land use activities’ having higher impacts. 4) Climate change was accounted for by all LCIAs. In LCIAs climate change effects were calculated by determining GHG production along the supply chain (including the effects of land use change) and converting this into an estimate of potential warming and subsequent change in biome distribution of species. Fixed warming scenarios are used in addition by LCIAs to help determine the severity of effects on biodiversity of emissions. The GBM simulates effects of fixed warming scenarios on biodiversity loss. The GBMs, BILBI, inSIGHTS, GLOBIOM and GLOBIO, all use warming scenarios which typically predict impacts on species distributions of warming. GLOBIOM additionally predicts impacts on yields changing the efficiency of certain ‘land use activities’. 5) IPBES pressures were mostly widely accounted in both LCIAs and GBMs. The only IPBES pressure not accounted for by either LCIAs or GBMs was the effects of invasive species. Linking food systems to biodiversity We compared the items included in the models to the food system components established by the FAO, described below. All models provide system-level impacts, but no models provide impacts disaggregated to the different components of the food system i.e. production, distribution/aggregation, consumption and waste. The LCIAs do not have explicit requirements on what information is needed to calculate the ‘midpoint’ making it highly dependent on the system boundary which components of the food system are included. LCIAs are therefore limited by the midpoints included e.g. ReCiPe can only calculate the effects of the pollution from toxicity, freshwater and marine pollution, freshwater ecotoxicity, ozone and terrestrial acidification. For GBMs overall, most of the ‘midpoints’ in the models explicitly accounted for were related to agricultural production, with few impacts measuring other components in the food system. Only one GBM model, GLOBIO explicitly mentions aggregation/distribution. GLOBIO did this by accounting for the presence of roads in the model. The GBM GLOBIOM was the only model which included the consumption components. GLOBIOM uses ‘price’ of food and ‘consumer preference’ through the effects they have on changes in production. Waste was only dealt with by the GBM inSIGHTS in the form of ‘average production losses’ and was applied as a co-efficient to yield to symbolize on farm losses. Considering Scenarios in biodiversity modelling At times, factors are indirectly considered into biodiversity models, as assumption that underline the dynamics or starting points. Scenarios are either embedded into the models or developed with the use of external models and constraints. The use of scenarios allows models to account for greater complexity and indirect biodiversity loss pressures in both GBMs and LCIAs. Indirect pressures driving biodiversity loss according to IPBES include aspects such as technological development, economic, socioeconomic interactions, the role of culture and policy in shifting the quantity and quality consumption, production and other component in the food system. The LCIA ReCiPe accounts for different scenarios in the future through value choices allowing users to choose from individualistic, hierarchism, egalitarian scenarios. These scenarios change the intensity of outcomes based on assumptions on the timeframe, socio-economic developments and the ability to adapt. The other LCIAs also have the option to account for future scenarios but are less detailed in their assumptions, only accounting for long term and short-term impacts. Aside from C-SAR (2018) and PREDICTS, all the GBMs also use scenarios such as Shared Socioeconomic Pathways (SSPs) to estimate the impacts of climate change on biodiversity. GLOBIOM accounts for climate change through scenarios not directly through GHG production of food items but from the influence warming will have on yield reductions and subsequent land use changes that would result to fulfill nutritional needs of the population. Indirect pressures were not accounted for by any of the models except GLOBIOM which linked consumer preference and economic considerations with impacts of biodiversity. Unlike other models GLOBIOM is also an Integrated Assessment Model (IAM), which means it is developed with the intention that other components will be paired with the GBM in scenarios – something similar happens with GLOBIO, normally used paired with IMAGE. Discussion In this review we identified ten models assessing biodiversity in global food system, focusing on two main approaches: LCIA and GBMs. We found that while LCIA and GBMs have similar aims, they tackle biodiversity loss from different perspectives (production/consumption focused), with different coverage of biodiversity pressures, indicators for reporting biodiversity outcomes, and part of the food system that was covered. GBMs and LCIAs are powerful tools in biodiversity modelling and have been used to understand the broadest range of effects of a certain action, including how to accomplish important aims of the international community, such as target 10 of the Kunming-Montreal Global Biodiversity Framework 27 (2030 targets). Combining these modelling tools could widen our understanding and help achieve these targets. Divide in LCIA/GBMs and their respective blind spots Both LCIA and GBMs have strengths and limitations in assessing biodiversity following from the aims of the models. LCIA are usually more focused on specific producers (e.g. footprint of beef production) and informing consumers about the impact of specific products. GBMs tend to be more focused on the production system-scale (e.g. field, landscape, country), and are often used to manage tradeoffs between different land uses. These differences are reflected in how biodiversity is reported in the models. In LCIAs, the primary indicator is PDF (potentially disappearing fraction), which is used as a biodiversity 'cost' of different activities. In contrast, GBMs use more landscape to regional to global indicators, such as MSA (mean species abundance) and BII (biodiversity intactness index), to indicate conservation in wider spatial areas. Overall GBMs were able to consider a much broader range of indicators and ‘biodiversity impacts’. LCIAs focus only on species extinction, while GBMs focus on species extinction, ecosystem diversity and quality, compositional intactness, intactness of local species composition, extinction risk and extent of suitable habitat. To calculate these 'biodiversity impacts', however, GBMs incorporate fewer ‘activities in the food system’ and ‘IPBES pressures’ than LCIAs do. Regarding ‘pollution’, LCIAs allowed a much wider range of impacts to be accounted for, such as accounting for ecotoxicity of compounds involved in the process and/or acidification, a point supported by previous literature assessing earlier generations of LCIAs and GBMs 18 . The ‘food system activity’ ‘land use activities’ played a large role in determining the effects of biodiversity loss, although these are predominately linked to pesticide and fertilizer use and in some cases irrigation. In turn, none of the ‘land use activities’ account for known biodiversity enhancing farming practices and/or agricultural systems 20,30,31,32,33 . For example, fragmentation has been shown to be mediated by certain land use activities such as agroforestry systems 34,35 . Previous research has suggested improvements could be made to GBMs by increasing the number of activities included, to provide a more holistic assessment of food system impacts on biodiversity 20 . Furthermore, none of the models included were able to account for the biodiversity pressure ‘invasive species’. Invasive species are listed as one of the five IPBES pressures 2 and are a key element of the Kunming-Montreal Global Biodiversity Framework 27 . Some models such as the Climex model are able to directly account for invasive species, but estimates have yet to be formally linked to food systems modelling 36 Click here to enter text.. In terms of addressing different components of the food system LCIAs are more flexible than GBMs as it is largely up to the user to decide what components are included and is dependent on the system boundaries they set. For example, if an LCIA had the ‘midpoint’ GHG emissions, these could be taken at any component of the food system including those related to production, transport or processing. While the intention of the LCA is to encompass food system activities from extraction to waste disposal or any point along the supply chain, they are still limited by the 'midpoints’ they include and the data that is available. This can have varying disruptions on the assessment, ranging from minor to significant 22 . If we were to take GHG emissions, for example, it is always better to account for all components but indeed the user is flexible and can decide to exclude some components. Conversely, other midpoints are indeed limiting as, for example, different types of land use are not accounted for in the conversion from the midpoint to the impact. Guidelines for LCA (such as ISO or environmental foot printing methods) are geared towards the standardization of LCA however these have been critiqued due to the value choices employed in carrying out the guidelines and lack of transparency they allow 37 Click here to enter text.. LCIAs have also been critiqued for the difficulty in usability due to the relative complexity of the programs 38 . In contrast, a strength of GBMs in addressing food systems aspects is that they can focus on spatial dynamics of the food system and can assess rebounds in the system, for example, how shifts in consumption of one food item may impact the diet and subsequent ramification for production. Due to the static nature of LCIAs they are not able to do this within the confines of its calculations, making capturing the intrinsically dynamic nature of the food system difficult. Additionally, no ‘activities in the food system’ were found for processing and few were found to populate consumption, waste and aggregation, and distribution component. This means that input activities such as energy use for processing or the land use implications from food waste are not explicitly required for either method of assessing biodiversity loss. LCIAs can include additional activities in the food system but it is highly dependent on the system boundaries and not consistent among studies. In GBMs the use of additional scenarios can aid in addressing aspects of the food system such as food waste and consumption shifts, however this is not always done, leaving blind spots for the impact of many food system activities. Furthermore, there were limited feedback loops observed in models which could result in underestimation of effects. One good example might come in the form of trophic web breakdown, or not accounting for interspecies dependence – e.g. the loss of certain taxa will have cascading effects on the losses of other taxa dependent or connected to them or the lack of accounting for synergistic/vicious effects across pressures - climate change and water scarcity and pollution 39,40 . While collectively the models form a more complete picture of biodiversity loss in food system including more diverse indicators and aspects of the food system and food system activities such as types of land use, this picture is still incomplete. Data limitations Lack of data plays a large part in the model's incomplete assessment of biodiversity, a known issue in the literature on biodiversity 31,32,41,42,43 . The different models have different data challenges. LCIA data is fragmented and often behind paywalls. Organizations such as HESTIA are working to resolve these problems through the creation of free online databases 44 . While databases such as these are growing, there is still much data needed. Meanwhile, GBMs require intense information on species distribution of the regions assessed and might also require empirical data on the specific coefficient of impacts of each activity in that geography. Including LCIA methods and calculations within GBMs would help address more production factors and better discriminate the impact of specific activities, while at the same time allowing the model to accommodate the changes of impact that might come from the spatial dynamics: e.g., increases in demand of a food item deemed sustainable could lead to unsustainable production of that item, which now would result in an increase of its impact coefficient and the model could readjust to a new equilibrium. Proper assessment of biodiversity and the effects of the multiple components of the food systems will also require close cooperation and interaction with data collection and monitoring around the world to allow for its feasibility. Outlook Despite its importance, biodiversity loss remains the Achilles heel of global food systems modelling. Biodiversity assessments in food systems have advanced greatly in past decades, with new understanding of the mechanisms affecting biodiversity loss also shown in multiple studies 4,7,16,45,46 . Current modelling approaches are very useful in that they can help assess and simulate or compound specific indicators or pressures important for international biodiversity conservation aims, such as with target 10, enhancement of biodiversity and sustainability in agriculture, aquaculture, fisheries and forestry, of the 2030 targets established by the Kunming-Montreal Global Biodiversity Framework. Nevertheless, they are still lacking in providing the holistic view and assessment needed for biodiversity. We still observe a disconnect between our field/theoretical knowledge of biodiversity loss and our current modelling tools. Our review adds to existing literature on the limitations of models in how they assess different agriculture/food system characteristics 20,22,47 , by depicting that combining both models would improve current biodiversity assessment and could be possible by integrating LCA calculations within GBMs runs – that would still not consider all changes in dynamics, but would add the spatial and supply changes to LCIA’s broader assessment. Our findings indicate that incorporating biodiversity loss into models currently simplifies a web of factors to single proxy coefficients (e.g. land use activities) and/or are often limited to specific biodiversity impacts, such as species distribution or extinction fractions. Many assumptions and simplifications must be made in order to make up for data gaps and approximate impacts at such a large scale. While it is clear LCIA and GBMs are stronger together, these results beg the question of how much information is enough? And what should be included? The field of LCA and GBMs while addressing similar issues, often do not intersect. In order to better understand how to identify and prioritize missing pieces and in global food systems modelling, future research utilizing the expertise of divers' fields including ones that work with LCA and GBMs is needed. However, even in these cases our overview is limited. What this shows is not that models are not useful, but rather that biodiversity modelling should be used with the awareness of its limitation and used to understand potential trends of biodiversity loss, rather than specific values for biodiversity loss. In a time of unprecedented loss of biodiversity and unregular data accessibility, at times limited exactly where biodiversity hotpots are, models can help us bridge the gaps where the data is non-existent. For this, going forward three steps could help us with biodiversity modelling: models should (1) be used and interpreted with the awareness of the factors currently not considered and (2) combined in their use, methods and interpretation for a more comprehensive interpretation of biodiversity trends. Lastly (3), future modelling research should focus on integrating feedback loops into biodiversity modelling, currently lacking, as well as better specify biodiversity impacts of systems and practices. To properly assess biodiversity in agroecosystems to remain within the established aims of the CBD or IPBES framework, biodiversity modelling needs to more specifically address the species affected and the practices driving the impact, as well as feedback dynamics in the system. This will require alignment across research programs for data collection and policy aims. Methods The conceptualization for this paper was developed in a series of working group meetings with co-authors and other biodiversity experts brought together by the Wageningen Biodiversity Initiative. To determine how biodiversity loss is assessed within global food systems models we conducted a literature review. In this literature review we refined the search strategy to search in titles, abstracts and keywords for relevant terms including biodiversity, food and or system, production, agriculture or consumption. We chose to focus our search in this way as we wanted to assess commonly used models looking at biodiversity loss in the food system and the results from the papers themselves. Hence, the structured literature review was built to identify and retrieve models. We also chose to limit the number of search engines as we were aiming to capture common approaches and not a comprehensive overview of the topic, linking to our exclusion criteria of models used more than once. From this pool of papers, a systematic approach was carried out in selecting and extracting data. Research questions The following research questions were developed to guide the review: What ‘food system activities’ are used to estimate biodiversity loss in food systems models? How are ‘activities in the food system’ distributed across different levels or components of the food system? How do different ‘activities in the food system’ connect to biodiversity loss pressures in selected models? What are the ‘biodiversity impacts’ and indicators used in global models estimating biodiversity loss in food systems? Search strategy Articles were identified for review in the search database Scopus. The query for our preliminary review was ( TITLE-ABS-KEY ( food OR system* AND biodiversity AND model* AND ( diet OR production OR agriculture OR consumption ) AND ( land* OR region* OR global OR chain ) ) + TITLE ( biodiversity ) + KEY ( assessment OR model ) ) + TITLE ( biodiversity ) + KEY ( assessment OR model ), which resulted in a total of 59 models/papers initially selected (Fig. 1). Eligibility Once the preliminary literature review was concluded, we applied two selection criteria: (1) documentation available, either as proper documentation or dedicated scientific article of their application and replication; and (2) global in their assessment. These criteria ensured that the models selected were possible to be used by different groups and in different contexts and data input, as well as aligned in terms of scale. C-SAR 2016 was removed from the final selection due to its similarities with C-SAR 2018. Data extraction and analysis In this review we have made a distinction between two model classes the LCIAs and the GBMs, however, things rarely fit into one category. For example, we consider C-SAR as an GBM, although there are studies that consider it a model class of its own 45,47,48 , with multiple ways to classify models. Regardless of classifications, C-SAR models are already used in a static way in multiple LCIA to determine land use impact. The fact that this is already done further emphasizes the merits of combining different models in assessing global biodiversity loss. Regardless of if they are in the LCIA or GBMs, models which look at different aspects of the biodiversity loss problem will be complementary towards each other. Data extraction was done using the documentation of the model papers and papers themselves reporting on the model components of the models which were defined as eligible for inclusion. Information on biodiversity impacts, indicators, food system activities, biodiversity pressures and additional data collected can be found in an excel file in supplementary material A. To create an overview of how biodiversity loss is determined in food systems we used the following building block. Categorization and description of models using literature as reference categories: IPBES considered the following biodiversity pressures: invasive species, climate change, land use change, natural resource use and exploitation and pollution. Indirect pressures include social and economic aspects of the food system. We have added models to a table depicting what indicators and ‘midpoints’ are used to model each of the biodiversity related impacts of global food systems. To categorize the aspect of biodiversity impacted by the food system, we refer to the ‘biodiversity impacts’ adapted from definitions of Pereira et al. 23 using the following seven biodiversity impacts: global species extinction, reginal species extinction, regional species extinction risk, reduced ecosystem diversity and quality, compositional intactness, intactness of local species composition, and the extent of suitable habitat. Declarations Acknowledgements. This project received funding from the AVINA foundation (www.circularfoodsystems.org) and CropMix Project (project number: NWA.1389.20.160) of the Dutch Research Agenda (NWA-ORC) of the Dutch Research Council (NWO). We would like to thank the Wageningen Biodiversity Initiative with regards to funding to gather all experts together. We are also very grateful to the reviewers and editor of NPJ Biodiversity, whose comments and suggestions improved the quality of this review. Ethics declaration. The authors declare no competing interest. Author contributions. FM (Felipe Cozim Melges), WJ (Wendy Jenkins), HZ (Hannah H.E. van Zanten, the idea for the study. FM, WJ and HZ contributed to the structuring of the manuscript. FM and WJ, conducted the literature search. FM, WJ and HZ helped conceptualize and design the figures, FM and WJ created them. All other co-authors commented and reviewed the design of the research and the manuscript. FM and WJ analyzed the data, and HZ helped interpret the data analysis. All authors helped review the results. FM, WJ and HZ wrote the main manuscript text. All authors contributed to, reviewed and approved the final manuscript submitted. Data availability. All additional data is available in the supplementary materials. References "Nature, biodiversity and health: an overview of interconnections. Copenhagen: WHO Regional Office for Europe; 2021. Licence: CC BY-NC-SA 3.0 IGO." Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). (2019). Summary for policymakers of the global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES secretariat, Bonn. Campbell, B. M., Beare, D. J., Bennett, E. M., Hall-Spencer, J. M., Ingram, J. S. I., Jaramillo, F., Ortiz, R., Ramankutty, N., Sayer, J. A., & Shindell, D. (2017). Agriculture production as a major driver of the Earth system exceeding planetary boundaries. Ecology and Society, 22(4), art8. https://doi.org/10.5751/ES-09595-220408 IPBES. (2022, April). Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (Version 1). Zenodo. https://doi.org/10.5281/zenodo.6417333. Brooks, T. M., Mittermeier, R. A., Mittermeier, C. G., Da Fonseca, G. A. B., Rylands, A. B., Konstant, W. R., Flick, P., Pilgrim, J., Oldfield, S., Magin, G., & Hilton‐Taylor, C. (2002). Habitat Loss and Extinction in the Hotspots of Biodiversity. Conservation Biology, 16(4), 909–923. https://doi.org/10.1046/j.1523-1739.2002.00530.x Meyfroidt, P., Rudel, T. K., & Lambin, E. F. (2010). Forest transitions, trade, and the global displacement of land use. Proceedings of the National Academy of Sciences, 107(49), 20917–20922. https://doi.org/10.1073/pnas.1014773107 Molotoks, A., Henry, R., Stehfest, E., Doelman, J., Havlik, P., Krisztin, T., Alexander, P., Dawson, T. P., & Smith, P. (2020). Comparing the impact of future cropland expansion on global biodiversity and carbon storage across models and scenarios. Philosophical Transactions of the Royal Society B: Biological Sciences, 375(1794), 20190189. https://doi.org/10.1098/rstb.2019.0189 Newbold, T., Hudson, L. N., Arnell, A. P., Contu, S., De Palma, A., Ferrier, S., Hill, S. L. L., Hoskins, A. J., Lysenko, I., Phillips, H. R. P., Burton, V. J., Chng, C. W. T., Emerson, S., Gao, D., Pask-Hale, G., Hutton, J., Jung, M., Sanchez-Ortiz, K., Simmons, B. I., … Purvis, A. (2016). Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353(6296), 288–291. https://doi.org/10.1126/science.aaf2201 Allan, E., Manning, P., Alt, F., Binkenstein, J., Blaser, S., Blüthgen, N., Böhm, S., Grassein, F., Hölzel, N., Klaus, V. H., Kleinebecker, T., Morris, E. K., Oelmann, Y., Prati, D., Renner, S. C., Rillig, M. C., Schaefer, M., Schloter, M., Schmitt, B., … Fischer, M. (2015). Land use intensification alters ecosystem multifunctionality via loss of biodiversity and changes to functional composition. Ecology Letters, 18(8), 834–843. https://doi.org/10.1111/ele.12469 Britschgi, A., Spaar, R., & Arlettaz, R. (2006). Impact of grassland farming intensification on the breeding ecology of an indicator insectivorous passerine, the Whinchat Saxicola rubetra: Lessons for overall Alpine meadowland management. Biological Conservation, 130(2), 193–205. https://doi.org/10.1016/j.biocon.2005.12.013 Carrie, R., Andrieu, E., Ouin, A., & Steffan-Dewenter, I. (2017). Interactive effects of landscape-wide intensity of farming practices and landscape complexity on wild bee diversity. Landscape Ecology, 32(8), 1631–1642. https://doi.org/10.1007/s10980-017-0530-y Agostini, M. G., Roesler, I., Bonetto, C., Ronco, A. E., & Bilenca, D. (2020). Pesticides in the real world: The consequences of GMO-based intensive agriculture on native amphibians. Biological Conservation, 241, 108355. https://doi.org/10.1016/j.biocon.2019.108355 Fromentin, J.-M. et al. Status, challenges and pathways to the sustainable use of wild species. Global Environmental Change 81, 102692 (2023). The Convention on Biological Diversity (CBD). (2022a). Final text of Kunming-Montreal Global Biodiversity Framework available in all languages. https://www.cbd.int/conferences/2021-2022/cop-15/documentsasdocument:CBD/COP/15/L25 European Parlament, Directorate-General for Internal Policies of the Union, & Nègre, F. (2020). The EU 2030 biodiversity strategy. European Parlament. https://doi.org/10.2861/545892 Leclère, D., Obersteiner, M., Barrett, M., Butchart, S. H. M., Chaudhary, A., De Palma, A., DeClerck, F. A. J., Di Marco, M., Doelman, J. C., Dürauer, M., Freeman, R., Harfoot, M., Hasegawa, T., Hellweg, S., Hilbers, J. P., Hill, S. L. L., Humpenöder, F., Jennings, N., Krisztin, T., … Young, L. (2020). Bending the curve of terrestrial biodiversity needs an integrated strategy. Nature, 585(7826), 551–556. https://doi.org/10.1038/s41586-020-2705-y Mommer, L., Nel, J., van Apeldoorn, D., van Hattum, T., Jones-Walters, L., Polman, N., Richter, A., Westerink, J., Alpízar, F., Candel, J., van Doorn, A., van Duinkerken, G., Maas, J., Fraanje, W., Groot, A., Jongeneel, R., Navarro-Matser, M., Ritzer, S., Siegmund-Schultze, M., ... de Wit, M. (2022). Nature-positive futures: Food systems as a catalyser for change. Wageningen University & Research. https://doi.org/10.18174/574286 Souza, D. M., Teixeira, R. F. M., & Ostermann, O. P. (2015). Assessing biodiversity loss due to land use with Life Cycle Assessment: Are we there yet? Global Change Biology, 21(1), 32–47. https://doi.org/10.1111/gcb.12709 Chopin, P., Bergkvist, G., & Hossard, L. (2019). Modelling biodiversity change in agricultural landscape scenarios—A review and prospects for future research. Biological Conservation, 235, 1–17. https://doi.org/10.1016/j.biocon.2019.03.046 López Rodríguez, S., Van Bussel, L. G. J., & Alkemade, R. (2024). Classification of agricultural land management systems for global modeling of biodiversity and ecosystem services. Agriculture, Ecosystems & Environment, 360, 108795. https://doi.org/10.1016/j.agee.2023.108795 Boakes, E. H., Dalin, C., Etard, A., & Newbold, T. (2024). Impacts of the global food system on terrestrial biodiversity from land use and climate change. Nature Communications, 15(1), 5750. https://doi.org/10.1038/s41467-024-49999-z Sanyé-Mengual, E., Biganzoli, F., Valente, A., Pfister, S., & Sala, S. (2023). What are the main environmental impacts and products contributing to the biodiversity footprint of EU consumption? A comparison of life cycle impact assessment methods and models. International Journal of Life Cycle Assessment, 28(9), 1194–1210. https://doi.org/10.1007/S11367-023-02169-7/FIGURES/5 Pereira, H. M., Martins, I. S., Rosa, I. M. D., Kim, H., Leadley, P., Popp, A., van Vuuren, D. P., Hurtt, G., Quoss, L., Arneth, A., Baisero, D., Bakkenes, M., Chaplin-Kramer, R., Chini, L., Di Marco, M., Ferrier, S., Fujimori, S., Guerra, C. A., Harfoot, M., … Alkemade, R. (2024). Global trends and scenarios for terrestrial biodiversity and ecosystem services from 1900 to 2050. Science, 384(6694), 458–465. https://doi.org/10.1126/science.adn3441 FAO. 2018. Sustainable food systems: concept and framework. Brief. Rome ISO. (2022). ISO 14044:2006 Environmental management — Life cycle assessment — Requirements and guidelines. Alkemade, R., van Bussel, L. G., Rodríguez, S. L. & Schipper, A. M. Global biodiversity assessments need to consider mixed multifunctional land-use systems. Current Opinion in Environmental Sustainability 56, 101174 (2022). 15/4. Kunming-Montreal Global Biodiversity Framework. (2024). The Convention on Biological Diversity (CBD). https://www.cbd.int/doc/decisions/cop-15/cop-15-dec-04-en.pdf European Parlament, Directorate-General for Internal Policies of the Union, & Nègre, F. (2020). The EU 2030 biodiversity strategy. European Parlament. https://doi.org/10.2861/545892 CBD (Convention on Biological Diversity) (2011) Strategic Plan for Biodiversity 2011-2020, Including Aichi Biodiversity Targets. https://www.cbd.int/sp/ Bengtsson, J., Ahnström, J., & Weibull, A.-C. (2005). The effects of organic agriculture on biodiversity and abundance: A meta-analysis: Organic agriculture, biodiversity and abundance. Journal of Applied Ecology, 42(2), 261–269. https://doi.org/10.1111/j.1365-2664.2005.01005.x Cozim-Melges, F., Ripoll-Bosch, R., Veen, G. F., Oggiano, P., Bianchi, F. J. J. A., Van Der Putten, W. H., & Van Zanten, H. H. E. (2024). Farming practices to enhance biodiversity across biomes: A systematic review. Npj Biodiversity, 3(1), 1. https://doi.org/10.1038/s44185-023-00034-2 Cozim-Melges, F., Ripoll-Bosch, R., Oggiano, P., van Zanten, H. H. E., van der Putten, W. H., & Veen, G. F. (2025). The effect of alternative agricultural practices on soil biodiversity of bacteria, fungi, nematodes and earthworms: A review. Agriculture, Ecosystems and Environment, 379, Article 109329. https://doi.org/10.1016/j.agee.2024.109329 Udawatta, R. P., Rankoth, L. M., & Jose, S. (2021). Agroforestry for biodiversity conservation. Agroforestry and Ecosystem Services, 245–274. https://doi.org/10.1007/978-3-030-80060-4_10/FIGURES/7 Vagge, I., Sgalippa, N., & Chiaffarelli, G. (2024). The role of agroforestry in solving the agricultural landscapes vulnerabilities in the Po Plain district. Community Ecology, 1–27. https://doi.org/10.1007/S42974-024-00203-8/TABLES/7 Kriticos, D. J., Maywald, G. F., Yonow, T., Zurcher, E. J., Herrmann, N. I., & Sutherst, R. W. (2015). CLIMEX Version 4: Exploring the Effects of Climate on Plants, Animals and Diseases. CSIRO, Canberra. 156 pp. Roßmann, M., Stratmann, M., Rötzer, N., Schäfer, P., & Schmidt, M. (2021). Comparability of lcas — review and discussion of the application purpose. Sustainable Production, Life Cycle Engineering and Management, 213–225. https://doi.org/10.1007/978-3-030-50519-6_15/FIGURES/1 Curran, M., Maia De Souza, D., Anton, A., Teixeira, R., Michelsen, O., Vidal Legaz, B., Sala, S. and Mila I Canals, L., How well does LCA model land use impacts on biodiversity—A comparison with approaches from ecology and conservation, ENVIRONMENTAL SCIENCE and TECHNOLOGY, ISSN 0013-936X, 50 (6), 2016, p. 2782–2795, JRC110785 Fatemi Nasrollahi, F.S., Campbell, C. & Albert, R. Predicting cascading extinctions and efficient restoration strategies in plant–pollinator networks via generalized positive feedback loops. Sci Rep 13, 902 (2023). https://doi.org/10.1038/s41598-023-27525-3 Luis Abdala-Roberts, Adriana Puentes, Deborah L Finke, Robert J Marquis, Marta Montserrat, Erik H Poelman, Sergio Rasmann, Arnaud Sentis, Celia C Symons, Nicole M van Dam, Gina Wimp, Christer Björkman, Kailen A Mooney (2025) Connecting the dots: Managing species interaction networks to mitigate the impacts of global change eLife https://doi.org/ https://doi.org/10.7554/eLife.98899 Wetzel, F. T., Bingham, H. C., Groom, Q., Haase, P., Kõljalg, U., Kuhlmann, M., Martin, C. S., Penev, L., Robertson, T., Saarenmaa, H., Schmeller, D. S., Stoll, S., Tonkin, J. D., & Häuser, C. L. (2018). Unlocking biodiversity data: Prioritization and filling the gaps in biodiversity observation data in Europe. Biological Conservation, 221, 78–85. https://doi.org/10.1016/j.biocon.2017.12.024 Cameron, E. K., Martins, I. S., Lavelle, P., Mathieu, J., Tedersoo, L., Gottschall, F., Guerra, C. A., Hines, J., Patoine, G., Siebert, J., Winter, M., Cesarz, S., Delgado-Baquerizo, M., Ferlian, O., Fierer, N., Kreft, H., Lovejoy, T. E., Montanarella, L., Orgiazzi, A., … Eisenhauer, N. (2018). Global gaps in soil biodiversity data. Nature Ecology & Evolution, 2(7), 1042–1043. https://doi.org/10.1038/s41559-018-0573-8 Bonfanti, J., Langridge, J., Avadí, A., Casajus, N., Chaudhary, A., Damour, G., Estrada-Carmona, N., Jones, S. K., Makowski, D., Mitchell, M., Seppelt, R., & Beillouin, D. (2024). Global review of meta-analyses reveals key data gaps in agricultural impact studies on biodiversity in croplands. https://doi.org/10.1101/2024.04.19.590051 Hestia. Available at: https://www.hestia.earth/explorer (Accessed: 23 May 2025). Kim, H., Rosa, I. M. D., Alkemade, R., Leadley, P., Hurtt, G., Popp, A., van Vuuren, D. P., Anthoni, P., Arneth, A., Baisero, D., Caton, E., Chaplin-Kramer, R., Chini, L., De Palma, A., Di Fulvio, F., Di Marco, M., Espinoza, F., Ferrier, S., Fujimori, S., … Pereira, H. M. (2018). A protocol for an intercomparison of biodiversity and ecosystem services models using harmonized land-use and climate scenarios. bioRxiv, 300632. https://doi.org/10.1101/300632 Adhikari, S., Adhikari, A., Weaver, D. K., Bekkerman, A., & Menalled, F. D. (2019). Impacts of Agricultural Management Systems on Biodiversity and Ecosystem Services in Highly Simplified Dryland Landscapes. Sustainability, 11(11), 3223. https://doi.org/10.3390/su11113223 Crenna, E., Marques, A., la Notte, A., & Sala, S. (2020). Biodiversity Assessment of Value Chains: State of the Art and Emerging Challenges. Environmental Science and Technology, 54(16), 9715–9728. https://doi.org/10.1021/ACS.EST.9B05153/ASSET/IMAGES/LARGE/ES9B05153_0002.JPEG Pereira, H. M., & Daily, G. C. (2006). MODELING BIODIVERSITY DYNAMICS IN COUNTRYSIDE LANDSCAPES. Ecology, 87(8), 1877–1885. https://doi.org/10.1890/0012-9658(2006)87[1877:MBDICL]2.0.CO;2 PEREIRA, H. M., ZIV, G., & MIRANDA, M. (2014). Countryside Species–Area Relationship as a Valid Alternative to the Matrix-Calibrated Species–Area Model. Conservation Biology, 28(3), 874–876. https://doi.org/10.1111/cobi.12289 Jaureguiberry, P., Titeux, N., Wiemers, M., Bowler, D. E., Coscieme, L., Golden, A. S., Guerra, C. A., Jacob, U., Takahashi, Y., Settele, J., Díaz, S., Molnár, Z., & Purvis, A. (2022). The direct drivers of recent global anthropogenic biodiversity loss. Science advances, 8(45), eabm9982. https://doi.org/10.1126/sciadv.abm9982 Tables Tables 1 to 4 are available in the Supplementary Files section Additional Declarations No competing interests reported. Supplementary Files SupplementaryMaterialA.7z Tables.docx Cite Share Download PDF Status: Under Revision Version 1 posted Editorial decision: Revision requested 31 Jul, 2025 Reviews received at journal 30 Jul, 2025 Reviews received at journal 22 Jul, 2025 Reviewers agreed at journal 04 Jun, 2025 Reviewers agreed at journal 02 Jun, 2025 Reviewers invited by journal 02 Jun, 2025 Editor assigned by journal 02 Jun, 2025 Submission checks completed at journal 30 May, 2025 First submitted to journal 23 May, 2025 You are reading this latest preprint version Research Square lets you share your work early, gain feedback from the community, and start making changes to your manuscript prior to peer review in a journal. As a division of Research Square Company, we’re committed to making research communication faster, fairer, and more useful. We do this by developing innovative software and high quality services for the global research community. Our growing team is made up of researchers and industry professionals working together to solve the most critical problems facing scientific publishing. Also discoverable on Platform About Our Team In Review Editorial Policies Advisory Board Help Center Resources Author Services Accessibility API Access RSS feed Manage Cookie Preferences © Research Square 2026 | ISSN 2693-5015 (online) Privacy Policy Terms of Service Do Not Sell My Personal Information {"props":{"pageProps":{"initialData":{"identity":"rs-6735266","acceptedTermsAndConditions":true,"allowDirectSubmit":false,"archivedVersions":[],"articleType":"Article","associatedPublications":[],"authors":[{"id":465200638,"identity":"4f9fc19c-b517-432a-82b8-fd0d3018e21a","order_by":0,"name":"Felipe Cozim Melges","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAABD0lEQVRIie3RsWqEMBjA8YigS6zrBzfcK3hTO8j5KorQySsHXW64IQ7NVHBV6FN0uKlDJGAXoavgVApO3QqFLkc/m+FaaDIXmj9EE8kPgxJis/3BfOYyoaYOw8sFDlc9ONMQKpwfBHB46dfaM5Dv65nQyEz8spTbHVmGoby5H3awrsL+fRE8kCstoS2TdU9WDWv5WPSQN/XmsKATudaRBDImA07SqC35uOGQR0NwGKkgGde9ZfmM5IhEOkiOkCdP/WQm4CBhSLqZMFhHpPDMhOLB6g5WzS2SooMUhsvzjzsBeuJL+bbdx/jFHqex2MdJWMmX6FXEWaUhc676HaqMqTv8vvdETiXGrTabzfYv+wQe81u6T6vQGAAAAABJRU5ErkJggg==","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":true,"prefix":"","firstName":"Felipe","middleName":"Cozim","lastName":"Melges","suffix":""},{"id":465200639,"identity":"7730abba-3967-470e-92ae-16c01c4efc7a","order_by":1,"name":"Wendy Margarida Nina Jenkins","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Wendy","middleName":"Margarida Nina","lastName":"Jenkins","suffix":""},{"id":465200640,"identity":"9a37cf54-7280-4f1a-a4a7-b368c5111669","order_by":2,"name":"Rob Alkemade","email":"","orcid":"","institution":"Netherlands Environmental Assessment Agency","correspondingAuthor":false,"prefix":"","firstName":"Rob","middleName":"","lastName":"Alkemade","suffix":""},{"id":465200641,"identity":"b067b7c7-a3a6-491e-b61d-d90d8d579789","order_by":3,"name":"Jeanne Nel","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Jeanne","middleName":"","lastName":"Nel","suffix":""},{"id":465200642,"identity":"6d6b4d5c-e708-4f7c-8e06-6f45e5026eb6","order_by":4,"name":"Patrick A. Jansen","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Patrick","middleName":"A.","lastName":"Jansen","suffix":""},{"id":465200643,"identity":"a203cdec-6b76-4609-ab84-8468b92d1caf","order_by":5,"name":"Argyris Kanellopoulos","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Argyris","middleName":"","lastName":"Kanellopoulos","suffix":""},{"id":465200644,"identity":"546b3da5-491d-42dc-b338-1305409af066","order_by":6,"name":"Dirk van Apeldoorn","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Dirk","middleName":"van","lastName":"Apeldoorn","suffix":""},{"id":465200645,"identity":"922c80fa-b864-420d-91d2-28f1793b3496","order_by":7,"name":"Jorad de Fries","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Jorad","middleName":"","lastName":"de Fries","suffix":""},{"id":465200646,"identity":"a525c037-606c-4467-ae8f-4469fd3e77ae","order_by":8,"name":"Liesje Mommer","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Liesje","middleName":"","lastName":"Mommer","suffix":""},{"id":465200648,"identity":"a403e8ce-8420-48c1-80db-27483c000592","order_by":9,"name":"Erik H. Poelman","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Erik","middleName":"H.","lastName":"Poelman","suffix":""},{"id":465200651,"identity":"7215f9ec-9129-4cb5-9ec6-5f2bab5bff8f","order_by":10,"name":"Raimon Ripoll-Bosch","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Raimon","middleName":"","lastName":"Ripoll-Bosch","suffix":""},{"id":465200654,"identity":"1424be3f-a02e-4dbb-98b5-4ee9233c949a","order_by":11,"name":"G. F. (Ciska) Veen","email":"","orcid":"","institution":"Netherlands Institute of Ecology","correspondingAuthor":false,"prefix":"","firstName":"G.","middleName":"F. (Ciska)","lastName":"Veen","suffix":""},{"id":465200658,"identity":"8ad48c4e-8024-4082-93ba-3cd38eec4afc","order_by":12,"name":"Arnold J.H. van Vliet","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Arnold","middleName":"J.H. van","lastName":"Vliet","suffix":""},{"id":465200663,"identity":"2b01e991-9f44-40eb-a571-f498b496db94","order_by":13,"name":"Michael Clark","email":"","orcid":"","institution":"University of Oxford","correspondingAuthor":false,"prefix":"","firstName":"Michael","middleName":"","lastName":"Clark","suffix":""},{"id":465200665,"identity":"acc6c015-ef60-4551-a3bf-b25db16c5114","order_by":14,"name":"Luisa M. Trindade","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Luisa","middleName":"M.","lastName":"Trindade","suffix":""},{"id":465200667,"identity":"cf1d3733-d63f-482f-b94b-3b5ca3df39a6","order_by":15,"name":"Hannah van Zanten","email":"","orcid":"","institution":"Wageningen University \u0026 Research","correspondingAuthor":false,"prefix":"","firstName":"Hannah","middleName":"van","lastName":"Zanten","suffix":""}],"badges":[],"createdAt":"2025-05-23 19:08:20","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6735266/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6735266/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":83880432,"identity":"46da8f43-22d2-4206-a738-38770a46861e","added_by":"auto","created_at":"2025-06-04 05:11:17","extension":"png","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":237778,"visible":true,"origin":"","legend":"\u003cp\u003eOverview of the results of the literature review of models and inclusion and exclusion criteria used.\u003c/p\u003e","description":"","filename":"1.png","url":"https://assets-eu.researchsquare.com/files/rs-6735266/v1/f0acaa26702192fc658eadfc.png"},{"id":83881850,"identity":"fd4c3086-346d-46c8-b2e1-f2f977e01774","added_by":"auto","created_at":"2025-06-04 05:43:17","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":734154,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6735266/v1/ab263e91-4788-42fb-8436-9ff30eff5180.pdf"},{"id":83880437,"identity":"2a16dda7-9bed-4b87-b8d3-be24b3c1b186","added_by":"auto","created_at":"2025-06-04 05:11:17","extension":"7z","order_by":1,"title":"","display":"","copyAsset":false,"role":"supplement","size":110253,"visible":true,"origin":"","legend":"","description":"","filename":"SupplementaryMaterialA.7z","url":"https://assets-eu.researchsquare.com/files/rs-6735266/v1/87e274fdbb43291b529cc9b2.7z"},{"id":83880438,"identity":"22a5c45e-f8fb-4677-9c5d-ba2c1ef7c512","added_by":"auto","created_at":"2025-06-04 05:11:17","extension":"docx","order_by":2,"title":"","display":"","copyAsset":false,"role":"supplement","size":233296,"visible":true,"origin":"","legend":"","description":"","filename":"Tables.docx","url":"https://assets-eu.researchsquare.com/files/rs-6735266/v1/53110888a61f8ed56ae710d9.docx"}],"financialInterests":"No competing interests reported.","formattedTitle":"Blind spots of modelling biodiversity loss in food systems","fulltext":[{"header":"Introduction","content":"\u003cp\u003eBeyond its intrinsic value, biodiversity is essential for human health and well-being through the provision and maintenance of ecosystem services such as crop pollination\u003csup\u003e1\u003c/sup\u003e.\u0026nbsp;Despite its importance global biodiversity is declining at unprecedented rates\u003csup\u003e2\u003c/sup\u003e. Food systems have been identified as the primary driver of global biodiversity loss\u003csup\u003e3,4\u003c/sup\u003e, caused by different activities, ranging from land expansion into natural areas\u003csup\u003e5,6,7,8\u003c/sup\u003e\u0026nbsp; to the intensification of production methods\u003csup\u003e9,10,11,12\u003c/sup\u003e\u0026nbsp; or over-harvesting of wild foods\u003csup\u003e13\u003c/sup\u003e. International policy agreements like the Kunming-Montreal Global Biodiversity Framework (GBF)\u003csup\u003e14\u003c/sup\u003e or the EU 2030 Biodiversity Strategy\u003csup\u003e15\u003c/sup\u003e, aim to reverse the trend of biodiversity loss. Effective nature conservation strategies combined with improvements to supply chains, from production to waste disposal, will be needed to bend the curve of biodiversity loss\u003csup\u003e16\u003c/sup\u003e. Capturing the effects of the whole supply chain on biodiversity however requires a food systems approach\u003csup\u003e17\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGlobal food systems models attempt to capture the effects of the whole supply chain on a variety of impacts. \u0026nbsp;Many different approaches to modelling biodiversity have been put forward in global food systems models, each with their own set of assumptions and simplifications to assess biodiversity\u003csup\u003e18,19,20\u003c/sup\u003e, e.g. different systems boundaries.\u0026nbsp;\u0026nbsp;The different assumptions from different models can yield different results for recommendations to reduce biodiversity loss in the food system\u003csup\u003e21,22\u003c/sup\u003e. This review aims to create an overview of what is currently being done and map how current approaches model biodiversity loss. We compared the models by looking at biodiversity pressures, indicators and food system components in each model, as per the definitions in the text box ‘Terminology Used’ below.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eFirst, we looked at how models are accounting for the most significant biodiversity pressures as outlined by the Global Assessment of the Intergovernmental Platform on Biodiversity and Ecosystem Services (IPBES)\u003csup\u003e2\u003c/sup\u003e, i.e., natural resource use and exploitation, pollution, climate change, land use, invasive species and occupation (table 2).\u0026nbsp;Second, we explored\u0026nbsp;which indicators were being used to assess biodiversity and clustering them based on the biodiversity impact they assess, as described by Pereira et al.\u003csup\u003e23\u003c/sup\u003e: species extinction, extent of suitable habitat, compositional intactness and species richness (table 3). Finally, we developed a comprehensive overview of how biodiversity is modelled across the components of the food system as defined by the FAO\u003csup\u003e24\u003c/sup\u003e as production, aggregation, distribution, processing, consumption and waste disposal. Multiple different venues to model biodiversity in food systems currently exist. Life Cycle Impact Assessments (LCIAs) are a step in the assessment method of the life cycle analysis, which is often employed when looking at the impact of a product\u003csup\u003e25\u003c/sup\u003e. LCIAs work by translating the effects of raw material use into environmental outcomes, such as climate change or water use which are then converted into biodiversity impacts\u003csup\u003e25\u003c/sup\u003eClick here to enter text.. Global Biodiversity Models (GBMs), using Integrated Assessment Models (IAM), are models which are used in tandem with designed scenarios to assess linkages between human activity and the biosphere, often focusing on land use activities, such as agricultural production\u003csup\u003e26\u003c/sup\u003e. Like LCIA, GBMs translate impacts such as climate change and land use and occupation into their impacts on biodiversity. Unlike LCIA, GBMs tend to make assessments clustering different midpoint outcomes into common impact coefficients factors, which are then used to estimate biodiversity impacts.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eOur results showed that LCIAs are able to encompass more specific impacts of the food system but lack the spatial dynamics and biodiversity indicator breadth of GBMs. We argue that using the models in tandem has potential for addressing some of the blind spots in both modelling approaches and increase the range of drivers covered, allowing for more robust and realistic considerations of biodiversity loss and potential interventions to bend the curve of biodiversity loss. This more comprehensive consideration is direly needed for the urgent action proposed in the Kunming-Montreal Global Biodiversity Framework mission\u003csup\u003e27\u003c/sup\u003e and the EU 2030 Biodiversity Strategy\u003csup\u003e28\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eBox 1. Terminology used\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eTo better merge terminology from different types of models we distinguish our definitions of the concepts below:\u003c/p\u003e\n\u003cp\u003e1) Biodiversity: “the variability and interaction among living organisms from all sources including, inter alia, terrestrial, marine, and freshwater ecosystems and the ecological complexes of which they are part of; this includes diversity within species, between species and of ecosystems”\u003csup\u003e29\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e2) Food systems: “encompass the entire range of actors and their interlinked value-adding activities involved in the production, aggregation, processing, distribution, consumption and disposal of food products that originate from agriculture, forestry or fisheries, and parts of the broader economic, societal and natural environments in which they are embedded”\u003csup\u003e24\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003e3) ‘Midpoints’ is a commonly used term in LCAs and are intermediate measures of impact of an activity known to, directly and/or indirectly, affect biodiversity loss e.g. the total amount of fertilizer used will contribute to the midpoints such as ‘GHG emissions’. We clustered ‘midpoints’ into broader categories of the ‘IPBES pressures’ (land use, natural resource use and exploitation, pollution and climate change). Clustering the midpoints from models into IPBES pressures allows us to see how each model is or is not addressing a biodiversity pressure. For example, for the IPBES pressure ‘land use’, some models may only account for the midpoint ‘land occupation’ while others may also account for ‘fragmentation’ or other land use related impacts. Each midpoint can be translated into a ‘biodiversity impact’.\u003c/p\u003e\n\u003cp\u003e4) ‘Biodiversity impact’ is the broad aspect of biodiversity encompassed by the indicator used in the model. ‘Biodiversity impact’ classifications were defined based on an adaptation of Pereira et al.\u003csup\u003e23\u003c/sup\u003e.\u0026nbsp;To address which aspect of biodiversity is being assessed in models, ‘biodiversity impact' classifications can be used to report the component of biodiversity being assessed across indicators. For example, the biodiversity intactness index and mean species abundance are two different indicators for the compositional intactness of biodiversity but compare different baselines and under different assumptions/perspectives. This classification\u0026nbsp;allowed us to assess how holistic/broad the biodiversity assessment of the combined models is\u003csup\u003e2\u003c/sup\u003e.\u0026nbsp;\u003c/p\u003e"},{"header":"Results","content":"\u003cp\u003eTo determine how biodiversity is accounted for in global food systems models, we conducted a literature review, retrieving a total of 126 papers, which after screening and filtering resulted in identifying ten distinct food-system models that assess global biodiversity (table 1, figure 1 methods). Two contrasting approaches emerged, Life Cycle Impact Assessment (LCIA), which captures product level impacts, and Global Biodiversity Models (GBM), which capture spatially dynamic impacts and tradeoffs assessing effects of the food system on biodiversity loss: Life Cycle Impact Assessments (LCIAs) (3 models: IMPACT world + ReCiPe 2016 and LC-impact) and Global Biodiversity Models (GBMs) (7 models: Map of Life, BILBI, PREDICTS, GLOBIOM, C-SAR 2018, inSIGHTS and GLOBIO \u0026ndash; table 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn the section below we will discuss the \u0026lsquo;IPBES pressures\u0026rsquo;, \u0026lsquo;midpoints\u0026rsquo;, \u0026lsquo;biodiversity indicators\u0026rsquo;, \u0026lsquo;biodiversity impacts\u0026rsquo; and \u0026lsquo;food system activities\u0026rsquo; that were found in the models. An overview of the blind spots and strengths of the models can be found in table 2. Broadly speaking, GBM models account for a more diverse range of indicators compared to the LCIAs, however in terms of \u0026lsquo;IPBES pressures\u0026rsquo;, the LCIAs more consistently included the pressures, particularly regarding pollution. The only IPBES pressure not included in any of the retrieved models was invasive species. In terms of food system activities production was the most represented with GBMs also explicitly including aggregation. It should be noted however that scenarios enable more diverse \u0026lsquo;food system activities\u0026rsquo; to be included in model outputs. For LCIAs the food system activities are dependent on the scenario chosen and its associated system boundaries, however like GBMs they are limited in the effects across the supply chain they can assess. The details of these results will be discussed in the sections below.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eBiodiversity indicators and metrics\u003c/p\u003e\n\u003cp\u003eOverall GBMs used six indicators for measuring biodiversity (PDF, SHI, BII, MSA, RLI and ESH) corresponding to all the \u0026lsquo;biodiversity impact\u0026rsquo; categories outlined by Pereira et al.\u003csup\u003e23\u003c/sup\u003e. LCIAs varied much less than GBMs using only the indicator PDF, corresponding to the biodiversity impact \u0026lsquo;species extinction\u0026rsquo;. The GBM inSIGHTS was the only model to account for multiple biodiversity indicators in the same model (RLI and ESH) covering the \u0026lsquo;biodiversity impacts\u0026rsquo; \u0026lsquo;species extinction\u0026rsquo; and \u0026lsquo;extent of suitable habitat\u0026rsquo; (table 3). C-SAR is the only GBM that uses the indicator PDF, and along with inSIGHTS is the only other GBM to account for the \u0026lsquo;biodiversity impact\u0026rsquo; \u0026lsquo;species extinction\u0026rsquo;. The \u0026lsquo;biodiversity impact\u0026rsquo; \u0026lsquo;extent of suitable habitat\u0026rsquo; was used by Map of Life with the indicator SHI, which is unique in that it uses GIS data to calculate ecosystem quality and diversity relative to a baseline year. The \u0026lsquo;biodiversity impact\u0026rsquo; \u0026lsquo;compositional intactness\u0026rsquo; represented by the indicator BII was used by BILBI, PREDICTS and GLOBIOM. BII looks specifically at changes in native terrestrial species to estimate the compositional intactness. Similarly, GLOBIO uses the \u0026lsquo;biodiversity impact\u0026rsquo; \u0026lsquo;compositional intactness\u0026rsquo; linked to the indicator MSA, the difference being, that while both the indicator BII and MSA look at all species native/original in a given region, BII looks at the percentage of loss by using total species abundance and a compositional similarity index, while GLOBIO does so referring to the mean species abundance of a given area. Table 3 depicts the pressures assessed and indicators used by the models retrieved.\u003c/p\u003e\n\u003cp\u003eFood systems\u0026rsquo; activities and IPBES\u0026rsquo; pressures\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn this section we present the results of the models retrieved in terms of the IPBES pressures. In table 4 we present the ways in which these pressures are represented in the models. We find that while all pressures are covered, they are mostly done with individual midpoints.\u003c/p\u003e\n\u003cp\u003e1) \u0026lsquo;Land use\u0026rsquo; was the most included \u0026lsquo;IPBES pressure\u0026rsquo; with all the GBMs and LCIAs including it. For LCIAs, only the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;land occupation and transformation\u0026rsquo; was included referring to the land use type, if it has been converted from an alternate land use activity e.g. nature areas to pasture. For the GBMs, both Map of life and GLOBIO included the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;habitat fragmentation\u0026rsquo; relating to the \u0026lsquo;IPBES pressure\u0026rsquo; \u0026lsquo;land use\u0026rsquo;. \u0026lsquo;Habitat fragmentation\u0026rsquo; causes a habitat to be split into a larger number of smaller fragments. This reduces the area and population size of each fragment, which reduces likelihood of persistence of the population in the fragment. \u0026lsquo;Agricultural area\u0026rsquo; was used to assess \u0026lsquo;fragmentation\u0026rsquo; and was dependent on location\u0026rsquo;, with land closer in proximity to natural areas and larger agricultural areas having higher impacts. The presence and type of road also impacted the severity of effects with larger roads having a bigger impact. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGLOBIO was the only GBM to account for the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;infrastructure disturbance\u0026rsquo; which looks at the effects of presence of infrastructure on neighboring species. \u0026lsquo;Land use activity\u0026rsquo; and \u0026lsquo;settlement population size\u0026rsquo; were both used by models to calculate the effects of the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;infrastructure disturbance\u0026rsquo;. In GLOBIO the impact zones related to \u0026lsquo;infrastructure disturbance\u0026rsquo; are predicted to change as the result of increasing human population sizes with the severity of impact is dependent on the type of land cover. The impacts of \u0026lsquo;infrastructure disturbance\u0026rsquo; differ from fragmentation because they arise from a species preference to avoid developed areas and roads as opposed to the effect of a physical barrier to movement. Similar to fragmentation, the \u0026lsquo;presence and type of roads\u0026rsquo; also impacted the effects of infrastructure on biodiversity. Other factors mediating the effects of infrastructure were factors including distance to the coast and distance from protected areas.\u003c/p\u003e\n\u003cp\u003e\u0026lsquo;Land occupation and transformation\u0026rsquo; was calculated using \u0026lsquo;land use activities\u0026rsquo; by all models. \u0026lsquo;Land use activities\u0026rsquo; varied from model to model with ReCiPe 2016 having the most different types of land use activities amongst the LCIAs (used forest, pasture and meadow, annual crops, permanent crops, mosaic agriculture) and GLOBIOM including the most amongst the GBMS (cropland (18 crops globally), managed grassland, short rotation plantations, intensity (subsistence, low input rainfed, high input rainfed, high input irrigated)).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e2) Natural resource use and exploitation was included in all LCIA models in the form of the \u0026lsquo;food system activity\u0026rsquo; \u0026lsquo;water use\u0026rsquo;. The impacts of \u0026lsquo;water use\u0026rsquo; were determined through assessing the reductions in availability for local plants and shifts in river discharge on terrestrial and marine species. GLOBIO was the only GBM to look at water use. GLOBIO calculates the effect of water limitations on \u0026lsquo;land use activities\u0026rsquo;, meaning that limited water would shift the \u0026lsquo;land use activity\u0026rsquo; to one that requires less as it is based on optimization. GLOBIOM was the only GBM to include the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;water use\u0026rsquo;. The impacts of \u0026lsquo;water use\u0026rsquo; were however differently accounted for than in the LCIAs as GLOBIOM assessed available water as a limitation for irrigation. In this way the effects on biodiversity loss were instead determined because of shifts in land use as available water limited what could be grown. The LCIAs instead accounted for the impacts of water shortage on terrestrial and marine species loss. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003eGLOBIO additionally looked at the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;encroachment impacts\u0026rsquo; which were linked to the effects of hunting on an area. \u0026lsquo;Distance to cropland\u0026rsquo; was used to estimate the effects of the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;encroachment\u0026rsquo; by determining an impact zone around agricultural/human land use areas under the assumption that hunting (disturbance), will take place at certain rates in this zone\u003c/p\u003e\n\u003cp\u003e3) Pollution was included in the greatest detail in the LCIAs. The IMPACT world + model had the highest number of \u0026lsquo;midpoints\u0026rsquo; related to \u0026lsquo;pollution\u0026rsquo; including seven different types of pollution categories. LC-impact and ReCiPe included the same number and type of \u0026lsquo;pollution\u0026rsquo; \u0026lsquo;midpoints\u0026rsquo; including four different types (i.e. toxicity, fresh water/marine eutrophication, ozone and terrestrial acidification). While the \u0026lsquo;midpoint\u0026rsquo; \u0026lsquo;nitrogen deposition\u0026rsquo; was not directly included in any LCIAs, the pollution \u0026lsquo;midpoints\u0026rsquo; which are included were predominately linked to pesticide and fertilizer production and application, and hence might be considered to indirectly account for the effects of nitrogen. \u0026lsquo;Pollution\u0026rsquo; was the least represented in all the GBM models, with only GLOBIO including it with the midpoint \u0026lsquo;nitrogen deposition\u0026rsquo;. \u0026lsquo;Nitrogen deposition\u0026rsquo; refers to the effect of a critical load of nitrogen on a given area and in the case of GLOBIO, refers specifically to the impacts on different land use types such as forests and grasslands GLOBIO uses the IMAGE model to create scenarios for present and future atmospheric \u0026lsquo;nitrogen deposition\u0026rsquo; based on \u0026lsquo;land use activity\u0026rsquo; e.g. pasture or cropland, with some \u0026lsquo;land use activities\u0026rsquo; having higher impacts.\u003c/p\u003e\n\u003cp\u003e4) Climate change was accounted for by all LCIAs. In LCIAs climate change effects were calculated by determining GHG production along the supply chain (including the effects of land use change) and converting this into an estimate of potential warming and subsequent change in biome distribution of species. Fixed warming scenarios are used in addition by LCIAs to help determine the severity of effects on biodiversity of emissions. The GBM simulates effects of fixed warming scenarios on biodiversity loss. The GBMs, BILBI, inSIGHTS, GLOBIOM and GLOBIO, all use warming scenarios which typically predict impacts on species distributions of warming. GLOBIOM additionally predicts impacts on yields changing the efficiency of certain \u0026lsquo;land use activities\u0026rsquo;.\u003c/p\u003e\n\u003cp\u003e5) IPBES pressures were mostly widely accounted in both LCIAs and GBMs. The only IPBES pressure not accounted for by either LCIAs or GBMs was the effects of invasive species.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eLinking food systems to biodiversity\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eWe compared the items included in the models to the food system components established by the FAO, described below. All models provide system-level impacts, but no models provide impacts disaggregated to the different components of the food system i.e. production, distribution/aggregation, consumption and waste. The LCIAs do not have explicit requirements on what information is needed to calculate the \u0026lsquo;midpoint\u0026rsquo; making it highly dependent on the system boundary which components of the food system are included. LCIAs are therefore limited by the midpoints included e.g. ReCiPe can only calculate the effects of the pollution from toxicity, freshwater and marine pollution, freshwater ecotoxicity, ozone and terrestrial acidification. \u0026nbsp;For GBMs overall, most of the \u0026lsquo;midpoints\u0026rsquo; in the models explicitly accounted for were related to agricultural production, with few impacts measuring other components in the food system. Only one GBM model, GLOBIO explicitly mentions aggregation/distribution. GLOBIO did this by accounting for the presence of roads in the model. \u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eThe GBM GLOBIOM was the only model which included the consumption components. GLOBIOM uses \u0026lsquo;price\u0026rsquo; of food and \u0026lsquo;consumer preference\u0026rsquo; through the effects they have on changes in production. Waste was only dealt with by the GBM inSIGHTS in the form of \u0026lsquo;average production losses\u0026rsquo; and was applied as a co-efficient to yield to symbolize on farm losses.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eConsidering Scenarios in biodiversity modelling\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eAt times, factors are indirectly considered into biodiversity models, as assumption that underline the dynamics or starting points. Scenarios are either embedded into the models or developed with the use of external models and constraints. The use of scenarios allows models to account for greater complexity and indirect biodiversity loss pressures in both GBMs and LCIAs. Indirect pressures driving biodiversity loss according to IPBES include aspects such as technological development, economic, socioeconomic interactions, the role of culture and policy in shifting the quantity and quality consumption, production and other component in the food system. The LCIA ReCiPe accounts for different scenarios in the future through value choices allowing users to choose from individualistic, hierarchism, egalitarian scenarios. These scenarios change the intensity of outcomes based on assumptions on the timeframe, socio-economic developments and the ability to adapt. The other LCIAs also have the option to account for future scenarios but are less detailed in their assumptions, only accounting for long term and short-term impacts. Aside from C-SAR (2018) and PREDICTS, all the GBMs also use scenarios such as Shared Socioeconomic Pathways (SSPs) to estimate the impacts of climate change on biodiversity. GLOBIOM accounts for climate change through scenarios not directly through GHG production of food items but from the influence warming will have on yield reductions and subsequent land use changes that would result to fulfill nutritional needs of the population. Indirect pressures were not accounted for by any of the models except GLOBIOM which linked consumer preference and economic considerations with impacts of biodiversity. Unlike other models GLOBIOM is also an Integrated Assessment Model (IAM), which means it is developed with the intention that other components will be paired with the GBM in scenarios \u0026ndash; something similar happens with GLOBIO, normally used paired with IMAGE.\u0026nbsp;\u003c/p\u003e"},{"header":"Discussion","content":"\u003cp\u003eIn this review we identified ten models assessing biodiversity in global food system, focusing on two main approaches: LCIA and GBMs. We found that while LCIA and GBMs have similar aims, they tackle biodiversity loss from different perspectives (production/consumption focused), with different coverage of biodiversity pressures, indicators for reporting biodiversity outcomes, and part of the food system that was covered. GBMs and LCIAs are powerful tools in biodiversity modelling and have been used to understand the broadest range of effects of a certain action, including how to accomplish important aims of the international community, such as target 10 of the Kunming-Montreal Global Biodiversity Framework\u003csup\u003e27\u003c/sup\u003e (2030 targets). Combining these modelling tools could widen our understanding and help achieve these targets.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eDivide in LCIA/GBMs and their respective blind spots\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eBoth LCIA and GBMs have strengths and limitations in assessing biodiversity following from the aims of the models. LCIA are usually more focused on specific producers (e.g. footprint of beef production) and informing consumers about the impact of specific products. GBMs tend to be more focused on the production system-scale (e.g. field, landscape, country), and are often used to manage tradeoffs between different land uses. These differences are reflected in how biodiversity is reported in the models. In LCIAs, the primary indicator is PDF (potentially disappearing fraction), which is used as a biodiversity 'cost' of different activities. In contrast, GBMs use more landscape to regional to global indicators, such as MSA (mean species abundance) and BII (biodiversity intactness index), to indicate conservation in wider spatial areas. Overall GBMs were able to consider a much broader range of indicators and ‘biodiversity impacts’. LCIAs focus only on species extinction, while GBMs focus on species extinction, ecosystem diversity and quality, compositional intactness, intactness of local species composition, extinction risk and extent of suitable habitat. To calculate these 'biodiversity impacts', however, GBMs incorporate fewer ‘activities in the food system’ and ‘IPBES pressures’ than LCIAs do.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eRegarding ‘pollution’, LCIAs allowed a much wider range of impacts to be accounted for, such as accounting for ecotoxicity of compounds involved in the process and/or acidification, a point supported by previous literature assessing earlier generations of LCIAs and GBMs\u003csup\u003e18\u003c/sup\u003e. The ‘food system activity’ ‘land use activities’ played a large role in determining the effects of biodiversity loss, although these are predominately linked to pesticide and fertilizer use and in some cases irrigation. In turn, none of the ‘land use activities’ account for known biodiversity enhancing farming practices and/or agricultural systems\u003csup\u003e20,30,31,32,33\u003c/sup\u003e. For example, fragmentation has been shown to be mediated by certain land use activities such as agroforestry systems\u003csup\u003e34,35\u003c/sup\u003e. Previous research has suggested improvements could be made to GBMs by increasing the number of activities included, to provide a more holistic assessment of food system impacts on biodiversity\u003csup\u003e20\u003c/sup\u003e.\u0026nbsp; Furthermore, none of the models included were able to account for the biodiversity pressure ‘invasive species’. Invasive species are listed as one of the five IPBES pressures\u003csup\u003e2\u003c/sup\u003e and are a key element of the Kunming-Montreal Global Biodiversity Framework\u003csup\u003e27\u003c/sup\u003e. Some models such as the Climex model are able to directly account for invasive species, but estimates have yet to be formally linked to food systems modelling\u003csup\u003e36\u003c/sup\u003eClick here to enter text..\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn terms of addressing different components of the food system LCIAs are more flexible than GBMs as it is largely up to the user to decide what components are included and is dependent on the system boundaries they set. For example, if an LCIA had the ‘midpoint’ GHG emissions, these could be taken at any component of the food system including those related to production, transport or processing. While the intention of the LCA is to encompass food system activities from extraction to waste disposal or any point along the supply chain, they are still limited by the 'midpoints’ they include and the data that is available. This can have varying disruptions on the assessment, ranging from minor to significant\u003csup\u003e22\u003c/sup\u003e. If we were to take GHG emissions, for example, it is always better to account for all components but indeed the user is flexible and can decide to exclude some components. Conversely, other midpoints are indeed limiting as, for example, different types of land use are not accounted for in the conversion from the midpoint to the impact. Guidelines for LCA (such as ISO or environmental foot printing methods) are geared towards the standardization of LCA however these have been critiqued due to the value choices employed in carrying out the guidelines and lack of transparency they allow\u003csup\u003e37\u003c/sup\u003eClick here to enter text.. LCIAs have also been critiqued for the difficulty in usability due to the relative complexity of the programs\u003csup\u003e38\u003c/sup\u003e.\u003c/p\u003e\n\u003cp\u003eIn contrast, a strength of GBMs in addressing food systems aspects is that they can focus on spatial dynamics of the food system and can assess rebounds in the system, for example, how shifts in consumption of one food item may impact the diet and subsequent ramification for production. Due to the static nature of LCIAs they are not able to do this within the confines of its calculations, making capturing the intrinsically dynamic nature of the food system difficult.\u0026nbsp;Additionally, no ‘activities in the food system’ were found for processing and few were found to populate consumption, waste and aggregation, and distribution component. This means that input activities such as energy use for processing or the land use implications from food waste are not explicitly required for either method of assessing biodiversity loss. LCIAs can include additional activities in the food system but it is highly dependent on the system boundaries and not consistent among studies.\u003c/p\u003e\n\u003cp\u003eIn GBMs the use of additional scenarios can aid in addressing aspects of the food system such as food waste and consumption shifts, however this is not always done, leaving blind spots for the impact of many food system activities. Furthermore, there were limited feedback loops observed in models which could result in underestimation of effects. One good example might come in the form of trophic web breakdown, or not accounting for interspecies dependence – e.g. the loss of certain taxa will have cascading effects on the losses of other taxa dependent or connected to them or the lack of accounting for synergistic/vicious effects across pressures - climate change and water scarcity and pollution\u003csup\u003e39,40\u003c/sup\u003e. While collectively the models form a more complete picture of biodiversity loss in food system including more diverse indicators and aspects of the food system and food system activities such as types of land use, this picture is still incomplete.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eData limitations\u0026nbsp;\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eLack of data plays a large part in the model's incomplete assessment of biodiversity, a known issue in the literature on biodiversity\u003csup\u003e31,32,41,42,43\u003c/sup\u003e. The different models have different data challenges. LCIA data is fragmented and often behind paywalls. Organizations such as HESTIA are working to resolve these problems through the creation of free online databases\u003csup\u003e44\u003c/sup\u003e. While databases such as these are growing, there is still much data needed. Meanwhile, GBMs require intense information on species distribution of the regions assessed and might also require empirical data on the specific coefficient of impacts of each activity in that geography. Including LCIA methods and calculations within GBMs would help address more production factors and better discriminate the impact of specific activities, while at the same time allowing the model to accommodate the changes of impact that might come from the spatial dynamics: e.g., increases in demand of a food item deemed sustainable could lead to unsustainable production of that item, which now would result in an increase of its impact coefficient and the model could readjust to a new equilibrium. Proper assessment of biodiversity and the effects of the multiple components of the food systems will also require close cooperation and interaction with data collection and monitoring around the world to allow for its feasibility.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cem\u003eOutlook\u003c/em\u003e\u003c/p\u003e\n\u003cp\u003eDespite its importance, biodiversity loss remains the Achilles heel of global food systems modelling. Biodiversity assessments in food systems have advanced greatly in past decades, with new understanding of the mechanisms affecting biodiversity loss also shown in multiple studies\u003csup\u003e4,7,16,45,46\u003c/sup\u003e. Current modelling approaches are very useful in that they can help assess and simulate or compound specific indicators or pressures important for international biodiversity conservation aims, such as with target 10, enhancement of biodiversity and sustainability in agriculture, aquaculture, fisheries and forestry, of the 2030 targets established by the Kunming-Montreal Global Biodiversity Framework. Nevertheless, they are still lacking in providing the holistic view and assessment needed for biodiversity. We still observe a disconnect between our field/theoretical knowledge of biodiversity loss and our current modelling tools. Our review adds to existing literature on the limitations of models in how they assess different agriculture/food system characteristics\u003csup\u003e20,22,47\u003c/sup\u003e, by depicting that combining both models would improve current biodiversity assessment and could be possible by integrating LCA calculations within GBMs runs – that would still not consider all changes in dynamics, but would add the spatial and supply changes to LCIA’s broader assessment.\u0026nbsp;Our findings indicate that incorporating biodiversity loss into models currently simplifies a web of factors to single proxy coefficients (e.g. land use activities) and/or are often limited to specific biodiversity impacts, such as species distribution or extinction fractions. Many assumptions and simplifications must be made in order to make up for data gaps and approximate impacts at such a large scale. While it is clear LCIA and GBMs are stronger together, these results beg the question of how much information is enough? And what should be included? The field of LCA and GBMs while addressing similar issues, often do not intersect. In order to better understand how to identify and prioritize missing pieces and in global food systems modelling, future research utilizing the expertise of divers' fields including ones that work with LCA and GBMs is needed. However, even in these cases our overview is limited. What this shows is not that models are not useful, but rather that biodiversity modelling should be used with the awareness of its limitation and used to understand potential trends of biodiversity loss, rather than specific values for biodiversity loss.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eIn a time of unprecedented loss of biodiversity and unregular data accessibility, at times limited exactly where biodiversity hotpots are, models can help us bridge the gaps where the data is non-existent. For this, going forward three steps could help us with biodiversity modelling: models should (1) be used and interpreted with the awareness of the factors currently not considered and (2) combined in their use, methods and interpretation for a more comprehensive interpretation of biodiversity trends. Lastly (3), future modelling research should focus on integrating feedback loops into biodiversity modelling, currently lacking, as well as better specify biodiversity impacts of systems and practices. To properly assess biodiversity in agroecosystems to remain within the established aims of the CBD or IPBES framework, biodiversity modelling needs to more specifically address the species affected and the practices driving the impact, as well as feedback dynamics in the system. This will require alignment across research programs for data collection and policy aims.\u003c/p\u003e"},{"header":"Methods","content":"\u003cp\u003eThe conceptualization for this paper was developed in a series of working group meetings with co-authors and other biodiversity experts brought together by the Wageningen Biodiversity Initiative. To determine how biodiversity loss is assessed within global food systems models we conducted a literature review. In this literature review we refined the search strategy to search in titles, abstracts and keywords for relevant terms including biodiversity, food and or system, production, agriculture or consumption. We chose to focus our search in this way as we wanted to assess commonly used models looking at biodiversity loss in the food system and the results from the papers themselves. Hence, the structured literature review was built to identify and retrieve models. We also chose to limit the number of search engines as we were aiming to capture common approaches and not a comprehensive overview of the topic, linking to our exclusion criteria of models used more than once. From this pool of papers, a systematic approach was carried out in selecting and extracting data.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eResearch questions\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eThe following research questions were developed to guide the review:\u0026nbsp;\u003c/p\u003e\n\u003col\u003e\n \u003cli\u003eWhat ‘food system activities’ are used to estimate biodiversity loss in food systems models?\u003c/li\u003e\n \u003cli\u003eHow are ‘activities in the food system’ distributed across different levels or components of the food system?\u003c/li\u003e\n \u003cli\u003eHow do different ‘activities in the food system’ connect to biodiversity loss pressures in selected models?\u0026nbsp;\u003c/li\u003e\n \u003cli\u003eWhat are the ‘biodiversity impacts’ and indicators used in global models estimating biodiversity loss in food systems?\u003c/li\u003e\n\u003c/ol\u003e\n\u003cp\u003e\u003cstrong\u003eSearch strategy\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eArticles were identified for review in the search database Scopus. The query for our preliminary review was ( TITLE-ABS-KEY ( food OR system* AND biodiversity AND model* AND ( diet OR production OR agriculture OR consumption ) AND ( land* OR region* OR global OR chain ) ) + TITLE ( biodiversity ) + KEY ( assessment OR model ) ) + TITLE ( biodiversity ) + KEY ( assessment OR model ), which resulted in a total of 59 models/papers initially selected (Fig. 1).\u0026nbsp;\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEligibility\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eOnce the preliminary literature review was concluded, we applied two selection criteria: (1) documentation available, either as proper documentation or dedicated scientific article of their application and replication; and (2) global in their assessment. These criteria ensured that the models selected were possible to be used by different groups and in different contexts and data input, as well as aligned in terms of scale. C-SAR 2016 was removed from the final selection due to its similarities with C-SAR 2018.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData extraction and analysis\u0026nbsp;\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eIn this review we have made a distinction between two model classes the LCIAs and the GBMs, however, things rarely fit into one category. For example, we consider C-SAR as an GBM, although there are studies that consider it a model class of its own\u003csup\u003e45,47,48\u003c/sup\u003e, with multiple ways to classify models. Regardless of classifications, C-SAR models are already used in a static way in multiple LCIA to determine land use impact. The fact that this is already done further emphasizes the merits of combining different models in assessing global biodiversity loss. Regardless of if they are in the LCIA or GBMs, models which look at different aspects of the biodiversity loss problem will be complementary towards each other.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eData extraction was done using the documentation of the model papers and papers themselves reporting on the model components of the models which were defined as eligible for inclusion. Information on biodiversity impacts, indicators, food system activities, biodiversity pressures and additional data collected can be found in an excel file in supplementary material A.\u003c/p\u003e\n\u003cp\u003eTo create an overview of how biodiversity loss is determined in food systems we used the following building block.\u0026nbsp;\u003c/p\u003e\n\u003cp\u003eCategorization and description of models using literature as reference categories: IPBES considered the following biodiversity pressures: invasive species, climate change, land use change, natural resource use and exploitation and pollution. Indirect pressures include social and economic aspects of the food system. We have added models to a table depicting what indicators and ‘midpoints’ are used to model each of the biodiversity related impacts of global food systems. To categorize the aspect of biodiversity impacted by the food system, we refer to the ‘biodiversity impacts’ adapted from definitions of Pereira et al.\u003csup\u003e23\u003c/sup\u003e using the following seven biodiversity impacts: global species extinction, reginal species extinction, regional species extinction risk, reduced ecosystem diversity and quality, compositional intactness, intactness of local species composition, and the extent of suitable habitat.\u0026nbsp;\u003c/p\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eAcknowledgements.\u0026nbsp;\u003c/strong\u003eThis project received funding from the AVINA foundation (www.circularfoodsystems.org) and CropMix Project (project number: NWA.1389.20.160) of the Dutch Research Agenda (NWA-ORC) of the Dutch Research Council (NWO). We would like to thank the Wageningen Biodiversity Initiative with regards to funding to gather all experts together. We are also very grateful to the reviewers and editor of NPJ Biodiversity, whose comments and suggestions improved the quality of this review.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eEthics declaration.\u003c/strong\u003e The authors declare no competing interest.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eAuthor contributions.\u0026nbsp;\u003c/strong\u003eFM (Felipe Cozim Melges), WJ (Wendy Jenkins), HZ (Hannah H.E. van Zanten, the idea for the study. FM, WJ and HZ contributed to the structuring of the manuscript. FM and WJ, conducted the literature search. FM, WJ and HZ helped conceptualize and design the figures, FM and WJ created them. All other co-authors commented and reviewed the design of the research and the manuscript. FM and WJ analyzed the data, and HZ helped interpret the data analysis. All authors helped review the results. FM, WJ and HZ wrote the main manuscript text. All authors contributed to, reviewed and approved the final manuscript submitted.\u003c/p\u003e\n\u003cp\u003e\u003cstrong\u003eData availability.\u003c/strong\u003e All additional data is available in the supplementary materials.\u0026nbsp;\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\n \u003cli\u003e\u0026quot;Nature, biodiversity and health: an overview of interconnections. Copenhagen: WHO Regional Office for Europe; 2021. Licence: CC BY-NC-SA 3.0 IGO.\u0026quot;\u003c/li\u003e\n \u003cli\u003eIntergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (IPBES). (2019). Summary for policymakers of the global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services. IPBES secretariat, Bonn.\u003c/li\u003e\n \u003cli\u003eCampbell, B. M., Beare, D. J., Bennett, E. M., Hall-Spencer, J. M., Ingram, J. S. I., Jaramillo, F., Ortiz, R., Ramankutty, N., Sayer, J. A., \u0026amp; Shindell, D. (2017). Agriculture production as a major driver of the Earth system exceeding planetary boundaries. Ecology and Society, 22(4), art8. https://doi.org/10.5751/ES-09595-220408\u003c/li\u003e\n \u003cli\u003eIPBES. (2022, April). Global assessment report on biodiversity and ecosystem services of the Intergovernmental Science-Policy Platform on Biodiversity and Ecosystem Services (Version 1). Zenodo. https://doi.org/10.5281/zenodo.6417333.\u003c/li\u003e\n \u003cli\u003eBrooks, T. M., Mittermeier, R. A., Mittermeier, C. G., Da Fonseca, G. A. B., Rylands, A. B., Konstant, W. R., Flick, P., Pilgrim, J., Oldfield, S., Magin, G., \u0026amp; Hilton‐Taylor, C. (2002). Habitat Loss and Extinction in the Hotspots of Biodiversity. Conservation Biology, 16(4), 909\u0026ndash;923. https://doi.org/10.1046/j.1523-1739.2002.00530.x\u003c/li\u003e\n \u003cli\u003eMeyfroidt, P., Rudel, T. K., \u0026amp; Lambin, E. F. (2010). Forest transitions, trade, and the global displacement of land use. Proceedings of the National Academy of Sciences, 107(49), 20917\u0026ndash;20922. https://doi.org/10.1073/pnas.1014773107\u003c/li\u003e\n \u003cli\u003eMolotoks, A., Henry, R., Stehfest, E., Doelman, J., Havlik, P., Krisztin, T., Alexander, P., Dawson, T. P., \u0026amp; Smith, P. (2020). Comparing the impact of future cropland expansion on global biodiversity and carbon storage across models and scenarios. Philosophical Transactions of the Royal Society B: Biological Sciences, 375(1794), 20190189. https://doi.org/10.1098/rstb.2019.0189\u003c/li\u003e\n \u003cli\u003eNewbold, T., Hudson, L. N., Arnell, A. P., Contu, S., De Palma, A., Ferrier, S., Hill, S. L. L., Hoskins, A. J., Lysenko, I., Phillips, H. R. P., Burton, V. J., Chng, C. W. T., Emerson, S., Gao, D., Pask-Hale, G., Hutton, J., Jung, M., Sanchez-Ortiz, K., Simmons, B. I., \u0026hellip; Purvis, A. (2016). Has land use pushed terrestrial biodiversity beyond the planetary boundary? A global assessment. Science, 353(6296), 288\u0026ndash;291. https://doi.org/10.1126/science.aaf2201\u003c/li\u003e\n \u003cli\u003eAllan, E., Manning, P., Alt, F., Binkenstein, J., Blaser, S., Bl\u0026uuml;thgen, N., B\u0026ouml;hm, S., Grassein, F., H\u0026ouml;lzel, N., Klaus, V. H., Kleinebecker, T., Morris, E. K., Oelmann, Y., Prati, D., Renner, S. C., Rillig, M. C., Schaefer, M., Schloter, M., Schmitt, B., \u0026hellip; Fischer, M. (2015). Land use intensification alters ecosystem multifunctionality via loss of biodiversity and changes to functional composition. Ecology Letters, 18(8), 834\u0026ndash;843. https://doi.org/10.1111/ele.12469\u003c/li\u003e\n \u003cli\u003eBritschgi, A., Spaar, R., \u0026amp; Arlettaz, R. (2006). Impact of grassland farming intensification on the breeding ecology of an indicator insectivorous passerine, the Whinchat Saxicola rubetra: Lessons for overall Alpine meadowland management. Biological Conservation, 130(2), 193\u0026ndash;205. https://doi.org/10.1016/j.biocon.2005.12.013\u003c/li\u003e\n \u003cli\u003eCarrie, R., Andrieu, E., Ouin, A., \u0026amp; Steffan-Dewenter, I. (2017). Interactive effects of landscape-wide intensity of farming practices and landscape complexity on wild bee diversity. Landscape Ecology, 32(8), 1631\u0026ndash;1642. https://doi.org/10.1007/s10980-017-0530-y\u003c/li\u003e\n \u003cli\u003eAgostini, M. G., Roesler, I., Bonetto, C., Ronco, A. E., \u0026amp; Bilenca, D. (2020). Pesticides in the real world: The consequences of GMO-based intensive agriculture on native amphibians. Biological Conservation, 241, 108355. https://doi.org/10.1016/j.biocon.2019.108355\u003c/li\u003e\n \u003cli\u003eFromentin, J.-M. et al. Status, challenges and pathways to the sustainable use of wild species. Global Environmental Change 81, 102692 (2023).\u003c/li\u003e\n \u003cli\u003eThe Convention on Biological Diversity (CBD). (2022a). Final text of Kunming-Montreal Global Biodiversity Framework available in all languages. https://www.cbd.int/conferences/2021-2022/cop-15/documentsasdocument:CBD/COP/15/L25\u003c/li\u003e\n \u003cli\u003eEuropean Parlament, Directorate-General for Internal Policies of the Union, \u0026amp; N\u0026egrave;gre, F. (2020). The EU 2030 biodiversity strategy. European Parlament. https://doi.org/10.2861/545892\u003c/li\u003e\n \u003cli\u003eLecl\u0026egrave;re, D., Obersteiner, M., Barrett, M., Butchart, S. H. M., Chaudhary, A., De Palma, A., DeClerck, F. A. J., Di Marco, M., Doelman, J. C., D\u0026uuml;rauer, M., Freeman, R., Harfoot, M., Hasegawa, T., Hellweg, S., Hilbers, J. P., Hill, S. L. L., Humpen\u0026ouml;der, F., Jennings, N., Krisztin, T., \u0026hellip; Young, L. (2020). Bending the curve of terrestrial biodiversity needs an integrated strategy. Nature, 585(7826), 551\u0026ndash;556. https://doi.org/10.1038/s41586-020-2705-y\u003c/li\u003e\n \u003cli\u003eMommer, L., Nel, J., van Apeldoorn, D., van Hattum, T., Jones-Walters, L., Polman, N., Richter, A., Westerink, J., Alpízar, F., Candel, J., van Doorn, A., van Duinkerken, G., Maas, J., Fraanje, W., Groot, A., Jongeneel, R., Navarro-Matser, M., Ritzer, S., Siegmund-Schultze, M., ... de Wit, M. (2022). Nature-positive futures: Food systems as a catalyser for change. Wageningen University \u0026amp; Research. https://doi.org/10.18174/574286\u003c/li\u003e\n \u003cli\u003eSouza, D. M., Teixeira, R. F. M., \u0026amp; Ostermann, O. P. (2015). Assessing biodiversity loss due to land use with Life Cycle Assessment: Are we there yet? Global Change Biology, 21(1), 32\u0026ndash;47. https://doi.org/10.1111/gcb.12709\u003c/li\u003e\n \u003cli\u003eChopin, P., Bergkvist, G., \u0026amp; Hossard, L. (2019). Modelling biodiversity change in agricultural landscape scenarios\u0026mdash;A review and prospects for future research. Biological Conservation, 235, 1\u0026ndash;17. https://doi.org/10.1016/j.biocon.2019.03.046\u003c/li\u003e\n \u003cli\u003eL\u0026oacute;pez Rodr\u0026iacute;guez, S., Van Bussel, L. G. J., \u0026amp; Alkemade, R. (2024). Classification of agricultural land management systems for global modeling of biodiversity and ecosystem services. Agriculture, Ecosystems \u0026amp; Environment, 360, 108795. https://doi.org/10.1016/j.agee.2023.108795\u003c/li\u003e\n \u003cli\u003eBoakes, E. H., Dalin, C., Etard, A., \u0026amp; Newbold, T. (2024). Impacts of the global food system on terrestrial biodiversity from land use and climate change. Nature Communications, 15(1), 5750. https://doi.org/10.1038/s41467-024-49999-z\u003c/li\u003e\n \u003cli\u003eSany\u0026eacute;-Mengual, E., Biganzoli, F., Valente, A., Pfister, S., \u0026amp; Sala, S. (2023). What are the main environmental impacts and products contributing to the biodiversity footprint of EU consumption? A comparison of life cycle impact assessment methods and models. International Journal of Life Cycle Assessment, 28(9), 1194\u0026ndash;1210. https://doi.org/10.1007/S11367-023-02169-7/FIGURES/5\u003c/li\u003e\n \u003cli\u003ePereira, H. M., Martins, I. S., Rosa, I. M. D., Kim, H., Leadley, P., Popp, A., van Vuuren, D. P., Hurtt, G., Quoss, L., Arneth, A., Baisero, D., Bakkenes, M., Chaplin-Kramer, R., Chini, L., Di Marco, M., Ferrier, S., Fujimori, S., Guerra, C. A., Harfoot, M., \u0026hellip; Alkemade, R. (2024). Global trends and scenarios for terrestrial biodiversity and ecosystem services from 1900 to 2050. Science, 384(6694), 458\u0026ndash;465. https://doi.org/10.1126/science.adn3441\u003c/li\u003e\n \u003cli\u003eFAO. 2018. Sustainable food systems: concept and framework. Brief. Rome\u003c/li\u003e\n \u003cli\u003eISO. (2022). ISO 14044:2006 Environmental management \u0026mdash; Life cycle assessment \u0026mdash; Requirements and guidelines.\u003c/li\u003e\n \u003cli\u003eAlkemade, R., van Bussel, L. G., Rodr\u0026iacute;guez, S. L. \u0026amp; Schipper, A. M. Global biodiversity assessments need to consider mixed multifunctional land-use systems. Current Opinion in Environmental Sustainability 56, 101174 (2022).\u003c/li\u003e\n \u003cli\u003e15/4. Kunming-Montreal Global Biodiversity Framework. (2024). The Convention on Biological Diversity (CBD). https://www.cbd.int/doc/decisions/cop-15/cop-15-dec-04-en.pdf\u003c/li\u003e\n \u003cli\u003eEuropean Parlament, Directorate-General for Internal Policies of the Union, \u0026amp; N\u0026egrave;gre, F. (2020). The EU 2030 biodiversity strategy. European Parlament. https://doi.org/10.2861/545892\u003c/li\u003e\n \u003cli\u003eCBD (Convention on Biological Diversity) (2011) Strategic Plan for Biodiversity 2011-2020, Including Aichi Biodiversity Targets. https://www.cbd.int/sp/\u003c/li\u003e\n \u003cli\u003eBengtsson, J., Ahnstr\u0026ouml;m, J., \u0026amp; Weibull, A.-C. (2005). The effects of organic agriculture on biodiversity and abundance: A meta-analysis: Organic agriculture, biodiversity and abundance. Journal of Applied Ecology, 42(2), 261\u0026ndash;269. https://doi.org/10.1111/j.1365-2664.2005.01005.x\u003c/li\u003e\n \u003cli\u003eCozim-Melges, F., Ripoll-Bosch, R., Veen, G. F., Oggiano, P., Bianchi, F. J. J. A., Van Der Putten, W. H., \u0026amp; Van Zanten, H. H. E. (2024). Farming practices to enhance biodiversity across biomes: A systematic review. Npj Biodiversity, 3(1), 1. https://doi.org/10.1038/s44185-023-00034-2\u003c/li\u003e\n \u003cli\u003eCozim-Melges, F., Ripoll-Bosch, R., Oggiano, P., van Zanten, H. H. E., van der Putten, W. H., \u0026amp; Veen, G. F. (2025). The effect of alternative agricultural practices on soil biodiversity of bacteria, fungi, nematodes and earthworms: A review. Agriculture, Ecosystems and Environment, 379, Article 109329. https://doi.org/10.1016/j.agee.2024.109329\u003c/li\u003e\n \u003cli\u003eUdawatta, R. P., Rankoth, L. M., \u0026amp; Jose, S. (2021). Agroforestry for biodiversity conservation. Agroforestry and Ecosystem Services, 245\u0026ndash;274. https://doi.org/10.1007/978-3-030-80060-4_10/FIGURES/7\u003c/li\u003e\n \u003cli\u003eVagge, I., Sgalippa, N., \u0026amp; Chiaffarelli, G. (2024). The role of agroforestry in solving the agricultural landscapes vulnerabilities in the Po Plain district. Community Ecology, 1\u0026ndash;27. https://doi.org/10.1007/S42974-024-00203-8/TABLES/7\u003c/li\u003e\n \u003cli\u003eKriticos, D. J., Maywald, G. F., Yonow, T., Zurcher, E. J., Herrmann, N. I., \u0026amp; Sutherst, R. W. (2015). CLIMEX Version 4: Exploring the Effects of Climate on Plants, Animals and Diseases. CSIRO, Canberra. 156 pp.\u003c/li\u003e\n \u003cli\u003eRo\u0026szlig;mann, M., Stratmann, M., R\u0026ouml;tzer, N., Sch\u0026auml;fer, P., \u0026amp; Schmidt, M. (2021). Comparability of lcas \u0026mdash; review and discussion of the application purpose. Sustainable Production, Life Cycle Engineering and Management, 213\u0026ndash;225. https://doi.org/10.1007/978-3-030-50519-6_15/FIGURES/1\u003c/li\u003e\n \u003cli\u003eCurran, M., Maia De Souza, D., Anton, A., Teixeira, R., Michelsen, O., Vidal Legaz, B., Sala, S. and Mila I Canals, L., How well does LCA model land use impacts on biodiversity\u0026acirc;\u0026euro;\u0026rdquo;A comparison with approaches from ecology and conservation, ENVIRONMENTAL SCIENCE and TECHNOLOGY, ISSN 0013-936X, 50 (6), 2016, p. 2782\u0026acirc;\u0026euro;\u0026ldquo;2795, JRC110785\u003c/li\u003e\n \u003cli\u003eFatemi Nasrollahi, F.S., Campbell, C. \u0026amp; Albert, R. Predicting cascading extinctions and efficient restoration strategies in plant\u0026ndash;pollinator networks via generalized positive feedback loops. Sci Rep 13, 902 (2023). https://doi.org/10.1038/s41598-023-27525-3\u003c/li\u003e\n \u003cli\u003eLuis Abdala-Roberts, Adriana Puentes, Deborah L Finke, Robert J Marquis, Marta Montserrat, Erik H Poelman, Sergio Rasmann, Arnaud Sentis, Celia C Symons, Nicole M van Dam, Gina Wimp, Christer Bj\u0026ouml;rkman, Kailen A Mooney (2025) Connecting the dots: Managing species interaction networks to mitigate the impacts of global change eLife https://doi.org/ https://doi.org/10.7554/eLife.98899\u003c/li\u003e\n \u003cli\u003eWetzel, F. T., Bingham, H. C., Groom, Q., Haase, P., K\u0026otilde;ljalg, U., Kuhlmann, M., Martin, C. S., Penev, L., Robertson, T., Saarenmaa, H., Schmeller, D. S., Stoll, S., Tonkin, J. D., \u0026amp; H\u0026auml;user, C. L. (2018). Unlocking biodiversity data: Prioritization and filling the gaps in biodiversity observation data in Europe. Biological Conservation, 221, 78\u0026ndash;85. https://doi.org/10.1016/j.biocon.2017.12.024\u003c/li\u003e\n \u003cli\u003eCameron, E. K., Martins, I. S., Lavelle, P., Mathieu, J., Tedersoo, L., Gottschall, F., Guerra, C. A., Hines, J., Patoine, G., Siebert, J., Winter, M., Cesarz, S., Delgado-Baquerizo, M., Ferlian, O., Fierer, N., Kreft, H., Lovejoy, T. E., Montanarella, L., Orgiazzi, A., \u0026hellip; Eisenhauer, N. (2018). Global gaps in soil biodiversity data. Nature Ecology \u0026amp; Evolution, 2(7), 1042\u0026ndash;1043. https://doi.org/10.1038/s41559-018-0573-8\u003c/li\u003e\n \u003cli\u003eBonfanti, J., Langridge, J., Avad\u0026iacute;, A., Casajus, N., Chaudhary, A., Damour, G., Estrada-Carmona, N., Jones, S. K., Makowski, D., Mitchell, M., Seppelt, R., \u0026amp; Beillouin, D. (2024). Global review of meta-analyses reveals key data gaps in agricultural impact studies on biodiversity in croplands. https://doi.org/10.1101/2024.04.19.590051\u003c/li\u003e\n \u003cli\u003eHestia. Available at: https://www.hestia.earth/explorer (Accessed: 23 May 2025).\u003c/li\u003e\n \u003cli\u003eKim, H., Rosa, I. M. D., Alkemade, R., Leadley, P., Hurtt, G., Popp, A., van Vuuren, D. P., Anthoni, P., Arneth, A., Baisero, D., Caton, E., Chaplin-Kramer, R., Chini, L., De Palma, A., Di Fulvio, F., Di Marco, M., Espinoza, F., Ferrier, S., Fujimori, S., \u0026hellip; Pereira, H. M. (2018). A protocol for an intercomparison of biodiversity and ecosystem services models using harmonized land-use and climate scenarios. bioRxiv, 300632. https://doi.org/10.1101/300632\u003c/li\u003e\n \u003cli\u003eAdhikari, S., Adhikari, A., Weaver, D. K., Bekkerman, A., \u0026amp; Menalled, F. D. (2019). Impacts of Agricultural Management Systems on Biodiversity and Ecosystem Services in Highly Simplified Dryland Landscapes. Sustainability, 11(11), 3223. https://doi.org/10.3390/su11113223\u003c/li\u003e\n \u003cli\u003eCrenna, E., Marques, A., la Notte, A., \u0026amp; Sala, S. (2020). Biodiversity Assessment of Value Chains: State of the Art and Emerging Challenges. Environmental Science and Technology, 54(16), 9715\u0026ndash;9728. https://doi.org/10.1021/ACS.EST.9B05153/ASSET/IMAGES/LARGE/ES9B05153_0002.JPEG\u003c/li\u003e\n \u003cli\u003ePereira, H. M., \u0026amp; Daily, G. C. (2006). MODELING BIODIVERSITY DYNAMICS IN COUNTRYSIDE LANDSCAPES. Ecology, 87(8), 1877\u0026ndash;1885. https://doi.org/10.1890/0012-9658(2006)87[1877:MBDICL]2.0.CO;2\u003c/li\u003e\n \u003cli\u003ePEREIRA, H. M., ZIV, G., \u0026amp; MIRANDA, M. (2014). Countryside Species\u0026ndash;Area Relationship as a Valid Alternative to the Matrix-Calibrated Species\u0026ndash;Area Model. Conservation Biology, 28(3), 874\u0026ndash;876. https://doi.org/10.1111/cobi.12289\u003c/li\u003e\n \u003cli\u003eJaureguiberry, P., Titeux, N., Wiemers, M., Bowler, D. E., Coscieme, L., Golden, A. S., Guerra, C. A., Jacob, U., Takahashi, Y., Settele, J., D\u0026iacute;az, S., Moln\u0026aacute;r, Z., \u0026amp; Purvis, A. (2022). The direct drivers of recent global anthropogenic biodiversity loss. Science advances, 8(45), eabm9982. https://doi.org/10.1126/sciadv.abm9982\u003c/li\u003e\n\u003c/ol\u003e"},{"header":"Tables","content":"\u003cp\u003eTables 1 to 4 are available in the Supplementary Files section\u003c/p\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":false,"highlight":"","institution":"","isAcceptedByJournal":false,"isAuthorSuppliedPdf":false,"isDeskRejected":"","isHiddenFromSearch":false,"isInQc":false,"isInWorkflow":false,"isPdf":false,"isPdfUpToDate":true,"isWithdrawnOrRetracted":false,"journal":{"display":true,"email":"[email protected]","identity":"npj-biodiversity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjbiodivers","sideBox":"Learn more about [npj Biodiversity](https://www.nature.com/npjbiodivers/)","snPcode":"44185","submissionUrl":"https://mts-npjbiodivers.nature.com/cgi-bin/main.plex","title":"npj Biodiversity","twitterHandle":"@npjbiodiversity","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"npj","inReviewEnabled":true,"inReviewRevisionsEnabled":true},"keywords":"Agroecosystems, LCIA, Land-use, GBM, Biodiversity, food systems, models ","lastPublishedDoi":"10.21203/rs.3.rs-6735266/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6735266/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eGlobal biodiversity is in crisis, with food systems identified as a major driver of its decline. Various modeling approaches are currently used to provide a critical step guiding policy making. In this paper, we reviewed ten global models assessing food systems, including seven global biodiversity models (GBMs) and three life cycle impact assessment (LCIA) models to map their capacities to capture the diverse effects on biodiversity both within and beyond agricultural fields. LCIA models more comprehensively focused on pollution and climate change, while GBMs captured spatial impacts like fragmentation better with a more diverse range of indicators such as compositional intactness, species richness, extinction, and habitat suitability. Gaps found in both GBMs and LCIAs highlight the need for complementary aspects to be included in current models to more holistically assess biodiversity in food systems.Combining GBMs and LCIAs could address their individual limitations and integrate consumption with shifting spatial dynamics.\u003c/p\u003e","manuscriptTitle":"Blind spots of modelling biodiversity loss in food systems","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-06-04 05:11:12","doi":"10.21203/rs.3.rs-6735266/v1","editorialEvents":[{"type":"communityComments","content":0},{"type":"decision","content":"Revision requested","date":"2025-07-31T06:04:58+00:00","index":"","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-31T00:05:43+00:00","index":"hide","fulltext":""},{"type":"editorInvitedReview","content":"","date":"2025-07-22T17:10:58+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"239874966800094812024440600383172490358","date":"2025-06-04T08:24:10+00:00","index":"hide","fulltext":""},{"type":"reviewerAgreed","content":"174466715975607071449132855451770187013","date":"2025-06-02T10:58:08+00:00","index":"hide","fulltext":""},{"type":"reviewersInvited","content":"","date":"2025-06-02T05:23:22+00:00","index":"","fulltext":""},{"type":"editorAssigned","content":"","date":"2025-06-02T05:14:02+00:00","index":"","fulltext":""},{"type":"checksComplete","content":"","date":"2025-05-30T06:53:10+00:00","index":"","fulltext":""},{"type":"submitted","content":"npj Biodiversity","date":"2025-05-23T19:06:37+00:00","index":"","fulltext":""}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"npj-biodiversity","isNatureJournal":false,"hasQc":false,"allowDirectSubmit":false,"externalIdentity":"npjbiodivers","sideBox":"Learn more about [npj Biodiversity](https://www.nature.com/npjbiodivers/)","snPcode":"44185","submissionUrl":"https://mts-npjbiodivers.nature.com/cgi-bin/main.plex","title":"npj Biodiversity","twitterHandle":"@npjbiodiversity","acdcEnabled":true,"dfaEnabled":true,"editorialSystem":"ejp","reportingPortfolio":"npj","inReviewEnabled":true,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"1c52942b-ece0-47da-8658-9c8f6175874b","owner":[],"postedDate":"June 4th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"in-revision","subjectAreas":[{"id":49382812,"name":"Earth and environmental sciences/Ecology"},{"id":49382813,"name":"Earth and environmental sciences/Environmental sciences"},{"id":49382814,"name":"Earth and environmental sciences/Planetary science"}],"tags":[],"updatedAt":"2026-03-16T08:12:12+00:00","versionOfRecord":[],"versionCreatedAt":"2025-06-04 05:11:12","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6735266","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6735266","identity":"rs-6735266","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}