Land Use Policy 126 (2023) 106558 Contents lists available at ScienceDirect Land Use Policy journal homepage: www.elsevier.com/locate/landusepol Modelling land use planning: Socioecological integrated analysis of metropolitan green infrastructures Joan Marull a, *, Roc Padró b, María José La Rota-Aguilera a, Joan Pino c, Annalisa Giocoli d, Jacob Cirera d, Núria Ruiz-Forés a, Francesc Coll a, Tarik Serrano-Tovar a, Raúl Velasco-Fernández a, e a Barcelona Institute of Regional and Metropolitan Studies, Autonomous University of Barcelona, 08193 Bellaterra, Spain b Department of Economy and Economic History, Autonomous University of Barcelona, 08193 Bellaterra, Spain c Center for Ecological Research and Forestry Applications, Autonomous University of Barcelona, 08193 Bellaterra, Spain d Redaction Service of the Barcelona Metropolitan Area Master Plan, Avinguda Josep Tarradellas 2-6, 08029 Barcelona, Spain e Institute of Environmental Science and Technology, Autonomous University of Barcelona, 08193 Bellaterra, Spain A R T I C L E I N F O A B S T R A C T Keywords: Land use planning of global metropolis is essential to meet the socioecological challenges of the next decades. Rural land use This paper aims to contribute to sustainable land use policies by proposing a socioecological integrated analysis Landscape planning of metropolitan green infrastructures, applying this landscape-metabolism model to the Urban Master Plan of the Agroecological transition Barcelona Metropolitan Area. The paper assesses the multiple functions and services of the green infrastructure Green infrastructure Multicriterial analysis in four land use scenarios and two types of agricultural management. The results show that the trending scenario Ecosystem services of applying the current land use planning would have a negative impact in the ecological functioning of the landscape, affecting biodiversity and the provision of ecosystem services to society. The results also show that certified organic farming is not enough to overcome some trends of industrial agriculture as low energy efficiency or greenhouse gas emissions. Finally, the results show different interactions between social metabolism and landscape ecology, as changes in the form of metabolism affect the functioning of the landscape, while changes in land cover specially affect resource use. Therefore, deeper changes are needed in land use policies that consider not only land covers, as has traditionally been done, but also metabolic flows to promote agroecological tran­ sitions towards more sustainable metropolitan green infrastructures under climate change scenarios. 1. Introduction Although agriculture has been the basis of subsistence for our societies, current agricultural systems have reached a critical transition point in The United Nations’ population estimate of 9.8 billion people their performance, environmental impacts, and energy patterns (Ging­ inhabiting the planet by 2050 is accompanied by a growth factor of 2–3 rich and Krausmann, 2018). Specifically, agricultural and food systems on the global energy and materials demand (Krausmann et al., 2008). are directly related to five of the nine planetary boundaries (Steffen We are reaching planetary limits, facing challenging global climate et al., 2015): global climate change, land-system changes, biosphere change scenarios (IPCC, 2022) and biodiversity loss, among others integrity, freshwater use, and biogeochemical flows. The predominantly (Steffen et al., 2015). As never before, a global consensus calls for the resource-intensive agricultural systems have impacted biogeochemical urgent need to change nature-society relationships and promote socio­ cycles and degraded soils (Tilman et al., 2011). Agro-industrial systems ecological transitions towards more sustainable land uses and related have also been closely related to biodiversity losses worldwide metabolisms in human-transformed landscapes (Tilman et al., 2002). (Tscharntke et al., 2005). Since the Green Revolution, agriculture went One essential dimension of social metabolism, both historically and at from being a provider to a net energy consumer (Pelletier et al., 2011). present, relies on agricultural systems. Inside these systems, energy, Today agri-food systems are responsible for nearly one-third of the materials, and information constantly enter, exit, and recirculate, not world’s greenhouse gas emissions (Crippa et al., 2021), mainly origi­ only altering but ultimately shaping the territories (Font et al., 2020). nated by the production procedures (i.e., land use change, fertilizer * Corresponding author. E-mail address: joan.marull@uab.cat (J. Marull). https://doi.org/10.1016/j.landusepol.2023.106558 Received 20 July 2022; Received in revised form 16 December 2022; Accepted 19 January 2023 Available online 23 January 2023 0264-8377/© 2023 Elsevier Ltd. All rights reserved. J. Marull et al. Land Use Policy 126 (2023) 106558 applications, irrigation, and machinery use). effect of each of its elements on its structure, functioning, and impact on Metropolitan areas are key cases to study these sustainability chal­ the metropolitan system are still under development (Xu et al., 2018; lenges. Today, near to 55.3% of the world’s population (4220 million Darvishi et al., 2022; La Rota-Aguilera and Marull, forthcoming). In this people) live in urban areas, and it is estimated that by 2030, 5167 article we will assess the contribution of agriculture to metropolitan million people (60.4% of the world’s population) will live in a city of at sustainability, and the possibilities that it offers to transform food sys­ least 500,000 inhabitants (UN, 2019). Urban growth poses great chal­ tems to mitigate climate change and biodiversity loss (FAO, 2019). The lenges that involve the peri-urban and rural matrix in which they are objective of this article is to evaluate the contribution of agriculture to usually integrated. These issues are especially concerning food security the sustainability of metropolitan areas, through the development of a (Satterthwaite et al., 2010), climate change mitigation and adaptation Socioecological Integrated Analysis (SIA) that enrich our capacity to (Demuzere et al., 2014), water quality and availability (Madrid et al., propose green infrastructure scenarios for sustainable land planning 2013), waste management (Chen, 2007), and habitat degradation and purposes. fragmentation (Fischer and Lindenmayer, 2007), compromising their ecological functionality and its ability to provide multiple ecosystem 2. Methodology services (Riley et al., 2003; McDonald et al., 2013; Liu et al., 2016). Under this scenario, metropolitan agriculture can play various roles both 2.1. Case study as a driver or mitigator of unsustainability (Cattaneo et al., 2018). Therefore, comprehensive, scientifically supported and socially viable The case study includes the whole Barcelona Metropolitan Area land use management and planning of metropolitan areas, is essential to (BMA), formed by 36 municipalities, a total area of 63,611 ha, and a meet the socio-ecological challenges of the next decades (Darvishi et al., population of 3.3 million inhabitants (IDESCAT, 2019). According to the 2020; Marull et al., 2021; Mendoza-Beltran et al., 2022; La latest land cover map (CREAF, 2016), open spaces are still the majority Rota-Aguilera et al., 2023). This implies the consideration of a myriad of (55%), divided between forests and woodlands (42%), agricultural areas interactions between ecological, economic, social, cultural, and tech­ (8%), pastures (3%), and other open spaces (2% river corridors, wet­ nological perspectives along the urban-rural gradient (Padró et al., lands). The remaining 45% of the BMA includes compact and disperse 2020). Specially, incorporating the agrarian systems to metropolitan urban areas, and transport infrastructures (Fig. 1). The study includes all landscape planning through the notion of the ‘green infrastructure’ the current open spaces, defined by the planning proposal as (Benedict and McMahon, 2002), as one specific example of nature-based non-developable urban land, as well as the urban park network, due to solutions to biological conservation, ecosystem service supply and ter­ their contribution in both structural and functional aspects of the green ritorial resilience, while warranting social and economic benefits for its infrastructure, as is also included in the Urban Master Plan (PDU) (AMB, population, improving climate change adaptation and mitigation (Maes 2019). The unit of analysis are cells of 500 × 500 m, obtaining a total of and Jacobs, 2017). 2326 cells covering the entire BMA. The scientific consensus states that a systemic change in the global food system is crucial to achieving the Paris Climate Agreement and the 2.2. Planning scenarios Sustainable Development Goals: only replacing non-renewable energy sources is not enough (Clark et al., 2020). As well, European strategies The analyses were carried out on four theoretical land use planning such as Farm to Fork or the Biodiversity Strategy call for a new agenda scenarios defined by the PDU (Fig. 2). Current scenario (S0): represents transforming these green infrastructures (European Commission, 2020). the current land cover situation of the BMA (CREAF, 2016), it offers the Although the concept of green infrastructure is widely recognized in baseline diagnosis to draw comparisons with the land planning sce­ urban development plans, the existing methodologies for assessing the narios. Trending scenario (S1): it exemplifies the strict application of the Fig. 1. Major land covers in the Barcelona Metropolitan Area (BMA). 2 J. Marull et al. Land Use Policy 126 (2023) 106558 Fig. 2. Theoretical scenarios defined by the Urban Master Plan (PDU) of the Barcelona Metropolitan Area (BMA) used in the Socioecological Integrated Analysis (SIA) model. current urban planning of all the metropolitan municipalities, implying transition) and a landscape ecology approach. The structural delimita­ an increase in urbanized areas and urban parks. Alternative scenario tion (such as the minimum surface area of a category) was adapted to the (S2): it considers that there is a transformation of the planned urban specific conditions of the BMA. Table 1 summarizes the 6 categories and parks into agricultural areas as well as the cultivation of certain sectors 17 subcategories established for the metropolitan green infrastructure. of doubtful urban consolidation, thus increasing the agricultural surface area by 65%. Potential scenario (S3): it implies the possible recovery of 2.4. Modeling application to the open spaces’ proposal agricultural areas in the BMA and is based on the 1956 land cover map (it is considered to correspond to the most suitable areas for agriculture, To evaluate the socioecological contribution of the green infra­ where it is still possible), which represents an increase of more than structure to the BMA, the SIA model was applied. A detailed description 150% compared to the current agricultural area (Giocoli, 2017). of the SIA model can be found in Marull et al. (2021). This model in­ The four planning scenarios (Fig. 2) correspond to a greater or lesser tegrates social metabolism and landscape ecology analyses to assess how degree of urban development and in the application of the land uses different land use or agricultural management scenarios can alter the defined in the current urban plan, as well as the application of measures contribution made by the green infrastructure to the overall system. The for the permeabilization to ecological connectivity of current and research presented in this article considerably improves previous ver­ planned transport infrastructures. Additionally, each of these land sions of the model, including water consumption (as a limiting factor in planning scenarios (S0-S3) was evaluated using the SIA model under two agriculture), biomass appropriation (due to its importance for biodi­ types of agricultural management (conventional vs organic), considering versity) and greenhouse gas emissions (due to its impact on climate changes in the way natural resources and external inputs are used inside change). This landscape-metabolism model has six interrelated di­ agricultural and livestock systems (Padró et al., 2019). Conventional mensions (Fig. 3): A Metabolic efficiency; B Biodiversity conservation; C agriculture scenarios reproduced the current main management condi­ Landscape functionality; D Climate change; E Ecosystem services; and F tions reflected in the 2009 agricultural census. This census was updated Social cohesion. taking 2015 as a reference year. To simulate the scenarios of organic The SIA assessment requires linking the different dimensions of the management, the study followed the guidelines to produce certified socioecological system accounted (Fig. 3) with the indicators that organic food and animals established by the legislation of the European evaluate them (Table 2). The SIA modelling is fed by two fundamental Commission (834/2007, 889/2008, 1235/2008) and the Catalan sources of information: land covers and socio-metabolic flows. The Council of Organic Agricultural Production (CCPAE, 2017). The pa­ socio-metabolic flows are established among the four funds of the rameters required for organic agricultural management are defined ac­ agroecosystem: land uses, livestock, landscape, and society. The six di­ cording to the following criteria: total removal of the use of non-mineral mensions of the metropolitan green infrastructure are characterized by chemical fertilizers; removal of the consumption of chemical pesticides 10 interrelated indicators: and herbicides; and limited and regulated use of external inputs. A shift A1A Energy efficiency, which measures the amount of energy ob­ to organic agricultural management also has other biophysical impli­ tained in the agricultural areas according to the amount of external cations: reduction in agricultural yields per unit area or animal output energy invested (Tello et al., 2016). This agricultural metabolic balance per unit time (De Ponti et al., 2012; Seufert et al., 2012), labor re­ determines the input and output energy flows of each crop per land unit quirements and machinery use (DAAR, 2007), the amount of unhar­ of analysis. This allows observing the relationships between the different vested biomass generated, or the management of animal manure fund components, especially between livestock and agricultural uses, (Villamor et al., 2020). and the degree of dependence on external inputs to produce useful biomass, while meeting the needs of soil nutrients replenishment, ani­ 2.3. Delimitation of the categories of open spaces mal nutrition and reproduction of human labor. A1B Water consumption, which estimates the theoretical amount of To define the categories of open spaces, the structural components of water used by the metropolitan green infrastructure (Madrid et al., the green infrastructure were considered. These components work 2013). It is based on meteorological data on observed precipitation, together to maintain a network that supports ecological processes and average temperature, and reference evapotranspiration. To calculate the provide ecosystem services (Benedict and McMahon, 2002). The defi­ potential evapotranspiration, maps of plant species coefficients are nition started from the proposal of categories of the PDU preliminary made. The calculation of all the estimated flows for each point in the reports (AMB, 2019). The proposal presented here is based on a delim­ territory is done: interception, effective precipitation, irrigation, real itation of the current structural characteristics of the open spaces, evapotranspiration and drainage. considering both the dynamics of the metropolis (i.e., the spaces in A1C Biomass appropriation, which estimates the percentage of current 3 J. Marull et al. Land Use Policy 126 (2023) 106558 Table 1 Criteria for the definition of categories and subcategories of open spaces proposed for the green infrastructure of the Barcelona Metropolitan Area (BMA). B1 Habitat suitability, which assesses the conditions for biodiversity based on the landscape complexity and energy efficiency of agricultural practices (Marull et al., 2019). It is based on the landscape patterns and processes and the energy flows of agrarian metabolism that imprint those land patterns and intervenes in those processes. The fraction of biomass left available for non-domesticated ecological trophic chains is obtained from the socio-metabolic balance of the agroecosystem (the unharvested fraction of the photosynthetic net primary production). C1 Landscape complexity, which values the patterns and processes of the landscape from the heterogeneity of the land covers and its ecological connectivity (Marull and Mallarach, 2005). It assesses the functional landscape structure from the conjunction between the habitat diversity, that determines the landscape patterns, and the ecological connectivity, that determines landscape processes. D1 Greenhouse gas emissions, which assesses the contribution of agriculture to global warming through greenhouse gas emissions (Aguilera et al., 2015a; b). It is based on the machinery, biomass burning, fertilizers, pesticides, water, greenhouses, carbon fixed in the soil and nitrogen balances. The calculations are limited to detailing Fig. 3. Socioecological Integrated Analysis (SIA) of the metropolitan green emissions from agriculture. They therefore do not consider the rest of infrastructure of Barcelona, considering six interrelated dimensions. open spaces of the green infrastructure, nor the role played by livestock (beyond the provision of excreta which is considered part of the emis­ net primary productivity that has been appropriated by human activities sions of fertilizers and N2O). (Haberl et al., 2007). It involves the appropriation of plant biomass, E1A Nutrients recirculation, which estimates the amount of phos­ through modification of land covers and use of natural resources (agri­ phorus that re-circulates within the agricultural system in interaction culture, forestry, etc.). This appropriation affects the amount of energy, with other uses and grazing (Marco et al., 2018). It is calculated in the form of biomass, available to other species. In this way, the considering the flows previously measured in the socio-metabolic biomass appropriation has great impacts on the landscape structure assessment. There are specific values of recirculation of nutrients for (land covers) and its functioning (biogeochemical cycles) to a degree each crop, all of them obtained from their corresponding N-P-K bal­ that can exceed its carrying capacity. ances. Since these balances have been accounted for the main 4 J. Marull et al. Land Use Policy 126 (2023) 106558 Table 2 Socioecological Integrated Analysis (SIA) main indicators for the land planning and agricultural management (conventional vs organic) evaluation of scenarios. macronutrients using homogeneous data, the most limiting nutrient (in the key factors characterizing the green infrastructure. Subsequently, an our case P) has been selected to account for soil nutrients recirculation. Exploratory Factor Analysis (EFA) was performed to visualize the dis­ E1B Carbon stock, which measures the total amount of carbon stored tribution of the indicators and the categories of the open spaces con­ to open spaces as soils, roots and air-filled structures (Doblas-Miranda cerning the main factors defined by the SIA. Finally, a Multiple Linear et al., 2013). The following fractions are considered: C in mineral soil, C Regression Analysis (MLRA) was applied to evaluate the contribution of in organic layer of the soil, and C in roots of plants and their woody each open space category to the SIA indicators. In this way, it was aerial structures. C in mineral soil is obtained from cropland databases possible to verify whether the open space categories are useful to (http://www.icgc.cat/). For forest soils, a regression model has been identify differentiated functions and services of the metropolitan green developed to estimate the belowground C stock according to forest cover infrastructure and thus facilitate the land use planning and management (total aboveground C) and soil (mineral composition, texture and thick of the territory. elements). The organic horizon in forests is estimated from coefficients of C stock into mineral soil, and in croplands is estimated from above­ 3. Results and discussion ground C and roots, which depends on species, ages and spatial patterns. E1C Agricultural production, which reports the production of each 3.1. Characterization of open spaces for the different planning scenarios cover (horticultural, winter crops, dry and irrigated grassland, dry and irrigated fruit trees, olive trees and vineyards). It accounts for food The open spaces constituting the BMA’s green infrastructure production leaving aside forestry and livestock production to focus on currently cover 55% of its surface area (35,032 ha). The predominant the edible products of agricultural activity, and because livestock feed is open space category for the current scenario (S0) are Conservation core mainly imported from external territories. It is obtained from the sta­ areas (51.0% of the AMB’s surface), followed by Mosaic or transition tistical records of the average productions of each municipality and areas (24.0%), and Specialized agricultural areas (7.9%). Finally, with less agricultural land cover. surface area are Urban interstitial areas (6.2%), Agricultural interstitial F1 Agricultural Jobs, which characterizes the number of Agricultural areas (5.6%), and finally Riparian Areas (5.3%) (Table 3). Work Units (AWU) required to keep running the open spaces (Padró The Fig. 4 shows the maps of the structural delimitation of the green et al., 2017). It measures the potential of full-time equivalent agricul­ infrastructure for the S0, S1, S2 and S3 scenarios. The most relevant tural jobs that would be required in the agrarian spaces of the metro­ differences between the S0 Current and the S1 Trending scenarios are the politan area, which are used as a proxy to the jobs required by the substitution of many transition areas by urban parks and the loss of metropolitan agroecosystems. The AWU working days required by each rainfed crops from specialized agricultural areas. In the case of the S2 land unit of the different existing crops, as well as for livestock and Alternative scenario, some specialized agricultural areas appeared and forestry, have been estimated. there was an expansion of mosaic landscapes. Finally, the effect of the increasing agricultural land if the S3 Potential scenario is materialized in 2.5. Statistical analysis of the SIA application a massive expansion of specialized agricultural areas but also of land­ scape mosaics. The 10 SIA indicators (Table 2) and the distribution and extension of open spaces categories (Table 1) were calculated for each one of the 3.2. Application of the model in different planning and management 500x500m cells of the BMA. The SIA assessment at cell level allows a scenarios pairwise comparison of the indicators for each scenario and their sta­ tistically significant differences based on a bilateral test-t for each cell We analyze the impact of land planning (S0, S1, S2, S3) and agri­ (n = 2467). This allows to find how strategies on land use changes or cultural management scenarios (conventional vs organic) in the BMA shifting management can suppose different green infrastructure’s per­ through the 10 indicators of the SIA model (Table 4). Since A1B Water formances for each SIA dimension. Principal Components Analysis consumption and C1 Landscape complexity were calculated based on land (PCA) with the SIA results for each scenario was performed to identify covers, there are no differences between conventional and organic 5 J. Marull et al. Land Use Policy 126 (2023) 106558 Table 3 Transformation matrix from land covers to open space categories of the green infrastructure of the Barcelona Metropolitan Area (BMA) in the current scenario (S0). values within the same land planning scenarios. In this sense, A1B tends management (Table 4). Despite this increase, it should be noted that to increase in scenarios S2 Alternative and S3 Potential, as the extension strict compliance with organic production regulations does not ensure a of agricultural areas and urban parks increases. C1 also increases in significant overall improvement in the contribution of the green infra­ scenarios with more landscape mosaics, but this difference is only structure to the socioecological functioning of the metropolitan area. significantly lower concerning the scenario S1 Trend, indicating the This is the case with the moderate reduction of D1 Greenhouse gas potential loss of landscape functionality, as built-up areas and infra­ emissions, which only improve by an average of 18% when compared to structure increase, causing fragmentation and habitat loss. This can as conventional production. This reduction only come from agricultural well be seen through B1 Habitat suitability which accounts for the production, without involving the entire agri-food system including the biodiversity conservation dimension, and slightly improving in organic food industries, packaging, transport, storage, wholesale and retail. It management scenarios. also does not include changes in diet. It is important to highlight here While a transition to organic agriculture following only CCPAE that the shift to organic farming is not enough, and the necessary criteria allows for a higher degree of autonomy by improving E1A changes to get higher climate change mitigation and biodiversity im­ Nutrient recirculation and F1 Agricultural labor, the process is associated provements must go towards a systemic agroecological transition. with a decrease in E1C Agricultural production and A1A Energy efficiency Finally, in terms of F1 Agricultural jobs, the average increase is 24% (Table 4). Previous studies have reported agricultural yield decreases AWU (Table 4). A socioecological transition would imply an increase under organic management, but this yield gap could be reduced over from the 640 AWU currently estimated to 2400 AWU in the S3 Potential time precisely because of improving nutrient recirculation (Schrama scenario. This increase in the volume of workers is mainly explained by et al., 2018). Regarding the contribution to climate change, there is an the increase in surface area, but also by the change to organic agriculture overall reduction of D1 Greenhouse gas emissions when shifting to organic and, to a lesser extent, by the expansion of agriculture to areas with management, although this improvement is only significant for land above-average productivity in the area. planning scenarios where agricultural land covers increase (S2, S3). The significant decrease in A1A Energy efficiency in organic management can be explained by the decrease in productivity, which implies higher 3.4. Exploratory analysis of the factors involved in the model external inputs per unit of product. These external inputs dependence also leads to a rise in D1 Greenhouse gas emissions when the agricultural Based on the results obtained for the SIA indicators in each of the area increases. spatial planning and agricultural management scenarios, a PCA was performed separately for conventional and organic management to identify the differences (Table 5). Under conventional management, two 3.3. Insights for an agroecological transition main factors explain 68.8% of the total variance of the model. Factor 1 (37.8% of explained variance) groups the indicators associated with The effect of an organic transition would significantly reduce E1C ’metabolic flows and system efficiency’: A1A Energy efficiency, A1C Agricultural production, with a drop of 17% (Table 4). As already Biomass appropriation, D1 Greenhouse gas emissions, and E1C Agricultural mentioned, however, this drop is greater than the fall in external inputs, production. Factor 2 (31.0% of explained variance), groups the indicators which under the assumptions established for the organic scenario corresponding to ’structure and functioning of the territorial matrix’: require in some cases the import of organic fertilizers to support agri­ E1B Carbon stock, C1 Landscape complexity, and B1 Habitat suitability. The cultural production, thus causing a drop in A1 Energy efficiency of be­ PCA for scenarios under organic management was also defined by the tween 9% and 20% at the aggregate level. In terms of the capacity to E1A same two factors but with explained variances of 37.0% and 32.3%, Nutrient recirculation, it is interesting that the S1 Trending scenario results respectively. in a higher capacity to close the nutrient cycles (where more than 35% of To graphically represent the EFA, in relation to the categories of open the phosphorus comes from the same BMA). In this case, the reduction of spaces proposed for the green infrastructure of the BMA (Table 1), the agricultural area implies a more balanced relationship between livestock PCA was considered based on the eight agricultural planning and and agriculture, which facilitates to close more nutrient cycles and ex­ management scenarios (Fig. 2). In the EFA (Fig. 5), Specialized agricul­ plains this difference, due to the ability of agriculture to reuse agricul­ tural areas have a positive association with ’metabolic flows and system tural resources and vice versa. efficiency’ (Factor 1). Mosaic or transition areas, Interstitial areas, and On the other hand, the difference between scenarios implies an in­ Riparian areas also have a positive relationship with Factor 1, but less crease of more than 30% of E1A between conventional and organic than in the case of Specialized agricultural areas. Finally, Conservation core 6 J. Marull et al. Land Use Policy 126 (2023) 106558 Fig. 4. Structural delimitation of the green infrastructure of the Barcelona Metropolitan Area (BMA) for the S0 Current, S1 Trending, S2 Alternative, and S3 Poten­ tial scenarios. areas have a negative relationship with Factor 1, as do the build-up and These latter categories have a mixed role that, in relative terms, is more non-open space areas and Landscaped or restored areas. Regarding Con­ positive for the functioning of the landscape than for the achievement of servation core areas and Mosaic or transition areas, they are positively metabolic flows. This fact highlights that the mosaics have multiple and associated with Factor 2. In contrast, Interstitial areas and Riparian areas complex functions and services. These results corroborate the need to are related to Factor 2 to a much lesser extent, and Specialized agricul­ understand in-depth the contribution of the different open spaces to the tural areas have almost no effect. Finally, Landscaped or restored areas functionality and services that the green infrastructure provides, or and Conservation core areas are negatively related to Factor 2. should provide, to society. There is a clear triad in the positive effect of the factors obtained by Finally, in general, the EFA shows a better performance of organic the PCA, which defines a boundary of possibilities for improvement. scenarios on Factor 2 while a slight decrease on the performance on Conservation core areas are key for the ecological functioning of the Factor 1 (Fig. 5). The S1 Trending scenario would imply a decrease in the territorial matrix, Specialized agricultural areas for the metabolic flows, structure and functioning of the territorial matrix (Factor 2), and the S3 and Mosaic or transition areas for both factors at the same time (Fig. 5). Potential scenario would imply an increase in the metabolic energy and 7 J. Marull et al. Land Use Policy 126 (2023) 106558 Table 4 Socioecological Integrated Analysis (SIA) of the Barcelona Metropolitan Area (BMB) green infrastructure. Mean comparison of the SIA indicators* values between land planning scenarios (S0 - S3) and agricultural management (conventional vs organic). * SIA indicators: A1A Energy efficiency; A1B Water consumption; A1C Biomass appropriation; B1 Habitat suitability; C1 Landscape complexity; D1 Greenhouse gas emissions; E1A Nutrient recirculation; E1B Carbon stock; E1C Agricultural production; F1 Agricultural jobs. The results are based on the two-tailed t-test assuming equal variances with a significance level of 0.05. For each significant pair, the key under the category (a, b, c, d, e, f, g, h) shows up beneath the category with a major average value. Using the Bonferroni correction, tests have been adjusted for all pair-wise comparisons. biomass flows, system efficiency, and greenhouse gas emissions (Factor areas. The open spaces that use less water are Mosaic or transition areas 1), compared to the S0 Current scenario. and Conservation core areas. The explanatory capacity of Interstitial areas for this model is not significant. 3.5. Contribution of the open space categories to the values of the indicators 3.5.3. A1C biomass appropriation With an explanatory capacity of 47% (Table 6), the model indicates To analyze the contribution of the different categories of open spaces that the open areas with the highest biomass appropriation are to the functioning and services provided by the green infrastructure in Specialized agricultural areas and, to a lesser extent, Riparian areas. the BMA and its role in a possible socioecological transition, a MLRA is Interestingly, a low but positive relationship is found for Mosaic or performed. In this analysis, the SIA indicators (Table 2) are related transition areas, where the biomass uptake is not very high. When con­ separately to the categories of the open spaces (Table 1). In this way, a trasting the management scenarios, the biomass uptake is lower in first validation of the correspondence between the structural criteria of organic scenarios for all agricultural areas. This could be explained due the delimitation of the open spaces proposed (Fig. 4) and their func­ to productive yields decrease (see E1C), which the greatest effects of this tionality (Fig. 5) is obtained. These functionalities are analyzed for each reduction are for Specialized agricultural areas. of the contributing dimensions of the green infrastructure to the metropolitan system (Fig. 3). To facilitate the interpretation of the re­ 3.5.4. B1 habitat suitability sults (Table 6), it is shown: (i) the percentage of explained variance by The model can explain 87% of the total variance (Table 6). The the regression model undertaken (R2) in each case; (ii) the non- values obtained in organic would be higher than those obtained in standardized coefficients (ß); and (iii) the predicted values for homo­ conventional. Particularly interesting are Specialized agricultural areas geneous cells of a single open space category (value of the constant of that have a significant, although small, relationship in conventional the multiple regression model). The results obtained for each indicator scenarios, but not in organic scenarios. On the other hand, Mosaic or are explained below. transition areas would present higher values than the rest of the agri­ cultural areas, while Riparian areas and Interstitial areas also contribute 3.5.1. A1A energy efficiency moderately. This indicator reinforces the land-sharing conservation In conventional scenarios, the model explains 29% of the variance strategy according to which the intervened spaces with intermediated (Table 6), according to which Specialized agricultural areas contribute the disturbance levels (such as agroforestry mosaics), can provide favorable most, followed by Riparian areas and Interstitial areas. On the other hand, conditions for biodiversity. Landscaped or restored areas and constructed the predicted value for a Mosaic or transition areas cell would be 2.528 or non-open space areas have a sustained negative effect. (ß), thus being below all the other categories in which agriculture is present. If the comparison is made with the scenarios with organic 3.5.5. C1 landscape complexity management, the estimated values for Specialized agricultural areas are In this case, the explanatory capacity of the model is 82% (Table 6). lower than in the conventional scenarios. However, it is interesting to It can be observed that the effect of Landscaped or restored areas and note that in Mosaic or transition areas the predicted indicator is higher in constructed or non-open space areas reduce the complexity of the organic than in conventional management while is lower for the other landscapes. Although with negative effect, Core conservation areas, being ones. This result highlights how an ecological transition is particularly large homogeneous masses, imply a small decrease in complexity. favorable in those mosaic areas, which historically have maintained a Mosaic or transition areas and Riparian areas have no specific estimator, more complex and efficient functioning of natural resources. thus resulting statistically insignificant in this model. 3.5.2. A1B water consumption 3.5.6. D1 greenhouse gas emissions The explanatory capacity of the model is very high (85%) (Table 6). This indicator presents higher absolute values in the conventional The model shows that Specialized agricultural areas are the open spaces scenarios and has an explanatory capacity of around 81% of the total that consume more water in the BMA, followed by Landscaped or restored variance (Table 6). The predicted value is similar in the different 8 J. Marull et al. Land Use Policy 126 (2023) 106558 Table 5 Principal Component Analysis (PCA) of the Socioecological Integrated Analysis (SIA) under the different land planning and agricultural management scenarios. Conventional Eigenvalues Sum of the square saturations after rotation Component Total Variance (%) Ac. variance (%) Total Variance (%) Ac. variance (%) 1 4.505 45.049 45.049 3.780 37.798 37.798 2 2.371 23.713 68.762 3.096 30.965 68.762 3 1.017 10.167 78.929 4 0.753 7.530 86.459 5 0.428 4.282 90.741 6 0.337 3.366 94.108 7 0.292 2.921 97.028 8 0.172 1.723 98.751 9 0.064 0.643 99.393 10 0.061 0.607 100.000 Organic Eigenvalues Sum of the square saturations after rotation Component Total Variance (%) Ac. variance (%) Total Variance (%) Ac. variance (%) 1 4.626 46.259 46.259 3.698 36.985 36.985 2 2.301 23.007 69.266 3.228 32.281 69.266 3 1.099 10.992 80.258 4 0.704 7.039 87.297 5 0.376 3.764 91.061 6 0.350 3.503 94.564 7 0.282 2.822 97.386 8 0.166 1.656 99.043 9 0.055 0.547 99.590 10 0.041 0.410 100.000 categories, with the highest values being obtained in Specialized agri­ maximum in Mosaic or transition areas (with a 27% reduction in emis­ cultural areas, followed to a much lesser extent by Riparian areas, where sions contribution). Again, agroforestry mosaics present particularly crops are usually irrigated. Mosaic or transition areas and Interstitial areas favorable contributions to reduce pressures over the environment. follow with a slightly lower contribution to emissions. The scenarios in organic production have predicted values below those of conventional 3.5.7. E1A nutrient recirculation production, with the minimum difference being in Specialized agricul­ In general, the percentages of nutrient recirculation are higher in tural areas (with a 16% reduction in emissions contribution) and the organic scenarios, especially in Interstitial areas, Riparian areas, and 9 J. Marull et al. Land Use Policy 126 (2023) 106558 Fig. 5. Exploratory Factor Analysis (EFA) of the Barcelona Metropolitan Area (BMA) green infrastructure. Open space categories (circles), SIA indicators* (triangles), and land planning and agricultural management scenarios (rhombi) * * are displayed. * SIA indicators: A1A Energy efficiency; A1B Water con­ sumption; A1C Biomass appropriation; B1 Habitat suit­ ability; C1 Landscape complexity; D1 Greenhouse gas emissions; E1A Nutrient recirculation; E1B Carbon stock; E1C Agricultural production; F1 Agricultural jobs. * * Land Planning scenarios: S0 Current, S1 Trend, S2 Alternative, and S3 Potential, under conventional (_C) or organic (_O) management. Mosaic or transition areas (Table 6). The latter is again the most beneficial 3.5.10. F1 agricultural jobs in terms of nutrient recirculation in the face of ecological transition. The For this indicator, a model has been obtained with an explanatory behavior of the models is similar in conventional and organic produc­ capacity of around 24% of the total variance for the conventional and tion, with an explained variance of only 23%. Riparian areas obtained organic scenarios (Table 6). As in the case of E1C and D1, the jobs the higher estimated value regardless of management, followed by generated are much higher in the case of Specialized agricultural areas. It Mosaic or transition areas. In contrast, Specialized agricultural areas, due to is followed by Interstitial areas, Mosaic or transition areas, and Riparian the strong pressure on the use of natural resources, has the lowest areas. The relationship between agricultural jobs and the different values, although much lower than areas without agricultural uses, on agricultural areas is stronger and more positive in both organic and which this indicator was not calculated directly. conventional Specialized agricultural areas. In other agricultural areas, the number of jobs increases considerably (up to 30%) in organic 3.5.8. E1B carbon stock settings. The multiple regression model explains 81% of the total variance (Table 6). The high values of this indicator might be explained because 4. Conclusions changes in land covers mean the loss of an important part of the accu­ mulated biomass (both aerial and belowground) and, as in the case of The results of applying a Socioecological Integrated Analysis (SIA) C1, does not present differences between organic and conventional model in the Barcelona Metropolitan Area (BMA) reinforce the hy­ scenarios. In this case, the highest values are obtained in Conservation pothesis that an operational green infrastructure is an essential struc­ core areas, followed by Mosaic or transition Areas and, far behind, by tural element of the metropolitan system, as it can provide multiple Specialized agricultural areas. functions and services. The establishment of categories of open spaces allows deepening the understanding of the socioecological interrelation 3.5.9. E1C agricultural production between the different components of the green infrastructure. These This model, with an explained variance of 72% (Table 6), reveals interrelations are crucial to maintaining an adequate cohesion and statistically significant associations for the categories containing agri­ integration between the different types of open spaces and with the rest cultural uses. It is interesting to see how, obviously, the most productive of the metropolitan system, a fundamental issue for landscape and urban areas are Specialized agricultural areas, but then there are Riparian areas planning. Thus, each of these categories of open spaces has a differential that also present high productivity. However, the expected value of role that can be developed from a multi-criteria, multi-scale, and sys­ these areas not differ much from the predicted value in Mosaic or tran­ temic perspective, necessary to understand the sustainability of the sition areas and Interstitial areas. Likewise, the only significant difference metropolis. between conventional and organic scenarios is found in Specialized The creation of four theoretical planning scenarios for the Urban agricultural areas, thus showing that these would be the most affected by Master Plan (PDU) allows us to analyze the territorial and metabolic an ecological transition. consequences of adopting differential forms of planning and 10 J. Marull et al. Land Use Policy 126 (2023) 106558 Table 6 Multivariant Linear Regression Analysis (MLRA) for each of the Socioecological Integrated Analysis (SIA) indicators* , according to the land planning and agricultural management (conventional vs organic) scenarios, and for each of the open space categories* * of the Barcelona Metropolitan Area (BMA) green infrastructure. * SIA indicators: A1A Energy efficiency; A1B Water consumption; A1C Biomass appropriation; B1 Habitat suitability; C1 Landscape complexity; D1 Greenhouse gas emissions; E1A Nutrient recirculation; E1B Carbon stock; E1C Agricultural production; F1 Agricultural jobs. * * Open space categories of the metropolitan green infrastructure: CCA Conservation core areas; SAA Specialized agricultural areas; MTA Mosaic or transition areas; RIA Riparian areas; INA Interstitial areas; LRA Landscaped or restored areas; CNA Constructed or non-open space areas. management of the green infrastructure. The Trending scenario of as those that refer to metabolic flows. These agroforestry mosaics have applying the current land-use planning would have a very important an important potential for surface area increase in the Alternative and negative impact, especially in the dimensions more related to the Potential scenarios. Based on the statistical models’ results, the agro­ ecological functioning of the landscape, affecting biodiversity and the forestry mosaics counteract the observed losses in agricultural produc­ provision of ecosystem services to society. On the other hand, the tion and energy efficiency. On the other hand, the interstitial areas play Alternative and Potential scenarios, where agricultural land would be a more important role than is often considered by introducing land cover recovered, are particularly favorable in terms of the circularity of diversity that favors certain patterns and processes in the landscape. In metabolic flows, agricultural production, and job creation, but also summary, the integration among livestock-agriculture and forests is increasing greenhouse gas emissions. needed to reduce external reluctance and to improve agroecosystems All these land-use planning scenarios have been evaluated concern­ functionality that would entail better energy efficiency performance as ing agricultural management (conventional vs organic) to evaluate the well as further greenhouse gas emission reduction or nutrient synergies and tradeoffs associated with a possible ecological transition. recirculation. Organic agriculture would be very effective to enhance nutrient recir­ culation and generating new jobs but would be less effective in other Data Availability dimensions such as biodiversity conservation or reduction of greenhouse gas emissions. On the other hand, organic management would be Data will be made available on request. particularly negative in terms of energy efficiency and agricultural production. All in all, limiting oneself to complying with the regulations Acknowledgements for organic agricultural production (CCPAE) is not enough for an ecological transition to promote dimensions such as climate change This work has been funded by the Barcelona metropolitan area to mitigation or energy efficiency. To achieve a significant improvement obtain criteria and methods for the Urban Master Plan of the Barcelona on these dimensions, it would be necessary to consider agroecological metropolitan area, and has also supported by the LIFE Urban Greening and not only organic management. Further research should evaluate the Plans project, funded by the European Commission, the MA4SURE impact of possible agroecological transitions, considering landscape- PRIMA-H2020 project funded by the European Commission, and the metabolism models, under climate change scenarios. BIOLANDSCAPEs project funded by the Spanish Ministry of Science, To identify the contribution of the different open space categories to Innovation and Universities (RTI2018–093970-B-C32). the green infrastructure, a Multivariate Linear Regression Analysis was used to determine the extent to which the categories themselves explain References the SIA indicator values. We can highlight the particularly multifunc­ tional contribution of the agroforestry mosaics with a rich and hetero­ Aguilera, E., Guzmán, G., Alonso, A., 2015a. Greenhouse gas emissions from conventional and organic cropping systems in Spain. I. Herbaceous crops. Agron. geneous mix of land uses, which have a positive impact both concerning Sustain. Dev. 35, 713–724. https://doi.org/10.1007/s13593-014-0267-9. the dimensions related to the landscape ecological functioning, as well 11 J. Marull et al. Land Use Policy 126 (2023) 106558 Aguilera, E., Guzmán, G., Alonso, A., 2015b. Greenhouse gas emissions from Madrid, C., Cabello, V., Giampietro, M., 2013. Water-use sustainability in socioecological conventional and organic cropping systems in Spain. II. Fruit. tree Orchards Agron. systems: /multiscale integrated approach. BioScience 63 (1), 14–24. Sustain. Dev. 35, 725–737. https://doi.org/10.1007/s13593-014-0265-y. Maes, J., Jacobs, S., 2017. Nature-based solutions for Europe’s sustainable development. AMB, 2019. Avanç del Pla Director Urbanístic Metropolità. Available at https:// Conserv. Lett. 10, 121–124. https://doi.org/10.1111/conl.12216. urbanisme.amb.cat/pdu-metropolita/documentacio-avanc. Marco, I., Padró, R., Cattaneo, C., et al., 2018. From vineyards to feedlots: a fund-flow Benedict, M.A., McMahon, E.T., 2002. Green infrastructure: smart conservation for the scanning of sociometabolic transition in the Vallès County (Catalonia) 21st century. Renewable resource. J. (Autumn) 12–17. 1860–1956–1999. Reg. Environ. Change 18 (4), 981–993. Cattaneo, C., Marull, J., Tello, E., 2018. Landscape agroecology. The dysfunctionalities of Marull, J., Mallarach, J.M., 2005. A GIS methodology for assessing ecological industrial agriculture and the loss of the circular bioeconomy in the Barcelona connectivity: application to the Barcelona Metropolitan Area. Landsc. Urban Plan. 71 Region, 1956-2009. Sustainability 10. https://doi.org/10.3390/su10124722. (2–4), 243–262. CCPAE, 2017. Ecological agriculture statistical book. Retrieved from: http://www.ccpae. Marull, J., Herrando, S., Brotons, L., et al., 2019. Building on Margalef: testing the links org/docs/estadistiques/espanya2017.pdf. between landscape structure, energy and information flows driven by farming and Chen, J., 2007. Rapid urbanization in China: a real challenge to soil protection and food biodiversity. Sci. Total Environ. 674, 603–614. https://doi.org/10.1016/j. security. CATENA 69, 1–15. https://doi.org/10.1016/J.CATENA.2006.04.019. scitotenv.2019.04.129. Clark, M.A., Domingo, N.G.G., Colgan, K., et al., 2020. Global food system emissions Marull, J., Padró, R., Cirera, J., et al., 2021. A socioecological integrated analysis of the could preclude achieving the 1.5◦ and 2◦ C climate change targets. Science 370, Barcelona metropolitan agricultural landscapes. Ecosyst. Serv. 51, 101350 https:// 705–708. https://doi.org/10.1126/science.aba7357. doi.org/10.1016/j.ecoser.2021.101350. CREAF, 2016. Mapa de Cobertes del Sòl de Catalunya. McDonald, R.I., Marcotullio, P.J., Güneralp, B., 2013. Urbanization and Global Trends in Crippa, M., Solazzo, E., Guizzardi, D., et al., 2021. Food systems are responsible for a Biodiversity and Ecosystem Services. Springer. third of global anthropogenic GHG emissions. Nat. Food 2, 198–209. https://doi. Mendoza-Beltran, A., Padró, R., La Rota-Aguilera, M.J., et al., 2022. Displaying org/10.1038/s43016–021-00225–9. geographic variability of peri-urban agriculture environmental impacts in the DAAR, 2007. Department of Agriculture, Food and Rural Development. White book on Metropolitan Area of Barcelona: a regionalized life cycle assessment. Sci. Total the organic agri-food production in Catalonia, Barcelona. Environ. 858, 159519 https://doi.org/10.1016/j.scitotenv.2022.159519. Darvishi, A., Yousefi, M., Marull, J., 2020. Modelling landscape ecological assessments of Padró, R., Marco, I., Cattaneo, C., et al., 2017. Does Your Landscape Mirror What You land use and cover change scenarios. Application to the Bojnourd Metropolitan Area Eat? Long-Term Socio-Metabolic Analysis of a Local Food System in the Vallès (NE Iran). Land Use Policy 99, 105098. https://doi.org/10.1016/j. County (Spain, 1860-1956-2000). In: Frankova, E., Haas, W., Singh, S.J. (Eds.), In landusepol.2020.105098. search of sustainable local food systems: Socio-metabolic perspectives, edited by. Darvishi, A., Yousefi, M., Marull, J., et al., 2022. Modelling ecological scarcity Springer. considering the long-term interaction between human and nature in dry agricultural Padró, R., Marull, J., Giocoli, A., et al., 2019. Anàlisi Socioecològica Integrada: aplicació landscapes. Application in Qazvin (Iran). Ecol. Model. 472, 110106 https://doi.org/ al planejament del territori metropolità. Anuari Metropolità de Barcelona 2018. Del 10.1016/j.ecolmodel.2022.110106. barri a la metròpoli. Institut d’Estudis Regionals i Metropolitans de Barcelona, De Ponti, T., Rijk, B., Van Ittersum, M.K., 2012. The crop yield gap between organic and Bellaterra. conventional agriculture. Agric. Syst. 108, 1–9. Padró, R., La Rota-Aguilera, M.J., Giocoli, A., et al., 2020. Assessing the sustainability of Demuzere, M., Orru, K., Heidrich, O., et al., 2014. Mitigating and adapting to climate contrasting land use scenarios through the socioecological integrated analysis (SIA) change: multi-functional and multi-scale assessment of green urban infrastructure. of the metropolitan green infrastructure in Barcelona. Landsc. Urban Plan. 203, J. Environ. Manag. 146, 107–115. https://doi.org/10.1016/j.jenvman.2014.07.025. 103905 https://doi.org/10.1016/j.landurbplan.2020.103905. Doblas-Miranda, E., Rovira, P., Brotons, Ll, et al., 2013. Soil carbon stocks and their Pelletier, N., Audsley, E., Brodt, S., et al., 2011. Energy intensity of agriculture and food variability across the forests, shrublands and grasslands of Peninsular Spain. systems. Annu. Rev. Environ. Resour. 36, 233–246. https://doi.org/10.1146/ Biogeosciences 10 (12), 8353–8361. annurev-environ-081710-161014. European Commission, 2020. Communication—A Farm to Fork strategy for a fair, Riley, S.P.D., Sauvajot, R.M., Fuller, T.K., et al., 2003. Effects of urbanization and habitat healthy and environmentally-friendly food system (+Annex). https://eur-lex.europa. fragmentation on bobcats and coyotes in southern California. Conserv. Biol. 17, eu/legal-content/EN/TXT/?uri=CELEX:52020DC0381. 566–576. FAO, 2019. TAPE Tool for Agroecology Performance Evaluation 2019- Process of Satterthwaite, D., Mcgranahan, G., Tacoli, C., 2010. Urbanization and its implications for Development and Tape. Test version. food and farming. Philos. Trans. R. Soc. B: Biol. Sci. 365, 2809–2820. https://doi. Fischer, J., Lindenmayer, D.B., 2007. Landscape modification and habitat fragmentation: org/10.1098/rstb.2010.0136. a synthesis. Glob. Ecol. Biogeogr. 16, 265–280. https://doi.org/10.1111/j.1466- Schrama, M., de Haan, J.J., Kroonen, M., et al., 2018. Crop yield gap and stability in 8238.2007.00287.x. organic and conventional farming systems. Agric., Ecosyst. Environ. 256, 123–130. Font, C., Padró, R., Cattaneo, C., et al., 2020. How farmers shape cultural landscapes. https://doi.org/10.1016/j.agee.2017.12.023. Dealing with information in farm systems (Vallès County, Catalonia, 1860). Ecol. Seufert, V., Ramankutty, N., Foley, J.A., 2012. Comparing the yields of organic and Indic. 112, 106104 https://doi.org/10.1016/j.ecolind.2020.106104. conventional agriculture. Nature 485 (7397), 229–232. Gingrich, S., Krausmann, F., 2018. At the core of the socio-ecological transition: Steffen, W., Richardson, K., Rockström, J., et al., 2015. Planetary boundaries: guiding Agroecosystem energy fluxes in Austria 1830–2010. Sci. Total Environ. 645, human development on a changing planet. Science 347. https://doi.org/10.1126/ 119–129. https://doi.org/10.1016/j.scitotenv.2018.07.074. science.1259855. Giocoli, A., 2017. L’activitat agrària a l’àrea metropolitana de Barcelona: reptes i Tello, E., Galán, E., Sacristán, V., et al., 2016. Opening the Black Box of Energy oportunitats per al planejament urbanístic des d’una visió agroecològica. Guillem Throughputs in Agroecosystems: A Decomposition Analysis of Final EROI into Its Tendero (coord.) La ciutat agrària. Agricultura urbana i sobirania alimentària. Xarxa Internal and External Returns (the Vallès County, Catalonia c. 1860 and 1999). Ecol. de Consum Solidari. Aliança per la Sobirania Aliment. De. Catalunya 81–96. Econ. 121, 160–74. Haberl, H., Erb, K.H., Krausmann, F., et al., 2007. Quantifying and mapping the human Tilman, D., Cassman, K.G., Matson, P.A., et al., 2002. Agricultural sustainability and appropriation of net primary production in earth’s terrestrial ecosystems. Proc. Natl. intensive production practices. Nature 418, 671–677. https://doi.org/10.1038/ Acad. Sci. USA 104, 12942–12947. https://doi.org/10.1073/pnas.0704243104. nature01014. IDESCAT, 2019. Anu. Estad. De. Catalunya 2019. Tilman, D., Balzer, C., Hill, J., et al., 2011. Global food demand and the sustainable IPCC, 2022. Climate Change 2022: Impacts, Adaptation and Vulnerability. https://www. intensification of agriculture. Proceedings of the National Academy of Sciences of ipcc.ch/. the United States of America 108, 20260–20264. https://doi.org/10.1073/ Krausmann, F., Fischer-Kowalski, M., Schandl, H., et al., 2008. The global sociometabolic pnas.1116437108. transition: past and present metabolic profiles and their future trajectories. J. Ind. Tscharntke, T., Klein, A.M., Kruess, I., et al., 2005. Landscape perspectives on Ecol. 12, 637–656. agricultural intensification and biodiversity: ecosystem service management. Ecol. La Rota-Aguilera, M.J., Marull, J., 2023. Towards a Landscape-Metabolism Model for the Lett. 857–874. https://doi.org/10.1111/j.1461-0248.2005.00782.x. Tropical Andes. Application in the Metropolitan Region of Cali (Colombia). Environ. UN, 2019. World population prospects 2019, Department of Economic and Social Affairs. Sci. Policy (forthcoming). World Population. Prospects 2019. La Rota-Aguilera, M.J., Zapata-Caldas, E., Buitrago-Bermudez, O., et al. 2023. New Villamor, G.B., Kliskey, A.D., Griffith, D.L., et al., 2020. Landscape social-metabolism in criteria and methods for sustainable land use planning of metropolitan green food-energy-water systems: agricultural transformation of the Upper Snake River infrastructures in the Tropical Andes. Land Use Policy (forthcoming). Basin. Sci. Total Environ. 705, 135817 https://doi.org/10.1016/j. Liu, Z., He, C., Wu, J., 2016. The relationship between habitat loss and fragmentation scitotenv.2019.135817. during urbanization: an empirical evaluation from 16 world cities. PLoS One 11, Xu, X., Yang, G., Tan, Y., et al., 2018. Ecosystem services trade-offs and determinants in e0154613. China’s Yangtze River Economic Belt from 2000 to 2015. Sci. Total Environ. 634, 1601–1614. https://doi.org/10.1016/j.scitotenv.2018.04.046. 12