A dynamic framework for calculating the biomass of fattening pigs with an application in estimating the burden of porcine reproductive and respiratory syndrome in the Netherlands

A dynamic framework for calculating the biomass of fattening pigs with an application in estimating the burden of porcine reproductive and respiratory syndrome in the Netherlands Marloes Boeters a,* , Wilma Steeneveld a , Beatriz Garcia-Morante b,c,d , Jonathan Rushton e,f , Gerdien van Schaik a,g a Department of Population Health Sciences, section Farm Animal Health, Faculty of Veterinary Medicine, Utrecht University, Utrecht, the Netherlands b IRTA, Programa de Sanitat Animal, Centre de Recerca en Sanitat Animal (CReSA), Campus de la Universitat Aut` onoma de Barcelona (UAB), Bellaterra, Catalonia 08193, Spain c WOAH Collaborating Centre for the Research and Control of Emerging and Re-Emerging Swine Diseases in Europe (IRTA-CReSA), Bellaterra 08193, Spain d Unitat Mixta d ′ Investigaci´ o IRTA-UAB en Sanitat Animal, Centre de Recerca en Sanitat Animal (CReSA), Campus de la Universitat Aut` onoma de Barcelona (UAB), Bellaterra 08193, Spain e Global Burden of Animal Diseases (GBADs) Programme, University of Liverpool, Liverpool, UK f Institute of Infection, Veterinary and Ecological Sciences, School of Health and Life Sciences, University of Liverpool, Liverpool, United Kingdom g Royal GD, Deventer, the Netherlands ARTICLE INFO Keywords: Swine Biomass Porcine reproductive and respiratory syndrome Disease burden ABSTRACT Gaining insight into the size and composition of national pig populations can support decisions on disease control, welfare, and environmental sustainability. However, if one needs to draw meaningful comparisons between the performance of various production systems or countries, a method for standardization is required. One approach to achieve this is by means of biomass estimation. The objective of this study was to develop a biomass estimation framework that can provide detailed and reliable estimates of fattening pig biomass disaggregated by pig life stage (suckling, weaning and fattening), while accounting for the dynamic nature of pig populations. The framework was developed on publicly accessible data pertaining to pig production in the Netherlands, and we additionally assessed availability of required data for several other European countries (Spain, Germany, and Great Britain). Three distinct life stages—suckling piglets, weaning pigs, and fattening pigs—are considered in the framework. Demographic and movement data, including yearly imports, exports, and slaughter numbers, along with standing populations, were collected from official governmental sources. Required production parameters were sourced from representative surveys, with missing parameters supplemented by private industry reports or expert elicitation. The results from the framework for the Netherlands yield insights into the Dutch pig sector. In 2020, 156 million kg, 552 million kg, and 1654 million kg of biomass were produced in the suckling, weaning, and fattening stages, respectively. The evaluation against census data indicated the framework’s reliability, with deviations mostly below 10 %. Data availability assessments for Spain, Germany and Great Britain reveal variations in data completeness and underscore the importance of local contacts and language expertise when extending the framework to other countries. The framework’s relevance was further demonstrated through an illustrative application, assessing the impact of porcine reproductive and respiratory syndrome on pig biomass in the Netherlands. In the most severe disease scenario, the produced biomass decreased by 13 %, 17 %, and 66 % in the suckling, weaning, and fattening stages, respectively. Beyond disease burden estimation, the biomass estimates can be used as a denominator for various purposes to provide efficiency metrics, such as the amount of antibiotics used or the volume of greenhouse gases emitted per kilogram of pig biomass produced. While the framework could benefit from further refinement regarding resource use and economic values, its current iteration provides a robust and unique foundation for estimating biomass disaggregated by pig life stage, aiding decision-makers in the agricultural and veterinary sector. * Correspondence to: Yalelaan 7, Utrecht, CL 3584, the Netherlands. E-mail addresses: [email protected] (M. Boeters), [email protected] (W. Steeneveld), [email protected] (B. Garcia-Morante), [email protected] (J. Rushton), [email protected] (G. van Schaik). Contents lists available at ScienceDirect Preventive Veterinary Medicine journal homepage: www.elsevier.com/locate/prevetmed https://doi.org/10.1016/j.prevetmed.2024.106383 Received 23 January 2024; Received in revised form 11 November 2024; Accepted 12 November 2024 Preventive Veterinary Medicine 234 (2025) 106383 Available online 17 November 2024 0167-5877/© 2024 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY license ( http://creativecommons.org/licenses/by/4.0/ ). 1. Introduction Pig farming is an important sector of the agricultural industry, providing a significant source of protein and contributing to the economy of many countries around the world. Gaining insight into the size and composition of national pig populations can support decisions on disease control, welfare, feed and nutrient management, as well as environmental sustainability. However, if one needs to draw meaningful comparisons between the performance of various production systems, pig breeds, age groups or countries, a systematic method is required to standardize the population for differing scales and production characteristics (e.g. target weights). One effective approach involves utilizing biomass estimates, a method that entails converting individual counts into their corresponding weights, measured in kilograms. Unlike conventional measures such as livestock units (LU), which provide a simplified metric that aggregates different livestock types into a standard unit to enable comparisons between livestock species, biomass estimation allows for a nuanced understanding of the structural dynamics within a single sector. Biomass estimates are already commonly used to standardize and compare antimicrobial use on a national scale (Bulut and Ivanek, 2022), often referred to as the population correction unit. Moreover, novel areas where biomass estimates are used, include the standardization of the environmental impact from livestock production (Akamati et al., 2022), the estimation of the economic value of livestock, and the burden of diseases and welfare problems (Huntington et al., 2021; Rushton et al., 2021). Pig biomass estimations are particularly challenging due to the dynamic nature of pig populations and disparities in production practices. Pigs can be classified into different age and weight categories, and they are commonly housed in distinct systems throughout their lives, transitioning through stages such as suckling, weaning, and fattening. These groups of animals are challenged by environmental, disease or welfare issues in different ways (Gebhardt et al., 2020; Sarrazin et al., 2019; USDA, 2015; van Staaveren et al., 2018). Addressing these challenges requires a systematic approach to assess issues both across the entire pig population and within specific age groups. Categorizing pigs into separate age classifications becomes crucial for gaining insights into how environmental, disease, and welfare concerns impact different stages of their development. This granularity enables the development of targeted interventions and management strategies tailored to the specific needs of each age group. Moreover, to enhance the applicability of the method beyond a single country, it is imperative that the methodology accommodates the diverse production practices and varying data availability across different countries. The development of a flexible and robust methodology is essential to ensure its effectiveness in accounting for discrepancies in farming practices and data collection systems across regions. This adaptability is vital for the successful utilization of the method in assessing pig biomass on a continental or even global scale. In recent years, several organizations and programs, including the World Organisation for Animal Health (WOAH) (World Organisation for Animal Health, 2023), the European Surveillance of Veterinary Antimicrobial Consumption (ESVAC) (European Medicines Agency, 2023), the Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) (Government of Canada, 2022), and the U.S. Food and Drug Administration (FDA) (Food and Drug Administration, 2023), have developed and applied biomass estimation methods for pigs. These methods primarily rely on data from livestock censuses or surveys and multiply the total number of pigs with an average pig weight derived from representative samples of the population. However, the methods vary in terms of whether imports and exports are considered, and which weight parameters and data sources are used, which results in varying levels of detail (Bulut and Ivanek, 2022). For instance, in line with their objective to standardize for different countries and species, WOAH and ESVAC rely on international public databases like FAOSTAT and Eurostat. These sources may provide easily accessible data but generally have a lower granularity and a lower accuracy than national public data sources. Although the methods of the FDA and CIPARS in turn use national data sources, all methods are lacking the ability to disaggregate national-level pig biomass by age. While biomass estimation also provides a crucial foundation for understanding pig farming dynamics and productivity, its true potential is realized when applied to specific challenges within the industry. One such challenge is posed by the porcine reproductive and respiratory syndrome virus (PRRSV), a pathogen with profound economic implications for pig production globally (Boeters et al., 2023). The virus induces reproductive impairment or failure in sows, as well as respiratory disease across all age groups of pigs (Zimmerman et al., 2019). While the disease affects both mortality and morbidity in each stage of a pig’s life, its severity is not uniform across all stages (Nathues et al., 2017). Although the disease has been endemic in many countries for decades, several countries have embarked on disease eradication programmes in order to improve national pig performance (Rathkjen and Dall, 2017; Szab´ o et al., 2019). Calculating the potential increase in produced biomass resulting from PRRSV eradication may support these developments and incentivize the implementation of eradication programmes in other countries. Hence, the objective of this study was to develop a generic and accurate biomass estimation framework that can provide detailed and reliable estimates of fattening pig biomass disaggregated by pig life stage (suckling, weaning and fattening) and that reflects the dynamic nature of pig populations. The framework was developed on publicly accessible data pertaining to pig production in the Netherlands, and we reviewed the availability of required data for several other European countries (Spain, Germany, and Great Britain). Additionally, we illustrate how biomass estimation improves our understanding of the burden imposed by the porcine reproductive and respiratory syndrome (PRRS) by evaluating changes in produced pig biomass in the Netherlands due to variation in disease severity. 2. Materials and methods 2.1. Generic biomass framework An overview of the flows considered within the framework, that are ultimately converted to biomass, is presented in Fig. 1. We developed the framework in Microsoft Excel (Microsoft Corporation, 2018) (available in Supplementary file S1). This framework focuses exclusively on the pig population raised for slaughter, as the breeding population was not included due to incomplete and insufficiently available data required for accurate calculations. Within these growing pigs, three life stages (S) were identified as common in Europe: suckling piglets (pre-- weaning), weaning pigs and fattening pigs. Each stage encompasses unique inflows and outflows of individuals, for which detailed calculations are provided in the next sections. The calculated quantities of individuals at each stage and within each flow serve as the foundation for the biomass estimations, as described in Section 2.1.7. The framework provides pig biomass estimations per year, at a national level. We built and evaluated the framework for the Netherlands, for the years 2019–2021. 2.1.1. Data and parameters used Table 1 provides an overview of all data and parameters needed to feed the framework. The framework requires demographic and movement data that includes yearly imports, exports and slaughters, as well as the standing population at the end of the current and the previous year, per life stage. These data are published monthly or bi-annually on official governmental websites and are freely accessible to the public. Required production parameters, such as piglets born alive per sow and mortality rates, are preferably obtained from public representative surveys from the national population. Compared to alternative sources, these surveys provide more accurate, transparent, and unbiased data, M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 2 ensuring that the results reflect the diversity of farm practices and are consistent across regions and time. For the Dutch pig sector, these surveys are published annually and are financially supported by the national ministry of agriculture, nature and food quality. Required parameters missing from these surveys were derived from private industry reports, or, in some cases, from expert elicitation. In general, the industry reports are published by companies in the pig sector, are often published irregularly or incidentally and are often only available in the national language. One especially useful report was published by the Agriculture and Horticulture Development Board (AHDB) in 2021 (Agriculture and Horticulture Development Board, 2021), in which several pig production parameters for seventeen European countries were published. The collaborating organisations are known collectively as InterPIG. 2.1.2. Flows from one stage to the next (FlownextS,t) Within a year, individuals may go through multiple life stages. For example, suckling piglets that are weaned within the year t, will flow to the weaning pig stage, whereas suckling piglets that are born at the end of the year will remain in the suckling stage until the next year. The probability of individuals moving from stage S to the next stage (Pnext,S), is calculated by: Pnext,S=1− (durationS/365) Where durationSdenotes the duration of the life stage S. Consequently, the number of individuals within a life stage S flowing to the next stage within year t(FlownextS,t)is calculated by: FlownextS,t=Pnext,S∗ (censusS,t−1+InflowS,t−OutflowS,t) Where censusS,t−1denotes the census population of stage S at the end of the previous year, InflowS,tdenotes all the inflows in stage S for year t, and OutflowS,tall the outflows from stage S for year t. 2.1.3. Suckling piglets The first life stage comprises the piglets that are still with the sow until they are weaned at a weight of on average 7 kg. At the start of the year t, the individuals in the stage consist of the census population at the end of the previous year, censussuckling,t−1.The only additional inflow in this stage are the new births in year t: Inflowsuckling,t=birthst birthst=piglets born alive per sowt∗no.of sowst The outflows consist of the deaths in the suckling life stage, both in the census population and the newborn suckling piglets. For some countries, the outflows may also include the sales of suckling piglets for slaughter (e.g., Spain). Outflowsuckling,t=deathssuckling,t deathsSuckling,t=deathsCensusSuckling,t+deathsNewborns,t deathsCensusSuckling,t=mortality ratesuckling ∗censussuckling,t−1 deathsNewborns,t= (piglets born alive per sowt−piglets weaned per sowt) ∗no.of sowst Where, censussuckling,t−1is the census population of suckling piglets at the end of the previous year. 2.1.4. Weaning pigs The second life stage contains the pigs from weaning until they reach a maximum bodyweight of 50 kg (on farrow-to finish farms) or until they are transferred to another facility or exported, often around 25 kg. To address this variability, we averaged the transition weight at 30 kg, based on industry reports (Agriculture and Horticulture Development Fig. 1. Overview of the considered flows of individuals and biomass within the framework. Table 1 Required data and parameters used within the framework, their notation as used in formulas, and the values and sources of the Dutch data and parameters. S1 = Supplementary file 1, G =Governmental database, S =Representative survey, R =Reports from the industry. Notation Data/ value Source Demographic and movement data Standing population data censusS,tS1 G (Centraal Bureau voor de Statistiek, n.d) Imports of live animals importsS,tS1 G (Rijksdienst voor Ondernemend Nederland, 2023) Exports of live animals exportsS,tS1 G (Rijksdienst voor Ondernemend Nederland, 2023) Slaughters slaughterS,tS1 G (Rijksdienst voor Ondernemend Nederland, 2023) Production parameters Piglets born alive per sow piglets born alive per sowtS1 S (Agrimatie, 2022) Piglets weaned per sow piglets weaned per sowtS1 S (Agrimatie, 2022) Pre-weaning mortality rate mortality ratesuckling 0.122 R (Blanken et al., 2019) Weaning mortality rate mortality rateweaning 0.025 R (Agriculture and Horticulture Development Board, 2021) Fattener mortality rate mortality ratefattening 0.025 R (Agriculture and Horticulture Development Board, 2021) Duration suckling stage durationsuckling 25 days S (The Netherlands Veterinary Medicines Institute, 2022) Duration weaning stage durationweaning 59 days R (Boehringer Ingelheim, 2018) Average daily gain Average daily gainfattening S1 S (Agrimatie, 2022) Replacement rate sows Replacementt0.5 Expert elicitation M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 3 Board, 2021). At the start of the year t, the individuals in the stage comprise the census population at the end of the previous year, censusweaning,t−1.In this stage, the inflows comprise the inflow from the suckling stage (Flownextsuckling,t), and the imports in year t (importsweaning,t): Inflowweaning,t=Flownextsuckling,t+importsweaning,t As outflows, we distinguish three flows: deaths (deathsweaning,t), exports (exportsweaning,t), and pigs selected to be reared as gilts instead of fattening pigs (giltst): Outflowweaning,t=deathsweaning,t+exportsweaning,t+giltst deathsweaning,t=mortality rateweaning ∗ (censusweaning,t−1+Inflowweaning,t) giltst=Replacementt∗no.of sowst Where, Replacementtdenotes the replacement rate of sows. 2.1.5. Fattening pigs The last life stage contains the fattening pigs that start with a body weight of on average 30 kg, and are kept in this stage until they are sent to slaughter at a weight of on average 125 kg. Again, at the start of the year t, the individuals in the stage comprise the census population at the end of the previous year, censusfattening,t−1. For this stage, the two inflows are the inflow from the weaning stage (Flownextweaning,t)and imports in year t (importsfattening,t): Inflowfattening,t=Flownextweaning,t+importsfattening,t The outflows from this stage are deaths, exports, and slaughters: Outflowfattening,t=deathsfattening,t+exportsfattening,t+slaughterfattening,t deathsfattening,t=mortality ratefattening ∗ (censusfattening,t−1+Inflowfattening,t) Where censusweaning,t−1denotes the census population of fattening pigs at the end of the previous year. The duration of the fattening stage (durationfattening)is dependent on the average daily gain and the kilograms the animal needs to gain from the start of the fattening stage until slaughter: durationfattening =Weight end fattening stage −weight start fattening stage Average daily gainfattening 2.1.6. Validation of results The framework results were validated by comparing the numbers of animals remaining in a certain life stage at the end of year t, with the census population reported for the end of year t, or censusS,t. This number of animals remaining in a life stage S for year t (RemainingS,t)is calculated as follows: RemainingS,t=censusS,t−1+InflowS,t−OutflowS,t−FlownextS,t The difference between the calculated number of individuals and the census population is expressed as a percentage deviation. 2.1.7. Biomass calculation To ultimately calculate the biomass of pigs during and flowing in or out of life stage S during year t, all numbers of pigs present in the standing population and in each stage-related flow are multiplied with their respective (standardized) weights in kilograms. Table 2 shows for all flows the assigned weight per life stage. Subsequently, we analysed biomass disaggregated by pig life stage, by calculating the produced biomass in each stage separately. This produced biomass reflects the biomass gained exclusively during the specified life stage, and, thus, enables the complete disaggregation of life stages. The produced biomass is calculated by adding the biomass that exited the phase (i.e. via exports, deaths, progression to the next life stage) and the biomass that remained in the population at the end of the year (remaining S,t ), then subtracting the biomass that entered the phase (i.e. imports, births, inflows from the previous life stage) and the biomass that was already present at the start of the year (census S,t ). As an example, the produced biomass of the suckling stage is calculated as: Produced biomasssuckling,t=(deathssuckling,t∗1.5+Flownextsuckling,t∗7 +remainingsuckling,t∗4.5)− (birthst∗1.3 +censussuckling,t−1∗4.5) As an illustration of what results are obtained from the framework, we report the outputs for the year 2020. 2.2. Data availability in other countries To explore the potential for applying the framework in other datarich countries, we reviewed the availability of the required demographic and movement data, production parameters, and weights per flow for Spain, Great Britain, and Germany. While we sought to identify relevant datasets, this review focused solely on data accessibility rather than a detailed assessment of the framework’s suitability for differences in production systems across these countries. Similar to the approach taken for the Netherlands, we attempted to retrieve all data on the census population, imports and exports, and slaughters from the websites of the respective ministry of agriculture per country. To obtain nation-specific production parameters and weights, publicly available results of representative national surveys were sought. The respective organisations and websites supplying this described information are given in Table 3. Where feasible, we supplemented missing information with private reports from the industry. Among these reports, we also considered the previously described report by the AHDB (Agriculture and Horticulture Development Board, 2021). Table 2 Average weights per flow per life stage of pigs in the Netherlands, used to convert the counts of animals to kilograms of biomass. Where possible, the used weights were averages for the year 2020 specifically. S =Representative survey, R=Reports from the industry. Flow Average weight (in kg) Source Suckling Weaning Fattening  Census, Remaining 4.5 17.5 70 S (The Netherlands Veterinary Medicines Institute, 2022) Flow Next 7 30 125 R (Agriculture and Horticulture Development Board, 2021), (Boehringer Ingelheim, 2018) Births 1.3 - - R (Boehringer Ingelheim, 2018) Deaths 1.5 17.5 70 S (The Netherlands Veterinary Medicines Institute, 2022)/ Expert elicitation Imports - 25 125 R (Agriculture and Horticulture Development Board, 2021)/ Expert elicitation Exports - 25 125 R (Agriculture and Horticulture Development Board, 2021)/ Expert elicitation Gilts - 30 - Expert elicitation Slaughter - - 125 R (Agriculture and Horticulture Development Board, 2021) M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 4 2.3. The impact of PRRS on pig biomass per life stage As an illustration, we showcase how the framework could be used to gain insight into the impact of PRRS on Dutch pig biomass across different life stages, in the year 2020. We considered the following scenarios: 1. Baseline scenario: As the baseline, we used our developed framework and assumed that the resulting biomass estimations already incorporate a slight impact of PRRS across all Dutch farms. 2. Decrease in disease impact and free-of-disease: These two scenarios involve calculating the potential increase in biomass by (partially) eradicating PRRSV in the Netherlands. 3. Moderate and severe increase in disease impact: In these two scenarios, we calculated the additional losses if the impact of PRRS intensifies in the Netherlands. In all scenarios, we assumed that pigs remain in the fattening stage until they reach their target weight. Additionally, we assumed that the quantities of imports and exports will remain constant, thereby capturing the impact of PRRS on produced biomass per stage through variations in the numbers of animals weaned, placed in fattening, and slaughtered. The parameters from our framework affected by the different disease scenarios include mortality rates per life stage, average piglets born alive per sow, and the duration of life stages (which reflects daily weight gain). The adjustments to these parameters are based on Nathues et al. (2017), who evaluated various PRRSV-infected farm scenarios. We used parameters from scenarios representing disease-free farms, moderately affected farms, and severely affected farms, expressing them as percentage changes relative to the "Repro &Respi – slightly" scenario described by Nathues et al., as we consider this scenario most comparable to our baseline (Table 4). Additionally, we introduced a scenario with a moderate decrease in disease impact, serving as an intermediate scenario between the baseline and the free-of-disease scenario. Ultimately, the framework was rerun with the affected parameters to obtain new estimations of pig biomass flows at each life stage, allowing the quantification of the impact of disease as differences in produced biomass. 3. Results 3.1. Outputs from the baseline framework: biomass calculations for the Netherlands Fig. 2 shows the calculated numbers of animals per inand outflow and throughout each pig life stage, for the year 2020. This figure shows, among others, the importance of export in the Dutch pig industry, as the number of pigs exported is roughly 25 % of total pigs born in a year. The mortality forms another 20 % of all pigs born in a year, which mainly occurred in the suckling stage (80 % of all deaths). The numbers of animals per flow and stage formed the basis for the biomass estimation. The amount of estimated biomass for each inand outflow is provided in Fig. 3 and Table 6. Although most mortality occurred in the suckling stage, Fig. 3a shows that the biomass lost due to mortality in this stage forms a relatively small proportion of all lost biomass due to mortality (23 %). Fig. 3b highlights the substantial exported biomass in both the weaning and fattening stages. Overall, the produced biomass per life stage was quite stable over the years, with only a slight decrease across all life stages in the year 2021 compared to 2020 (Fig. 3c). While most flows show relatively minor year-to-year changes, a notable exception is the clear decrease in exports of fattening pigs observed over time. When we evaluated the calculations of the framework by comparing the numbers of animals remaining in a certain stage at the end of the year with the data on the census population at that time, the deviations remained below 10 % except for the fattening stage in 2021. 3.2. Assessing the availability of required data in other countries Table 5 provides an overview of the availability of full data and parameters needed to use the baseline biomass framework for three other European countries: Spain, Great Britain, and Germany. Of these countries, Great Britain showed the highest availability of data and Germany the lowest. Required data that were available in all cases were the number of slaughters per year. Although population census data is also complete for Spain and Germany, slight differences exist in how life stages are defined (in terms of age and weights) that could complicate their use in the framework. Data on imports and exports of live animals were not published on the governmental websites, for none of the countries. Most of the required parameters were obtained from the published results from representative national surveys (Table 5) or from the report by AHDB (Agriculture and Horticulture Development Board, 2021). Parameters from sow farms (i.e., pre-weaning mortality, piglets born alive per sow, sow replacement rate) are not included in this report by AHDB but could in most cases be recalculated from other data or be found published elsewhere. The parameters that were not retrieved by the authors, were the duration of the weaning stage in Spain and Germany, the replacement rate of sows in Germany, standardized live weights of the census population, and the live weight at the start of fattening in Spain. 3.3. Illustration of utilizing the framework: The impact of PRRS scenarios on pig biomass per life stage Table 6 presents the changes in estimated biomass for affected flows per life stage under four PRRS severity scenarios, compared to the baseline for the year 2020. The most significant impacts on pig biomass stem from changes in mortality rates across all life stages and reductions in the number of live births. The decreased number of live births affects the entire production chain, and when combined with increased mortality and altered stage durations, the most pronounced effects are observed on fattening farms, resulting in a decrease in produced biomass of up to 66 % under the most severe disease scenario. 4. Discussion This study has introduced a novel and comprehensive biomass estimation framework. An essential aspect that sets this biomass estimation framework apart is its ability to offer detailed insights into the dynamics and productivity of pig populations at different life stages. We additionally partially evaluated the generalizability of the framework by Table 3 Governmental databases and representative national samples identified for three European countries (Spain, Great Britain and Germany), supplying data and parameters required for the biomass estimation framework. Spain Great Britain Germany Government database (G) Ministerio de agricultura, pesca y alimentaci´ on www.mapa. gob.es Department for Environment, Food &Rural Affairs www.gov.uk Bundesinformationszentrum Landwirtschaft www.bzl-datenzentrum.de Representative national sample (S) BDporc https://bdp orc.irta.es/ Agriculture and Horticulture Development Board (AHDB) porktools. ahdb.org.uk 1 Not found 1 Although the work of AHDB is not funded by the government, quarterly statistics are published based on representative samples from the population. M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 5 identifying sources of data and production parameters for several European countries. Lastly, our study sheds light on the potential applications of biomass estimation by means of the framework, particularly in assessing the impact of PPRS on the national sector. 4.1. Key insights from estimating pig biomass in the Netherlands The biomass framework was developed on data specific for the Netherlands and the results illustrate the size and variations in produced biomass during the suckling, weaning and fattening stages over the years. These outputs facilitate comparisons of productivity across years within a specific farm system. Notably, the export of weaning pigs constitutes a substantial component of the Dutch pig sector, accounting for roughly 25 % of all pigs born in a year. The other primary outflow of individual pigs consisted of mortality, with approximately 80 % of all mortality occurring in the suckling stage. However, this number constituted less than 25 % of the total biomass lost due to mortality. While smaller in number, deaths occurring in the fattening stage contributed most to the overall biomass lost due to mortality. These deaths can also be considered the most financially burdensome for a farmer, as larger investments are made up until the point of death compared to mortality in the suckling stage. Consequently, while our framework highlights mortality in the fattening stage due to the higher weight of the pigs, this focus could inadvertently overshadow the welfare needs of pigs in the suckling stage, where mortality rates are significantly higher, even though the biomass losses are lower. This duality highlights a conflict between focusing on biomass loss and addressing welfare concerns, and underscores the importance of adopting a balanced approach that prioritizes both economic viability and animal well-being in the management of pig populations. 4.2. Challenges in data collection and utilization The developed framework currently operates independently of the sow population dynamics, which encompasses reproductive performance metrics like the return-to-oestrus and abortion rates. When attempting to integrate this component into the framework, we encountered several challenges due to the lack of accessible data, particularly pertaining to the census records of various age groups of sows and the records of sow slaughters. Although the framework does not require this component, applications such as estimating the burden of PRRS could be more complete if effects on sow performance would also be considered. In general, the framework relies on accurate data collection and can be affected by underreporting or incomplete data. One challenge we encountered in the collection of governmental census data, was the ambiguity regarding the transition threshold between weaning and fattening pigs. Official census data classifies pigs under 50 kg as weaning pigs, but in practice, farmers often consider pigs heavier than 25 kg as fattening pigs, especially when they are moved to separate facilities. This discrepancy can affect the accuracy of the framework’s outputs, particularly when estimating and comparing the number of pigs and produced biomass across the weaning and fattening stages. The framework uses 30 kg as the end weight of the weaning stage, based on European industry norms reported by AHDB (Agriculture and Horticulture Development Board, 2021), to address this ambiguity. However, the absence of a clear, standardized definition in the available data complicates the classification of pigs between these two stages, potentially causing overlap between them. Given the influence of weight on the total produced biomass per stage, this ambiguity could affect the accuracy of the framework’s outputs. Further investigation into how farmers Table 4 Dutch baseline production parameters affected by PRRS and their percentage changes across four scenarios, adapted from Nathues et al. (2017), with an additional moderate decrease in disease impact scenario by the authors. Scenario Affected life stage Affected parameter Baseline value for the Netherlands (2020) Free of disease (%) Moderate decrease (%) Moderate increase (%) Severe increase (%) Suckling Mortality rate 0.122 −8.3 −4.2 +12.5 +25.0 Piglets born alive per sow 35.7 +5.0 +2.5 −5.8 −10.7 Weaning Mortality rate 0.02 −40.0 −20 +100.0 +200.0 Duration stage (days) 59 −6.3 −3.2 +4.2 +14.6 Fattening Mortality rate 0.025 −25.0 −12.5 +50.0 +150.0 Duration stage (days) 109.7 −2.5 −1.3 +4.1 +99.9 Fig. 2. Flows of pigs (in millions of heads) through or out of three pig life stages in the Netherlands (2020). Inflows from imports were excluded from the figure, as numbers of pigs were negligibly small. NB: for simplicity and clarity, standing populations are excluded from the figure, and outflows due to mortality are depicted as occurring halfway during a life stage, although the framework accounts for deaths occurring throughout the entire life stage. M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 6 classify their animals in practice could provide valuable insights and help refine these estimates. Although all demographic and movement data were sourced from governmental databases, and most production parameters were derived from these representative surveys, some parameters had to be retrieved from industry reports. While industry reports can offer valuable insights, they may be less reliable due to several factors. These reports often represent smaller or more selective samples that do not fully capture the breadth of the national population. Additionally, they can be outdated or influenced by commercial interests, which may introduce biases that compromise their accuracy. In contrast, national surveys and governmental data collection are systematically designed to be more representative of the entire farming sector, as they are often mandatory and subject to regulatory oversight. This makes them a more consistent and reliable source for annual, national-level data. Therefore, it is advisable to expand these national surveys to include currently unavailable parameters, particularly average stage durations, sow replacement rates, mortality rates, and average weights of pigs at birth, death, import, export, and slaughter. In order to partially validate the estimates generated by the framework, we compared the calculated animal counts at the end of the year to the census population at the start of the following year. Overall, the deviations were relatively small, mostly remaining below 10 %. All deviations were positive, meaning the framework tended to overestimate slightly, but never underestimated. The deviations could be attributed to the ambiguity in the classification of life stages in census data, used production parameters derived from industry reports not being completely representative for the sector or outdated, underreporting, or missing outflows of individuals. As an example for the latter, a small fraction of the deviation could be explained by an outflow of weaning piglets sold for slaughter. This flow was not incorporated in the framework as no data or parameters are publicly available on this. One notably large deviation occurred in the year 2021 for the fattening pig stage, where the calculated deviation was 25.8 %. In the same year, the data showed a large unexplained decrease in exported fattening pigs (-40 % compared to 2020). While underreporting might be a potential cause, this decline could also be influenced by other factors of which the authors are not aware of, such as market disruptions, changes in trade policies, or shifts in demand due to external circumstances. These variations in export levels are critical to consider as they alter the estimated domestic pig population dynamics and, thus, impact biomass estimations and disease burden assessments. To ensure the model’s robustness, we opted to present outputs for the year 2020, which showed more consistent trends, while acknowledging the need for further investigation into the causes of such fluctuations. 4.3. Availability of required data in other European countries While the framework was initially developed using data from the Netherlands, we also explored the potential for applying the framework in other data-rich countries, by reviewing data accessibility. For the countries Spain, Germany and Great Britain, most demographic and movement data, and production parameters were publicly available, although in several cases they were not in the same format as the Dutch data. An issue that stood out was the lack of data published on governmental websites on the imports and exports of live animals. Additionally, disparities were observed in the publication of production parameters across the countries, both in terms of direct accessibility and Fig. 3. (a-c). Annual biomass dynamics and productivity, showing biomass losses due to mortality for 2020 (Panel a), biomass contributions from births, imports, exports, and gilt rearing for 2019–2021 (Panel b), and total produced biomass across pig life stages for 2019–2021 (Panel c). M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 7 retrievability from governmental databases or representative surveys. Mortality rates, stage durations, and (standardized) live weights proved particularly elusive to obtain. The absence of several key parameters in both representative surveys and industry reports necessitates the reliance on assumptions or expert elicitation to bridge these informational gaps. It is crucial to acknowledge that this approach introduces an element of uncertainty, potentially impacting the precision and reliability of the results. Overall, Germany exhibited the lowest data availability, which can be attributed to the absence of an identified representative survey for the German pig sector, both at the national level and state levels. It remains unclear whether the absence of such a survey in Germany is due to its nonexistence or if the authors, hindered by language constraints, were unable to locate it. The same ambiguity applies to the identification of private reports from the German industry. This highlights the significance of engaging with local contacts to obtain country-specific parameters for implementing a framework of this level of detail. However, this task naturally becomes progressively more challenging when attempting a comparative analysis on a continental or global scale. Lastly, to fully evaluate the replicability of the framework across other countries, such an assessment would also need to encompass a review of the differences in production systems and their implications for the structure of the framework. This would involve a deeper analysis of how varying production practices and systems across countries may affect the framework’s applicability. Given our current scope and resource limitations, we were unable to perform such an in-depth review, but we recognize this as an important area for future research. 4.4. Using the biomass framework to estimate the burden of PRRS As an illustration of further use for the framework, we utilized it to assess the impact of PRRS on pig biomass in the Netherlands. It is important to stress that this is only intended as an illustration of the potential of the framework, and, thus, parameters were adapted directly from the study by Nathues et al. (2017). It is also important to note that in our scenarios, we assumed there are no maximum or minimum capacity constraints on housing and slaughtering animals, and that imports and exports remain constant. We hypothesize that, in practice, most farmers will sell their pigs to Dutch slaughterhouses first. Therefore, any surplus or deficit in produced biomass would primarily be reflected in imports and exports, rather than in the number of animals slaughtered. For simplicity and lack of concrete evidence, we have not considered this mechanism in our calculations. The results from the illustrative scenarios depicting varying PRRS severity demonstrate the framework’s capability to capture the disease’s impact on each life stage individually, as well as the cumulative effects across the entire production chain. For instance, in a free-of-disease scenario, the produced biomass increased by 6 %, 7 %, and 12 % in the suckling, weaning, and fattening stages, respectively. In contrast, under the most severe disease scenario, the produced biomass decreased significantly by 13 %, 17 %, and 66 % across these stages. This highlights how PRRS severely affects the fattening stage, where mortality and reduced growth culminate, showcasing the importance of targeting Table 5 Accessibility of data and parameters essential for the baseline biomass framework: A comparative overview for three European countries. Where, A = Directly available in the same structure as the Dutch data; I =Not directly available, but could be requested from governmental databases, or could be calculated by using other available data/parameters, and; NA =data or parameter is completely unavailable and should be obtained via expert elicitation. Additionally, for all (in-) directly available data/parameters, the source is indicated, where G =Governmental database, S =Representative survey, R = Reports from the industry. Spain Great Britain Germany Demographic and movement data    Standing population data A (G) I (G) A (G) Imports of live animals A (R) I (R) 1 I (R) Exports of live animals I (R) I (R) 1 I (G) Slaughters A (G) A (R) 1 A (G) Production parameters    Piglets born alive per sow I (S) A (S) 1 I (R) Piglets weaned per sow A (R) A (S) 1 A (R) Pre-weaning mortality rate A (S) A (S) 1 I (G) Weaning mortality rate A (R) A (S) 1 A (R) Fattener mortality rate A (R) A (S) 1 A (R) Duration pre-weaning stage A (S) A (R) I (G) Duration weaning stage NA A (S) 1 NA Average daily gain A (R) A (S) 1 A (R) Replacement rate sows A (S) A (S) 1 NA Live weights    Standardized for standing population NA I (R) NA At birth A (G) A (S) 1 I (G) At weaning I (R) A (S) 1 I (G) At start fattening NA A (S) 1 I (G) At slaughter A (R) A (R) A (R) 1 Although the source provides quarterly statistics from representative samples from the population, the funding for these surveys is sourced from farmers and other stakeholders in the supply chain, as opposed to government funding. Table 6 Biomass estimates for all flows per life stage included in the framework for the year 2020, with percentage changes under four different PRRS severity scenarios compared to the baseline. NA =not applicable, meaning these flows are assumed to be unaffected. Percent change in estimated biomass per scenario Life stage Flow / total produced biomass Baseline biomass in 2020 (in million kg) Free of disease (%) Moderate decrease (%) Moderate increase (%) Severe increase (%) Suckling Census population 8.6 NA NA NA NA Births 40.4 +2.5 +5.0 −5.8 −10.7 Deaths 14.8 −3.6 −1.7 +5.7 +11.1 Pigs weaned 181.9 +6.2 +3.1 −7.5 −14.0 Produced biomass 156.3 +6.0 +3.0 −7.1 −13.2 Weaning Census population 50.5 NA NA NA NA Imports 1.2 NA NA NA NA Exports 175.1 NA NA NA NA Gilts 10.9 NA NA NA NA Deaths 12.1 −36.6 −17.8 +86.4 +162.2 Pigs moved to fattening 536.4 +9.9 +4.9 −12.6 −24.1 Produced biomass 551.7 +7.1 +3.5 −8.8 −16.7 Fattening Census population 291.4 NA NA NA NA Imports 1.3 NA NA NA NA Exports 156.9 NA NA NA NA Deaths 38.6 −19.0 −9.0 +34.7 +101.1 Calculated slaughters 1923.2 +9.5 +4.7 −12.3 −44.8 Produced biomass 1653.5 +12.1 +4.3 −23.0 −65.8 M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 8 disease control interventions effectively. This is particularly interesting for diseases like PRRS, where vaccination and other interventions are often applied early in the production chain (Nathues et al., 2018). 4.5. Expanded applications of the biomass framework Results from a disease burden analysis can be further utilized to estimate an economic burden due to disease by considering impacted revenues and costs. In these calculations, one should account for potential price fluctuations resulting from changes in pig supply, as well as market dynamics in response to variations in biomass at the national and European levels, which could influence prices and, by extension, revenues. For a more comprehensive estimation of farm-level financial impacts due to disease, we recommend extending the framework to include shifts in inputs and outputs across each pig life stage (Huntington et al., 2021), while also reconsidering the farmer’s selling strategy. In our analysis, we were constrained by the input parameters and assumed that fattening pigs remain on the farm until they reach their target weight. However, in practice, farmers may choose to sell pigs at a lower weight to respect their all-in-all-out management. Incorporating these factors would offer a clearer understanding of resource use and production efficiency under various disease scenarios. In addition to aiding in disease burden assessment, our method could also enhance existing standardization methods for antimicrobial usage and greenhouse gas emissions (Akamati et al., 2022; Bulut and Ivanek, 2022). It is, then, important to assess whether using produced biomass as a denominator offers the desired interpretation. When applied to antibiotic usage and greenhouse gas emissions, produced biomass as a denominator can provide efficiency metrics, such as the amount of antibiotics used or the volume of greenhouse gases emitted per kilogram of pig biomass produced. The advantage of our approach, compared to other existing biomass estimation methods, is that it includes only the biomass that has directly contributed to antibiotic use or greenhouse gas emissions within a specific life stage, offering a more accurate reflection and comparison of resource use. As a final remark, we feel it is important to stress that for purposes related to animal welfare, such as estimating the welfare burden per life stage, it is preferable to utilize the population framework underlying the biomass calculations, referring to individual animals rather than their mass or production efficiency. 5. Concluding remarks In conclusion, this study presents an innovative biomass estimation framework - the first to differentiate among three distinct pig life stages on a national level - by detailing the specific inflows and outflows of individuals and calculating the annual biomass produced during each stage. The framework was initially applied to the Dutch pig sector, and the availability of required data was explored for three pork-producing European countries. As an additional illustration, we calculated changes in produced biomass under various PRRS scenarios. This illustration underscores the framework’s potential in disease burden estimates, offering crucial insights for safeguarding both sector welfare and financial stability. While the framework could benefit from further refinement regarding resource use and economic values, its current iteration provides a robust and unique foundation for estimating biomass disaggregated by pig life stage. This foundation can aid decision-makers in the agricultural and veterinary sector, ensuring the continued sustainability and prosperity of the industry. CRediT authorship contribution statement Marloes Boeters: Writing –original draft, Visualization, Methodology, Formal analysis, Data curation. Wilma Steeneveld: Writing – review &editing, Supervision, Conceptualization. Beatriz GarciaMorante: Writing –review &editing, Data curation. Jonathan Rushton: Writing –review &editing, Supervision, Conceptualization. Gerdien van Schaik: Writing –review &editing, Supervision, Conceptualization. Declaration of Generative AI and AI-assisted technologies in the writing process During the preparation of this work the authors used ChatGPT 3.4 in order to improve readability and language. After using this tool, the authors reviewed and edited the content as needed and take full responsibility for the content of the publication. Declaration of Competing Interest Gerdien van Schaik is partly employed at Royal GD; the other authors declare no conflict of interest. Acknowledgements This work has received funding from the European Union’s Horizon 2020 research and innovation programme under grant agreement No. 101000494 (DECIDE). We express our gratitude to Dr William Gilbert, Dr Jo˜ ao Afonso and Dr Beat Thomann, for their thoughtful discussions and valuable suggestions, which influenced the conceptualization of the framework. Additionally, we extend our thanks to Karlijn Eenink (DVM) for her expertise in the Dutch pig sector, which was valuable in shaping the methodology. Supplementary files S1. The biomass framework, filled in for the Netherlands. Appendix A. Supporting information Supplementary data associated with this article can be found in the online version at doi:10.1016/j.prevetmed.2024.106383. References Agriculture and Horticulture Development Board. (2021). 2021 pig cost of production in selected countries. AHDB. 〈https://projectblue.blob.core.windows.net/media/Defa ult/Pork/Pork%20MI%20files/CostPigProduction2021_221117_WEB.pdf〉. Agrimatie. (2022). Technisch resultaat, prijzen en saldo (bedragen excl. BTW) van de zeugenhouderij - Varkensbedrijven. 〈https://agrimatie.nl/binternet.aspx?ID=16&bedr ijfstype=5〉. Akamati, K., Laliotis, G.P., Bizelis, I., 2022. Comparative Assessment of Greenhouse Gas Emissions in Pig Farming Using Tier Inventories. Environments 9 (5), 59. Blanken, K., de Buisonje, F., Evers, A., Ouweltjes, W., Verkaik, J., Vermeij, I., & Wemmenhove, H. (2019). KWIN 2019-2020: Kwantitatieve Informatie Veehouderij. Boehringer Ingelheim. (2018). Zakboekje Varkens. Boehringer Ingelheim Netherlands b.v. 〈https://magazines.digital-minds.nl/BI_zakboekjevarkens/#!/inhoudsopgave〉. Boeters, M., Garcia-Morante, B., van Schaik, G., Segal´ es, J., Rushton, J., Steeneveld, W., 2023. The economic impact of endemic respiratory disease in pigs and related interventions-a systematic review. Porc. Health Manag. 9 (1), 45. Bulut, E., Ivanek, R., 2022. Comparison of different biomass methodologies to adjust sales data on veterinary antimicrobials in the USA. J. Antimicrob. Chemother. 77 (3), 827–842. Centraal Bureau voor de Statistiek. (n.d.). StatLine 〈https://opendata.cbs.nl/statline/ #/CBS/nl/〉. European Medicines Agency. (2023). European Surveillance of Veterinary Antimicrobial Consumption (ESVAC) 〈https://www.ema.europa.eu/en/veterinary-regulatory/ove rview/antimicrobial-resistance/european-surveillance-veterinary-antimicrobial-con sumption-esvac#standardised-units-of-measurement-section〉. Food and Drug Administration. (2023). Biomass-Adjusted Antimicrobial Sales and Distribution Data in Food-Producing Animals: Interactive Summary.〈https://www.fda. gov/animal-veterinary/antimicrobial-resistance/biomass-adjusted-antimicrobial-sa les-and-distribution-data-food-producing-animals-interactive〉. Gebhardt, J.T., Tokach, M.D., Dritz, S.S., DeRouchey, J.M., Woodworth, J.C., Goodband, R.D., Henry, S.C., 2020. Postweaning mortality in commercial swine production. I: review of non-infectious contributing factors. Transl. Anim. Sci. 4 (2), 462–484. Government of Canada. (2022). Canadian Integrated Program for Antimicrobial Resistance Surveillance (CIPARS) 2019: Pigs. Huntington, B., Bernardo, T.M., Bondad-Reantaso, M., Bruce, M., Devleesschauwer, B., Gilbert, W., Grace, D., Havelaar, A., Herrero, M., Marsh, T.L., 2021. Global Burden of M. Boeters et al. Preventive Veterinary Medicine 234 (2025) 106383 9