Ecosystem Services 51 (2021) 101350 Contents lists available at ScienceDirect Ecosystem Services journal homepage: www.elsevier.com/locate/ecoser Full Length Article A socioecological integrated analysis of the Barcelona metropolitan agricultural landscapes Joan Marull a, *, Roc Padró a, Jacob Cirera b, Annalisa Giocoli b, Manel Pons a, Enric Tello c a Barcelona Institute of Regional and Metropolitan Studies, Autonomous University of Barcelona, 08193 Bellaterra, Spain b Barcelona Metropolitan Area, Street C 62 num. 16-18, Zona Franca, 08040 Barcelona, Spain c Department of Economic History, Institutions, Policy and World Economy at the University of Barcelona, Faculty of Economics and Business, Diagonal Avenue 690, 08034 Barcelona, Spain A R T I C L E I N F O A B S T R A C T Keywords: To implement the Milan Urban Food Policy Pact while enhancing ecosystem services, the Barcelona Metropolitan Socioecological systems Authority has commissioned a socioecological assessment of the metropolitan agriculture. A Socioecological Resilient agroecosystems Integrated Analysis has been carried out to incorporate ecosystem services as fundamental elements of the new Periurban agriculture Master Land Use Plan. This analysis involves assessing the functioning of metropolitan agricultural landscapes Energy-landscape integrated analysis Metropolitan areas focusing on different dimensions: socio-metabolic efficiency (energy return on investment), biodiversity con­ Land-use planning servation, landscape functionality, global change, ecosystem services (supporting, regulation, provisioning) and social cohesion. After accounting indicators separately, they have been grouped through a principal component analysis into three factors: socio-metabolic flows, landscape functionality and system efficiency. An exploratory factor analysis has revealed trade-offs and synergies between these factors, with relevant implications for land- use policy. Finally, a correlation analysis has evaluated how the indicators interrelate among them. The results show that the improvement of complex socioecological systems requires new multi-criterial management where the different interrelated dimensions are jointly addressed. A new sustainability-oriented land-use planning combined with agricultural and environmental policies aimed at integrating farming with forestry and livestock activities would help reduce the dependence on non-renewable external energy inputs and, indirectly, the greenhouse gas emissions stemming from agri-food chains. 1. Introduction Metropolitan agricultural landscapes also represent the greenbelts needed for more efficient agri-food supply chains, as outlined in the Mapping and Assessing Ecosystem Services (MAES), an activity Milan Urban Food Policy Pact and the EU Farm to Fork Strategy included in the European Union’s (EU) Biodiversity Strategy for 2030, (Moragues-Faus and Morgan, 2014), and also deliver many regulatory aims to help inform policy decisions that affect the environment (e.g. and cultural ES provided by associated biodiversity (Haines-Young and within a framework of nature, biodiversity, climate change, agriculture, Potschin, 2010). Therefore, MAES can act as a useful framework for forestry and water policies). In this paper, we use Maes et al. (2020) to assessing the urban agroecosystems that are relevant components of the highlight Ecosystem Services (ES) as key parameters in urban planning overall set of Ecosystem Service Providers for city dwellers in metro­ and to identify priorities when restoring and fomenting the deployment politan areas. With this in mind, the new Urban Master Plan of the of green infrastructures (European Commission, 2014; Harrison et al., Barcelona Metropolitan Area, to be approved in 2021, has commis­ 2014). We propose that ‘metropolitan agricultural landscapes’ should be sioned from the Metropolitan Laboratory of Ecology and Territory of identified as a specific reference case since they represent the largest Barcelona an initial assessment of how urban agroecosystems function. surface area of non-urbanized open spaces in metropolitan areas. They These spaces are regarded as the most important element of the green usually consist of a mosaic of forests, farmland and pastures that exist as infrastructures in the metropolis that this plan aims to foster as natural protected areas and productive or abandoned agricultural areas Ecosystem Service Providers (Benedict and McMahon, 2002; Lafortezza (Maruani and Amit-Cohen, 2007). et al., 2013). * Corresponding author. E-mail address: Joan.Marull@uab.cat (J. Marull). https://doi.org/10.1016/j.ecoser.2021.101350 Received 16 March 2019; Received in revised form 26 July 2021; Accepted 28 July 2021 Available online 17 August 2021 2212-0416/© 2021 The Author(s). Published by Elsevier B.V. This is an open access article under the CC BY-NC-ND license (http://creativecommons.org/licenses/by-nc-nd/4.0/). J. Marull et al. Ecosystem Services 51 (2021) 101350 This assessment covers all the 36 municipalities included in the socioecological indicators, we obtained an Integrated Socio-Ecological Barcelona Metropolitan Area, in which 43% of the Catalan population Analysis (SIA) of the metropolitan agricultural landscape. The concep­ live in just 2% of its land surface at a population density of 236 inhab./ tual approach underpinning this multidimensional evaluation (SIA) km2. The study takes into account all the productive soils that remain, combines the views, methods and indicators of Ecological Economics which still cover 56% of this metropolitan area and act as Ecosystem (MEFA) and Landscape Agroecology (ELIA), together with Land-Use Service Providers in this metropolitan area’s green infrastructures Planning aimed at enhancing ES (Hansen and Pauleit, 2014). (Marull et al., 2010). A novel approach in the new Urban Master Plan This multidisciplinary assessment of the biophysical structures of was the inclusion of all abandoned areas that had been cultivated or metropolitan agriculture seeks to contribute to research on the mecha­ used for forestry in 1956 – but as yet are not occupied by urban- nisms of ES provision. This assessment highlights the key nexuses be­ industrial areas – as productive agroecosystems that could potentially tween the way farmers manage the land, the type of landscapes they be restored in the future. With this objective in mind, and under the create, the aboveground and belowground biodiversity they maintain, auspices of the Milan Urban Food Policy Pact signed and co-directed by and the provisioning, regulating and cultural ES provided as an end the Barcelona City Council, the MAES evaluation carried out by the product (Oteros-Rozas et al., 2019). This approach supports land-use Metropolitan Laboratory of Ecology and Territory of Barcelona focused planners in the identification of key drivers linking land-uses with ag­ on metropolitan agroecosystems. Consequently, the evaluation pre­ roecological processes to be able to improve their multifunctional nature sented here is a first step towards a broader MAES for the whole Bar­ and enhance their roles as Ecosystem Service Providers within metro­ celona Metropolitan Area. politan green infrastructures (Davies et al., 2006; de Groot et al., 2010; By focusing on urban agroecosystems as Ecosystem Service Pro­ Bastian et al., 2012). This approach considers not only land-use patterns viders, this study uses a novel way of exploring the role of ES in but also the ecological processes underpinning ES (Palmer and Febria, metropolitan areas. Although it is consistent with the cascade model of 2012; Maes et al., 2016) to explain how farming activities, despite the ES (Haines-Young and Potschin, 2018), the conceptual and spatial ecological disturbance they cause, can contribute to the functioning of framework of this study remains focused on the biophysical structures of complex bio-cultural landscapes capable of supplying provisioning as agroecosystems as providers of material and energy flows. Interacting well as regulating and cultural ES (Boumans et al., 2015; Hamann et al., with natural processes at landscape level, these flows give rise to 2015; Santos-Martín et al., 2019). different final ES that extend human well-being when they enter into the social and economic system from the environment (Potschin-Young 2. Conceptual approach et al., 2018; Costanza et al., 2017). The principal reason for proceeding in this way was the assessment commissioned by Barcelona Metropol­ 2.1. Socio-metabolic approach to metropolitan agriculture itan Area aimed at providing useful criteria and indicators for planning these metropolitan agricultural landscapes in different ways (e.g. Green infrastructure planning implies the adoption of a complex restoring, protecting and incentivizing) according to their current state socioecological perspective. The cascade model connects the func­ (e.g. abandoned, degraded, intensively cultivated or included in the tioning of the biophysical structures of ecosystems with human well- network of natural protected areas). The overall objective was to help being as a means of highlighting the sequence of causes and effects transform these areas into green infrastructures and enhance and pro­ that are transmitted from the ecological functioning of the land matrix to mote ES provision. It is clear that the Urban Master Plan cannot simply the satisfying of citizens’ needs (Tzoulas et al., 2007; Haines-Young and hope to preserve a series of green areas and assume that natural func­ Potschin, 2010; Haines-Young and Potschin, 2018; Baró et al., 2016; tioning will lead to an optimal provision of ES. Potschin-Young et al., 2018). However, this cascade approach is also Agroecosystems are natural ecosystems transformed by the material under debate since the marginalizing of ‘supporting’ (MEA, 2005) or and energy flows driven by human labour and information whose ‘habitat’ (TEEB, 2008; TEEB, 2010) ES for being only indirectly related objective is to increase food, fodder and energy supply. The key issue, to human wellbeing can lead to the underestimation of the importance however, is to what extent do increases in provisioning ES take place at of ecosystem functions such as biodiversity, soil fertility and primary the expense of regulating cultural ES. On the other hand, a synergistic production that underlie provisioning, regulating and cultural ES (Cos­ improvement of all ES could be possible (Hansen and Pauleit, 2014; tanza et al., 2017). TEEB, 2018). In 2005, the first Millennium Ecosystem Assessment report According to the fund-flow bio-economic approach, all biophysical brought this question to the fore by including intermediate supporting flows that sustain societal metabolism are provided by live funds that, in ES within its framework (MEA, 2005). In this paper, we do not discuss order to continue providing, need to be nourished and reproduced the consensus reached subsequently to suppress intermediate supporting (Georgescu-Roegen, 1971; Giampietro et al., 2011; Giampietro et al., ES (Potschin-Young et al., 2018); however, this consensus also means 2013). This approach leads to a circular bioeconomic view that assesses replacing these ES with an evaluation of the socioecological state of the the sustainability of socioecological systems in terms of the reproduc­ underlying biophysical structures, whose functioning leads to an ibility of the fund components of the biophysical structures of our so­ adequate level of provisioning, regulating and cultural ES. A good cieties (Tello et al., 2016; Gingrich et al., 2018; Marco et al., 2020). In socioecological state cannot be taken for granted, above all in agro­ agroecosystems, this means taking into account both the reproduction of ecosystems that are hybrid biophysical structures where society and live funds such as soil fertility, which requires the replenishment of nature constantly interact to co-produce different ES (Haberl et al., organic matter and nutrients (Maeder et al., 2002; Watson et al., 2006), 2004; Van der Ploeg, 2014; Gliessman and Engles, 2015). The question is and aboveground biodiversity, which requires differentiated habitats in how to assess the socioecological state of metropolitan agricultural heterogeneous landscapes with sufficient Net Primary Production free of landscapes in order to improve their capacity to provide ES. human appropriation (Loreau et al., 2003; Hole et al., 2005; Gabriel Thus, our study adopts the Material and Energy Flow Accounting et al., 2006; Tscharntke et al., 2012). Instead of an unidirectional, tiered (MEFA) of the social metabolism that takes place within agroecosystems relationship that goes from ecosystems to the ES received by society, the (González de Molina and Toledo, 2014; Guzmán and González de socioecological processes at stake should be understood and planned as Molina, 2017). We use it as the basis for carrying out an Energy- a set of interrelated biophysical cycles in which the live fund compo­ Landscape Integrated Analysis (ELIA) on how these matter-energy nents of the land matrix requires maintaining an investment of socio- flows interact with natural processes and create agroecological terri­ metabolic flows in order to maintain their ecological functioning (Ho tories with different capacities for hosting biodiversity and associated ES and Ulanowicz, 2005; Marull et al., 2016). (Altieri and Nicholls, 2012; Wezel et al., 2016; FAO, 2018). By This becomes even more important when agroecosystems and agri- combining quantitative values and geospatial layers with other food chains are taken into account to compensate for the lack of 2 J. Marull et al. Ecosystem Services 51 (2021) 101350 attention paid to urban agriculture from an ES point of view (Sandhu agro-ecosystems in metropolitan areas from a reproductive point of view et al., 2008; Tancoigne et al., 2014; TEEB, 2018). Adopting an eco-agri- (Padró et al., 2019). A SIA is a socio-metabolic-territorial assessment food system approach, ES research and planning have to consider how designed to be applied in land-use planning. It is fuelled above all by agricultural multi-functionality give rise to synergies and trade-offs databases on agricultural, forestry and livestock socio-metabolic flows between different dimensions and components of the biophysical and land-uses. Its nodal point is the idea that via farming society invests structures of the land matrix such as the live funds of agroecosystems, in a set of biophysical flows in the agricultural system to obtain ES which in turn can lead to different patterns of Ecosystem Service Pro­ indirectly. These ES can only be guaranteed by conserving the socio- viders (Hansen and Pauleit, 2014). The multifunctional approach to the metabolic agricultural flows that replicate a set of vital live funds such different ES provided by each landscape unit is an important step to­ as the agrarian community, livestock, soil fertility and the functional wards facilitating the integration of nature conservation from a land- landscape structure (Fig. 3). The closer the functioning of these funds to sharing approach (Perfecto and Vandermeer, 2010; Phalan et al., natural processes, the more sustainable the agroecosystem will be 2011), an agri-food transition towards agroecological territories and Gliessmann (1998). culturally meaningful foodscapes (Altieri and Nicholls, 2012; FAO, The preservation of the ecological integrity of the biophysical 2018; Moragues-Faus and Morgan, 2014), and land-use planning of structures that are Ecosystem Service Providers is a necessary condition green infrastructures from an ES perspective (Tzoulas et al., 2007; for the maintenance of regulating, provisioning and cultural ES over Burkhard et al., 2012; Hansen and Pauleit, 2014). time (MEA 2005; de Groot et al., 2010; Burkhard et al., 2012; Haines- Young and Potschin, 2018; Maes et al., 2020). However, this condition 2.2. Conceptual outline of the socioecological integrated analysis is often stated from a traditional nature conservation standpoint, which ignores how society has to intervene through eco-efficient forms of The SIA assessment presented here focuses on the multiple di­ natural resource management to maintain the ecological integrity and mensions of the contribution to social welfare made by agroecosystems functioning of biophysical structures whenever cultural landscapes are in metropolitan agricultural landscapes. The set of integrated indicators taken in account instead of natural ecosystems (Agnoletti, 2014; Agno­ generated enable strategic land-use planning to improve agroecosystem letti and Rotherham, 2015; Agnoletti and Emanueli, 2016). Given that functioning, make agroecosystems part of the green infrastructure, and agroecosystems are maintained due to a continuous society-nature help them move towards more sustainable agro-futures. This conceptual interaction, these cultural landscapes should be assessed from a socio­ approach is outlined in Fig. 2, adapted from Cardinale et al. (2012), in an ecological point of view (Antrop, 2006). In agricultural landscapes this attempt to highlight the society-nature interactions that take place in interaction takes place through the imprint of farm metabolism on the Fig. 1. Land-cover map of the Barcelona Metropolitan Area (2015). Source: Centre for Ecological Research and Forestry Applications (CREAF, https://www.creaf. uab.es/mcsc/). 3 J. Marull et al. Ecosystem Services 51 (2021) 101350 Fig. 2. Outline of the Socioecological Integrated Databases I: Analysis (SIA) carried out as a metabolic- Metabolic fluxes agroecology model applied to land-use planning feed by databases on agricultural socio-metabolic flows (I) and land-uses (II). Six dimensions are A. Metabolic accounted: A. Agricultural Metabolic Efficiency; B. Efficiency Biodiversity Conservation; C. Agroecology Land­ scape Functionality; D. Global Change; E. Ecosystem Services (regulation, provisioning, cultural); F. So­ cial Cohesion. It highlights the importance of the D. Global B. Biodiversity E. Ecosystem green infrastructure in the functional structure of the Change Conservation Services network between cities and agriculture within metropolitan areas. C. Landscape Functionality Databases II: Land uses F. Social Cohesion Fig. 3. Outline of the model run to carry out the Socioecological Integrated Analysis (SIA) according to the bioeconomy circular fund-flow approach adopted. Arrows show the main socio-metabolic flows of agroecosystems, and the ellipses the four live funds considered (society-agrarian community, livestock, soil fertility and landscape functional structure) that must be preserved to guarantee the sustainability of the socioecological system over time. Below are listed the primary models (MEEA, GHGA, ELIA, SAFRA, SERA, NBA, WMA) used to account for the different indicators (Table 1), according to the diverse di­ mensions to be finally assessed (Fig. 2). functional structure of the territory, which ends up affecting the Change, which range from biodiversity loss (Benedict and McMahon, ecological patterns and processes that take place there (Marull et al., 2002; Mell, 2009) to climate change mitigation and adaptation (Gill 2010). This embraces cultural ES, which includes many aesthetic and et al., 2012)). SIA can address these issues within an agrarian meta­ identity values that can be seen as a societal acknowledgement of the bolism framework by examining the characteristics of the biophysical bio-cultural heritage present in cultural landscapes (Tenberg et al., structure of the land matrix that supports biodiversity (landscape het­ 2012; Agnoletti and Emanueli, 2016), although there are also other erogeneity, unharvested Net Primary Production), maintains soil factors that, for example, characterize leisure and recreational services fertility (organic matter replenishment and nutrient cycling through (Plieninger et al., 2015). All these ES require the maintenance of the biomass reused), mitigates climate change (non-renewable energy ecological functions of cultural landscapes. Whether measured directly consumed as external inputs and carbon sequestration) and improves or indirectly, biodiversity is key in the functioning of green in­ the ecological status of water bodies (waste and pollution generated and frastructures (Mell, 2009); precisely because society plays a relevant water-use). It does so by using the set of models depicted in Fig. 3, role in the reproduction of the live funds that support the ecological although only some of these models and indicators are employed in this functioning of cultural landscapes, in addition to the ES obtained in article. return SIA includes job creation as a proxy for social cohesion. Sustainability oriented land-use planning has to link all scales from local to global (Lafortezza et al., 2013). This includes the role agricul­ tural landscapes play within metropolitan green infrastructures as a means of facing up to the current and future challenges of Global 4 J. Marull et al. Ecosystem Services 51 (2021) 101350 3. Methodological approach production, the yearbook of annual statistics on crops, fertilizers, the five-year statistics of phytosanitary products consumed, and the annual 3.1. Case study and data sources statistics for forestry and the management balances of cereals. We took from IDESCAT (https://www.idescat.cat/?lang=es) data on agricultural The Barcelona Metropolitan Area (63,611 ha) has 3.24 million in­ machinery in terms of ownership and, finally, used trade data from habitants spread out over 36 municipalities (Fig. 1). This SIA assessment DATACOMEX (http://datacomex.comercio.es/) to obtain the origin of was performed on the full set of open spaces mapped in 2015 by the foreign imports to calculate the embodied energy. CREAF digital land-cover map (https://www.creaf.uab.es/mcsc/). Aside from the built-up area that occupies approximately 44% of the Barce­ 3.2. Main indicators of the socioecological integrated analysis lona Metropolitan Area, forest and scrub predominate and occupy, respectively, in 27% and 15% of the total surface area of the remaining A SIA assessment links the different dimensions of the socio­ open spaces. Croplands cover 8% of the total surface area and are mainly ecological system (Fig. 2) to the indicators that evaluate them (Table 1). concentrated in the Llobregat river valley and delta, predominantly in To this end, a selection of these indicators is made in a hierarchical the form of horticulture (mostly in fields with some greenhouses). A few sequence: main indicators (transversal in all the applications of the patches of arable land, vineyards and arboriculture still form mosaics assessment), secondary indicators (variables according to the case of with forests in the Vallès plain and on the slopes of sparsely populated study), and specific indicators (that assess more acutely the relationship areas. There are also some meadows and pastures (3%) and, to a lesser between agricultural open spaces and built-up spaces). These indicators extent, fluvial corridors and wetlands, mainly concentrated in the Llo­ were obtained from the SIA modelling, which is fed by two fundamental bregat delta and representing 2 and 1%, respectively. sources of information: land-cover, and agricultural, forestry and live­ The SIA was calculated with the same grid of 500 × 500-m cells as stock socio-metabolic flows (Fig. 2). The socio-metabolic flows were used in the Urban Master Plan of the Barcelona Metropolitan Area. established between the four funds of the agroecosystem: i) land-uses; ii) Socio-metabolic flows were calculated using land-cover and farming livestock; iii) landscape functional structure; and iv) the agrarian com­ databases for agriculture, livestock, forestry and trade. Land surfaces munity and society at large (Fig. 3). We defined an integrated system of were taken from the DARPA (http://agricultura.gencat.cat/ca/inici), indicators to evaluate the contribution of each different dimension to the and the production and yields of the main crops in each municipality functioning of the agricultural landscapes that are to be promoted as from the DUN (http://agricultura.gencat.cat/ca/ambits/desenvolupa part of the overall green infrastructure in the metropolitan system. This ment-rural/declaracio-unica-agraria/) and SIGPAC (https://www.ma integrated assessment also allowed us to identify synergies and trade- pa.gob.es/es/agricultura/temas/sistema-de-informacion-geografica-de- offs between indicators. parcelas-agricolas-sigpac-/default.aspx) databases, which also provide farm and livestock census, and information on annual wildfires. We took 3.2.1. Energy socio-metabolic efficiency provincial data from the MAPA livestock survey (https://www.mapa. Socio-metabolic efficiency (A) is represented by the main indicator gob.es/en/), as well as data for dairy, egg, honey and wool A1 that accounts for the energy efficiency of metropolitan agriculture Table 1 Multi-criteria set of indicators used to evaluate quantitatively the contribution of the metropolitan green infrastructure in the provision of ecosystem services. Source: Metropolitan Laboratory of Ecology and Territory of Barcelona (LET). 5 J. Marull et al. Ecosystem Services 51 (2021) 101350 (Table 1). Energy efficiency (A1) measures the energy return on the renewable external energy inputs (D1), which measures in MJ per inputs spent (EROI), calculated on the basis of the final product obtained hectare the total amount of non-renewable external energy spent on per unit of external inputs originating from outside the agroecosystem maintaining forestry, livestock and agricultural production according to (EFEROI) (Tello et al., 2016). This agricultural socio-metabolic balance the socio-metabolic balance (Tello et al., 2015). Higher values of non- determines the input and output flows of each crop per land unit of renewable external energy consumed (D1) are associated with more analysis. This also allows us to observe the relationships between intensive industrial farming that depends on external fossil-fuel inputs different agroecosystem fund components, above all between livestock and therefore is more likely to generate more greenhouse gas emissions. and agricultural uses, and the dependence on external inputs for pro­ In the case of the energy flows from agriculture or forestry, their site- ducing biomass (provisioning ES), while bearing in mind the needs of specific georeferencing requires calculating their value for each land- soil nutrient replenishment, animal nutrition and the reproduction of use (machinery, fertilizers, greenhouse facilities, herbicides and pesti­ human labour. The energy efficiency value (A1) is determined in each cides, and energy consumption in water pumping). Conversely, for cell by the following formula: livestock the georeferencing of flows is measured by the energy costs of ∑i=k the transport of feed, together with the energy consumed in stables or FPi industrial feed-lots. Finally, costs of livestock maintenance are imputed A1 = ∑i=1i=k i=1 EIi to the land-covers where animal excreta are spread following the pro­ FP is the amount of final useful energy produced (MJ) from all cedure described in Marull et al. (2016). agricultural uses in the sample cell, regardless of its destination (human consumption or animal feed). EI is the investment in external inputs 3.2.5. Ecosystem services (artificial fertilizers, herbicides and pesticides, agricultural machinery, Biophysical integrity (support) of ecosystem services providers (EA) greenhouse facilities where they exist, water pumping, seeds and human are assessed by a principal indicator, soil nutrient recirculation (E1A). labour) of each agricultural use in the sample cell, based on the calcu­ Soil nutrient recirculation measures the relative amount of Phosphorus lation of the socio-metabolic energy balance. For external inputs, all the (P) that recirculates within the agricultural system in interaction with energy incorporated into the production processes is considered livestock uses. High values of E1A represent a higher capacity of the (embodied energy), together with the useful internal energy (enthalpy). farm system to deal with the closure of nutrient cycles. The recirculation of nutrients is calculated for each farm land-use by taking into account 3.2.2. Biodiversity conservation the flows previously calculated in the socio-metabolic balance (A1) ac­ Biodiversity conservation (B) is represented by the principal indi­ cording to the nutrient flows considered in the framework defined by cator B1 (energy-landscape integration in Table 1) that measures the González de Molina et al. (2010). Thus, there are specific values for the energy-information-landscape relationship (Marull et al., 2016). recirculation of nutrients in the specific cases of horticulture vegetable Energy-landscape integration (B1) evaluates the socio-ecological con­ production, greenhouses, arable land and irrigated herbaceous crops, ditions for hosting biodiversity (Marull et al., 2019), based on landscape irrigated and non-irrigated fruit trees, irrigated and non-irrigated olive patterns and processes (landscape complexity, C1) and the flows of trees, and vineyards, all obtained from their corresponding N-P-K bal­ agrarian metabolism that imprint the land-use patterns and intervene in ances. Since these balances account for the main macronutrients using the landscape ecological processes (energy reused, A2; and energy homogeneous data, the most limiting nutrient (in our case P) was redistributed, A3). The fraction of biomass left that is available for non- selected to account for soil nutrient recirculation (E1A). The units (%P) domesticated ecological trophic chains (A2) is obtained from the socio- from agricultural, forest and pasture lands in the Barcelona Metropolitan metabolic balance of the agroecosystem (the unharvested fraction of the Area divided by the total circulating in any land-use give the following photosynthetic Net Primary Production). Then, the distribution pattern formula: for these biomass flows (A3, energy redistributed) can be calculated. internal P cyclingi Once these secondary indicators are obtained, we calculate the energy- E1Ai = ⋅100 total P cyclingi landscape integration (B1) for each cell: √̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅̅ The origin of each source of fertilizers (biomass burial, animal 3 A2i ⋅A3i manure or synthetic fertilizer) is evaluated using the socio-metabolic B1i = ⋅C1i k nutrient balance (Marco et al., 2018) in the Barcelona Metropolitan Area. The origins of biomass burial and synthetic fertilizers are internal where k allows us to scale factor A2 ⋅ A3 to the maximum theoretical and external to the metropolitan agroecosystems, respectively. For an­ value. The values of B1 lie between 0 and 1. imal excreta, the percentage of livestock feed derived from biomass produced internally is estimated, from which the share of the amount P 3.2.3. Landscape functionality originating from the Barcelona Metropolitan Area agroecosystems can Landscape functionality (C) is represented by the principal indicator be obtained. (C1), which is the landscape complexity of agroecosystems (Marull The indicator of the Carbon stock (E1B) is selected to represent part et al., 2016). Landscape complexity (C1) evaluates the functional of the Regulating ecosystem services (EB). E1B measures the amount of landscape structure (patterns and processes) based on the nexus be­ total Carbon (C) stored in the vegetated open spaces of the agricultural tween land-cover heterogeneity (C2) determining landscape patterns, metropolitan system; the different fractions of the C stock are calculated and ecological connectivity (C3) determining landscape processes for two differentiated agricultural spaces, croplands and other farmland (Marull and Mallarach, 2005). Its value per cell can be calculated thus: uses (forest, scrubland and pastures). Four fractions are considered: C in ( C3i ) the mineral soil, C in the organic layer of the soil, and C in the roots of C1i = C2i + 10 /2 the plants and in their woody aerial structures. The accumulated C in herbaceous crops was not considered due to its low persistence. In The ecological connectivity (C3) is divided by 10 since it is the cropland areas the average depth of soil samples taken was 30 cm but maximum value that ecological connectivity can take; on the other was 100 cm for forest soils. The value of C in mineral soils was taken in hand, landscape heterogeneity (C2) ranges between 0 and 1 to allow a the case of croplands from the ICGC database (http://www.icgc.cat/). value for landscape complexity (C1) between 0 and 1 to be obtained. For forest soils, a regression model was developed to estimate the belowground C stock according to the characteristics of the forest cover 3.2.4. Global change (total aboveground C) and soils (mineral composition and texture) Global change (D) is represented by the principal indicator of non- (Doblas-Miranda et al., 2013). The organic horizon in forest soils was 6 J. Marull et al. Ecosystem Services 51 (2021) 101350 estimated from coefficients of the total C stock in the mineral soil. For and roots through factors used in other studies on climate change carried aboveground C, a map developed by CREAF was available for non- out in Catalonia, which vary according to the species, age and spatial agricultural uses, while for plant roots the proportion between the pattern of tree planting (Funes et al., 2015). aerial and the belowground C fractions was estimated after a literature Provisioning ecosystem services (EC) are assessed by the principal review. In the case of cropland, an estimate was made of aboveground C indicator of agricultural production (E1C) measured in weight units of Fig. 4. Socio-ecological Indicators: A1 Energy Efficiency (Energy Return to External Inputs –EFEROI); B1 Energy-Landscape Integrated Analysis (ELIA); C1 Land­ scape Complexity (Le); D1 Non-Renewable External Inputs (MJ); E1A Soil Nutrients Recirculation (% rP); E1B Carbon Stock (t of C); E1C Agricultural Production (Kg d.m.); F1 Agricultural Jobs (AWU). 7 J. Marull et al. Ecosystem Services 51 (2021) 101350 edible dry matter (Marco et al., 2018). We accounted for food produc­ conditions for energy-landscape integration where the traditional tion – leaving aside forestry and livestock production – and focused on mosaic pattern created by mixed farming of crops, forestry and extensive the edible products of agricultural activity because livestock feed is livestock grazing still remains. They are located at the eastern and mainly imported in the study area. Moreover, food crop production western margins of Barcelona Metropolitan Area, as well as within the predominates compared to other types of agricultural production forested area in its centre where the Collserola Natural Park acts as a (forestry and livestock raising). High E1C values correspond to highly green belt embracing the northern part of the city of Barcelona. productive zones within the study area. The value of the indicator – The landscape complexity indicator (C1) shows a normal pattern of corresponding to the absolute value in kg of dry matter per cell – was distribution, with fewer than 10% of the cells in the Barcelona Metro­ obtained from the statistical records of the average yield in each land- politan Area with values over 0.5 (Fig. 4). These areas are clustered in cover type and municipality. Given the great variability in moisture three areas: in the north, on the edges of the natural protected areas in between products, water content was discounted from production data the Garraf massif, and, to a lesser extent, in the eastern mountains of to obtain comparable values of dry matter. Serralada de Marina. These results highlight the currently low values of C1 in the municipalities in the centre of Barcelona Metropolitan Area, 3.2.6. Social cohesion where the Collserola Natural Park has lower values of C1 than those of Social cohesion (F) is represented by the principal indicator F1 that the rest of the natural protected areas, with very few cells exceeding 0.5. measures the total number of agricultural jobs. F1 takes into account the The indicator of non-renewable external inputs (D1) shows a strong potential for full-time equivalent agricultural jobs in Agricultural polarization between open spaces with very low external energy inputs Working Units (AWU) that is required in the agricultural area of the (56% of cells with values below 400 MJ/ha) and spaces specializing in metropolis (Padró et al., 2017). The most labour-intensive crops are horticultural production along the Llobregat river and in its delta awarded higher values, while more extensive agriculture or more (Fig. 4). In all, 51% of these external, fossil-fuelled energy inputs derive mechanized activities have lower F1 values. Through the data obtained from four municipalities with the highest agricultural use of external from the Xarxa Comptable Agrària de Catalunya (XCAC), the AWU energy flows (16% of the surface area of the Barcelona Metropolitan working days required by the different crops, as well as by livestock and Area). For example, the municipality of Castellbisbal (at the northern tip forestry, were estimated for each land unit. At cell level only the AWUs of the metropolis) is notable for its agriculture that has a great depen­ associated with agricultural activity were calculated. dence on external industrial inputs (D1) but also for its more spatially homogeneous cultural landscapes. 3.3. Statistical analysis The indicator of soil nutrient recirculation (E1A) shows how only some municipalities in the north-east Vallès plain and the north-west Once the values of the socioecological indicators were obtained for mountains have high levels of self-sufficiency, and how in 13 munici­ the 500 × 500-m cells, a statistical Principal Component Analysis (PCA) palities P recirculation does not exceed 10% (Fig. 4). The degree of was performed to identify the key SIA factors that characterize this case integration between livestock and crops plays a significant role in the P study. Then, an Exploratory Factor Analysis (EFA) was performed to replenishment in the soil, and reveals that the industrial horticulture visualize the distribution of indicators and land-cover categories in that specializes in the large-scale vegetable production in the Llobregat relation to the factors defined by the SIA modelling. Finally, Pearson and Agrarian Park currently has a strong dependence on imports of nutrients Spearman Correlation Analyses (PA and SA, respectively) enabled us to from outside the metropolitan system. Approximately 80% of the cells identify not only the presence of linear but also of non-linear relation­ have soil P recirculation values below 30% (that is, 70% of P comes from ships, and were performed to verify whether or not there are synergies outside the study area). and trade-offs between the indicators and the various dimensions The Carbon stock indicator (E1B) shows that the wooded area within considered. We used both coefficients because PA identifies linear re­ the Collserola Natural Park in the centre of the Barcelona Metropolitan lationships, while SA measures the concordance between indicators; Area possesses the greatest number of cells with high values, along with comparing them allowed us to infer whether or not a linear relationship certain municipalities in the Ordal mountains, as well as Begues, Tor­ exists. relles and Sant Climent de Llobregat, where arboriculture (cherry and other fruit trees) reinforce the role of forests. However, much of the 4. Results Barcelona Metropolitan Area has low C stock values and the categories ranging from 0 to 2000 t C cover 51% of the total area. 4.1. Socioecological Integrated Analysis The indicator of agricultural production (E1C) demonstrates that 51% of this provisioning service is concentrated in the Llobregat delta The agricultural energy returns on the external energy inputs used (17% of surface of the Barcelona Metropolitan Area), and in the flat land (A1) show high values in the north-west plains and in some riverside along the two banks of the Llobregat river in the protected Baix Llo­ municipalities due to the presence of energy-efficient herbaceous crops bregat Agricultural Park, where intensive horticulture is mainly carried (Fig. 4). By contrast, along most of the Llobregat river axis, and in out as conventional industrial farming (Fig. 4). Of the 6630 tonnes of dry coastal municipalities, the A1 values are low due to the farming of crops matter a year produced in the Barcelona Metropolitan Area, two thirds requiring very intensive external inputs that include, for example, large- are produced in the municipalities in the Llobregat valley and delta. The scale horticulture produced in the open or very small-scale flower pro­ farmland located on the slopes of this valley and other smaller tributary duction in greenhouses with energy efficiencies ranging between 0.12 valleys also plays a role, together with land in the south-western Ordal and 0.25. Two thirds of the cells with agricultural crops have values massif and on the western slopes of Collserola. The distribution of crop below 0.4, with the most common category being 0–0.2. This clearly intensities by cells has an annual mean value within the category of demonstrates how industrial agricultural management predominates 2000–4000 kg of dry matter. over low-input farming. Finally, the indicator of agricultural jobs (F1) shows that the inten­ The indicator of energy-landscape integration (B1) shows that the sive production of vegetables in the Baix Llobregat Agricultural Park best conditions for hosting farm-associated biodiversity in the bio­ (13% of the Barcelona Metropolitan Area) concentrates 28% of the physical structures of the matrix of agriculture, forestry and pastures are area’s whole AWUs (Fig. 4). Arboriculture located in the Ordal mountain found near the western hills in the Garraf massif (Fig. 4). The eastern range along the western axis also has a significant agricultural work mountains of Serralada de Marina also have high values of B1. In the potential. Overall, the distribution of agricultural jobs shows a strong whole Barcelona Metropolitan Area only 3% of cells have values over concentration in the intensive horticulture in the Llobregat valley and 0.7. These results also allow us to identify the areas with the best delta, in contrast to other areas where extensive herbaceous croplands 8 J. Marull et al. Ecosystem Services 51 (2021) 101350 are highly mechanized and have a lower labour intensity. energy-landscape interaction (B1) in agroecosystems. It highlights a group of indicators linked to the ecological processes that take place in 4.2. Principal component analysis agricultural landscapes and reveals their role as providers of a set of ES that can be enhanced to improve their functioning as green in­ The PCA shows that only three main components (which we call frastructures. From a circular bioeconomic point of view, the biophysi­ factors) account for 82% of the total variability (Table 2). These statis­ cal structures supporting ES provision are a set of live funds whose tical associations between indicators are relevant to sustainable strategic integration can be improved to generate more complex cultural land­ land-use planning in agroecosystems orientated towards improving ES scapes in the Barcelona Metropolitan Area. Factor 2 (‘landscape func­ as green infrastructures in the Barcelona Metropolitan Area. A first tionality’) accounts for 31% of the variability (Table 2). factor groups together the indicators corresponding to the most relevant Finally, a third factor reveals the importance of how society interacts energy flows that enter and exit from the metropolitan agroecosystems: with the land-matrix and eco-efficiency in the management of the nat­ non-renewable external inputs (D1), number of agricultural jobs (F1) ural resources that represent the biophysical structures that underpin and agricultural production per unit of land (E1C). Factor 1 (the pattern most ES. This factor consists of two closely related indicators: farming of agrarian ‘socio-metabolic flows’) accounts for 32% of the variability. energy efficiency (A1) and soil nutrient recirculation (E1A). The current A second factor, however, points to the functioning and properties of low capacity for recirculating soil nutrients creates a great dependence the land-matrix: Carbon stock (E1B), landscape complexity (C1) and in agriculture on external imports possessing large amounts of embodied Table 2 Principal Component Analysis in the relationship between socio-ecological indicators and land-covers in the Barcelona Metropolitan Area (cells of 500 × 500 m). Note: The Principal Component Analysis has been calculated using the socio-ecological indicators (coloured). Land-covers are supplementary variables. 9 J. Marull et al. Ecosystem Services 51 (2021) 101350 energy and significant Carbon footprints. In turn, this high dependence our starting conceptual assumptions, as well as the differentiated and on external inputs in industrial farming decreases energy and material complementary contribution of these factors to the provision of ES by eco-efficiency in the current management of metropolitan agro­ metropolitan agroecosystems (Fig. 5). D1 (‘dependence on non- ecosystems, a trait that underlines the considerable room for improve­ renewable external inputs’), F1 (‘agricultural job creation’) and E1C ment that exists if farmers – and society at large – are able to enhance the (‘agricultural production’) are associated with Factor 1 (‘socio-meta­ green infrastructure for Ecosystem Services Providers by increasing the bolic flows’). E1B (‘Carbon stock’), C1 (‘landscape complexity’) and B1 sustainability of the eco-agri-food system. Factor 3 (‘system efficiency’ (‘energy-landscape integration’) are associated with Factor 2 (‘land­ for the provision of ES) accounts for 19% of the variability (Table 2), scape functionality’), while A1 (‘farming energy efficiency’) and E1A which confirms the relevance of considering the circularity and closure (‘soil nutrient recirculation’) are associated with Factor 3 (‘system of the matter-energy flows required for the reproducibility of the live efficiency’). funds of metropolitan agroecosystems. It also attests to how the capacity Next, we examined how these factors are related to different land of Ecosystem Services Providers will improve as they advance towards uses in the metropolitan area. Compact urban areas have an obvious agroecology territories. negative role in Factor 2 (‘landscape functionality’) but not so much in Factor 1 (‘socio-metabolic flows’) and very little in Factor 3 (‘system efficiency’) (Fig. 5). By contrast, forests play a key role in Factor 2 4.3. Exploratory factor analysis (‘landscape functionality’) but their role is anecdotal in relation to the other factors, as is to be expected. To a lesser extent, the same happens The EFA shows that the different indicators are distributed orthog­ with scrubland. This is mainly due to the state of abandonment of many onally in relation to the three factors defined by the SIA. This confirms Fig. 5. Exploratory Factor Analysis. The socio-ecological indicators* and land-cover classes** are expressed according to their association to Factor 1 (metabolic flows), Factor 2 (landscape functionality) and Factor 3 (system efficiency). 10 J. Marull et al. Ecosystem Services 51 (2021) 101350 forests and pastures in the area and their lack of integration with 4.4. Pearson and spearman correlation analysis croplands as part of a complex mixed farming agroecosystem, a char­ acteristic that indicates once again that there is room for improvement The PA and SA allowed us to assess how certain socioecological in­ in their role as a green infrastructure. dicators affect others (Table 3). If we consider the indicators that are It is interesting to note that each type of crop has a different profile part of Factor 1 (‘socio-metabolic flows’) to be configurators of farming (Fig. 5). While irrigated herbaceous crops and fruit trees play a major socio-metabolic patterns such as the dependence on external energy role mainly in Factor 1 (‘socio-metabolic flows’), dry grain-farming is inputs (D1), agricultural production (E1C) and social cohesion (F1), we key in Factor 3 (‘system efficiency’) and dry arboriculture is relatively found a clear positive relationship between these indicators with a important in Factor 1 (‘socio-metabolic flows’) and also in Factor 2 Pearson coefficient that exceeds 0.6 in all cases. In the case of D1 (‘landscape functionality’). These results have important implications (external inputs) and E1C (agricultural production) there was a unam­ for sustainable land-use planning, since these type of land-covers have biguous relationship, which indicates that any increase in agricultural been hither to regarded as belonging to the same ‘agricultural’ class, production is currently highly associated with external energy inputs. even though they actually provide very different contributions to the This dependence on external energy inputs is also statistically related to supply of ES. These differences allow us to identify the priorities in each more labour-intensive crops per unit of land. The maximum values of F1 case depending on the social values assigned to the different combina­ (>5 AWU per cell) appear to be associated with intermediate levels of tions of ES that ought to be obtained from the metropolitan agro­ imports of external non-renewable flows. ecosystems acting as green infrastructures in the territory. Of the indicators associated with the configuration of landscape Finally, some land covers seem to possess certain counterintuitive functionality in Factor 2 (‘landscape functionality’) – i.e. C stock (E1B), characteristics (Fig. 5), especially in relation to Factor 2 (‘landscape landscape complexity (C1) and energy-landscape integration (B1) – the functionality’). This is the case of river corridors and wetlands or, most correlated ones are B1 and C1, with a Pearson coefficient above conversely, low-density urban areas. It is important to note that these 0.92 (Table 3). At the same time, these indicators also maintain a clear statistical analyses were calculated for 500 × 500-m cells. As metro­ relationship with C stock, although with a progressive increase in politan wetlands are often adjacent to urban areas, the relationship dispersion, both with Pearson coefficients above 0.64. between Factor 2 (‘landscape functionality’) and wetlands is virtually The main indicators determining the efficiency of the agricultural null if all cells are considered. However, if only cells with 10% or more of system as a provider of ES in Factor 3 (‘system efficiency’), farming the surface area actually covered by wetlands are accounted for, a clear energy efficiency (A1) and soil P recirculation (E1A) also have a certain upwards linear relationship with respect to B1 (‘energy-landscape statistical relationship, although the Pearson correlation is lower integration’) or C1 (‘landscape complexity’) appears. The same rela­ (Table 3). The recirculation of nutrients is much more affected than tionship, albeit negative, is observed between low density urban areas energy efficiency by the lack of integration between livestock and and the B1 or C1 indicators, as was to be expected. agriculture. This is the case, for example, of the town of Sant Boi de Llobregat where, despite very high values of A1 (1.51 of EFEROI on Table 3 Pearson and Spearman Correlation Analysis between the socio-ecological indicators considered in Barcelona Metropolitan Area (cells of 500 × 500 m). 11 J. Marull et al. Ecosystem Services 51 (2021) 101350 average), its capacity for soil replenishment of nutrients is very low agroecosystems that need to be enhanced as green infrastructures. (around 6% of total P) because its agricultural specialization in livestock The results also show the poor capacity for dealing with the closure feeding does not match the local number of heads of livestock, which are of soil nutrient cycles due to the prevalence of industrial farming and, in proportionally low (only 15% of the animal feed produced is consumed particular, the high proportion of P coming from outside the territory. locally, which implies that there is an exchange of animal feed with This is a fundamental issue since P is currently the most limiting of all neighbouring areas). This example demonstrates the need for a balanced macronutrients (Garcia-Ruiz et al., 2012). Once more, the key is the lack integration of all the different fund components of the agroecosystem of integration between livestock and crop production in their industrial (soil-crop-livestock) to guarantee high values of socio-metabolic effi­ management. Although the closure of metabolic cycles at landscape ciency in the management of agricultural landscapes. scale implies a long-term vision, greater rationality in the use of soil The three main factors (‘socio-metabolic flows’, ‘landscape func­ nutrients, as well as the change to organic sources, are fundamental tionality’ and ‘system efficiency’) are not independent of each other and, improvements not only for the sustainable development of agri-food as can be seen from the relationships between the indicators that make chains but also for avoiding air and water pollution by nitrates and up Factor 1 and Factor 2, there are some trade-offs (Table 3). Thus, if we soil degradation. This is a clear example of synergistic agroecological look at the relationship between non-renewable external energy inputs improvement in line with the EU Farm to Fork Strategy. and landscape functionality, we can see that high values of landscape The recovery of metropolitan agriculture can also foster more land­ complexity, energy-landscape integration and C stock are related to low scape heterogeneity via the land-use mosaics established by organic levels of non-renewable external inputs. This highlights the importance mixed farming. Recent studies explain biodiversity conservation in of low-input mixed farming and organic management of agro­ cultural landscapes through the interaction between the energy flows ecosystems. However, the higher values of these external inputs are also available for trophic chains and the landscape complexity of agro­ related to intermediate levels in landscape complexity and integration, ecosystems (Tscharntke et al., 2005; Pierce, 2014; Marull et al., 2018; and with C stock. This means that a point can be found at which inter­ Marull et al., 2019). The SIA results support a proposal for land-use mediate levels of landscape transformation and high levels of socio- planning in metropolitan green spaces that encompasses all landscape metabolic flows could lead to interesting combinations in terms of units, including agricultural and forest areas, in order to achieve joint agricultural jobs and production. This result indicates that there are improvements in the metabolic efficiency and ecological functioning of possibilities for agroecological intensification of farming but always the whole land matrix. This implies the need to recover old abandoned within limits. agricultural areas and promote a more circular pattern of biophysical Factor 1 (‘socio-metabolic flows’) is also related to Factor 3 (‘system flows within agroecological territories, combined with the positive ef­ efficiency’). This is a trade-off relationship and, although the Pearson fect of restoring cultural landscapes on biodiversity maintenance. coefficients do not exceed 0.27, Spearman coefficients were between The landscape complexity assessment shows that Mediterranean 0.77 and 0.94 (Table 3). Accordingly, high values of energetic socio- land-cover mosaics (i.e. a heterogeneous and well-connected land-cover metabolic efficiency or P recirculation reach an upper limit in regard pattern) still exist and must be preserved and extended to take advan­ to the reduction of non-renewable external inputs, job creation and tage of their contribution to the ecological processes that enhance the agricultural production. This means that improving agroecosystem ef­ landscape’s ability to host biodiversity in metropolitan areas (Santos ficiency in the use of resources through the closure of socio-metabolic et al., 2008; Marull et al., 2014). They must be identified and cared for in cycles at local level will lead – if above a maximum threshold – to the metropolitan policies. According to the SIA results, there is a trade- detrimental impacts both on production and agricultural jobs, at least as off between the increase in agricultural land-uses and the heterogeneity long as the current model of conventional industrial farming is main­ of the resulting landscapes, which must be cared for when taking de­ tained and other possible agroecology synergies remain unexplored. cisions on land-use planning in these open vegetated spaces. Lastly, it should be noted that there is no clear relationship between Within the metropolitan system, food provisioning remains an Factor 2 (‘landscape functionality’) and Factor 3 (‘system efficiency’), important ES, despite the shrinking of urban agricultural activity due to since all the Spearman coefficients were less than 0.44 (Table 3). This urban expansion into flat land and the abandonment of steep terrains. In means that, although efficiency in resource use influences socio- total, about 6630 t of dry matter a year are produced that represent metabolic flows in farming, and that these are affected in turn by the nearly 10% of the fresh vegetables consumed in the metropolitan area ecological landscape functioning of the land-matrix, there is only a and a smaller part of its animal protein intake. It is clear that the food rather indirect and weak relationship. Therefore, there could be high needs of the 3.24 million people who live in the Barcelona Metropolitan quality levels of land-matrix functioning under a low level of efficiency Area cannot be satisfied internally. However, prioritizing those crops in resource use. This finding helps explain why farming socio-metabolic and agricultural activities that, beyond food provisioning, maximize flows have so often been neglected in analyses of ES provision, as well as other functions and ES for city dwellers could be a crucial way of in analyses of the role agroecosystems play as green infrastructures highlighting the importance of metropolitan agroecosystems as high­ where traditional nature conservation approaches are usually adopted. lighted by the Milan Food Policy Pact. In the same vein, Factor 1 (‘socio-metabolic flows’) has often been The agrarian sector generates at present over 900 full-time agricul­ addressed by paying attention only to agricultural production as a tural jobs (AWU), all of which are linked to the management of the provisioning ES without taking note of its dependence on non-renewable vegetated areas of this territory and, especially, to farming (70% of these external inputs or job creation. jobs). The ES derived from restoring forestry in certain areas to reduce the accumulation of dead biomass and the risk of wildfires could provide 5. Discussion a supply of fuel for the biomass energy sector and create agricultural jobs in the metropolitan agroecosystems (Cervera et al., 2019). A more The C stock is equivalent to the amount of organic C stored in substantial change would come if there was a systemic change in the metropolitan agricultural areas, forests and pastures, which is a funda­ agricultural model, i.e. the recovery of extensive livestock grazing in­ mental ES in climate change mitigation and adaptation (Doblas-Miranda tegrated with farming and forest activities, both under new types of et al., 2013; Aguilera et al., 2013). It is worth noting that forest areas agroecological management that are more labour-intensive and much play a significant role in the total C stock in the Barcelona Metropolitan less dependent on external inputs from fossil fuels. Further SIA research Area and that more than 55% of C is estimated to be found in soils. should incorporate the offer-demand vision of ES (Baró et al., 2016) and Average differences between forest and agricultural soils are not that look deeper into rural–urban interactions. large (40 t of C/ha in soils of certain crops vs. 68 t of C/ha in oak forests). Finally, perhaps the most important result obtained from this first This key ES has to be balanced out by the recovery of metropolitan SIA assessment of agroecosystems’ capacity for multiple ES provision in 12 J. Marull et al. Ecosystem Services 51 (2021) 101350 the Barcelona Metropolitan Area is the joint statistical analysis that systemic approach to the restoration of agroecosystems and to revealed which are the most important drivers, as well as their trade-offs increasing food system sustainability, which can be developed via a and possible synergies. PCA grouped the SIA indicators into three main Socioecological Integrated Analysis of agricultural landscapes. factors: the pattern of farming ‘metabolic flows’, the resulting ‘landscape functionality’, and the ‘system efficiency’ of multiple ES that increase Declaration of Competing Interest citizens’ welfare. In relation to the strategic planning of new green in­ frastructures aimed at increasing the supply of all ES in metropolitan The authors declare that they have no known competing financial areas, the main conclusion to be drawn is that the statistical correlations interests or personal relationships that could have appeared to influence found between the different dimensions confirm that a new systemic the work reported in this paper. vision of metropolitan land-use planning is required, based on a multi- criteria analysis and integrated indicators that imply setting to one Acknowledgements side the outmoded sectoral public policies of the past. Consequently, new multidimensional approaches are needed to un­ This work has been funded by the Barcelona Metropolitan Area in derstand and plan agroecosystems as part of metropolitan green in­ order to obtain criteria and methods for the Metropolitan Land Use frastructures orientated towards enhancing the provision of a full array Master Plan of Barcelona, and has also received funding by the LIFE of ES (Padró et al., 2020). Furthermore, facing up to the emerging global UrbanGreeningPlans project, financed by the European Commission, socioecological crisis involves relocating the current globalized socio- and the BIOLANDSCAPEs project financed by the Spanish Ministry of metabolic flows (Levidow et al., 2014). To minimize the negative Science, Innovation and Universities (RTI2018-093970-B-C32). socio-ecological impacts of industrial agriculture, any new strategic planning of metropolitan areas aimed at improving green infrastructures References as ES providers must take into account the fresh synergies and trade-offs that more sustainable forms of agriculture would imply. It is also Agnoletti, M., 2014. Rural landscape, nature conservation and culture: some notes on research trends and management approaches from a (southern) European important to consider what this implies for urban–rural relationships on perspective. Landscape Urban Plan. 126, 66–73. https://doi.org/10.1016/j. a broader scale, a subject that will require additional research that could landurbplan.2014.02.012. be carried out with the SIA model (Table 1). An integrated enlarged SIA Agnoletti, M., Rotherham, I.D., 2015. Landscape and biocultural diversity. Biodivers. Conserv. 24, 3155–3165. https://doi.org/10.1007/s10531-015-1003-8. would end up linking the sustainable planning of larger territories in a Agnoletti, M., Emanueli, F., 2016. In: Biocultural Diversity in Europe. Springer, Cham. multi-scalar way as a means of exploring possible pathways towards https://doi.org/10.1007/978-3-319-26315-1. feasible agroecological territories and culturally meaningful foodscapes. Aguilera, E., Lassaletta, L., Gattinger, A., Gimeno, B.S., 2013. Managing soil carbon for climate change mitigation and adaptation in Mediterranean cropping systems: A meta-analysis. Agric. Ecosyst. Environ. 168, 25–36. https://doi.org/10.1016/j. 6. Conclusions agee.2013.02.003. Altieri, M.A., Nicholls, C.I., 2012. Agroecology scaling up for food sovereignty and A Socioecological Integrated Analysis of the Barcelona Metropolitan resiliency. In: Lichtfouse, E. (Ed.), Sustainable Agriculture Reviews. Springer, pp. 1–29. https://doi.org/10.1007/978-94-007-5449-2_1. Area was carried out to identify strategic factors in sustainable land-use Antrop, M., 2006. Sustainable landscapes: contradiction, fiction or utopia? Landscape planning aimed at strengthening the ecosystem services provided by Urban Plan. 75 (3–4), 187–197. https://doi.org/10.1016/j. agricultural systems and landscape mosaics. The results show that the landurbplan.2005.02.014. Baró, F., Palomo, I., Zulian, G., Vizcaino, P., Haase, D., Gómez-Baggethun, E., 2016. current industrial agriculture model that characterizes the Barcelona Mapping ecosystem service capacity, flow and demand for landscape and urban Metropolitan Area has very low energy efficiency. The prevailing agri­ planning: a case study in the Barcelona metropolitan region. Land Use Policy 57, cultural model depends on huge external inputs with large carbon 405–417. https://doi.org/10.1016/j.landusepol.2016.06.006. Bastian, O., Haase, D., Grunewald, K., 2012. Ecosystem properties, potentials and footprints (Tello et al., 2016; Cattaneo et al., 2018). To start an agri-food services – the EPPS conceptual framework and an urban application example. Ecol. transition towards agroecological territories (FAO, 2018), energy and Ind. 21, 7–16. https://doi.org/10.1016/j.ecolind.2011.03.014. material efficiency must be significantly increased through farming Benedict, M.A., McMahon, E.T., 2002. Green infrastructure: smart conservation for the 21st century. Renewable Resour. J. 12–17 https://doi.org/10.4135/ management strategies based on increased biomass recirculation within 9781412973816.n70. the metropolitan system that will replace the external inputs. The high Boumans, R., Roman, J., Altman, I., Kaufman, L., 2015. The multiscale integrated model dependence on non-renewable external energy inputs can be linked to of ecosystem services (MIMES): Simulating the interactions of coupled human and natural systems. Ecosyst. Serv. 12, 30–41. https://doi.org/10.1016/j. the greenhouse gases emitted by industrial agriculture (Sanz-Covena ecoser.2015.01.004. et al., 2017; González de Molina et al., 2020). Novel sustainable land-use Burkhard, B., Kroll, F., Nedkov, S., Müller, F., 2012. Mapping ecosystem service supply, planning combined with agricultural and environmental policies aimed demand and budgets. Ecol. Ind. 21, 17–29. https://doi.org/10.1016/j. at integrating farming with forestry and livestock activities would help ecolind.2011.06.019. Cattaneo, C., Marull, J., Tello, E., 2018. Landscape agroecology. The dysfunctionalities of reduce this dependence on external inputs and, indirectly, the green­ industrial agriculture and the loss of the circular bioeconomy in the Barcelona house gas emissions stemming from agri-food chains. Region, 1956–2009. Sustainability 10. https://doi.org/10.3390/su10124722. The new Master Urban Plan of the Barcelona Metropolitan Area is Cardinale, B.J., Duffy, J.E., Gonzalez, A., Hooper, D.U., Perrings, C., Venail, P., Narwani, A., Mace, G.M., Tilman, D., Wardle, D.A., Kinzig, A.P., Daily, G.C., being designed using these indicators and assessment in order to comply Loreau, M., Grace, J.B., Larigauderie, A., Srivastava, D.S., Naeem, S., 2012. with the C40 commitment for climate action (https://www.c40.org/) Biodiversity loss and its impact on humanity. Nature 486 (7401), 59–67. https://doi. and the Milan Urban Food Policy Pact (http://www.milanurban org/10.1038/nature11148. Cervera, T., Pino, J., Marull, J., Padró, R., Tello, E., 2019. Understanding the long-term foodpolicypact.org/), both of which have been signed and co-led by dynamics of Forest Transition: from deforestation to afforestation in a Mediterranean the Barcelona City Council. This Plan aims to recover many former landscape (Catalonia, 1868–2005). Land Use Policy 80, 318–331. https://doi.org/ agricultural areas that are currently abandoned. Even if this implies 10.1016/j.landusepol.2016.10.006. Costanza, R., de Groot, R., Braat, L., et al., 2017. Twenty years of ecosystem services: strengthening the local biophysical flows that require some energy how far have we come and how far do we still need to go? Ecosyst. Serv. 28, 1–16. expenditure, we must consider the net balance with the external energy https://doi.org/10.1016/j.ecoser.2017.09.008. consumption and the greenhouse gas emissions that are avoided. Davies, C., MacFarlane, R., McGloin, C., 2006. Green Infrastructure. In: Cohen, N., Robbins, P. (Eds.), Green Cities: An A-to-Z Guide. Thousand Oaks (California). SAGE. Concurrently, the EU Biodiversity Strategy for 2030 (https://ec.europa. https://doi.org/10.4135/9781412973816.n70. eu/environment/nature/biodiversity/strategy/index_en.htm), the Farm de Groot, R.S., Alkemade, R., Braat, L., Hein, L., Willemen, L., 2010. Challenges in to Fork Strategy (https://ec.europa.eu/food/farm2fork_en) and the integrating the concept of ecosystem services and values in landscape planning, Green Deal (https://ec.europa.eu/info/strategy/priorities management and decision making. Ecol. Complex. 7 (3), 260–272. https://doi.org/ 10.1016/j.ecocom.2009.10.006. -2019–2024/european-green-deal_en) have all been adopted by the Doblas-Miranda, E., Rovira, P., Brotons, L., Martínez-Vilalta, J., Retana, J., Pla, M., EU. All constitute a strong policy framework, requiring a holistic and Vayreda, J., 2013. Soil carbon stocks and their variability across the forests, 13 J. Marull et al. Ecosystem Services 51 (2021) 101350 shrublands and grasslands of peninsular Spain. Biogeosciences 10 (12), 8353–8361. Maes, J., Liquete, C., Teller, A., et al., 2016. An indicator framework for assessing https://doi.org/10.5194/bg-10-8353-201310.5194/bg-10-8353-2013-supplement. ecosystem services in support of the EU Biodiversity Strategy to 2020. Ecosyst. Serv. Commission, E., 2014. In: Mapping and Assessment of Ecosystems and their Services 17 (2016), 14–23. https://doi.org/10.1016/j.ecoser.2015.10.023. Indicators for ecosystem assessments under Action 5 of the EU Biodiversity Strategy Maes, J., Teller, A., Erhard, M., et al., 2020. In: Mapping and Assessment of Ecosystems to 2020 2nd Report – Final. European Union Publication Office. https://doi.org/ and their Services: An EU Ecosystem Assessment. Publications Office of the European 10.2779/75203. Union, Luxembourg. https://doi.org/10.2760/757183. FAO, 2018. Scaling Up Agroecology to Achieve the Sustainable Development Goals. FAO, Marco, I., Padró, R., Cattaneo, C., et al., 2018. From vineyards to feedlots : A fund-flow Rome. scanning of sociometabolic transitions in the Vallès County (Catalonia) 1860–1956- Funes I., Savé R., Vayreda J., et al., 2015. Evaluación de la biomasa de cultivos leñosos 1999. Reg. Environ. Change 18, 981–993. https://doi.org/10.1007/s10113-017- como stock de carbono en la cuenca mediterránea. In: V Workshop Remedia sobre 1172-y. Mitigación de Gases de Efecto Invernadero Procedentes del Sector Agroforestal. Marco, I., Padró, R., Tello, E., 2020. Labour, nature, and exploitation: Social metabolism Madrid. . and inequality in a farming community in mid-19th century Catalonia. J. Agrarian Gabriel, D., Roschewitz, I., Tscharntke, T., et al., 2006. Beta diveristy at different spatial Change 20 (3), 408–436. https://doi.org/10.1111/joac.v20.310.1111/joac.12359. scales: Plant communities in organic and conventional agriculture. Ecol. Appl. 16 Maruani, T., Amit-Cohen, I., 2007. Open space planning models: A review of approaches (5), 2011–2021. http://www.jstor.org/stable/40061770?origin=JSTOR-pdf. and methods. Landscape Urban Plann. 81 (1–2), 1–13. https://doi.org/10.1016/j. Garcia-Ruiz, R., González de Molina, M., Guzmán, G., Soto, D., Infante-Amate, J., 2012. landurbplan.2007.01.003. Guidelines for constructing nitrogen, phosphorus, and potassium balances in Marull, J., Mallarach, J.M., 2005. A GIS methodology for assessing ecological historical agricultural systems. J. Sustain. Agr. 36 (6), 650–682. https://doi.org/ connectivity: application to the Barcelona Metropolitan Area. Landscape Urban Plan. 10.1080/10440046.2011.648309. 71 (2–4), 243–262. https://doi.org/10.1016/j.landurbplan.2004.03.007. Georgescu-Roegen, N., 1971. The Entropy Law and the Economic Process. Harvard Marull, J., Pino, J., Tello, E., Cordobilla, M.J., 2010. Social metabolism, landscape University Press, Cambridge. change and land-use planning in the Barcelona Metropolitan Region. Land Use Giampietro, M., Mayumi, K., Sorman, A.H., 2011. The Metabolic Pattern of Societies. Policy 27 (2), 497–510. https://doi.org/10.1016/j.landusepol.2009.07.004. Where Economists Fall Short. Routledge, Abingdon and New York. Marull, J., Tello, E., Wilcox, P.T., et al., 2014. Recovering the landscape history behind a Giampietro, M., Mayumi, K., Sorman, A.H., 2013. Energy Analysis for Sustainable Mediterranean edge environment (The Congost Valley, Catalonia, 1854e2005): The Future: Multi-Scale Integrated Analysis of Societal and Ecosystem Metabolism. importance of agroforestry systems in biological conservation. Appl. Geogr. 2014, Routledge, Abingdon and New York. 1–17. https://doi.org/10.1016/j.apgeog.2014.06.030. Gill, S.E., Handley, J.F., Ennos, A.R., Pauleit, S., 2012. Adapting cities for climate change: Marull, J., Font, C., Padró, R., Tello, E., Panazzolo, A., 2016. Energy-Landscape the role of the green infrastructure. Built Environ. 33 (1), 115–133. https://doi.org/ Integrated Analysis: a proposal for measuring complexity in internal agroecosystem 10.2148/benv.33.1.115. processes (Barcelona Metropolitan Region, 1860–2000). Ecol. Ind. 66, 30–46. Gingrich, S., Marco, I., Aguilera, E., et al., 2018. Agroecosystem energy transitions in the https://doi.org/10.1016/j.ecolind.2016.01.015. old and new worlds: trajectories and determinants at the regional scale. Reg. Marull, J., Tello, E., Bagaria, G., Font, X., Cattaneo, C., Pino, J., 2018. Exploring the links Environ. Change 18, 1089–1101. https://doi.org/10.1007/s10113-017-1261-y. between social metabolism and biodiversity distribution across landscape gradients: Gliessmann, S., 1998. Agroecology: Ecological Processes in Sustainable Agriculture. a regional-scale contribution to the land-sharing versus land-sparing debate. Sci. Lewis Publishers, London. Total Environ. 619-620, 1272–1285. https://doi.org/10.1016/j. Gliessman, S.R., Engles, E.W., 2015. Agroecology: The ecology of sustainable food scitotenv.2017.11.196. systems. CRC Press, Boca Raton. Marull, J., Herrando, S., Brotons, L., Melero, Y., Pino, J., Cattaneo, C., Pons, M., González de Molina, M., García Ruiz, R., Guzmán, G.I., et al., 2010. Guideline for Llobet, J., Tello, E., 2019. Building on Margalef: testing the links between landscape constructing nutrient balance in historical agricultural systems (and its application structure, energy and information flows driven by farming and biodiversity. Sci. to three case-studies in southern Spain) (No. 1008), DT-SEHA. Madrid. Total Environ. 674, 603–614. https://doi.org/10.1016/j.scitotenv.2019.04.129. González de Molina, M., Toledo, V., 2014. The Social Metabolism. Springer, Cham. Mell, I.C., 2009. Can green infrastructure promote urban sustainability? Proc. Inst. Civil González de Molina, M., Soto-Fernández, D., Guzmán-Casado, G., et al., 2020. The Social Eng.-Eng. Sustainability 162 (1), 23–34. https://doi.org/10.1680/ Metabolism of Spanish Agriculture, 1900–2008. Springer, Cham. ensu.2009.162.1.23. Guzmán, G.I., González de Molina, M., 2017. Energy in Agroecosystems. A Tool for MEA, 2005. Ecosystems and Human Well-being: Current State and Trends. Island Press, Assessing Sustainability. CRC Press, Boca Raton. Washington DC. Haberl, H., Fischer-Kowalski, M., Krausmann, F., Weisz, H., Winiwarter, V., 2004. Moragues-Faus, A., Morgan, K., 2014. Reframing the foodscape: the emergent world of Progress towards sustainability? What the conceptual framework of material and urban food policy. Environ. Plann. A: Eco. Space 1558–1573. https://doi.org/ energy flow accounting (MEFA) can offer. Land Use Policy 21 (3), 199–213. https:// 10.1177/0308518X15595754. doi.org/10.1016/j.landusepol.2003.10.013. Oteros-Rozas, E., Ruiz-Almeida, A., Aguado, M., González, J.A., Rivera-Ferre, M.G., Haines-Young, R., Potschin, M., 2010. The links between biodiversity, ecosystem services 2019. A social–ecological analysis of the global agri-food system. Proc. Natl. Acad. and human well-being. In: Raffaelli, D.G., Frid, C.L.J. (Eds.), Ecosystem Ecology: A Sci. U.S.A. 116 (52), 26465–26473. https://doi.org/10.1073/pnas.1912710116. New Synthesis. Cambridge University Press, Cambridge, pp. 110–139. https://doi. Padró, R., La Rota, M.J., Giocoli, A., Cirera, J., Coll, F., Pons, M., Pino, J., Pili, S., org/10.1017/CBO9780511750458.007. Serrano, T., Villalba, G., Marull, J., 2020. Assessing the sustainability of contrasting Haines-Young, R., Potschin M., 2018. Common International Classification of Ecosystem land use scenarios through the Socioecological Integrated Analysis (SIA) of the Services (CICES) V5.1 and Guidance on the Application of the Revised Structure. metropolitan green infrastructure in Barcelona. Landscape and Urban Planning 203, Available from . 103905. https://doi.org/10.1016/j.landurbplan.2020.103905. Hamann, M., Biggs, R., Reyers, B., 2015. Mapping social-ecological systems: Identifying Padró, R., Marco, I., Font, C., et al., 2019. Beyond Chayanov: a sustainable farm “green-loop” and “red-loop” dynamics based on characteristic bundles of ecosystem reproductive analysis of peasant domestic units and rural communities (Sentmenat; service use. Global Environ. Change 34, 218–226. https://doi.org/10.1016/j. Catalonia, 1860). Ecol. Econ. 160, 227–239. https://doi.org/10.1016/j. gloenvcha.2015.07.008. ecolecon.2019.02.009. Hansen, R., Pauleit, S., 2014. From multifunctionality to multiple ecosystem services? A Palmer, M.A., Febria, C.M., 2012. Ecology: the heartbeat of ecosystems. Science 336 conceptual framework for multifunctionality in green infrastructure planning for (6087), 1393–1394. https://doi.org/10.1126/science.1223250. Urban Areas. Ambio 43 (4), 516–529. https://doi.org/10.1007/s13280-014-0510-2. Perfecto, I., Vandermeer, J., 2010. The agroecological matrix as alternative to the land- Harrison, P.A., Berry, P.M., Simpson, G., Haslett, J.R., Blicharska, M., Bucur, M., sparing/agriculture intensification model. Proc. Natl. Acad. Sci. U.S.A. 107 (13), Dunford, R., Egoh, B., Garcia-Llorente, M., Geamănă, N., Geertsema, W., 5786–5791. https://doi.org/10.1073/pnas.090545510710.1073/pnas.0905455107. Lommelen, E., Meiresonne, L., Turkelboom, F., 2014. Linkages between biodiversity Phalan, B., Onial, M., Balmford, A., et al., 2011. Reconciling food production and attributes and ecosystem services: a systematic review. Ecosyst. Serv. 9, 191–203. biodiversity conservation: land sharing and land sparing compared. Science 333, https://doi.org/10.1016/j.ecoser.2014.05.006. 1289–1291. https://science.sciencemag.org/content/333/6047/1289. Ho, M.-W., Ulanowicz, R., 2005. Sustainable systems as organisms? Bio Syst. 82 (1), Pierce, S., 2014. Implications for biodiversity conservation of the lack of consensus 39–51. https://doi.org/10.1016/j.biosystems.2005.05.009. regarding the humped-back model of species richness and biomass production. Hole, D.G., Perkins, A.J., Wilson, J.D., et al., 2005. Does organic farming benefit Funct. Ecol. 28, 253–257. https://doi.org/10.1111/1365-2435.12147. biodiversity? Biol. Conserv. 122, 113–130. https://doi.org/10.1016/j. Plieninger, T., Bieling, C., Fagerholm, N., et al., 2015. The role of cultural ecosystem biocon.2004.07.018. services in landscape management and planning. Curr. Opin. Environ. Sustainability Lafortezza, R., Davies, C., Sanesi, G., Konijnendijk, C., 2013. Green infrastructure as a 14, 28–33. https://doi.org/10.1016/j.cosust.2015.02.006. tool to support spatial planning in European urban regions. IForest 6 (3), 102–108. Potschin-Young, M., Haines-Young, R., Görg, C., et al., 2018. Understanding the role of https://doi.org/10.3832/ifor0723-006. conceptual frameworks: reading the ecosystem service cascade. Ecosyst. Serv. 29, Levidow, L., Pimbert, M., Vanloqueren, G., 2014. Agroecological research: 428–440. https://doi.org/10.1016/j.ecoser.2017.05.015. conforming—or transforming the dominant agro-food regime? Agroecol. Sustainable Sandhu, H.S., Wratten, S.D., Cullen, R., et al., 2008. The future of farming: the value of Food 38 (10), 1127–1155. https://doi.org/10.1080/21683565.2014.951459. ecosystem services in conventional and organic arable land. An experimental Loreau, M., Mouquet, N., Gonzalez, A., 2003. Biodiversity as spatial insurance in approach. Ecol. Econ. 64, 835–848. https://doi.org/10.1016/j.ecoser.2014.07.004. heterogeneous landscapes. Proc. Natl. Acad. Sci. U.S.A. 100 (22), 12765–12770. Santos, K.C., Pino, J., Rodà, F., et al., 2008. Beyond the reserves: the role of non- https://doi.org/10.1073/pnas.223546510010.1073/pnas.2235465100. protected rural areas for avifauna conservation in the area of Barcelona (NE of Maeder, P., Fliessbach, A., Dubois, D., et al., 2002. Soil fertility and biodiversity in Spain). Landscape Urban Plan. 84, 140–151. https://doi.org/10.1016/j. organic farming. Science 296 (1694), 1694–1697. https://science.sciencemag. landurbplan.2007.07.004. org/content/296/5573/1694.full. Santos-Martín, F., Zorrilla-Miras, P., Palomo-Ruiz, I., et al., 2019. Protecting nature is necessary but not sufficient for conserving ecosystem services: a comprehensive 14 J. Marull et al. Ecosystem Services 51 (2021) 101350 assessment along a gradient of land-use intensity in Spain. Ecosyst. Serv. 35, 43–51. internal and external returns (the Vallès county, Catalonia c. 1860 and 1999). Ecol. https://doi.org/10.1016/j.ecoser.2018.11.006. Econ. 121, 160–174. https://doi.org/10.1016/j.ecolecon.2015.11.012. Sanz-Covena, A., Lassaletta, L., Aguilera, E., et al., 2017. Strategies for greenhouse gas Tenberg, A., Fredholm, S., Eliasson, I., et al., 2012. Cultural ecosystem services provided emissions mitigation in Mediterranean agriculture: a review. Agric. Ecosyst. Environ. by landscapes: assessment of heritage values and identity. Ecosyst. Serv. 2, 14–26. 238, 5–24. https://doi.org/10.1016/j.agee.2016.09.038. https://doi.org/10.1016/j.ecoser.2012.07.006. Tancoigne, E., Barbier, M., Cointet, J.-P., et al., 2014. The place of agricultural sciences Tscharntke, T., Klein, A.M., Kruess, A., et al., 2005. Landscape perspectives on in the literature on ecosystem services. Ecosyst. Serv. 10, 35–48. https://doi.org/ agricultural intensification and biodiversity-ecosystem service management. Ecol. 10.1016/j.ecoser.2014.07.004. Lett. 8, 857–874. https://doi.org/10.1111/j.1461-0248.2005.00782.x. TEEB, 2008. The Economics of Ecosystem & Biodiversity: An Interim Report. Cambridge: Tscharntke, T., Tylianakis, J.M., Rand, T.A., et al., 2012. Landscape moderation of Banson Pub. . biodiversity patterns and processes—eigth hypotheses. Biol. Rev. 87, 661–685. TEEB, 2010. Mainstreaming the Economics of Nature: A Synthesis of the Approach, https://doi.org/10.1111/j.1469-185X.2011.00216.x10.1111/j.1469- Conclusions and Recommendations of TEEB. Malta: Banson Pub. . urban areas using Green Infrastructure: a literature review. Landscape and Urban TEEB, 2018. Measuring what matters in agriculture and food systems: a synthesis of the Plan. 81 (3), 167–178. https://doi.org/10.1016/j.landurbplan.2007.02.001. results and recommendations of TEEB for Agriculture and Food’s Scientific and Van der Ploeg, J.D., 2014. Peasants and the Art of Farming: A Chayanovian Manifesto. Economic Foundations report. Geneva: UN Environment. Download report. . budgets for organic farms as a tool for management of soil fertility. Soil Use Manag. Tello E., Galán E., Cunfer G., et al., 2015. A proposal for a workable analysis of Energy 18 (1), 264–273. https://doi.org/10.1111/j.1475-2743.2002.tb00268.x. Return On Investment (EROI) in agroecosystems. Part I : Analytical approach (Social Wezel, A., Brives, H., Casagrande, M., et al., 2016. Agroecology territories: places for Ecology No. 156). . sustainable agricultural and food systems and biodiversity conservation. Agroecol. Tello, E., Galán, E., Sacristán, V., et al., 2016. Opening the black box of energy Sustainable Food Syst. 40 (2), 132–144. https://doi.org/10.1080/ throughputs in agroecosystems: a decomposition analysis of final EROI into its 21683565.2015.1115799. 15