Investigating the Carbon Footprint of Italian Specialized Livestock Farms and its Drivers

Investigating the Carbon Footprint of Italian Specialized Livestock Farms and its Drivers | 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 Research Article Investigating the Carbon Footprint of Italian Specialized Livestock Farms and its Drivers Vincenzo Giaccio, Diana Salottolo, Luca Romagnoli, Maria Bonaventura Forleo, and 1 more This is a preprint; it has not been peer reviewed by a journal. https://doi.org/ 10.21203/rs.3.rs-6959911/v1 This work is licensed under a CC BY 4.0 License Status: Posted Version 1 posted You are reading this latest preprint version Abstract The livestock sector is a significant contributor to climate change, as it is a direct source of greenhouse gases and an indirect source of gas and carbon emissions through the processes of land use and feed production. The objective of the present study is to contribute to the ongoing debate concerning the economic and environmental performance of the specialised livestock sector. The study analyses the relationship between economic variables associated with the management of livestock farms and the value of the carbon footprint. The latter is a variable representing environmental sustainability at the level of individual farms. The analysis is based on data extracted from the Italian section of the EU Farm Accountancy Data Network (FADN) database and through the implementation of a multilevel mixed-effects regression model on a balanced panel dataset. The correlation between utilisation of agricultural inputs and emissions demonstrated a positive elasticity, which proved to be statistically significant. The analysis by farming types indicates that the most significant environmental impact is attributed to beef cattle farming, while poultry farming is found to be the least contributing factor to the carbon footprint of production units. Interactions with the age of farmers reveal that, all other things being equal, farms run by young people do not have significantly different average emissions compared to traditional farms. However, interactions with the main production factors are highly significant, suggesting differential carbon footprint performances according to the type of inputs. livestock sector carbon footprint Linear Mixed-Effects Regression Model type of farming FADN Figures Figure 1 Figure 2 Figure 3 Figure 4 Figure 5 Figure 6 Figure 7 Figure 8 1. Introduction and literature background Greenhouse gases (GHGs) are natural elements in the atmosphere and are produced by natural phenomena, such as the respiration of plants, animals and microbes, and the natural emissions of volcanoes and wildfires. However, the disproportionate accumulation of greenhouse gases and consequent global warming are a result of human activity, which must be monitored (Allen et al., 2018 ). The agri-food system impacts climate change through various mechanisms, including the emission of greenhouse gases (GHGs). Indeed, agricultural production is a significant contributor to anthropogenic global warming, and reducing agricultural emissions — largely methane and nitrous oxide — could play a significant role in mitigating climate change (Alem, 2033; Lynch et al., 2021 ). Livestock farming directly emits climate-altering gases, including methane, nitrous oxide and carbon dioxide, and indirectly emits gases and carbon through land use and feed production (Contini et al., 2020 ). The livestock sector in particular is a driver of climate change, generating around 15% of global GHG emissions from anthropogenic sources (Pais et al., 2022 ; Cheng et al., 2022 ; FAO, 2024 ). In 2021, Italy was the fourth EU country most responsible for the greenhouse effect, after Germany, Poland and France (Eurostat, 2024 ). According to the Italian Institute for Environmental Protection and Research (ISPRA), in 2022, 7.4% of GHG emissions in Italy came from the agricultural sector. Therefore, agriculture is the third largest source of emissions after the energy sector (55.0%) and transport (24.7%). Livestock farming has been identified as a significant contributor to air pollution, with particular concern regarding methane (CH 4 ), nitrogen oxides (NO 2 ), and carbon dioxide (CO 2 ) emissions (Castillo-Díaz et al., 2024 ; Fouda et al., 2021 ). However, the impact on the atmosphere varies according to the type of greenhouse gas. For instance, methane, the primary GHG emitted by livestock farms, and carbon dioxide, the most significant, have different effects. Methane has a limited effect over time, and thus, it is less responsible for the greenhouse effect compared to carbon dioxide, which persists in the atmosphere for a longer duration (Huntingford et al., 2015 ). As is well known, carbon dioxide is produced when fossil fuels are burned to generate energy (Schröder et al., 2012). The literature agrees that methane is the main cause of anthropogenic greenhouse gas emissions on livestock farms (Velarde-Guillén et al., 2022 ; Mazzetto et al., 2020 ; Rice et al., 2017 ). Regarding the country analysed in the study, it should be noted that Italian livestock farms are responsible for 79% of overall emissions in the national agricultural sector (ISPRA, 2022 ). Enteric fermentation is the main cause of methane emissions in the agricultural sector, representing 47% of emissions into the atmosphere. It is mainly generated by reactions in the digestive systems of livestock, particularly cattle. Animal waste and its management, as well as spreading livestock manure on agricultural land and grazing, are instead at the basis of nitrous oxide emissions (de Vries et al., 2015 ). Methane, the gas with the greatest impact on the carbon footprint of Italian livestock farms from enteric fermentation, is the primary contributor, while emissions of other greenhouse gases are comparatively lower. Furthermore, 21% of emissions derive from manure management, and 10% from the management of agricultural soils that could be traced back to livestock farms themselves (ISPRA, 2022 ). The impact of cattle (68.1%) and pigs (12.7%) represents approximately 80% of the total emissions from Italian livestock farms in 2022 (ISPRA, 2022 ). This observation is particularly pertinent when considering the type of livestock in question. Feed production has been identified as the primary contributor to greenhouse gas (GHG) emissions, particularly within the context of the poultry supply chain, where the production of ingredients, such as corn and soy, plays a significant role (Thoma et al., 2015 ). In the context of poultry farming, the predominant organisational model is characterised by an intensive livestock, with a clear demarcation between production categories for meat and eggs. This is further compounded by the high human consumption of products derived from this sector (Dogbe et al., 2024 ). These products are comparatively limited in size and are able to guarantee the levels of proteins necessary for optimal nutrition (Herrero et al., 2009 ; Mottet and Tempio, 2017 ). The carbon footprint (CF) is the most widely used indicator, usually expressed in terms of carbon dioxide equivalent (CO₂eq), to quantify the impact of livestock farming on global warming (Mancilla-Leytón et al., 2023 ). A comprehensive review of the extant literature reveals a wide range of variables for estimating the carbon footprint, including the production costs of on-farm or off-farm inputs and external services. As Andrade et al. ( 2024 ) and Bonnin et al. ( 2021 ) note, the production costs of raw materials (e.g. animal feeds, veterinary products, fuel, electricity, water and services such as veterinary consultants) are some of the variables considered by the literature to measure the carbon footprint of livestock farms. The financial implications of mechanisation have been identified as a key variable in extant literature. Mechanisation in livestock farming has been demonstrated to enhance the efficiency of waste disposal from dairy farming and poultry, thereby effectively reducing the discharge of pollutants. This is a significant variable, given its application in the provision of sustenance and hydration for animals, as well as the establishment of climate-controlled stabling conditions (Singh and Pagare, 2024 ). Conversely, the process of mechanization gives rise to the emission of carbon dioxide into the atmosphere. A number of studies have indicated that farms managed by young people demonstrate higher levels of sustainability from an environmental perspective (Sau et al., 2023 ). The objective of the study is to analyse the relationship between the most significant variables related to livestock management and the carbon footprint, as calculated by CREA (Coderoni et al., 2013) and intended as a variable representing the environmental sustainability of specialised livestock farming over a two-year period. The input variables are: feed and fodder (FF), water, energy and fuel (WEF), veterinary care (VET), mechanization (MEC). All of these are measured in terms of cost. The goal is to shed some light on the effect that economic variables have on CO₂ equivalent emissions at level of farm unit. The methodology employed is a multilevel mixed-effects regression model (Linear Mixed-Effects Regression Model) on a balanced panel dataset (Gelman and Hill, 2006 ). This model is suitable for managing hierarchical data structures and for taking into account unobserved variability between farm groups by type of farming and over time (the last two years with available data, 2021–2022), including an interaction analysis with the variable "Young entrepreneur" (< 40 years). The panel data structure enables the control of both observable and unobservable heterogeneity across units and over time, in accordance with the panel econometrics framework (Sul, 2019 ). The rationale behind the analysis at the farm level is substantiated by existing literature on the environmental impact of the livestock sector, which attributes the majority of impacts to this phase of production (Kang et al., 2025 ; Notarnicola et al., 2017 ; Treml et al., 2025 ). 2. Dataset and methods The analysis was applied to a balanced panel of 1861 livestock farms, which were observed for two consecutive years (2021–2022). The data were extracted from the RICA database, Italian section of the EU Farm Accountancy Data Network (FADN) database collected by CREA. The FADN is a sample survey carried out in all the Member States of the European Union. It is the only harmonised European source on farm management. The information framework of the Italian FADN allowas analysis at the farm, sectoral and territorial levels. The FADN dataset is a comprehensive set of data that classifies specialised livestock holdings into five primary Type of Farming (ToF) categories and eleven orientations. This classification system enables both aggregate and sectoral analysis of the emission footprint in the livestock sector. With regard to the carbon footprint, the Italian FADN dataset collects information on the quantities of production factors used in the processes within the farm (natural processes, production methods, resource management), which are useful for estimating agricultural greenhouse gas emissions. Consequently, it is feasible to concentrate on the individual farm as a unit of analysis by readily collecting activity data and integrating the three distinct sectors (Agriculture, Land Use and Energy). This approach enables the analysis of the emission impacts attributable to the individual farm in a unified manner. In addition, the RICA database facilitates the calculation of emission intensity indicators with reference to economic and structural company parameters. Consequently, assessments are formulated in terms of the efficiency of mitigation practices. As of 2025, the Farm Sustainability Data Network (FSDN) is set to replace the Farm Accountancy Data Network (FADN). The FSDN will build on the FADN's legacy, expanding its scope to include not only farms' income and business activities, but also information on their environmental and social sustainability performance. With regard to environmental considerations, the FSDN will collate data pertaining to agricultural management practices and the utilisation and management of nutrients (manure management, ration), in addition to more detailed information on the use of plant protection products and antibiotics. The collection of data will also encompass aspects pertaining to the management of water resources, the consumption and production of energy, and the implementation of environmental certification schemes. A multilevel model was adopted in order to account for the hierarchical structure of the data and to capture intra-sector and time variability through the inclusion of random intercepts. In order to enhance model convergence, ensure robustness, and reduce multicollinearity, particularly in instances where interaction terms are present (Gelman & Hill, 2006 ), it was necessary to transform all continuous variables using a log-transformation and mean-centring process. The log-linear structure of the estimated model incorporates fixed effects for each cost item, in addition to random effects on the intercept per group and year. Furthermore, a dichotomous variable was incorporated to identify farms managed by young people (i.e. the entrepreneur's age being less than 40 years), in conjunction with interactions with individual input components, to ascertain potential managerial disparities associated with youth entrepreneurship. The structure of the dataset, which incorporates repeated farm-level observations, is consistent with recent developments in panel data econometrics. These developments emphasise the importance of cross-sectional dependence and latent heterogeneity (Sul, 2019 ). The theoretical underpinnings of these models substantiate the utilisation of multilevel frameworks, which possess the capacity to discern unobserved sectoral and temporal variations. The multilevel model employed in this study assumes the following specification: $$\:{y}_{ijt}=\:{\varvec{x}{\prime\:}}_{ijt}\varvec{\beta\:}+\delta\:{d}_{ijt}+({{\varvec{x}}^{{\prime\:}}}_{ijt}\times\:{d}_{ijt})\varvec{\gamma\:}+\:{u}_{j}^{ToF}+{v}_{t}^{Year}+{\epsilon\:}_{ijt}$$ 1 where: \(\:{y}_{ijt}\) the log-transformed and mean-centred carbon footprint emissions ( \(\:{CO}_{2}eq)\) of farm i , belonging to farming type j , in year t ; \(\:{\varvec{x}\varvec{{\prime\:}}}_{\varvec{i}\varvec{j}\varvec{t}}\) the K -dimensional log-transformed and mean-centred input costs of farm i , belonging to farming type j , in year t ; \(\:\varvec{\beta\:}\) the corresponding K -dimensional vector of fixed effects parameters; \(\:{d}_{ijt}\) dummy variable indicating the presence of a young farmer ( farmer < 40 years old ), and \(\:\delta\:\) the associated parameter; \(\:{\varvec{x}\varvec{{\prime\:}}}_{ijt}\:\times\:\:{d}_{ijt}\) the interactions between input costs and the young farmer dummy; $$\:\varvec{\gamma\:}:\:the\:associated\:K-dimensional\:vector;$$ \(\:{u}_{j}^{ToF}\sim\:N\left(0,{\sigma\:}_{u}^{2}\right)\:\) random intercept capturing unobserved heterogeneity across types of farming. \(\:{v}_{t}^{Year}\sim\:N\left(0,{\sigma\:}_{v}^{2}\right)\) random intercept capturing unobserved heterogeneity over time. \(\:{\epsilon\:}_{ijt}\sim\:N(0,{\sigma\:}_{\epsilon\:}^{2})\) residual error term. In accordance with Wooldridge ( 2010 ), when employing log-log models, estimated coefficients can be interpreted as elasticities. That is to say, the percentage change in the dependent variable resulting from a 1% change in the regressor. While this interpretation is strictly valid under the assumption of independence between the error term and the covariates, it remains a commonly accepted approximation when only mean-independence is assumed ( \(\:\text{{\rm\:E}}\left[\epsilon\:|x\right]=0\) ). The model can also be written as: $$\:\varvec{y}\:=\varvec{M}\theta\:+\varvec{U}\varvec{u}+\varvec{V}\varvec{v}+\epsilon\:$$ 2 where: \(\:\varvec{y}\) the \(\:(NT,1)\) vector dependent variable of log-transformed mean-centred dependent variable ( \(\:{CO}_{2}eq\) emissions), with \(\:N=1861\) farms observed over \(\:T=2\:\) years (2021 and 2022), for a total of \(\:NT=3722\) observations; \(\:\varvec{M}=\left[\varvec{X}\:\varvec{d}\:\varvec{Z}\right]\) is the \(\:(NT,2K+1)\) matrix combining the fixed effects; \(\:\varvec{\theta\:}=\left[{\varvec{\beta\:}}^{{\prime\:}}\:\delta\:\:{\varvec{\gamma\:}}^{\varvec{{\prime\:}}}\right]\varvec{{\prime\:}}\) is the \(\:(2K+1,\:1)\) vector of fixed-effect coefficients combining \(\:\varvec{\beta\:}\) , \(\:\delta\:\) , and \(\:\varvec{\gamma\:}\) . \(\:\varvec{X}\) the \(\:(NT,K+1)\) matrix of fixed effects covariates with \(\:K=4\) variables (cost items); \(\:\varvec{d}\) : \(\:NT\) -dimensional (dummy) vector for young farmers; \(\:\varvec{Z}\:\) the ( \(\:NT,K\) ) matrix of interactions between inputs and the young farmer’s dummy; \(\:\varvec{U}\:\left(NT,G\right)\) and \(\:\varvec{V}\:(NT,T)\) : incidence matrices, with \(\:G=11\:\) and \(\:T=2\) for the random effects related to the second-level aggregation in type of farming (ToF) and in time period (year), respectively; with stochastic assumptions: \(\:\varvec{u}\mathcal{\:}\sim\:\mathcal{\:}\mathcal{N}\left(0,\:\:{\sigma\:}_{u}^{2}\:{\varvec{I}}_{G}\right)\) random intercept capturing unobserved heterogeneity across types of farming \(\:\varvec{v}\mathcal{\:}\sim\:\mathcal{\:}\mathcal{N}\left(0,\:{\sigma\:}_{v}^{2}\:{\varvec{I}}_{T}\right)\) random intercept capturing unobserved heterogeneity over time \(\:\varvec{\epsilon\:}\mathcal{\:}\sim\:\mathcal{\:}\mathcal{N}\left(0,\:{\sigma\:}^{2}{\varvec{I}}_{NT}\right)\) residual error term. The aforementioned assumptions imply that all random effects and residuals are independently and identically distributed with constant variance and no cross-correlations. 3. Results and discussion This section reports the main findings of the analyses. The results are presented according to the classification of livestock farming activities by type of farming (ToF) which is structured in two levels: the main level (MToF, comprising five types) and the particular type of farming (PToF, comprising 11 types). 3.1 Results of the model at level of Main Type of Farming The application of the multilevel model to the classification of farms according to MToF provides a robust estimate of the relationship between external inputs and emission values. It is evident that all the coefficients of the economic variables are positive and highly significant \(\:(p\:<\:0.001)\) . The results (Table 1 ) demonstrate that the elasticity of emissions in relation to the inputs is both positive and statistically significant. Specifically, an increase of 1% in feed and fodder costs results in an average increase of 0.134% in emissions \(\:({\widehat{\beta\:}}_{1}=0.134,p\:<\:0.001)\) , while costs of water, electricity and fuel are associated with an average increase of 0.121% \(\:({\widehat{\beta\:}}_{3}=\:0.121,\:p\:<\:0.001)\) . Mechanization expenditure shows the highest elasticity among all input items analysed: an increase of 1% is related to a 0.159% increase in CO₂eq. emissions \(\:({\widehat{\beta\:}}_{2}=\:0.159,\:p\:<\:0.001)\) . Veterinary expenses have been found to have a positive effect, yet this effect is less pronounced in comparison to that of other input components. The estimated elasticity of veterinary expenses is 0.030% \(\:({\widehat{\beta\:}}_{4}=\:0.034,\:p\:<\:0.001)\) . Table 1 Results for MToF model estimation Random effects Groups Name Variance Std. Dev Main ToF (Intercept) 0.447 0.668 Year (Intercept) 0.001 0.024 Residuals 0.682 0.826 Fixed Effects Variable Coeff. Std. Error p-value FF 0.134 0.004 < 0.001 MEC 0.159 0.009 < 0.001 WEF 0.121 0.009 < 0.001 VET 0.034 0.005 < 0.001 YOUNG 0.074 0.039 0.060 FF × YOUNG -0.039 0.012 < 0.001 MEC × YOUNG -0.001 0.020 0.978 WEF × YOUNG -0.075 0.020 < 0.001 VET × YOUNG -0.030 0.013 < 0.010 In order to capture the moderating role of youth entrepreneurship in the relationship between farm input management and environmental impact, a dummy variable was introduced. This variable, "Young," denoted enterprises managed by farmers under 40 years of age according to the EU definition (Article 50 of Regulation (EU) 1307/2013). The model specification incorporates interaction terms between this dummy and each of the four cost categories in order to estimate potential age-related differences in the emissions response. The findings of the model (see Table 1 ) demonstrate that the direct impact of the 'Young' variable on the intercept is marginally significant, indicating a tendency for higher average levels of emissions in farms managed by young individuals, when input costs are taken into account \(\:(\widehat{\delta\:}=0.074)\) . However, it is in the interaction terms that significant managerial differences emerge. All interactions, with the exception of mechanization \(\:(p=0.978)\) , demonstrate that the presence of a young entrepreneur serves to moderate the relationship between costs and emissions, thereby reducing the emission elasticity. A significant discrepancy becomes apparent in the interaction with the WEF input \(\:{(\widehat{\gamma\:}}_{3}=\:\--0.075,\:p<\:0.001)\) . In farms managed by young entrepreneurs, a 1% increase in these costs leads to a 0.046% rise in CO 2 eq. emissions, compared to a 0.121% increase observed in other farms. This finding corresponds to a 62% reduction in the emission impact, suggesting a significantly more efficient use of energy-related inputs in farms managed by younger farmers. A similar pattern is observed for feed and fodder (FF) costs. The elasticity of emissions with respect to the costs for feed and forage is significantly attenuated in young farms, with a reduction of 0.039 percentage points. In farms under the management of young entrepreneurs, a one per cent increase in fuel costs is associated with a 0.095 per cent increase in emissions \(\:{(\widehat{\gamma\:}}_{1}=\:-0.039,\:p\:<\:0.001)\) , compared to 0.134% in other farms. This indicates a 29% reduction in emissions. The interaction with VET expenses demonstrates a significant discrepancy, exhibiting a 0.030-point reduction \(\:{(\widehat{\gamma\:}}_{4}=-0.030,\:p<\:0.010)\) . This corresponds to an emission increase of merely 0.004% in younger farms, in contrast to 0.034% in other categories. It is important to note, however, that the interaction with mechanization costs (MEC × YOUNG) was not statistically significant. The model demonstrates a more "efficient" approach to input management by entrepreneurs under 40, with regard to both feed and energy use and veterinary services. This approach could also be indicative of a greater focus on sustainability among young enterprises. Moreover, this finding indicates that the generational shift in agriculture, besides being favourable from a demographic perspective, particularly in inland regions, can serve as a strategic factor for the economic resilience of the sector. The promotion of youth entrepreneurship is therefore not only a demographic imperative, but also a key enabler of climate and ecological transition, contributing to the integration of environmental, economic, and social competitiveness. The random effect pertaining to the general category of farming returns a statistically significant intercept variance ( \(\:{\widehat{\sigma\:}}_{u}^{2}=\:0.447\) ), indicating that about 45% of the variability in CO₂eq. emissions is attributable to structural differences between types of rearing, while the random effect of time ( \(\:{\widehat{\sigma\:}}_{v}^{2}\) < 0,001) is negligible. The random effects for each group of MToF \(\:{(\widehat{u}}_{j})\) are estimated to represent deviations from the population-level intercept, subsequent to the accounting for the fixed effects in the model. These reflect how each group differs from the global average expectation, conditional on the other predictors (see Fig. 1). The analysis of the estimated random effects for the five MToF reveals that poultry farms exhibit the strongest negative deviation from the overall intercept, suggesting lower conditional emission levels, net of fixed effects \(\:{(\widehat{u}}_{1}=-1.399)\) . This is probably due to a higher feed conversion efficiency and to the absence of enteric fermentation. Conversely, cattle farms demonstrate a positive deviation \(\:{(\widehat{u}}_{4}=0.161)\) , indicating above-average emissions when controlling for other predictors. The reduced environmental impact of specialised poultry farms, as evidenced by the findings of this study, could be further elucidated by examining the characteristics of this farming method and the features associated with the broiler chicken production contract between the processor, who is the owner of the chickens, and the breeder (Peng et al, 2024 ). The processor provides the production inputs, including chicks, feed, medicines, health services and technical assistance. Conversely, the financial burden of expenses such as electricity, labour, water, chicken coop maintenance, and disinfection are shouldered by the breeder (Taraman et al., 2023 ). Consequently, certain inputs utilised in the production process, notably feed, are not charged to the poultry holding but are regarded as off-farm (Wiedemann et al., 2012 ). This is exemplified by veterinary services, whose minimal impact on the carbon footprint might be attributable to vertical integration contracts. With regard to the pig farms, which demonstrate a random effect below the mean ( \(\:{\widehat{u}}_{2}=-0.331\) ), earlier research findings indicated that their contribution to the formation of carbon footprint is predominantly associated with manure and its management. Pig rearing is also essentially intensive, and is concerned with forms of vertical integration, through contracts with suppliers of inputs and processing companies. These contracts effectively exclude current production costs, which therefore do not contribute to the formation of the carbon footprint at the enterprise level (Reimer, 2006 ). The analysis indicates that sheep farms demonstrate positive deviations \(\:{(\widehat{u}}_{5}=0.210)\) with higher average emissions. These small ruminants, in conjunction with cattle, are responsible for the primary production of greenhouse gases, predominantly through enteric methane emissions, and, to a lesser extent, through coal and manure management systems (Recktenwald and Ehrhardt, 2024 ). According to Marino et al. ( 2016 ), the socio-economic role of sheep and goat farming is crucial, in that it contributes to landscape management, ecosystem preservation and biodiversity conservation, whilst also providing niche products to the market. Consequently, there is a robust interest in evaluating and enhancing the environmental performance of the sheep farming sector. However, Opio et al. ( 2013 ) claimed that global emissions from sheep are significantly higher than those from goats, due to discrepancies in meat production. Furthermore, goats have been shown to exhibit a lower emission intensity on average than sheep, due to their higher yields compared to sheep milk. It has been observed that goats on extensive farms exhibit elevated methane emissions; however, carbon sequestration has been demonstrated to substantially mitigate net emissions (Horrillo et al., 2024 ). In support of the aforementioned evidence, the boxplots of the carbon footprint by MToF (Fig. 2) provide additional visual confirmation of the model estimates. Poultry farms have been found to display the lowest median and interquartile range of CO₂eq. emissions, thus confirming the lower conditional impact that has been estimated by the model. Conversely, cattle and sheep farms exhibited wider distributions and higher median values, indicative of their positive random effects and elevated emission intensities. These visual deviations serve to reinforce the model's findings. It is noteworthy that the sheep and cattle sectors exhibited higher median CO₂eq values in comparison to the poultry sector, in accordance with the estimated random effects. Similar distributional patterns occur in both years of observation. The standardized residual plot (see Fig. 3) demonstrates a symmetrical distribution around zero, thus indicating a satisfactory model fit. In addition, the negligible Pearson correlation coefficient (r̂) between the fitted values and residuals ( \(\:\widehat{\rho\:}\:=\:0.001\) ) substantiates the absence of linear dependence, thereby satisfying the assumption of uncorrelated residuals and supporting the validity of the model. The distribution of residuals (see Fig. 4) approximates a normal bell-shaped curve, thereby suggesting that the assumption of residual normality is reasonably satisfied. This finding lends further support to the appropriateness of the log-transformed linear model that was utilised in the analysis. The estimated model demonstrates a clear correlation between a 1% increase in input costs (including feed, energy, mechanization and veterinary care) and a statistically significant rise in CO₂eq emissions. The analysis of the data according to farm type reveals the underlying factors that are responsible for the observed differences. These factors are found to be specific to each sector, and are both physiological and managerial in nature. The findings of the present study are consistent with those of INEA (Coderoni et al., 2016), which identified ruminant operations, particularly those involving cattle and sheep, as the primary source of enteric methane emissions. 3.2 Results of the model by Particular Type of Farming The analysis at the more detailed PToF classification stage enabled the specification of the model that most accurately reflected the production complexity of the sector. The model results are set out in Table 2 . Table 2 Results for PToF model estimation Random effects Groups Name Variance Std. Dev Particular ToF (Intercept) 0.556 0.746 Year (Intercept) 0.001 0.031 Residuals 0.679 0.825 Fixed effects Variable Coeff. Std. Error p-value FF 0.140 0.004 < 0.001 MEC 0.161 0.008 < 0.001 WEF 0.138 0.009 < 0.001 VET 0.037 0.005 < 0.001 YOUNG 0.046 0.039 0.232 FF × YOUNG -0.041 0.012 < 0.001 MEC × YOUNG 0.001 0.020 0.966 WEF × YOUNG -0.075 0.020 < 0.001 VET × YOUNG -0.027 0.012 0.032 The estimates of the coefficients of the economic variables fully confirm the observations made in the model by MToF (see Table 1 ), with all effects remaining statistically positive, highly significant and consistent (Table 2 ). It is noteworthy that the elasticity with respect to mechanisation expenditure is once again the most pronounced \(\:({\widehat{\beta\:}}_{2}=0.161,\:p<0.001),\:\) followed by FF costs \(\:({\widehat{\beta\:}}_{1}=0.140,\:p<0.001)\) and by WEF costs \(\:{\widehat{\beta\:}}_{3}=0.138,\:\:p<0.001)\) . A 1% increase in input costs has been shown to result in an average rise in emissions of 0.161% for mechanisation, 0.140% for feed and fodder, and 0.138% for water, electricity, and fuel. Veterinary care expenses were found to be the least impactful, though still significant \(\:({\widehat{\beta\:}}_{4}=0.037,\:p<0.001)\) . In order to further explore the management heterogeneity related to farmers' age, we estimated the specification of the model with interactions by particular type of farming (see Table 2 ). The model incorporates random intercepts at both the particular ToF and year level, thereby facilitating the capture of unobserved heterogeneity within a highly disaggregated structure. The model demonstrates a higher explanatory capacity than previous specifications, as evidenced by the reduction in AIC (9142.636 versus 9193.424) and a 56% share of variance explained at the group level (Table 3 ). The direct effect of the young entrepreneur variable on the intercept is not statistically significant, indicating that farms run by young farmers do not have average emissions that are significantly different from those produced by adult farms. However, interactions with the main production factors are highly significant, suggesting a differentiated management behaviour. Specifically, the negative interaction between FF costs and YOUNG \(\:{(\widehat{\gamma\:}}_{1}=-0.041,\:p<0.001)\) demonstrates a reduced impact of emissions relative to the increase for this input, thereby indicating a more efficient or technically advanced management of this item. A similar effect is observed for WEF costs, with a reduction in elasticity \(\:{(\widehat{\gamma\:}}_{3}=-0.075,\:p<0.001)\) . A similar effect is also observed for VET \(\:{(\widehat{\gamma\:}}_{4}=-0.027,\:p<0.05)\) . It is evident that no substantial interaction was identified with MEC input, a finding that is consistent with previous observations. The present study sets out to explore the relationship between WEF costs and CO₂eq. emissions in adult-run farms. The findings reveal that an increase of 1% in WEF costs is associated with a 0.138% increase in CO₂eq. emissions in farms run by adults, co mpared to only 0.063% in farms run by young entrepreneurs. This suggests a 54% lower emission elasticity for this input. In addition, an increase of 1% in FF costs results in a 0.140% rise in CO₂eq. emissions in traditional farms as opposed to 0.099% in young-run ones. This indicates a 29.3% reduction in the emission impact. With regard to VET, the estimated coefficient decreases from 0.037% in adult-managed farms to just 0.010% in those led by young farmers, corresponding to a 73% reduction in emission impact. The field of youth management has been shown to employ more sustainable management strategies, maybe as a result of a greater propensity for the adoption of innovative technologies and optimised practices in the domains of animal nutrition, energy management and livestock health. These recommendations, which require further investigation, underscore the correlation between environmental sustainability and generational renewal in the livestock sector. They also bear pertinent consequences for policies that support ecological transition and innovation. With regard to the outcomes of the model incorporating interactions, the age of farmers exerts a less significant influence on the global average carbon footprint. However, this relationship becomes apparent in contexts pertaining to production inputs, with the exception of mechanisation. A survey of young farmers has revealed that the primary motivations for adopting genetic selection of the herd are the reduction of emissions and the enhancement of manure, fodder, pasture, and animal housing infrastructure management. It has been demonstrated that these individuals tend to prioritise research, knowledge sharing and innovation in the agricultural sector (Gómez-Limón et al., 2010). The stability of these results, despite the adoption of a more granular sectoral classification, highlights the robustness of the estimated economic relationships across models. A comparison of information criteria (IC) reveals a marked enhancement in the quality of the model. The model based on the Particular Type of Farming demonstrates a lower AIC in comparison to the model based on the Main Type of Farming, thus indicating a superior degree of explanatory power in the more detailed classification. In a similar fashion, the log-likelihood undergoes an improvement from − 4584.700 to − 4559.300, thereby suggesting a superior fit to the observed data (see Table 3 ). The incorporation of both marginal and conditional \(\:{R}^{2}\) values serves to provide supplementary evidence for the robustness and appropriateness of the multilevel structure that has been adopted (see Table 3.1 ). As posited by Nakagawa and Schielzeth ( 2013 ), the conditional \(\:{R}^{2}\) is defined as the proportion of variance explained by both fixed and random effects. Conversely, the marginal \(\:{R}^{2}\) is defined as the variance explained by fixed effects alone. The null model demonstrates that 40.8% of the variance can be attributed to group-level differences ( \(\:{{R}^{2}}_{c}=\:\text{I}\text{C}\text{C}\:=\:0.408\) ), thereby validating the implementation of a hierarchical model structure (Gelman and Hill, 2006 ). In addition, the PToF model demonstrates superiority over the MToF model in terms of classification accuracy, as evidenced by a higher conditional \(\:{R}^{2}\) (0.649 vs. 0.606), higher ICC (0.457 vs. 0.396), and lower AIC. These enhancements signify a more precise modelling of group-level heterogeneity and enhanced explanatory power of the more granular sectoral classification. Table 3 IC measures of models Information criteria and variance of random intercepts for the MToF and PToF models. Model AIC BIC Log-likelihood Variance (ToF) MToF model 9193.424 9268.089 -4584.700 0.447 PToF model 9142.636 9217.300 -4559.300 0.562 Table 3.1 Model Comparison based on IC measures Model AIC BIC \(\:{{\varvec{R}}^{2}}_{\varvec{c}}\) \(\:{{\varvec{R}}^{2}}_{\varvec{m}}\) ICC RMSE Null model 11248.930 11273.818 0.408 0.000 0.408 1.087 MToF model 9193.424 9268.089 0.606 0.347 0.396 0.825 PToF model 9142.636 9217.300 0.649 0.354 0.457 0.817 As with the MToF, the inclusion of random effects on production patterns allowed further structural differentiations between farms. The intercept variance associated with the PToF group is 56% ( \(\:{\widehat{\sigma\:}}_{u}^{2}\) = 0.562) which is higher than that estimated for the main ToF ( \(\:{\widehat{\sigma\:}}_{u}^{2}\) = 0.447). This indicates greater between-group heterogeneity when a more detailed classification is adopted. The random effect of time ( \(\:{\widehat{\sigma\:}}_{v}^{2}<0.001)\) remains negligible even in this specification. The estimated random effects represent the deviations of the specific intercept for each ToF group from the implicit mean of the model, providing a measure of the emissive intensity "net" of the observed economic components. Negative values indicate groups with below-average emission levels at the same cost; conversely, positive values indicate a higher propensity to emissions. The findings of the analysis indicate that broiler poultry farms, in addition to combined-type and laying-hen farms, demonstrate a significant negative impact. This finding is indicative of a trend towards lower emission intensity, which is presumably attributable to high food efficiency and concentrated production structures. In contrast, the analysis of beef and sheep farms has revealed a positive effect, suggesting a greater emission impact. This finding is further substantiated by the distribution of estimated random effects \(\:\left({\widehat{u}}_{j}\right)\) , as illustrated in Fig. 5. As demonstrated in Fig. 1, the poultry farming sector is significantly below the overall mean. These deviations are indicative of structural differences in emission intensity across farming sectors, after controlling for all other covariates in the model. As demonstrated by the figure, the internal consistency of specialisations within each production orientation is highlighted. Poultry farms, which have been divided according to their production orientations, show relatively close values of random effects for meat \(\:{(\widehat{u}}_{1}=-1.460)\) . The findings of the study demonstrated a correlation between the utilisation of eggs \(\:{(\widehat{u}}_{3}=-1.145)\) and the production of a combined product \(\:{(\widehat{u}}_{2}=-1.369)\) , thus confirming a common production structure characterised by high feed conversion efficiency and the absence of enteric fermentation. Furthermore, the discrepancies in carbon footprint impacts between broiler and egg production are reflected in feed consumption, associated manure production, and material and energy use. For instance, the production of eggs incurs higher energy costs than that of broiler farms, in addition to generating greater quantities of manure. Consequently, the production cycle of broiler meat is comparatively brief, and the feed consumption and manure production per unit are at their lowest (Coderoni et al., 2013; Kyriazakis, 2013 ). The impacts of two types of farms belonging to the pig sector, related to rearing ( \(\:{\widehat{u}}_{4}=-0.399\) ) and meat ( \(\:{\widehat{u}}_{5}=-0.356\) ) are close (De Cuyper et al., 2024 ), but a notably high value is graphically observed for mixed pig farms ( \(\:{\widehat{u}}_{10}=0.340\) ). The latter case may be indicative of the presence of a small number of large-scale mixed pig farms within the sample. Such farms are distinguished by their high production intensity, which involves the consolidation of multiple production cycles. These farms are associated with elevated levels of feed input, energy consumption and manure output, consequently resulting in a more substantial carbon footprint. In the cattle sector, dairy and mixed farms demonstrate analogous yet modestly favourable outcomes ( \(\:{\widehat{u}}_{7}=0.042\) and \(\:{\widehat{u}}_{8}=0.098\) , respectively), whilst meat-oriented farming exhibits distinct characteristics, manifesting the most pronounced positive impact of the entire distribution by PToF \(\:{(\widehat{u}}_{11}=0.371)\) . It is evident that cattle farms have the greatest impact on the carbon footprint of agriculture, primarily due to enteric emissions from these ruminants. However, a distinction is made in the literature between dairy cows and beef farms. It is evident that beef cattle farms exert a significantly greater influence on the soil, given their requirement for a substantially larger area of land. In addition, it is important to note that emissions from dairy cattle are also related to their slaughter and the subsequent meat production (Levasseur et al., 2024 ; Beauchemin et al., 2011 ). Sheep farms have also been observed to demonstrate a positive effect, albeit to a lesser extent than that observed for beef cattle. This observation is consistent with the physiology of small ruminants and the high incidence of enteric emissions characteristic of livestock farming systems. Finally, the findings of this study demonstrate that goat farms \(\:{(\widehat{u}}_{6}= -0.254\) ) remain below average, in accordance with the results obtained in the MToF model. The presented patterns confirm that our PToF classification effectively captures meaningful intra-group heterogeneity in emissive intensity. The analysis at the PToF level reveals internal differences between the various types of farms according to the destination of the product (e.g. milk versus meat, or eggs versus meat). This is likely to reflect higher structural emissions from slow-growing ruminants with lower feed yields, as documented in some technical reports (Coderoni and Bonati, 2013 ). The boxplots in Fig. 6 visually corroborate these findings, confirming the structural differences in emission intensity across PToF categories. Median emissions are consistently lowest in poultry broilers, combined poultry, and laying hens, in line with the model-based estimates of negative random effects for these specialisations. The distribution of carbon footprint values across PToF confirms the structural differences observed in the model estimates. Emission levels in poultry farms, especially broiler farms, combined poultry farms, and laying hen farms, are consistently the lowest. In contrast, beef cattle, sheep, and pig fattening farms exhibit wider variability and higher median values. Among cattle operations, dairy and combined systems demonstrate moderate emissions. These trends demonstrate stability throughout the observation period, aligning with the estimated random effects. This finding lends support to the robustness of the adopted sectoral classification. It is important to note that the combined pig category includes a very limited number of exceptionally large farms, which likely explains the presence of extreme values in the upper tail of the distribution. Due to their limited representativeness, these observations were not emphasised in the discussion, as they may distort the typical emission profile of this sector. The validity of the statistical assumptions was evaluated through model diagnostics. The scatterplot of standardized residuals by PToF (Fig. 7) displays a symmetrical cloud centred on zero with no discernible pattern, visually corroborating the assumption of homoscedasticity. In order to assess potential multicollinearity among the independent variables, Variance Inflation Factors (VIFs) were calculated. As demonstrated in Fig. 8, all values are found to be well below the commonly accepted threshold of 5, thus indicating a low level of multicollinearity. This finding lends further support to the reliability and stability of the coefficient estimates in the model. 4. Conclusions The results presented in this study may have implications from several perspectives. Interventions aimed at curtailing the use of feed and enhancing energy efficiency represent pivotal mechanisms for mitigating the carbon footprint of Italian livestock farms, particularly in light of projections regarding future meat consumption by consumers. On the other hand, they could encourage the adoption of circular approaches, particularly in the most impactful areas of farming. With regard to the energy requirements for the management of livestock, the implementation of photovoltaic panels for self-consumption has the potential to enhance both the economic and environmental sustainability of livestock farms. Furthermore, it is well established that adequate nutrition and hydration are pivotal to the successful husbandry of livestock, exerting a substantial impact on the quality of breeding stock and the final product. Ruminants have been identified as the primary contributors to pollution on livestock farms, with their digestive processes being the main source of methane emissions. As is evident, the feeding of animals is subject to stringent regulation in accordance with the principles of quality production. However, the recommended dietary regime involves the substitution of carbohydrates with unsaturated lipids, or alternatively, modifications to the ration, including a reduction in dietary fibre and an increase in the concentrate/forage ratio. These measures are designed to reduce enteric emissions and nitrogen excretion. The analysis indicates that mechanisation exhibits the highest degree of elasticity among all of the input items that were analysed in the econometric study, both in terms of the overall level and of the specific type of farming. Indeed, the mechanisation of farming processes has been demonstrated to increase energy costs (e.g. milking equipment; production of farm crops for livestock) and fuel consumption, as well as to cause carbon dioxide emissions. The propagation of digital technologies is conducive to the mitigation of deleterious effects and the enhancement of the efficiency of production processes in the context of livestock farming. This is attributable to the capacity of digital systems to encompass all technological stages of breeding, care and fattening. Finally, the sustainable approach observed among young farmers is a further motivation for strengthening the existing policies for the recruitment of young people in agriculture. In this regard, policy measures could be implemented to encourage the education of young entrepreneurs within specific livestock courses and to facilitate the ongoing training of older entrepreneurs. In conclusion, it is evident that enterprises must prioritise sustainability as a strategic imperative, rather than considering it an optional extra. A potential limitation of this study is the short temporal span of the panel, which includes only two years. However, using the dataset allowed us to capture substantial cross-sectional heterogeneity, thereby contributing to the robustness of the model. Despite the temporal constraint, we are confident in the goodness of the results, as they seem to comply with all underlying statistical assumptions. Future research could be useful to improve the comprehension of how production systems can evolve to address the pressing environmental and socio-economic challenges in the coming decades. Abbreviations CF Carbon footprint CH 4 Methan CO 2 carbon dioxide CO 2 eq Carbon dioxide equivalent CREA Council for Agricultural Research and Economics EU European Union FADN Farm Accountancy Data Network FF Feed and fodder GHGs Greenhouse gases ISPRA Italian Institute for Environmental Protection and Research MEC Mechanization MToF Main Type of Farming NO 2 nitrogen oxides PToF Particular Type of Farming VET Veterinary care WEF Water, energy and fuel Declarations Ethics approval and consent to participate Non applicable Consent for publication Not applicable Funding Not applicable Author Contribution V.G. and M.B.F. conceptualize the study and wrote the initial draft; D.S. and L.R. designed the methodology; A.S. carried out the data analysis, V.G. and M.B.F. provided supervision. All authors reviewed and edited the manuscript. Data Availability Data that support the findings of this study come from FADN (Farm Accountancy Data Network), Italian section References Alem H (2023) The role of green total factor productivity to farm-level performance: evidence from Norwegian dairy farms. Agric Econ 11:2. https://doi.org/10.1186/s40100-023-00243-0 Allen MR, Shine KP, Fuglestved J et al (2018) A solution to the misrepresentations of CO₂ -equivalent emissions of short-lived climate pollutants under ambitious mitigation. npj Clim Atmos Sci 1:16. https://doi.org/10.1038/s41612-018-0026-8 Andrade HJ, Vega A, Martínez-Salinas A, Villanueva C, Jiménez-Trujillo JA, Betanzos-Simon JE, Pérez E, Ibrahim M, Sepúlveda CJ (2024) The carbon footprint of livestock farms under conventional management and silvopastoral systems in Jalisco, Chiapas, and Campeche (Mexico). Front Sustain Food Syst 8. https://doi.org/10.3389/fsufs.2024.1363994 Beauchemin KA, Janzen HH, Little SM, McAllister TA, McGinn SM (2011) Mitigation of greenhouse gas emissions from beef production in western Canada – Evaluation using farm-based life cycle assessment, Anim Feed Sci. Technol., 166–167 . https://doi.org/10.1016/j.anifeedsci.2011.04.047 Bonnin D, Tabacco E, Borreani G (2021) Variability of greenhouse gas emissions and economic performances on 10 Piedmontese beef farms in North Italy. Agricultural Syst 194. 10.1016/j.agsy.2021.103282 Castillo-Díaz FJ, Belmonte-Ureña LJ, López-Serrano MJ et al (2024) Quantifying sustainability in the agri-food system: a comprehensive methodological framework and expert consensus approach. Agric Econ 12:20. https://doi.org/10.1186/s40100-024-00314-w Coderoni S, Bonati D (eds) (2013) Impronta Carbonica Aziende Agricole Italiane – ICAAI. Inea, Roma Coderoni S, Pontrandolfi A (eds) (2016) Analisi delle potenzialità e delle prospettive CREA Cheng M, McCarl B, Fei C (2022) Climate change and livestock production: a literature review. Atmosphere 13(1):140. 10.3390/atmos13010140 Contini C, Marotta G, Torquati B (2020) Multi-actor approaches to implement cooperative strategies and value chains based on sustainability. Agric Econ 8:7. https://doi.org/10.1186/s40100-019-0147-3 de Vries M, van Middelaar CE, de Boer IJM (2015) Comparing environmental impacts of beef production systems: A review of life cycle assessments. Livest Sci 178. https://doi.org/10.1016/j.livsci.2015.06.020 De Cuyper C, Van linden V, Leen F, Aluwé M, Van Meensel J, Millet S (2024) The impact of slaughter weight and sex on the carbon footprint of pig feed intake. Animal 18:4. https://doi.org/10.1016/j.animal.2024.101119 Dogbe W, Wang Y, Revoredo-Giha C (2024) Nutritional implications of substituting plant-based proteins for meat: evidence from home scan data. Agric Econ 12:30. https://doi.org/10.1186/s40100-024-00324-8 Eurostat (2024) Air emissions accounts by NACE Rev. 2 activity. https://data.europa.eu/data/datasets/j3ywu7pahvs5s1ejw4a3fw?locale=en . (Accessed 10 December 2024) FAO (2024) Greenhouse gas emissions from agrifood systems. Global, regional and country trends, 2000–2022. https://openknowledge.fao.org/items/74bfebdb-3272-4e6a-98f4-ee36c7146d44 . (Accessed 10 March 2025) Fouda T, Awny A, Darwish M, Kassab N, Ghoname M (2021) Carbon footprint estimation in poultry production farms. Sci Pap Ser Manag Econ Eng Agric Rural Dev 21:4. 10.13140/RG.2.2.31479.34726 Gelman A, Hill J (2006) Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press, New York, NY, USA Gómez-Limón JA, Sanchez-Fernandez G (2010) Empirical Evaluation of Agricultural Sustainability Using Composite Indicators. Ecol Econ 69:1062–1075. https://doi.org/10.1016/j.ecolecon.2009.11.027 Herrero M, Thornton FK, Gerber P, Reid RS (2009) Livestock, livelihoods and the environment: understanding the trade-offs. Curr Opin Environ Sustain 1:2 Horrillo A, Gaspar P, Rodríguez-Ledesma A, Escribano M (2024) Assessment of Greenhouse Gas Emissions and Carbon Sequestration in Dairy Goat Farming Systems in Northern Extremadura, Spain. Animals 14(23):3501 Huntingford C, Lowe JA, Howarth N, Bowerman N, Gohar LK, Otto A, Lee D, Smith S, den Elzen M, van Vuuren D, Millar R, Allen M (2015) The implications of carbon dioxide and methane exchange for the heavy mitigation RCP2.6 scenario under two metrics. Environ Sci Policy 51. http://dx.doi.org/10.1016/j.envsci.2015.03.013 ISPRA (2022) Le emissioni nazionali di gas serra. Settore Agricoltura – anno 2022. https://emissioni.sina.isprambiente.it/inventario-nazionale/#Report . (Accessed 10 December 2024) Kang HY, Hwang YW, Lee JH, Cho SJ, Jeon YS, Kim NS, Kim J (2025) Evaluating the Greenhouse Gas Emissions Footprint of Chicken Meat Production in South Korea: A Life Cycle Perspective. Food Bioprod Process. https://doi.org/10.1016/j.fbp.2025.01.017 Kyriazakis I (2013) Quantifying the environmental impacts of UK broiler and egg production systems. Lohmann Inform 48:2 Levasseur P, Mariotti F, Denis I et al (2024) Potential unexpected effects of meat reduction in diet: Could educational attainment influence meat substitution strategies? Agric Econ 12:4. https://doi.org/10.1186/s40100-024-00298-7 Lynch J, Cain M, Frame D, Pierrehumbert R (2021) Agriculture's contribution to climate change and role in mitigation is distinct from predominantly fossil CO2-emitting sectors. Front Sust Food Syst 4:518039. https://doi.org/10.3389/fsufs.2020.518039 Mancilla-Leytón JM, Morales-Jerrett E, Muñoz-Vallés S, Mena Y (2023) A Comparative Analysis of Carbon Footprint in the Andalusian Autochthonous Dairy Goat Production Systems. Animals 13(18):2864. 10.3390/ani13182864 Marino R, Atzori AS, D'Andrea M, Iovane G, Trabalza-Marinucci M, Rinaldi L (2016) Climate change: Production performance, health issues, greenhouse gas emissions and mitigation strategies in sheep and goat farming. Small Rumin Res 135. https://doi.org/10.1016/j.smallrumres.2015.12.012 Mazzetto AM, Bishop G, Styles D, Arndt C, Brook R, Chadwick D (2020) Comparing the environmental efficiency of milk and beef production through life cycle assessment of interconnected cattle systems. J Clean Prod 277:124108. https://doi.org/10.1016/j.jclepro.2020.124108 Mottet A, Tempio G (2017) Global poultry production: current state and future outlook and challenges. World’s Poult Sci J 73:245–256. 10.1017/S0043933917000071 Nakagawa S, Schielzeth H (2013) A general and simple method for obtaining R² from generalized linear mixed-effects models. Methods Ecol Evol 4(2):133–142. https://doi.org/10.1111/j.2041-210X.2012.00261.x Notarnicola B, Tassielli G, Renzulli PA, Castellani V, Sala S (2017) Environmental impacts of food consumption in Europe. J Clean Prod 140:753–765. https://doi.org/10.1016/j.jclepro.2016.06.080 Opio C, Gerber P, Mottet A, Falcucci A, Tempio G, MacLeod M, Vellinga T, Henderson B, Steinfeld H (2013) Greenhouse Gas Emissions from Ruminant Supply Chains – A Global Life Cycle Assessment. Food and Agriculture Organization of the United Nations (FAO), Rome Pais DF, Marques AC, Fuinhas JA (2022) The cost of healthier and more sustainable food choices: Do plant-based consumers spend more on food? Agric Econ 10:18 Peng B, Melnikiene R, Balezentis T et al (2024) Structural dynamics and sustainability in the agricultural sector: the case of the European Union. Agric Econ 12:31. https://doi.org/10.1186/s40100-024-00321-x Recktenwald EB, Ehrhardt RA (2024) Greenhouse gas emissions from a diversity of sheep production systems in the United States. Agric Syst 217. 10.1016/j.agsy.2024.103915 Reimer J (2006) Vertical Integration in the Pork Industry. Am J Agric Econ 88:234–248. https://doi.org/10.1111/j.1467-8276.2006.00850.x Rice P, O'Brien D, Shalloo L, Holden NM (2017) Evaluation of allocation methods for calculation of carbon footprint of grass-based dairy production. J Environ Manag 202:311–319. https://doi.org/10.1016/j.jenvman.2017.06.071 Sau P, Arru B, Cisilino F, Furesi R, Pulina P, Madau FA (2023) Economic and Environmental Sustainability Trade-Off Analysis in Sheep Farming Using the Farm Accountancy Data Network Database. Sustainability 15(24):16621. https://doi.org/10.3390/su152416621 Schroeder R, Aguiar L, Baines RN (2012) Carbon Footprint in Meat Production and Supply Chains. J Food Sci Eng 652–665. 2 https://doi.org/10.17265/2159-5828/2012.11.005 Singh S, Pagare V (2024) Status and Scope of Mechanization in Animal Husbandry. RASSA J Sci Soc 6:75–83. https://doi.org/10.5958/2583-3715.2024.00012.X Sul D (2019) Panel Data Econometrics: Common Factor Analysis for Empirical Researchers. New. Routledge, York, NY Taraman S, Taşcıoğlu Y, Bulut OD (2023) Profitability and Cost Analysis for Contract Broiler Production in Turkey. Animals , 13(13), 2072. 10.3390/ani13132072 Thoma G, Wiedemann S, Palhares J, Saleque M, Kanzaki M, Missohou A, Kyriazakis I, Gittins J, Burr J, Mitloehner F (2015) Greenhouse gas emissions and fossil energy use from poultry supply chains. Food and Agricultural Organization of the United Nations Treml N, Rudi A, Schultmann F (2025) Evaluating environmental impacts of pork production: A life cycle assessment of seven case studies in Germany. J Clean Prod 503:145408. https://doi.org/10.1016/j.jclepro.2025.145408 Velarde-Guillén J, Arndt C, Gómez CA (2022) Carbon footprint in Latin American dairy systems. Trop Anim Health Prod 54:1–7. https://doi.org/10.1007/s11250-021-03021-6 Wiedemann S, McGahan E, Poad G (2012) Using life cycle assessment to quantify the environmental impact of chicken meat production. RIRDC Publication No. 12/029. (Rural Industries Research and Development Corporation (RIRDC), Australia). https://agrifutures.com.au/product/using-life-cycle-assessment-to-quantify-the-environmental-impact-of-chicken-meat-production/ Wooldridge JM (2010) Econometric Analysis of Cross Section and Panel Data, second edn. MIT Press, Cambridge, MA Additional Declarations No competing interests reported. Cite Share Download PDF Status: Posted Version 1 posted 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-6959911","acceptedTermsAndConditions":true,"allowDirectSubmit":true,"archivedVersions":[],"articleType":"Research Article","associatedPublications":[],"authors":[{"id":489654477,"identity":"581f7975-482c-4f6e-a7e1-7daf61847df4","order_by":0,"name":"Vincenzo Giaccio","email":"data:image/png;base64,iVBORw0KGgoAAAANSUhEUgAAAZAAAAAyAQMAAABI0h/eAAAABlBMVEX///8AAABVwtN+AAAACXBIWXMAAA7EAAAOxAGVKw4bAAAAuklEQVRIiWNgGAWjYDACZhBhwCDHwMDYwMBQwMBDtBZjiBYDBh4i9EBAYgMDRC9ha+TdmR9++FFgk77hdnPrhg8GDDL2hLQYHmYzluwxSMvdcOdg280ZxDjMsJnBjJnB4HDuhhuJbbd5iNPC/g2kJd2AaC3yzDxgWxKI12LAzFMM8ovhTKAWoF8keHgOELKl//jGDz/+2Mjz3Uh/duNDhY09ewMhW9DMlCDkLKAthMwcBaNgFIyCUcAAANJkOHYO+PMsAAAAAElFTkSuQmCC","orcid":"","institution":"University of Molise","correspondingAuthor":true,"prefix":"","firstName":"Vincenzo","middleName":"","lastName":"Giaccio","suffix":""},{"id":489654478,"identity":"161506d2-5e5a-471e-ad0a-ddb4a7cf72ed","order_by":1,"name":"Diana Salottolo","email":"","orcid":"","institution":"University of Molise","correspondingAuthor":false,"prefix":"","firstName":"Diana","middleName":"","lastName":"Salottolo","suffix":""},{"id":489654479,"identity":"c26574b1-d605-4d49-809f-8761436816ae","order_by":2,"name":"Luca Romagnoli","email":"","orcid":"","institution":"University of Molise","correspondingAuthor":false,"prefix":"","firstName":"Luca","middleName":"","lastName":"Romagnoli","suffix":""},{"id":489654480,"identity":"5c3c24c1-0bbc-4b03-b802-26e0070583d8","order_by":3,"name":"Maria Bonaventura Forleo","email":"","orcid":"","institution":"University of Molise","correspondingAuthor":false,"prefix":"","firstName":"Maria","middleName":"Bonaventura","lastName":"Forleo","suffix":""},{"id":489654481,"identity":"8e5eef80-95cf-4fbc-868d-35187e3105ed","order_by":4,"name":"Alfonso Scardera","email":"","orcid":"","institution":"Council for Agricultural Research and Economics (CREA)","correspondingAuthor":false,"prefix":"","firstName":"Alfonso","middleName":"","lastName":"Scardera","suffix":""}],"badges":[],"createdAt":"2025-06-23 21:38:14","currentVersionCode":1,"declarations":"","doi":"10.21203/rs.3.rs-6959911/v1","doiUrl":"https://doi.org/10.21203/rs.3.rs-6959911/v1","draftVersion":[],"editorialEvents":[],"editorialNote":"","failedWorkflow":false,"files":[{"id":87565717,"identity":"4e9df05f-d644-47d1-805a-2e303b30d54b","added_by":"auto","created_at":"2025-07-25 09:25:40","extension":"jpeg","order_by":1,"title":"Figure 1","display":"","copyAsset":false,"role":"figure","size":157453,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEstimated Random Effects by Main Type of Farming (MToF)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRandom intercept deviations for each MToF from overall average\u003c/p\u003e","description":"","filename":"floatimage1.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/ea2af8c1ea412a3fedb57847.jpeg"},{"id":87565712,"identity":"f84e62ad-aa0e-441f-85ec-201ffbdf12e3","added_by":"auto","created_at":"2025-07-25 09:25:40","extension":"png","order_by":2,"title":"Figure 2","display":"","copyAsset":false,"role":"figure","size":33578,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBoxplots of CO₂eq emissions by MToF – Years 2021-2022\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDistribution of CO\u003csub\u003e2\u003c/sub\u003e equivalent emissions by MToF for the years 2021 and 2022\u003c/p\u003e","description":"","filename":"floatimage2.png","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/02deb5d90aeec2803ae83113.png"},{"id":87567090,"identity":"e24d9568-7428-4b07-b5b7-4fbed8046c98","added_by":"auto","created_at":"2025-07-25 09:41:40","extension":"png","order_by":3,"title":"Figure 3","display":"","copyAsset":false,"role":"figure","size":59294,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eResidual cloud\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eScatterplot of residuals versus fitted values by MToF\u003c/p\u003e","description":"","filename":"floatimage3.png","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/8f035c37f3d76ce19297b996.png"},{"id":87565869,"identity":"adb87343-da29-4e96-baea-9e56d5638cc1","added_by":"auto","created_at":"2025-07-25 09:33:40","extension":"png","order_by":4,"title":"Figure 4","display":"","copyAsset":false,"role":"figure","size":15379,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eHistogram of residuals\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDistribution of residuals with overlaid normal distribution curve\u003c/p\u003e","description":"","filename":"floatimage4.png","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/d3f2bb85306fc2f657e3fd33.png"},{"id":87565715,"identity":"25bbd533-8bc5-44b3-994f-a595d071ea99","added_by":"auto","created_at":"2025-07-25 09:25:40","extension":"jpeg","order_by":5,"title":"Figure 5","display":"","copyAsset":false,"role":"figure","size":243881,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eEstimated Random Effects by Particular Type of Farming (PToF)\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eRandom intercept deviations for each PToF from overall average\u003c/p\u003e","description":"","filename":"floatimage5.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/89c70aaddf1182608c7a1431.jpeg"},{"id":87565870,"identity":"e9bb5d29-d7ba-45ef-aabb-e0aed9037405","added_by":"auto","created_at":"2025-07-25 09:33:40","extension":"jpeg","order_by":6,"title":"Figure 6","display":"","copyAsset":false,"role":"figure","size":201478,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eBoxplots of CO₂eq emissions by PToF – Years 2021-2022\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eDistribution of CO\u003csub\u003e2\u003c/sub\u003e equivalent emissions by PToF for the years 2021 and 2022\u003c/p\u003e","description":"","filename":"floatimage6.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/2395011c33e58b5aefe820d6.jpeg"},{"id":87565720,"identity":"276d4ef1-9778-443c-b236-07d15d8ae6f2","added_by":"auto","created_at":"2025-07-25 09:25:40","extension":"png","order_by":7,"title":"Figure 7","display":"","copyAsset":false,"role":"figure","size":70715,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eResidual cloud\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eScatterplot of residuals versus fitted values by PToF\u003c/p\u003e","description":"","filename":"floatimage7.png","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/5c9a6f259bc5766e02d003a0.png"},{"id":87565722,"identity":"1c7e7b89-83be-486d-bf01-317acd190a15","added_by":"auto","created_at":"2025-07-25 09:25:40","extension":"jpeg","order_by":8,"title":"Figure 8","display":"","copyAsset":false,"role":"figure","size":34957,"visible":true,"origin":"","legend":"\u003cp\u003e\u003cstrong\u003eCollinearity\u003c/strong\u003e\u003c/p\u003e\n\u003cp\u003eVariance Inflation Factors (VIFs) for the Independent Variables\u003c/p\u003e","description":"","filename":"floatimage8.jpeg","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/9646f43696fad7d9d9f8b6ca.jpeg"},{"id":87567273,"identity":"1cdf2c4b-b51b-4651-a0f7-c5ebbedf159b","added_by":"auto","created_at":"2025-07-25 09:49:42","extension":"pdf","order_by":0,"title":"","display":"","copyAsset":false,"role":"manuscript-pdf","size":1704683,"visible":true,"origin":"","legend":"","description":"","filename":"manuscript.pdf","url":"https://assets-eu.researchsquare.com/files/rs-6959911/v1/1830f169-9949-4780-81af-3a4b81c594fa.pdf"}],"financialInterests":"No competing interests reported.","formattedTitle":"Investigating the Carbon Footprint of Italian Specialized Livestock Farms and its Drivers","fulltext":[{"header":"1. Introduction and literature background","content":"\u003cp\u003eGreenhouse gases (GHGs) are natural elements in the atmosphere and are produced by natural phenomena, such as the respiration of plants, animals and microbes, and the natural emissions of volcanoes and wildfires. However, the disproportionate accumulation of greenhouse gases and consequent global warming are a result of human activity, which must be monitored (Allen et al., \u003cspan citationid=\"CR2\" class=\"CitationRef\"\u003e2018\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe agri-food system impacts climate change through various mechanisms, including the emission of greenhouse gases (GHGs). Indeed, agricultural production is a significant contributor to anthropogenic global warming, and reducing agricultural emissions \u0026mdash; largely methane and nitrous oxide \u0026mdash; could play a significant role in mitigating climate change (Alem, 2033; Lynch et al., \u003cspan citationid=\"CR26\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). Livestock farming directly emits climate-altering gases, including methane, nitrous oxide and carbon dioxide, and indirectly emits gases and carbon through land use and feed production (Contini et al., \u003cspan citationid=\"CR10\" class=\"CitationRef\"\u003e2020\u003c/span\u003e). The livestock sector in particular is a driver of climate change, generating around 15% of global GHG emissions from anthropogenic sources (Pais et al., \u003cspan citationid=\"CR34\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Cheng et al., \u003cspan citationid=\"CR9\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; FAO, \u003cspan citationid=\"CR15\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). In 2021, Italy was the fourth EU country most responsible for the greenhouse effect, after Germany, Poland and France (Eurostat, \u003cspan citationid=\"CR14\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). According to the Italian Institute for Environmental Protection and Research (ISPRA), in 2022, 7.4% of GHG emissions in Italy came from the agricultural sector. Therefore, agriculture is the third largest source of emissions after the energy sector (55.0%) and transport (24.7%).\u003c/p\u003e\u003cp\u003eLivestock farming has been identified as a significant contributor to air pollution, with particular concern regarding methane (CH\u003csub\u003e4\u003c/sub\u003e), nitrogen oxides (NO\u003csub\u003e2\u003c/sub\u003e), and carbon dioxide (CO\u003csub\u003e2\u003c/sub\u003e) emissions (Castillo-D\u0026iacute;az et al., \u003cspan citationid=\"CR6\" class=\"CitationRef\"\u003e2024\u003c/span\u003e; Fouda et al., \u003cspan citationid=\"CR16\" class=\"CitationRef\"\u003e2021\u003c/span\u003e). However, the impact on the atmosphere varies according to the type of greenhouse gas. For instance, methane, the primary GHG emitted by livestock farms, and carbon dioxide, the most significant, have different effects. Methane has a limited effect over time, and thus, it is less responsible for the greenhouse effect compared to carbon dioxide, which persists in the atmosphere for a longer duration (Huntingford et al., \u003cspan citationid=\"CR21\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). As is well known, carbon dioxide is produced when fossil fuels are burned to generate energy (Schr\u0026ouml;der et al., 2012). The literature agrees that methane is the main cause of anthropogenic greenhouse gas emissions on livestock farms (Velarde-Guill\u0026eacute;n et al., \u003cspan citationid=\"CR46\" class=\"CitationRef\"\u003e2022\u003c/span\u003e; Mazzetto et al., \u003cspan citationid=\"CR29\" class=\"CitationRef\"\u003e2020\u003c/span\u003e; Rice et al., \u003cspan citationid=\"CR38\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eRegarding the country analysed in the study, it should be noted that Italian livestock farms are responsible for 79% of overall emissions in the national agricultural sector (ISPRA, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). Enteric fermentation is the main cause of methane emissions in the agricultural sector, representing 47% of emissions into the atmosphere. It is mainly generated by reactions in the digestive systems of livestock, particularly cattle. Animal waste and its management, as well as spreading livestock manure on agricultural land and grazing, are instead at the basis of nitrous oxide emissions (de Vries et al., \u003cspan citationid=\"CR11\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). Methane, the gas with the greatest impact on the carbon footprint of Italian livestock farms from enteric fermentation, is the primary contributor, while emissions of other greenhouse gases are comparatively lower. Furthermore, 21% of emissions derive from manure management, and 10% from the management of agricultural soils that could be traced back to livestock farms themselves (ISPRA, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). The impact of cattle (68.1%) and pigs (12.7%) represents approximately 80% of the total emissions from Italian livestock farms in 2022 (ISPRA, \u003cspan citationid=\"CR22\" class=\"CitationRef\"\u003e2022\u003c/span\u003e). This observation is particularly pertinent when considering the type of livestock in question.\u003c/p\u003e\u003cp\u003eFeed production has been identified as the primary contributor to greenhouse gas (GHG) emissions, particularly within the context of the poultry supply chain, where the production of ingredients, such as corn and soy, plays a significant role (Thoma et al., \u003cspan citationid=\"CR44\" class=\"CitationRef\"\u003e2015\u003c/span\u003e). In the context of poultry farming, the predominant organisational model is characterised by an intensive livestock, with a clear demarcation between production categories for meat and eggs. This is further compounded by the high human consumption of products derived from this sector (Dogbe et al., \u003cspan citationid=\"CR13\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). These products are comparatively limited in size and are able to guarantee the levels of proteins necessary for optimal nutrition (Herrero et al., \u003cspan citationid=\"CR19\" class=\"CitationRef\"\u003e2009\u003c/span\u003e; Mottet and Tempio, \u003cspan citationid=\"CR30\" class=\"CitationRef\"\u003e2017\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe carbon footprint (CF) is the most widely used indicator, usually expressed in terms of carbon dioxide equivalent (CO₂eq), to quantify the impact of livestock farming on global warming (Mancilla-Leyt\u0026oacute;n et al., \u003cspan citationid=\"CR27\" class=\"CitationRef\"\u003e2023\u003c/span\u003e). A comprehensive review of the extant literature reveals a wide range of variables for estimating the carbon footprint, including the production costs of on-farm or off-farm inputs and external services. As Andrade et al. (\u003cspan citationid=\"CR3\" class=\"CitationRef\"\u003e2024\u003c/span\u003e) and Bonnin et al. (\u003cspan citationid=\"CR5\" class=\"CitationRef\"\u003e2021\u003c/span\u003e) note, the production costs of raw materials (e.g. animal feeds, veterinary products, fuel, electricity, water and services such as veterinary consultants) are some of the variables considered by the literature to measure the carbon footprint of livestock farms. The financial implications of mechanisation have been identified as a key variable in extant literature. Mechanisation in livestock farming has been demonstrated to enhance the efficiency of waste disposal from dairy farming and poultry, thereby effectively reducing the discharge of pollutants. This is a significant variable, given its application in the provision of sustenance and hydration for animals, as well as the establishment of climate-controlled stabling conditions (Singh and Pagare, \u003cspan citationid=\"CR41\" class=\"CitationRef\"\u003e2024\u003c/span\u003e). Conversely, the process of mechanization gives rise to the emission of carbon dioxide into the atmosphere.\u003c/p\u003e\u003cp\u003eA number of studies have indicated that farms managed by young people demonstrate higher levels of sustainability from an environmental perspective (Sau et al., \u003cspan citationid=\"CR39\" class=\"CitationRef\"\u003e2023\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe objective of the study is to analyse the relationship between the most significant variables related to livestock management and the carbon footprint, as calculated by CREA (Coderoni et al., 2013) and intended as a variable representing the environmental sustainability of specialised livestock farming over a two-year period. The input variables are: feed and fodder (FF), water, energy and fuel (WEF), veterinary care (VET), mechanization (MEC). All of these are measured in terms of cost. The goal is to shed some light on the effect that economic variables have on CO₂ equivalent emissions at level of farm unit. The methodology employed is a multilevel mixed-effects regression model (Linear Mixed-Effects Regression Model) on a balanced panel dataset (Gelman and Hill, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2006\u003c/span\u003e). This model is suitable for managing hierarchical data structures and for taking into account unobserved variability between farm groups by type of farming and over time (the last two years with available data, 2021\u0026ndash;2022), including an interaction analysis with the variable \"Young entrepreneur\" (\u0026lt;\u0026thinsp;40 years).\u003c/p\u003e\u003cp\u003eThe panel data structure enables the control of both observable and unobservable heterogeneity across units and over time, in accordance with the panel econometrics framework (Sul, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The rationale behind the analysis at the farm level is substantiated by existing literature on the environmental impact of the livestock sector, which attributes the majority of impacts to this phase of production (Kang et al., \u003cspan citationid=\"CR23\" class=\"CitationRef\"\u003e2025\u003c/span\u003e; Notarnicola et al., \u003cspan citationid=\"CR32\" class=\"CitationRef\"\u003e2017\u003c/span\u003e; Treml et al., \u003cspan citationid=\"CR45\" class=\"CitationRef\"\u003e2025\u003c/span\u003e).\u003c/p\u003e"},{"header":"2. Dataset and methods","content":"\u003cp\u003eThe analysis was applied to a balanced panel of 1861 livestock farms, which were observed for two consecutive years (2021\u0026ndash;2022). The data were extracted from the RICA database, Italian section of the EU Farm Accountancy Data Network (FADN) database collected by CREA. The FADN is a sample survey carried out in all the Member States of the European Union. It is the only harmonised European source on farm management. The information framework of the Italian FADN allowas analysis at the farm, sectoral and territorial levels.\u003c/p\u003e\u003cp\u003eThe FADN dataset is a comprehensive set of data that classifies specialised livestock holdings into five primary Type of Farming (ToF) categories and eleven orientations. This classification system enables both aggregate and sectoral analysis of the emission footprint in the livestock sector. With regard to the carbon footprint, the Italian FADN dataset collects information on the quantities of production factors used in the processes within the farm (natural processes, production methods, resource management), which are useful for estimating agricultural greenhouse gas emissions. Consequently, it is feasible to concentrate on the individual farm as a unit of analysis by readily collecting activity data and integrating the three distinct sectors (Agriculture, Land Use and Energy). This approach enables the analysis of the emission impacts attributable to the individual farm in a unified manner.\u003c/p\u003e\u003cp\u003eIn addition, the RICA database facilitates the calculation of emission intensity indicators with reference to economic and structural company parameters. Consequently, assessments are formulated in terms of the efficiency of mitigation practices. As of 2025, the Farm Sustainability Data Network (FSDN) is set to replace the Farm Accountancy Data Network (FADN). The FSDN will build on the FADN's legacy, expanding its scope to include not only farms' income and business activities, but also information on their environmental and social sustainability performance. With regard to environmental considerations, the FSDN will collate data pertaining to agricultural management practices and the utilisation and management of nutrients (manure management, ration), in addition to more detailed information on the use of plant protection products and antibiotics. The collection of data will also encompass aspects pertaining to the management of water resources, the consumption and production of energy, and the implementation of environmental certification schemes.\u003c/p\u003e\u003cp\u003eA multilevel model was adopted in order to account for the hierarchical structure of the data and to capture intra-sector and time variability through the inclusion of random intercepts. In order to enhance model convergence, ensure robustness, and reduce multicollinearity, particularly in instances where interaction terms are present (Gelman \u0026amp; Hill, \u003cspan citationid=\"CR17\" class=\"CitationRef\"\u003e2006\u003c/span\u003e), it was necessary to transform all continuous variables using a log-transformation and mean-centring process. The log-linear structure of the estimated model incorporates fixed effects for each cost item, in addition to random effects on the intercept per group and year.\u003c/p\u003e\u003cp\u003eFurthermore, a dichotomous variable was incorporated to identify farms managed by young people (i.e. the entrepreneur's age being less than 40 years), in conjunction with interactions with individual input components, to ascertain potential managerial disparities associated with youth entrepreneurship.\u003c/p\u003e\u003cp\u003eThe structure of the dataset, which incorporates repeated farm-level observations, is consistent with recent developments in panel data econometrics. These developments emphasise the importance of cross-sectional dependence and latent heterogeneity (Sul, \u003cspan citationid=\"CR42\" class=\"CitationRef\"\u003e2019\u003c/span\u003e). The theoretical underpinnings of these models substantiate the utilisation of multilevel frameworks, which possess the capacity to discern unobserved sectoral and temporal variations. The multilevel model employed in this study assumes the following specification:\u003cdiv id=\"Equ1\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ1\" name=\"EquationSource\"\u003e\n$$\\:{y}_{ijt}=\\:{\\varvec{x}{\\prime\\:}}_{ijt}\\varvec{\\beta\\:}+\\delta\\:{d}_{ijt}+({{\\varvec{x}}^{{\\prime\\:}}}_{ijt}\\times\\:{d}_{ijt})\\varvec{\\gamma\\:}+\\:{u}_{j}^{ToF}+{v}_{t}^{Year}+{\\epsilon\\:}_{ijt}$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e1\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere:\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{y}_{ijt}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe log-transformed and mean-centred carbon footprint emissions (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{CO}_{2}eq)\\)\u003c/span\u003e\u003c/span\u003e of farm \u003cem\u003ei\u003c/em\u003e, belonging to farming type \u003cem\u003ej\u003c/em\u003e, in year \u003cem\u003et\u003c/em\u003e;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{x}\\varvec{{\\prime\\:}}}_{\\varvec{i}\\varvec{j}\\varvec{t}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe \u003cem\u003eK\u003c/em\u003e-dimensional log-transformed and mean-centred input costs of farm \u003cem\u003ei\u003c/em\u003e, belonging to farming type \u003cem\u003ej\u003c/em\u003e, in year \u003cem\u003et\u003c/em\u003e;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\beta\\:}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe corresponding \u003cem\u003eK\u003c/em\u003e-dimensional vector of fixed effects parameters;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{d}_{ijt}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003edummy variable indicating the presence of a young farmer (\u003cem\u003efarmer\u0026thinsp;\u0026lt;\u0026thinsp;40 years old\u003c/em\u003e), and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\delta\\:\\)\u003c/span\u003e\u003c/span\u003e the associated parameter;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\varvec{x}\\varvec{{\\prime\\:}}}_{ijt}\\:\\times\\:\\:{d}_{ijt}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe interactions between input costs and the young farmer dummy;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cdiv id=\"Equa\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equa\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{\\gamma\\:}:\\:the\\:associated\\:K-dimensional\\:vector;$$\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{u}_{j}^{ToF}\\sim\\:N\\left(0,{\\sigma\\:}_{u}^{2}\\right)\\:\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003erandom intercept capturing unobserved heterogeneity across types of farming.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{v}_{t}^{Year}\\sim\\:N\\left(0,{\\sigma\\:}_{v}^{2}\\right)\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003erandom intercept capturing unobserved heterogeneity over time.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\epsilon\\:}_{ijt}\\sim\\:N(0,{\\sigma\\:}_{\\epsilon\\:}^{2})\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003eresidual error term.\u003c/p\u003e\u003c/p\u003e\u003cp\u003eIn accordance with Wooldridge (\u003cspan citationid=\"CR48\" class=\"CitationRef\"\u003e2010\u003c/span\u003e), when employing log-log models, estimated coefficients can be interpreted as elasticities. That is to say, the percentage change in the dependent variable resulting from a 1% change in the regressor. While this interpretation is strictly valid under the assumption of independence between the error term and the covariates, it remains a commonly accepted approximation when only mean-independence is assumed (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\text{{\\rm\\:E}}\\left[\\epsilon\\:|x\\right]=0\\)\u003c/span\u003e\u003c/span\u003e).\u003c/p\u003e\u003cp\u003eThe model can also be written as:\u003cdiv id=\"Equ2\" class=\"Equation\"\u003e\u003cdiv format=\"TEX\" class=\"mathdisplay\" id=\"FileID_Equ2\" name=\"EquationSource\"\u003e\n$$\\:\\varvec{y}\\:=\\varvec{M}\\theta\\:+\\varvec{U}\\varvec{u}+\\varvec{V}\\varvec{v}+\\epsilon\\:$$\u003c/div\u003e\u003cdiv class=\"EquationNumber\"\u003e2\u003c/div\u003e\u003c/div\u003e\u003c/p\u003e\u003cp\u003ewhere:\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{y}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(NT,1)\\)\u003c/span\u003e\u003c/span\u003e vector dependent variable of log-transformed mean-centred dependent variable (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{CO}_{2}eq\\)\u003c/span\u003e\u003c/span\u003e emissions), with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:N=1861\\)\u003c/span\u003e\u003c/span\u003e farms observed over \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:T=2\\:\\)\u003c/span\u003e\u003c/span\u003e years (2021 and 2022), for a total of \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:NT=3722\\)\u003c/span\u003e\u003c/span\u003e observations;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{M}=\\left[\\varvec{X}\\:\\varvec{d}\\:\\varvec{Z}\\right]\\)\u003c/span\u003e\u003c/span\u003e is the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(NT,2K+1)\\)\u003c/span\u003e\u003c/span\u003e matrix combining the fixed effects;\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\theta\\:}=\\left[{\\varvec{\\beta\\:}}^{{\\prime\\:}}\\:\\delta\\:\\:{\\varvec{\\gamma\\:}}^{\\varvec{{\\prime\\:}}}\\right]\\varvec{{\\prime\\:}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003eis the \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(2K+1,\\:1)\\)\u003c/span\u003e\u003c/span\u003e vector of fixed-effect coefficients combining \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\beta\\:}\\)\u003c/span\u003e\u003c/span\u003e, \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\delta\\:\\)\u003c/span\u003e\u003c/span\u003e, and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\gamma\\:}\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{X}\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(NT,K+1)\\)\u003c/span\u003e\u003c/span\u003e matrix of fixed effects covariates with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:K=4\\)\u003c/span\u003e\u003c/span\u003e variables (cost items);\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{d}\\)\u003c/span\u003e\u003c/span\u003e : \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:NT\\)\u003c/span\u003e\u003c/span\u003e-dimensional (dummy) vector for young farmers;\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{Z}\\:\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003ethe (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:NT,K\\)\u003c/span\u003e\u003c/span\u003e) matrix of interactions between inputs and the young farmer\u0026rsquo;s dummy;\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{U}\\:\\left(NT,G\\right)\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{V}\\:(NT,T)\\)\u003c/span\u003e\u003c/span\u003e: incidence matrices, with \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:G=11\\:\\)\u003c/span\u003e\u003c/span\u003eand \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:T=2\\)\u003c/span\u003e\u003c/span\u003e for the random effects related to the second-level aggregation in type of farming (ToF) and in time period (year), respectively;\u003c/p\u003e\u003cp\u003ewith stochastic assumptions:\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{u}\\mathcal{\\:}\\sim\\:\\mathcal{\\:}\\mathcal{N}\\left(0,\\:\\:{\\sigma\\:}_{u}^{2}\\:{\\varvec{I}}_{G}\\right)\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003erandom intercept capturing unobserved heterogeneity across types of farming\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{v}\\mathcal{\\:}\\sim\\:\\mathcal{\\:}\\mathcal{N}\\left(0,\\:{\\sigma\\:}_{v}^{2}\\:{\\varvec{I}}_{T}\\right)\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003erandom intercept capturing unobserved heterogeneity over time\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\varvec{\\epsilon\\:}\\mathcal{\\:}\\sim\\:\\mathcal{\\:}\\mathcal{N}\\left(0,\\:{\\sigma\\:}^{2}{\\varvec{I}}_{NT}\\right)\\)\u003c/span\u003e\u003c/span\u003e\u003c/strong\u003e\u003cp\u003eresidual error term.\u003c/p\u003e\u003c/p\u003e\u003cp\u003eThe aforementioned assumptions imply that all random effects and residuals are independently and identically distributed with constant variance and no cross-correlations.\u003c/p\u003e"},{"header":"3. Results and discussion","content":"\u003cp\u003eThis section reports the main findings of the analyses. The results are presented according to the classification of livestock farming activities by type of farming (ToF) which is structured in two levels: the main level (MToF, comprising five types) and the particular type of farming (PToF, comprising 11 types).\u003c/p\u003e\n\u003cdiv id=\"Sec4\" class=\"Section2\"\u003e\n \u003ch2\u003e3.1 Results of the model at level of Main Type of Farming\u003c/h2\u003e\n \u003cp\u003eThe application of the multilevel model to the classification of farms according to MToF provides a robust estimate of the relationship between external inputs and emission values. It is evident that all the coefficients of the economic variables are positive and highly significant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(p\\:\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e. The results (Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e) demonstrate that the elasticity of emissions in relation to the inputs is both positive and statistically significant. Specifically, an increase of 1% in feed and fodder costs results in an average increase of 0.134% in emissions \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{1}=0.134,p\\:\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e, while costs of water, electricity and fuel are associated with an average increase of 0.121% \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{3}=\\:0.121,\\:p\\:\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e. Mechanization expenditure shows the highest elasticity among all input items analysed: an increase of 1% is related to a 0.159% increase in CO₂eq. emissions \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{2}=\\:0.159,\\:p\\:\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e. Veterinary expenses have been found to have a positive effect, yet this effect is less pronounced in comparison to that of other input components. The estimated elasticity of veterinary expenses is 0.030% \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{4}=\\:0.034,\\:p\\:\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e.\u0026nbsp;\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab1\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eResults for MToF model estimation\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRandom effects\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eGroups\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eName\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eVariance\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eStd. Dev\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMain ToF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e(Intercept)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.447\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.668\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e(Intercept)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.024\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eResiduals\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.682\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.826\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eFixed Effects\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eVariable\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eCoeff.\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eStd. Error\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep-value\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.134\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMEC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.159\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWEF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.121\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVET\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.034\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.074\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.060\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFF \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMEC \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.978\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWEF \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.075\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVET \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.030\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.013\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.010\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eIn order to capture the moderating role of youth entrepreneurship in the relationship between farm input management and environmental impact, a dummy variable was introduced. This variable, \u0026quot;Young,\u0026quot; denoted enterprises managed by farmers under 40 years of age according to the EU definition (Article 50 of Regulation (EU) 1307/2013). The model specification incorporates interaction terms between this dummy and each of the four cost categories in order to estimate potential age-related differences in the emissions response. The findings of the model (see Table \u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e) demonstrate that the direct impact of the \u0026apos;Young\u0026apos; variable on the intercept is marginally significant, indicating a tendency for higher average levels of emissions in farms managed by young individuals, when input costs are taken into account \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(\\widehat{\\delta\\:}=0.074)\\)\u003c/span\u003e\u003c/span\u003e. However, it is in the interaction terms that significant managerial differences emerge.\u003c/p\u003e\n \u003cp\u003eAll interactions, with the exception of mechanization \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:(p=0.978)\\)\u003c/span\u003e\u003c/span\u003e, demonstrate that the presence of a young entrepreneur serves to moderate the relationship between costs and emissions, thereby reducing the emission elasticity. A significant discrepancy becomes apparent in the interaction with the WEF input \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{\\gamma\\:}}_{3}=\\:\\--0.075,\\:p\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e. In farms managed by young entrepreneurs, a 1% increase in these costs leads to a 0.046% rise in CO\u003csub\u003e2\u003c/sub\u003eeq. emissions, compared to a 0.121% increase observed in other farms. This finding corresponds to a 62% reduction in the emission impact, suggesting a significantly more efficient use of energy-related inputs in farms managed by younger farmers. A similar pattern is observed for feed and fodder (FF) costs. The elasticity of emissions with respect to the costs for feed and forage is significantly attenuated in young farms, with a reduction of 0.039 percentage points. In farms under the management of young entrepreneurs, a one per cent increase in fuel costs is associated with a 0.095 per cent increase in emissions \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{\\gamma\\:}}_{1}=\\:-0.039,\\:p\\:\u0026lt;\\:0.001)\\)\u003c/span\u003e\u003c/span\u003e, compared to 0.134% in other farms. This indicates a 29% reduction in emissions. The interaction with VET expenses demonstrates a significant discrepancy, exhibiting a 0.030-point reduction \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{\\gamma\\:}}_{4}=-0.030,\\:p\u0026lt;\\:0.010)\\)\u003c/span\u003e\u003c/span\u003e. This corresponds to an emission increase of merely 0.004% in younger farms, in contrast to 0.034% in other categories.\u003c/p\u003e\n \u003cp\u003eIt is important to note, however, that the interaction with mechanization costs (MEC \u0026times; YOUNG) was not statistically significant. The model demonstrates a more \u0026quot;efficient\u0026quot; approach to input management by entrepreneurs under 40, with regard to both feed and energy use and veterinary services. This approach could also be indicative of a greater focus on sustainability among young enterprises. Moreover, this finding indicates that the generational shift in agriculture, besides being favourable from a demographic perspective, particularly in inland regions, can serve as a strategic factor for the economic resilience of the sector. The promotion of youth entrepreneurship is therefore not only a demographic imperative, but also a key enabler of climate and ecological transition, contributing to the integration of environmental, economic, and social competitiveness.\u003c/p\u003e\n \u003cp\u003eThe random effect pertaining to the general category of farming returns a statistically significant intercept variance (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{\\sigma\\:}}_{u}^{2}=\\:0.447\\)\u003c/span\u003e\u003c/span\u003e), indicating that about 45% of the variability in CO₂eq.\u0026nbsp;emissions is attributable to structural differences between types of rearing, while the random effect of time (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{\\sigma\\:}}_{v}^{2}\\)\u003c/span\u003e\u003c/span\u003e\u0026lt; 0,001) is negligible.\u003c/p\u003e\n \u003cp\u003eThe random effects for each group of MToF \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{j})\\)\u003c/span\u003e\u003c/span\u003e are estimated to represent deviations from the population-level intercept, subsequent to the accounting for the fixed effects in the model. These reflect how each group differs from the global average expectation, conditional on the other predictors (see Fig. 1).\u003c/p\u003e\n \u003cp\u003eThe analysis of the estimated random effects for the five MToF reveals that poultry farms exhibit the strongest negative deviation from the overall intercept, suggesting lower conditional emission levels, net of fixed effects \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{1}=-1.399)\\)\u003c/span\u003e\u003c/span\u003e. This is probably due to a higher feed conversion efficiency and to the absence of enteric fermentation. Conversely, cattle farms demonstrate a positive deviation \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{4}=0.161)\\)\u003c/span\u003e\u003c/span\u003e, indicating above-average emissions when controlling for other predictors.\u003c/p\u003e\n \u003cp\u003eThe reduced environmental impact of specialised poultry farms, as evidenced by the findings of this study, could be further elucidated by examining the characteristics of this farming method and the features associated with the broiler chicken production contract between the processor, who is the owner of the chickens, and the breeder (Peng et al, \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe processor provides the production inputs, including chicks, feed, medicines, health services and technical assistance. Conversely, the financial burden of expenses such as electricity, labour, water, chicken coop maintenance, and disinfection are shouldered by the breeder (Taraman et al., \u003cspan class=\"CitationRef\"\u003e2023\u003c/span\u003e). Consequently, certain inputs utilised in the production process, notably feed, are not charged to the poultry holding but are regarded as off-farm (Wiedemann et al., \u003cspan class=\"CitationRef\"\u003e2012\u003c/span\u003e). This is exemplified by veterinary services, whose minimal impact on the carbon footprint might be attributable to vertical integration contracts.\u003c/p\u003e\n \u003cp\u003eWith regard to the pig farms, which demonstrate a random effect below the mean (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{u}}_{2}=-0.331\\)\u003c/span\u003e\u003c/span\u003e), earlier research findings indicated that their contribution to the formation of carbon footprint is predominantly associated with manure and its management. Pig rearing is also essentially intensive, and is concerned with forms of vertical integration, through contracts with suppliers of inputs and processing companies. These contracts effectively exclude current production costs, which therefore do not contribute to the formation of the carbon footprint at the enterprise level (Reimer, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe analysis indicates that sheep farms demonstrate positive deviations \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{5}=0.210)\\)\u003c/span\u003e\u003c/span\u003e with higher average emissions. These small ruminants, in conjunction with cattle, are responsible for the primary production of greenhouse gases, predominantly through enteric methane emissions, and, to a lesser extent, through coal and manure management systems (Recktenwald and Ehrhardt, \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e). According to Marino et al. (\u003cspan class=\"CitationRef\"\u003e2016\u003c/span\u003e), the socio-economic role of sheep and goat farming is crucial, in that it contributes to landscape management, ecosystem preservation and biodiversity conservation, whilst also providing niche products to the market. Consequently, there is a robust interest in evaluating and enhancing the environmental performance of the sheep farming sector. However, Opio et al. (\u003cspan class=\"CitationRef\"\u003e2013\u003c/span\u003e) claimed that global emissions from sheep are significantly higher than those from goats, due to discrepancies in meat production. Furthermore, goats have been shown to exhibit a lower emission intensity on average than sheep, due to their higher yields compared to sheep milk. It has been observed that goats on extensive farms exhibit elevated methane emissions; however, carbon sequestration has been demonstrated to substantially mitigate net emissions (Horrillo et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eIn support of the aforementioned evidence, the boxplots of the carbon footprint by MToF (Fig.\u0026nbsp;2) provide additional visual confirmation of the model estimates.\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003ePoultry farms have been found to display the lowest median and interquartile range of CO₂eq.\u0026nbsp;emissions, thus confirming the lower conditional impact that has been estimated by the model. Conversely, cattle and sheep farms exhibited wider distributions and higher median values, indicative of their positive random effects and elevated emission intensities. These visual deviations serve to reinforce the model\u0026apos;s findings. It is noteworthy that the sheep and cattle sectors exhibited higher median CO₂eq values in comparison to the poultry sector, in accordance with the estimated random effects. Similar distributional patterns occur in both years of observation.\u003c/p\u003e\n \u003cp\u003eThe standardized residual plot (see Fig.\u0026nbsp;3) demonstrates a symmetrical distribution around zero, thus indicating a satisfactory model fit. In addition, the negligible Pearson correlation coefficient (r̂) between the fitted values and residuals (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\widehat{\\rho\\:}\\:=\\:0.001\\)\u003c/span\u003e\u003c/span\u003e) substantiates the absence of linear dependence, thereby satisfying the assumption of uncorrelated residuals and supporting the validity of the model.\u003c/p\u003e\n \u003cp\u003eThe distribution of residuals (see Fig.\u0026nbsp;4) approximates a normal bell-shaped curve, thereby suggesting that the assumption of residual normality is reasonably satisfied. This finding lends further support to the appropriateness of the log-transformed linear model that was utilised in the analysis. The estimated model demonstrates a clear correlation between a 1% increase in input costs (including feed, energy, mechanization and veterinary care) and a statistically significant rise in CO₂eq emissions. The analysis of the data according to farm type reveals the underlying factors that are responsible for the observed differences. These factors are found to be specific to each sector, and are both physiological and managerial in nature. The findings of the present study are consistent with those of INEA (Coderoni et al., 2016), which identified ruminant operations, particularly those involving cattle and sheep, as the primary source of enteric methane emissions.\u003c/p\u003e\n\u003c/div\u003e\n\u003cdiv id=\"Sec5\" class=\"Section2\"\u003e\n \u003ch2\u003e3.2 Results of the model by Particular Type of Farming\u003c/h2\u003e\n \u003cp\u003eThe analysis at the more detailed PToF classification stage enabled the specification of the model that most accurately reflected the production complexity of the sector. The model results are set out in Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e.\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab2\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 2\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eResults for PToF model estimation\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRandom effects\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003cth align=\"left\"\u003e\u0026nbsp;\u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eGroups\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eName\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eVariance\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eStd. Dev\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eParticular ToF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e(Intercept)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.556\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.746\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYear\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003e(Intercept)\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.031\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eResiduals\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.679\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.825\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cstrong\u003eFixed effects\u003c/strong\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003ctd align=\"left\"\u003e\u0026nbsp;\u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eVariable\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eCoeff.\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003eStd. Error\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u003cem\u003ep-value\u003c/em\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.140\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.004\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMEC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.161\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.008\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWEF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.138\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.009\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVET\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.037\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.005\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eYOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.046\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.039\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.232\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eFF \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.041\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMEC \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.966\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eWEF \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.075\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.020\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e\u0026lt;\u0026thinsp;0.001\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eVET \u0026times; YOUNG\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e-0.027\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.012\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003e0.032\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eThe estimates of the coefficients of the economic variables fully confirm the observations made in the model by MToF (see Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e1\u003c/span\u003e), with all effects remaining statistically positive, highly significant and consistent (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). It is noteworthy that the elasticity with respect to mechanisation expenditure is once again the most pronounced \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{2}=0.161,\\:p\u0026lt;0.001),\\:\\)\u003c/span\u003e\u003c/span\u003efollowed by FF costs \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{1}=0.140,\\:p\u0026lt;0.001)\\)\u003c/span\u003e\u003c/span\u003e and by WEF costs \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{\\beta\\:}}_{3}=0.138,\\:\\:p\u0026lt;0.001)\\)\u003c/span\u003e\u003c/span\u003e. A 1% increase in input costs has been shown to result in an average rise in emissions of 0.161% for mechanisation, 0.140% for feed and fodder, and 0.138% for water, electricity, and fuel. Veterinary care expenses were found to be the least impactful, though still significant \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:({\\widehat{\\beta\\:}}_{4}=0.037,\\:p\u0026lt;0.001)\\)\u003c/span\u003e\u003c/span\u003e.\u003c/p\u003e\n \u003cp\u003eIn order to further explore the management heterogeneity related to farmers\u0026apos; age, we estimated the specification of the model with interactions by particular type of farming (see Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e2\u003c/span\u003e). The model incorporates random intercepts at both the particular ToF and year level, thereby facilitating the capture of unobserved heterogeneity within a highly disaggregated structure. The model demonstrates a higher explanatory capacity than previous specifications, as evidenced by the reduction in AIC (9142.636 versus 9193.424) and a 56% share of variance explained at the group level (Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe direct effect of the young entrepreneur variable on the intercept is not statistically significant, indicating that farms run by young farmers do not have average emissions that are significantly different from those produced by adult farms. However, interactions with the main production factors are highly significant, suggesting a differentiated management behaviour.\u003c/p\u003e\n \u003cp\u003eSpecifically, the negative interaction between FF costs and YOUNG \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{\\gamma\\:}}_{1}=-0.041,\\:p\u0026lt;0.001)\\)\u003c/span\u003e\u003c/span\u003e demonstrates a reduced impact of emissions relative to the increase for this input, thereby indicating a more efficient or technically advanced management of this item. A similar effect is observed for WEF costs, with a reduction in elasticity \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{\\gamma\\:}}_{3}=-0.075,\\:p\u0026lt;0.001)\\)\u003c/span\u003e\u003c/span\u003e. A similar effect is also observed for VET \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{\\gamma\\:}}_{4}=-0.027,\\:p\u0026lt;0.05)\\)\u003c/span\u003e\u003c/span\u003e. It is evident that no substantial interaction was identified with MEC input, a finding that is consistent with previous observations.\u003c/p\u003e\n \u003cp\u003eThe present study sets out to explore the relationship between WEF costs and CO₂eq.\u0026nbsp;emissions in adult-run farms. The findings reveal that an increase of 1% in WEF costs is associated with a 0.138% increase in CO₂eq.\u0026nbsp;emissions in farms run by adults, \u003csub\u003eco\u003c/sub\u003empared to only 0.063% in farms run by young entrepreneurs. This suggests a 54% lower emission elasticity for this input. In addition, an increase of 1% in FF costs results in a 0.140% rise in CO₂eq.\u0026nbsp;emissions in traditional farms as opposed to 0.099% in young-run ones. This indicates a 29.3% reduction in the emission impact. With regard to VET, the estimated coefficient decreases from 0.037% in adult-managed farms to just 0.010% in those led by young farmers, corresponding to a 73% reduction in emission impact.\u003c/p\u003e\n \u003cp\u003eThe field of youth management has been shown to employ more sustainable management strategies, maybe as a result of a greater propensity for the adoption of innovative technologies and optimised practices in the domains of animal nutrition, energy management and livestock health. These recommendations, which require further investigation, underscore the correlation between environmental sustainability and generational renewal in the livestock sector. They also bear pertinent consequences for policies that support ecological transition and innovation.\u003c/p\u003e\n \u003cp\u003eWith regard to the outcomes of the model incorporating interactions, the age of farmers exerts a less significant influence on the global average carbon footprint. However, this relationship becomes apparent in contexts pertaining to production inputs, with the exception of mechanisation. A survey of young farmers has revealed that the primary motivations for adopting genetic selection of the herd are the reduction of emissions and the enhancement of manure, fodder, pasture, and animal housing infrastructure management. It has been demonstrated that these individuals tend to prioritise research, knowledge sharing and innovation in the agricultural sector (G\u0026oacute;mez-Lim\u0026oacute;n et al., 2010).\u003c/p\u003e\n \u003cp\u003eThe stability of these results, despite the adoption of a more granular sectoral classification, highlights the robustness of the estimated economic relationships across models.\u003c/p\u003e\n \u003cp\u003eA comparison of information criteria (IC) reveals a marked enhancement in the quality of the model. The model based on the Particular Type of Farming demonstrates a lower AIC in comparison to the model based on the Main Type of Farming, thus indicating a superior degree of explanatory power in the more detailed classification. In a similar fashion, the log-likelihood undergoes an improvement from \u0026minus;\u0026thinsp;4584.700 to \u0026minus;\u0026thinsp;4559.300, thereby suggesting a superior fit to the observed data (see Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3\u003c/span\u003e). The incorporation of both marginal and conditional \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e values serves to provide supplementary evidence for the robustness and appropriateness of the multilevel structure that has been adopted (see Table\u0026nbsp;\u003cspan class=\"InternalRef\"\u003e3.1\u003c/span\u003e). As posited by Nakagawa and Schielzeth (\u003cspan class=\"CitationRef\"\u003e2013\u003c/span\u003e), the conditional \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e is defined as the proportion of variance explained by both fixed and random effects. Conversely, the marginal \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e is defined as the variance explained by fixed effects alone. The null model demonstrates that 40.8% of the variance can be attributed to group-level differences (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{R}^{2}}_{c}=\\:\\text{I}\\text{C}\\text{C}\\:=\\:0.408\\)\u003c/span\u003e\u003c/span\u003e), thereby validating the implementation of a hierarchical model structure (Gelman and Hill, \u003cspan class=\"CitationRef\"\u003e2006\u003c/span\u003e). In addition, the PToF model demonstrates superiority over the MToF model in terms of classification accuracy, as evidenced by a higher conditional \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{R}^{2}\\)\u003c/span\u003e\u003c/span\u003e (0.649 vs. 0.606), higher ICC (0.457 vs. 0.396), and lower AIC. These enhancements signify a more precise modelling of group-level heterogeneity and enhanced explanatory power of the more granular sectoral classification.\u0026nbsp;\u003c/p\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003ctable id=\"Tab3\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003e\u003cstrong\u003eIC measures of models\u003c/strong\u003e Information criteria and variance of random intercepts for the MToF and PToF models.\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAIC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eBIC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eLog-likelihood\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eVariance (ToF)\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMToF model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9193.424\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9268.089\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-4584.700\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.447\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePToF model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9142.636\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9217.300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e-4559.300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.562\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cdiv class=\"gridtable\"\u003e\n \u003cdiv class=\"colspec\" align=\"left\"\u003e\u0026nbsp;\u003c/div\u003e\n \u003ctable id=\"Tab4\" border=\"1\"\u003e\n \u003ccaption\u003e\n \u003cdiv class=\"CaptionNumber\"\u003eTable 3.1\u003c/div\u003e\n \u003cdiv class=\"CaptionContent\"\u003e\n \u003cp\u003eModel Comparison based on IC measures\u003c/p\u003e\n \u003c/div\u003e\n \u003c/caption\u003e\n \u003cthead\u003e\n \u003ctr\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eModel\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eAIC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eBIC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\varvec{R}}^{2}}_{\\varvec{c}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{{\\varvec{R}}^{2}}_{\\varvec{m}}\\)\u003c/span\u003e\u003c/span\u003e\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eICC\u003c/p\u003e\n \u003c/th\u003e\n \u003cth align=\"left\"\u003e\n \u003cp\u003eRMSE\u003c/p\u003e\n \u003c/th\u003e\n \u003c/tr\u003e\n \u003c/thead\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eNull model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11248.930\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e11273.818\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.408\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.000\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.408\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e1.087\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003eMToF model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9193.424\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9268.089\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.606\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.347\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.396\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.825\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd align=\"left\"\u003e\n \u003cp\u003ePToF model\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9142.636\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e9217.300\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.649\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.354\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.457\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd align=\"char\"\u003e\n \u003cp\u003e0.817\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n \u003c/table\u003e\n \u003c/div\u003e\n \u003cp\u003eAs with the MToF, the inclusion of random effects on production patterns allowed further structural differentiations between farms. The intercept variance associated with the PToF group is 56% (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{\\sigma\\:}}_{u}^{2}\\)\u003c/span\u003e\u003c/span\u003e = 0.562) which is higher than that estimated for the main ToF (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{\\sigma\\:}}_{u}^{2}\\)\u003c/span\u003e\u003c/span\u003e = 0.447). This indicates greater between-group heterogeneity when a more detailed classification is adopted. The random effect of time \u003cem\u003e(\u003c/em\u003e\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{\\sigma\\:}}_{v}^{2}\u0026lt;0.001)\\)\u003c/span\u003e\u003c/span\u003e remains negligible even in this specification.\u003c/p\u003e\n \u003cp\u003eThe estimated random effects represent the deviations of the specific intercept for each ToF group from the implicit mean of the model, providing a measure of the emissive intensity \u0026quot;net\u0026quot; of the observed economic components. Negative values indicate groups with below-average emission levels at the same cost; conversely, positive values indicate a higher propensity to emissions. The findings of the analysis indicate that broiler poultry farms, in addition to combined-type and laying-hen farms, demonstrate a significant negative impact. This finding is indicative of a trend towards lower emission intensity, which is presumably attributable to high food efficiency and concentrated production structures. In contrast, the analysis of beef and sheep farms has revealed a positive effect, suggesting a greater emission impact. This finding is further substantiated by the distribution of estimated random effects \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:\\left({\\widehat{u}}_{j}\\right)\\)\u003c/span\u003e\u003c/span\u003e, as illustrated in Fig.\u0026nbsp;5.\u003c/p\u003e\n \u003cp\u003eAs demonstrated in Fig.\u0026nbsp;1, the poultry farming sector is significantly below the overall mean. These deviations are indicative of structural differences in emission intensity across farming sectors, after controlling for all other covariates in the model. As demonstrated by the figure, the internal consistency of specialisations within each production orientation is highlighted. Poultry farms, which have been divided according to their production orientations, show relatively close values of random effects for meat \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{1}=-1.460)\\)\u003c/span\u003e\u003c/span\u003e. The findings of the study demonstrated a correlation between the utilisation of eggs \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{3}=-1.145)\\)\u003c/span\u003e\u003c/span\u003e and the production of a combined product \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{2}=-1.369)\\)\u003c/span\u003e\u003c/span\u003e, thus confirming a common production structure characterised by high feed conversion efficiency and the absence of enteric fermentation. Furthermore, the discrepancies in carbon footprint impacts between broiler and egg production are reflected in feed consumption, associated manure production, and material and energy use. For instance, the production of eggs incurs higher energy costs than that of broiler farms, in addition to generating greater quantities of manure. Consequently, the production cycle of broiler meat is comparatively brief, and the feed consumption and manure production per unit are at their lowest (Coderoni et al., 2013; Kyriazakis, \u003cspan class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe impacts of two types of farms belonging to the pig sector, related to rearing (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{u}}_{4}=-0.399\\)\u003c/span\u003e\u003c/span\u003e) and meat (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{u}}_{5}=-0.356\\)\u003c/span\u003e\u003c/span\u003e) are close (De Cuyper et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e), but a notably high value is graphically observed for mixed pig farms (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{u}}_{10}=0.340\\)\u003c/span\u003e\u003c/span\u003e). The latter case may be indicative of the presence of a small number of large-scale mixed pig farms within the sample. Such farms are distinguished by their high production intensity, which involves the consolidation of multiple production cycles. These farms are associated with elevated levels of feed input, energy consumption and manure output, consequently resulting in a more substantial carbon footprint.\u003c/p\u003e\n \u003cp\u003eIn the cattle sector, dairy and mixed farms demonstrate analogous yet modestly favourable outcomes (\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{u}}_{7}=0.042\\)\u003c/span\u003e\u003c/span\u003e and \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{\\widehat{u}}_{8}=0.098\\)\u003c/span\u003e\u003c/span\u003e, respectively), whilst meat-oriented farming exhibits distinct characteristics, manifesting the most pronounced positive impact of the entire distribution by PToF \u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{11}=0.371)\\)\u003c/span\u003e\u003c/span\u003e. It is evident that cattle farms have the greatest impact on the carbon footprint of agriculture, primarily due to enteric emissions from these ruminants. However, a distinction is made in the literature between dairy cows and beef farms. It is evident that beef cattle farms exert a significantly greater influence on the soil, given their requirement for a substantially larger area of land. In addition, it is important to note that emissions from dairy cattle are also related to their slaughter and the subsequent meat production (Levasseur et al., \u003cspan class=\"CitationRef\"\u003e2024\u003c/span\u003e; Beauchemin et al., \u003cspan class=\"CitationRef\"\u003e2011\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eSheep farms have also been observed to demonstrate a positive effect, albeit to a lesser extent than that observed for beef cattle. This observation is consistent with the physiology of small ruminants and the high incidence of enteric emissions characteristic of livestock farming systems.\u003c/p\u003e\n \u003cp\u003eFinally, the findings of this study demonstrate that goat farms\u003cspan class=\"InlineEquation\"\u003e\u003cspan class=\"mathinline\"\u003e\\(\\:{(\\widehat{u}}_{6}= -0.254\\)\u003c/span\u003e\u003c/span\u003e) remain below average, in accordance with the results obtained in the MToF model.\u003c/p\u003e\n \u003cp\u003eThe presented patterns confirm that our PToF classification effectively captures meaningful intra-group heterogeneity in emissive intensity. The analysis at the PToF level reveals internal differences between the various types of farms according to the destination of the product (e.g. milk versus meat, or eggs versus meat). This is likely to reflect higher structural emissions from slow-growing ruminants with lower feed yields, as documented in some technical reports (Coderoni and Bonati, \u003cspan class=\"CitationRef\"\u003e2013\u003c/span\u003e).\u003c/p\u003e\n \u003cp\u003eThe boxplots in Fig.\u0026nbsp;6 visually corroborate these findings, confirming the structural differences in emission intensity across PToF categories.\u003c/p\u003e\n \u003cp\u003e\u0026nbsp;\u003c/p\u003e\n \u003cp\u003eMedian emissions are consistently lowest in poultry broilers, combined poultry, and laying hens, in line with the model-based estimates of negative random effects for these specialisations. The distribution of carbon footprint values across PToF confirms the structural differences observed in the model estimates. Emission levels in poultry farms, especially broiler farms, combined poultry farms, and laying hen farms, are consistently the lowest. In contrast, beef cattle, sheep, and pig fattening farms exhibit wider variability and higher median values.\u003c/p\u003e\n \u003cp\u003eAmong cattle operations, dairy and combined systems demonstrate moderate emissions. These trends demonstrate stability throughout the observation period, aligning with the estimated random effects. This finding lends support to the robustness of the adopted sectoral classification. It is important to note that the combined pig category includes a very limited number of exceptionally large farms, which likely explains the presence of extreme values in the upper tail of the distribution. Due to their limited representativeness, these observations were not emphasised in the discussion, as they may distort the typical emission profile of this sector. The validity of the statistical assumptions was evaluated through model diagnostics. The scatterplot of standardized residuals by PToF (Fig.\u0026nbsp;7) displays a symmetrical cloud centred on zero with no discernible pattern, visually corroborating the assumption of homoscedasticity.\u003c/p\u003e\n \u003cp\u003eIn order to assess potential multicollinearity among the independent variables, Variance Inflation Factors (VIFs) were calculated. As demonstrated in Fig.\u0026nbsp;8, all values are found to be well below the commonly accepted threshold of 5, thus indicating a low level of multicollinearity. This finding lends further support to the reliability and stability of the coefficient estimates in the model.\u003c/p\u003e\n\u003c/div\u003e"},{"header":"4. Conclusions","content":"\u003cp\u003eThe results presented in this study may have implications from several perspectives.\u003c/p\u003e\u003cp\u003eInterventions aimed at curtailing the use of feed and enhancing energy efficiency represent pivotal mechanisms for mitigating the carbon footprint of Italian livestock farms, particularly in light of projections regarding future meat consumption by consumers. On the other hand, they could encourage the adoption of circular approaches, particularly in the most impactful areas of farming.\u003c/p\u003e\u003cp\u003eWith regard to the energy requirements for the management of livestock, the implementation of photovoltaic panels for self-consumption has the potential to enhance both the economic and environmental sustainability of livestock farms.\u003c/p\u003e\u003cp\u003eFurthermore, it is well established that adequate nutrition and hydration are pivotal to the successful husbandry of livestock, exerting a substantial impact on the quality of breeding stock and the final product. Ruminants have been identified as the primary contributors to pollution on livestock farms, with their digestive processes being the main source of methane emissions. As is evident, the feeding of animals is subject to stringent regulation in accordance with the principles of quality production. However, the recommended dietary regime involves the substitution of carbohydrates with unsaturated lipids, or alternatively, modifications to the ration, including a reduction in dietary fibre and an increase in the concentrate/forage ratio. These measures are designed to reduce enteric emissions and nitrogen excretion.\u003c/p\u003e\u003cp\u003eThe analysis indicates that mechanisation exhibits the highest degree of elasticity among all of the input items that were analysed in the econometric study, both in terms of the overall level and of the specific type of farming. Indeed, the mechanisation of farming processes has been demonstrated to increase energy costs (e.g. milking equipment; production of farm crops for livestock) and fuel consumption, as well as to cause carbon dioxide emissions. The propagation of digital technologies is conducive to the mitigation of deleterious effects and the enhancement of the efficiency of production processes in the context of livestock farming. This is attributable to the capacity of digital systems to encompass all technological stages of breeding, care and fattening.\u003c/p\u003e\u003cp\u003eFinally, the sustainable approach observed among young farmers is a further motivation for strengthening the existing policies for the recruitment of young people in agriculture. In this regard, policy measures could be implemented to encourage the education of young entrepreneurs within specific livestock courses and to facilitate the ongoing training of older entrepreneurs.\u003c/p\u003e\u003cp\u003eIn conclusion, it is evident that enterprises must prioritise sustainability as a strategic imperative, rather than considering it an optional extra.\u003c/p\u003e\u003cp\u003eA potential limitation of this study is the short temporal span of the panel, which includes only two years. However, using the dataset allowed us to capture substantial cross-sectional heterogeneity, thereby contributing to the robustness of the model. Despite the temporal constraint, we are confident in the goodness of the results, as they seem to comply with all underlying statistical assumptions.\u003c/p\u003e\u003cp\u003eFuture research could be useful to improve the comprehension of how production systems can evolve to address the pressing environmental and socio-economic challenges in the coming decades.\u003c/p\u003e"},{"header":"Abbreviations","content":"\u003ctable border=\"0\" cellspacing=\"0\" cellpadding=\"0\"\u003e\n \u003ctbody\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCarbon footprint\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCH\u003csub\u003e4\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMethan\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ecarbon dioxide\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCO\u003csub\u003e2\u003c/sub\u003eeq\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCarbon dioxide equivalent\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCREA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eCouncil for Agricultural Research and Economics\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEU\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eEuropean Union\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFADN\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFarm Accountancy Data Network\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eFeed and fodder\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eGHGs\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eGreenhouse gases\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eISPRA\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eItalian Institute for Environmental Protection and Research\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMEC\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMechanization\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMToF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eMain Type of Farming\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eNO\u003csub\u003e2\u003c/sub\u003e\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003enitrogen oxides\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003ePToF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eParticular Type of Farming\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eVET\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eVeterinary care\u0026nbsp;\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003ctr\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWEF\u003c/p\u003e\n \u003c/td\u003e\n \u003ctd valign=\"top\"\u003e\n \u003cp\u003eWater, energy and fuel\u003c/p\u003e\n \u003c/td\u003e\n \u003c/tr\u003e\n \u003c/tbody\u003e\n\u003c/table\u003e"},{"header":"Declarations","content":"\u003cp\u003e\u003cstrong\u003eEthics approval and consent to participate\u003c/strong\u003e\u003cp\u003eNon applicable\u003c/p\u003e\u003c/p\u003e\u003cp\u003e\u003cstrong\u003eConsent for publication\u003c/strong\u003e\u003cp\u003eNot applicable\u003c/p\u003e\u003c/p\u003e\u003ch2\u003eFunding\u003c/h2\u003e\u003cp\u003eNot applicable\u003c/p\u003e\u003ch2\u003eAuthor Contribution\u003c/h2\u003e\u003cp\u003eV.G. and M.B.F. conceptualize the study and wrote the initial draft; D.S. and L.R. designed the methodology; A.S. carried out the data analysis, V.G. and M.B.F. provided supervision. All authors reviewed and edited the manuscript.\u003c/p\u003e\u003ch2\u003eData Availability\u003c/h2\u003e\u003cp\u003eData that support the findings of this study come from FADN (Farm Accountancy Data Network), Italian section\u003c/p\u003e"},{"header":"References","content":"\u003col\u003e\u003cli\u003e\u003cspan\u003eAlem H (2023) The role of green total factor productivity to farm-level performance: evidence from Norwegian dairy farms. Agric Econ 11:2. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-023-00243-0\u003c/span\u003e\u003cspan address=\"10.1186/s40100-023-00243-0\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAllen MR, Shine KP, Fuglestved J et al (2018) A solution to the misrepresentations of CO₂ -equivalent emissions of short-lived climate pollutants under ambitious mitigation. npj Clim Atmos Sci 1:16. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1038/s41612-018-0026-8\u003c/span\u003e\u003cspan address=\"10.1038/s41612-018-0026-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eAndrade HJ, Vega A, Mart\u0026iacute;nez-Salinas A, Villanueva C, Jim\u0026eacute;nez-Trujillo JA, Betanzos-Simon JE, P\u0026eacute;rez E, Ibrahim M, Sep\u0026uacute;lveda CJ (2024) The carbon footprint of livestock farms under conventional management and silvopastoral systems in Jalisco, Chiapas, and Campeche (Mexico). Front Sustain Food Syst 8. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fsufs.2024.1363994\u003c/span\u003e\u003cspan address=\"10.3389/fsufs.2024.1363994\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBeauchemin KA, Janzen HH, Little SM, McAllister TA, McGinn SM (2011) Mitigation of greenhouse gas emissions from beef production in western Canada \u0026ndash; Evaluation using farm-based life cycle assessment, \u003cem\u003eAnim Feed Sci. Technol., 166\u0026ndash;167\u003c/em\u003e. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.anifeedsci.2011.04.047\u003c/span\u003e\u003cspan address=\"10.1016/j.anifeedsci.2011.04.047\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eBonnin D, Tabacco E, Borreani G (2021) Variability of greenhouse gas emissions and economic performances on 10 Piedmontese beef farms in North Italy. Agricultural Syst 194. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.agsy.2021.103282\u003c/span\u003e\u003cspan address=\"10.1016/j.agsy.2021.103282\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCastillo-D\u0026iacute;az FJ, Belmonte-Ure\u0026ntilde;a LJ, L\u0026oacute;pez-Serrano MJ et al (2024) Quantifying sustainability in the agri-food system: a comprehensive methodological framework and expert consensus approach. Agric Econ 12:20. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-024-00314-w\u003c/span\u003e\u003cspan address=\"10.1186/s40100-024-00314-w\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCoderoni S, Bonati D (eds) (2013) Impronta Carbonica Aziende Agricole Italiane \u0026ndash; ICAAI. Inea, Roma\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCoderoni S, Pontrandolfi A (eds) (2016) Analisi delle potenzialit\u0026agrave; e delle prospettive CREA\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eCheng M, McCarl B, Fei C (2022) Climate change and livestock production: a literature review. Atmosphere 13(1):140. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/atmos13010140\u003c/span\u003e\u003cspan address=\"10.3390/atmos13010140\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eContini C, Marotta G, Torquati B (2020) Multi-actor approaches to implement cooperative strategies and value chains based on sustainability. Agric Econ 8:7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-019-0147-3\u003c/span\u003e\u003cspan address=\"10.1186/s40100-019-0147-3\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ede Vries M, van Middelaar CE, de Boer IJM (2015) Comparing environmental impacts of beef production systems: A review of life cycle assessments. Livest Sci 178. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.livsci.2015.06.020\u003c/span\u003e\u003cspan address=\"10.1016/j.livsci.2015.06.020\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDe Cuyper C, Van linden V, Leen F, Aluw\u0026eacute; M, Van Meensel J, Millet S (2024) The impact of slaughter weight and sex on the carbon footprint of pig feed intake. Animal 18:4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.animal.2024.101119\u003c/span\u003e\u003cspan address=\"10.1016/j.animal.2024.101119\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eDogbe W, Wang Y, Revoredo-Giha C (2024) Nutritional implications of substituting plant-based proteins for meat: evidence from home scan data. Agric Econ 12:30. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-024-00324-8\u003c/span\u003e\u003cspan address=\"10.1186/s40100-024-00324-8\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eEurostat (2024) Air emissions accounts by NACE Rev. 2 activity. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://data.europa.eu/data/datasets/j3ywu7pahvs5s1ejw4a3fw?locale=en\u003c/span\u003e\u003cspan address=\"https://data.europa.eu/data/datasets/j3ywu7pahvs5s1ejw4a3fw?locale=en\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (Accessed 10 December 2024)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFAO (2024) Greenhouse gas emissions from agrifood systems. Global, regional and country trends, 2000\u0026ndash;2022. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://openknowledge.fao.org/items/74bfebdb-3272-4e6a-98f4-ee36c7146d44\u003c/span\u003e\u003cspan address=\"https://openknowledge.fao.org/items/74bfebdb-3272-4e6a-98f4-ee36c7146d44\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (Accessed 10 March 2025)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eFouda T, Awny A, Darwish M, Kassab N, Ghoname M (2021) Carbon footprint estimation in poultry production farms. Sci Pap Ser Manag Econ Eng Agric Rural Dev 21:4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.13140/RG.2.2.31479.34726\u003c/span\u003e\u003cspan address=\"10.13140/RG.2.2.31479.34726\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eGelman A, Hill J (2006) Data Analysis Using Regression and Multilevel/Hierarchical Models. Cambridge University Press, New York, NY, USA\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eG\u0026oacute;mez-Lim\u0026oacute;n JA, Sanchez-Fernandez G (2010) Empirical Evaluation of Agricultural Sustainability Using Composite Indicators. Ecol Econ 69:1062\u0026ndash;1075. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.ecolecon.2009.11.027\u003c/span\u003e\u003cspan address=\"10.1016/j.ecolecon.2009.11.027\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHerrero M, Thornton FK, Gerber P, Reid RS (2009) Livestock, livelihoods and the environment: understanding the trade-offs. Curr Opin Environ Sustain 1:2\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHorrillo A, Gaspar P, Rodr\u0026iacute;guez-Ledesma A, Escribano M (2024) Assessment of Greenhouse Gas Emissions and Carbon Sequestration in Dairy Goat Farming Systems in Northern Extremadura, Spain. Animals 14(23):3501\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eHuntingford C, Lowe JA, Howarth N, Bowerman N, Gohar LK, Otto A, Lee D, Smith S, den Elzen M, van Vuuren D, Millar R, Allen M (2015) The implications of carbon dioxide and methane exchange for the heavy mitigation RCP2.6 scenario under two metrics. Environ Sci Policy 51. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttp://dx.doi.org/10.1016/j.envsci.2015.03.013\u003c/span\u003e\u003cspan address=\"10.1016/j.envsci.2015.03.013\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eISPRA (2022) Le emissioni nazionali di gas serra. Settore Agricoltura \u0026ndash; anno 2022. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://emissioni.sina.isprambiente.it/inventario-nazionale/#Report\u003c/span\u003e\u003cspan address=\"https://emissioni.sina.isprambiente.it/inventario-nazionale/#Report\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e. (Accessed 10 December 2024)\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKang HY, Hwang YW, Lee JH, Cho SJ, Jeon YS, Kim NS, Kim J (2025) Evaluating the Greenhouse Gas Emissions Footprint of Chicken Meat Production in South Korea: A Life Cycle Perspective. Food Bioprod Process. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.fbp.2025.01.017\u003c/span\u003e\u003cspan address=\"10.1016/j.fbp.2025.01.017\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eKyriazakis I (2013) Quantifying the environmental impacts of UK broiler and egg production systems. Lohmann Inform 48:2\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLevasseur P, Mariotti F, Denis I et al (2024) Potential unexpected effects of meat reduction in diet: Could educational attainment influence meat substitution strategies? Agric Econ 12:4. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-024-00298-7\u003c/span\u003e\u003cspan address=\"10.1186/s40100-024-00298-7\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eLynch J, Cain M, Frame D, Pierrehumbert R (2021) Agriculture's contribution to climate change and role in mitigation is distinct from predominantly fossil CO2-emitting sectors. Front Sust Food Syst 4:518039. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3389/fsufs.2020.518039\u003c/span\u003e\u003cspan address=\"10.3389/fsufs.2020.518039\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMancilla-Leyt\u0026oacute;n JM, Morales-Jerrett E, Mu\u0026ntilde;oz-Vall\u0026eacute;s S, Mena Y (2023) A Comparative Analysis of Carbon Footprint in the Andalusian Autochthonous Dairy Goat Production Systems. Animals 13(18):2864. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ani13182864\u003c/span\u003e\u003cspan address=\"10.3390/ani13182864\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMarino R, Atzori AS, D'Andrea M, Iovane G, Trabalza-Marinucci M, Rinaldi L (2016) Climate change: Production performance, health issues, greenhouse gas emissions and mitigation strategies in sheep and goat farming. Small Rumin Res 135. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.smallrumres.2015.12.012\u003c/span\u003e\u003cspan address=\"10.1016/j.smallrumres.2015.12.012\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMazzetto AM, Bishop G, Styles D, Arndt C, Brook R, Chadwick D (2020) Comparing the environmental efficiency of milk and beef production through life cycle assessment of interconnected cattle systems. J Clean Prod 277:124108. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jclepro.2020.124108\u003c/span\u003e\u003cspan address=\"10.1016/j.jclepro.2020.124108\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eMottet A, Tempio G (2017) Global poultry production: current state and future outlook and challenges. World\u0026rsquo;s Poult Sci J 73:245\u0026ndash;256. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1017/S0043933917000071\u003c/span\u003e\u003cspan address=\"10.1017/S0043933917000071\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNakagawa S, Schielzeth H (2013) A general and simple method for obtaining R\u0026sup2; from generalized linear mixed-effects models. Methods Ecol Evol 4(2):133\u0026ndash;142. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.2041-210X.2012.00261.x\u003c/span\u003e\u003cspan address=\"10.1111/j.2041-210X.2012.00261.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eNotarnicola B, Tassielli G, Renzulli PA, Castellani V, Sala S (2017) Environmental impacts of food consumption in Europe. J Clean Prod 140:753\u0026ndash;765. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jclepro.2016.06.080\u003c/span\u003e\u003cspan address=\"10.1016/j.jclepro.2016.06.080\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eOpio C, Gerber P, Mottet A, Falcucci A, Tempio G, MacLeod M, Vellinga T, Henderson B, Steinfeld H (2013) Greenhouse Gas Emissions from Ruminant Supply Chains \u0026ndash; A Global Life Cycle Assessment. Food and Agriculture Organization of the United Nations (FAO), Rome\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePais DF, Marques AC, Fuinhas JA (2022) The cost of healthier and more sustainable food choices: Do plant-based consumers spend more on food? Agric Econ 10:18\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003ePeng B, Melnikiene R, Balezentis T et al (2024) Structural dynamics and sustainability in the agricultural sector: the case of the European Union. Agric Econ 12:31. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1186/s40100-024-00321-x\u003c/span\u003e\u003cspan address=\"10.1186/s40100-024-00321-x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRecktenwald EB, Ehrhardt RA (2024) Greenhouse gas emissions from a diversity of sheep production systems in the United States. Agric Syst 217. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.1016/j.agsy.2024.103915\u003c/span\u003e\u003cspan address=\"10.1016/j.agsy.2024.103915\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eReimer J (2006) Vertical Integration in the Pork Industry. Am J Agric Econ 88:234\u0026ndash;248. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1111/j.1467-8276.2006.00850.x\u003c/span\u003e\u003cspan address=\"10.1111/j.1467-8276.2006.00850.x\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eRice P, O'Brien D, Shalloo L, Holden NM (2017) Evaluation of allocation methods for calculation of carbon footprint of grass-based dairy production. J Environ Manag 202:311\u0026ndash;319. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jenvman.2017.06.071\u003c/span\u003e\u003cspan address=\"10.1016/j.jenvman.2017.06.071\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSau P, Arru B, Cisilino F, Furesi R, Pulina P, Madau FA (2023) Economic and Environmental Sustainability Trade-Off Analysis in Sheep Farming Using the Farm Accountancy Data Network Database. Sustainability 15(24):16621. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.3390/su152416621\u003c/span\u003e\u003cspan address=\"10.3390/su152416621\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSchroeder R, Aguiar L, Baines RN (2012) Carbon Footprint in Meat Production and Supply Chains. J Food Sci Eng 652\u0026ndash;665. 2\u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.17265/2159-5828/2012.11.005\u003c/span\u003e\u003cspan address=\"10.17265/2159-5828/2012.11.005\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSingh S, Pagare V (2024) Status and Scope of Mechanization in Animal Husbandry. RASSA J Sci Soc 6:75\u0026ndash;83. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.5958/2583-3715.2024.00012.X\u003c/span\u003e\u003cspan address=\"10.5958/2583-3715.2024.00012.X\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eSul D (2019) Panel Data Econometrics: Common Factor Analysis for Empirical Researchers. New. Routledge, York, NY\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTaraman S, Taşcıoğlu Y, Bulut OD (2023) Profitability and Cost Analysis for Contract Broiler Production in Turkey. \u003cem\u003eAnimals\u003c/em\u003e, 13(13), 2072. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003e10.3390/ani13132072\u003c/span\u003e\u003cspan address=\"10.3390/ani13132072\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eThoma G, Wiedemann S, Palhares J, Saleque M, Kanzaki M, Missohou A, Kyriazakis I, Gittins J, Burr J, Mitloehner F (2015) Greenhouse gas emissions and fossil energy use from poultry supply chains. Food and Agricultural Organization of the United Nations\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eTreml N, Rudi A, Schultmann F (2025) Evaluating environmental impacts of pork production: A life cycle assessment of seven case studies in Germany. J Clean Prod 503:145408. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1016/j.jclepro.2025.145408\u003c/span\u003e\u003cspan address=\"10.1016/j.jclepro.2025.145408\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eVelarde-Guill\u0026eacute;n J, Arndt C, G\u0026oacute;mez CA (2022) Carbon footprint in Latin American dairy systems. Trop Anim Health Prod 54:1\u0026ndash;7. \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://doi.org/10.1007/s11250-021-03021-6\u003c/span\u003e\u003cspan address=\"10.1007/s11250-021-03021-6\" targettype=\"DOI\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWiedemann S, McGahan E, Poad G (2012) Using life cycle assessment to quantify the environmental impact of chicken meat production. RIRDC Publication No. 12/029. (Rural Industries Research and Development Corporation (RIRDC), Australia). \u003cspan class=\"ExternalRef\"\u003e\u003cspan class=\"RefSource\"\u003ehttps://agrifutures.com.au/product/using-life-cycle-assessment-to-quantify-the-environmental-impact-of-chicken-meat-production/\u003c/span\u003e\u003cspan address=\"https://agrifutures.com.au/product/using-life-cycle-assessment-to-quantify-the-environmental-impact-of-chicken-meat-production/\" targettype=\"URL\" class=\"RefTarget\"\u003e\u003c/span\u003e\u003c/span\u003e\u003c/span\u003e\u003c/li\u003e\u003cli\u003e\u003cspan\u003eWooldridge JM (2010) Econometric Analysis of Cross Section and Panel Data, second edn. MIT Press, Cambridge, MA\u003c/span\u003e\u003c/li\u003e\u003c/ol\u003e"}],"fulltextSource":"","fullText":"","funders":[],"hasAdminPriorityOnWorkflow":false,"hasManuscriptDocX":true,"hasOptedInToPreprint":true,"hasPassedJournalQc":"","hasAnyPriority":false,"hideJournal":true,"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":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true},"keywords":"livestock sector, carbon footprint, Linear Mixed-Effects Regression Model, type of farming, FADN","lastPublishedDoi":"10.21203/rs.3.rs-6959911/v1","lastPublishedDoiUrl":"https://doi.org/10.21203/rs.3.rs-6959911/v1","license":{"name":"CC BY 4.0","url":"https://creativecommons.org/licenses/by/4.0/"},"manuscriptAbstract":"\u003cp\u003eThe livestock sector is a significant contributor to climate change, as it is a direct source of greenhouse gases and an indirect source of gas and carbon emissions through the processes of land use and feed production. The objective of the present study is to contribute to the ongoing debate concerning the economic and environmental performance of the specialised livestock sector. The study analyses the relationship between economic variables associated with the management of livestock farms and the value of the carbon footprint. The latter is a variable representing environmental sustainability at the level of individual farms. The analysis is based on data extracted from the Italian section of the EU Farm Accountancy Data Network (FADN) database and through the implementation of a multilevel mixed-effects regression model on a balanced panel dataset. The correlation between utilisation of agricultural inputs and emissions demonstrated a positive elasticity, which proved to be statistically significant. The analysis by farming types indicates that the most significant environmental impact is attributed to beef cattle farming, while poultry farming is found to be the least contributing factor to the carbon footprint of production units. Interactions with the age of farmers reveal that, all other things being equal, farms run by young people do not have significantly different average emissions compared to traditional farms. However, interactions with the main production factors are highly significant, suggesting differential carbon footprint performances according to the type of inputs.\u003c/p\u003e","manuscriptTitle":"Investigating the Carbon Footprint of Italian Specialized Livestock Farms and its Drivers","msid":"","msnumber":"","nonDraftVersions":[{"code":1,"date":"2025-07-25 09:25:36","doi":"10.21203/rs.3.rs-6959911/v1","editorialEvents":[{"type":"communityComments","content":0}],"status":"published","journal":{"display":true,"email":"[email protected]","identity":"researchsquare","isNatureJournal":false,"hasQc":true,"allowDirectSubmit":true,"externalIdentity":"","sideBox":"","snPcode":"","submissionUrl":"/submission","title":"Research Square","twitterHandle":"researchsquare","acdcEnabled":true,"dfaEnabled":false,"editorialSystem":"","reportingPortfolio":"","inReviewEnabled":false,"inReviewRevisionsEnabled":true}}],"origin":"","ownerIdentity":"c51a25db-95b7-4de5-a5b0-6d4891886949","owner":[],"postedDate":"July 25th, 2025","published":true,"recentEditorialEvents":[],"rejectedJournal":[],"revision":"","amendment":"","status":"posted","subjectAreas":[],"tags":[],"updatedAt":"2026-04-30T13:53:18+00:00","versionOfRecord":[],"versionCreatedAt":"2025-07-25 09:25:36","video":"","vorDoi":"","vorDoiUrl":"","workflowStages":[]},"version":"v1","identity":"rs-6959911","journalConfig":"researchsquare"},"__N_SSP":true},"page":"/article/[identity]/[[...version]]","query":{"redirect":"/article/rs-6959911","identity":"rs-6959911","version":["v1"]},"buildId":"8U1c8b4HqxoKbykW_rLl7","isFallback":false,"isExperimentalCompile":false,"dynamicIds":[84888],"gssp":true,"scriptLoader":[]}