Marginal Agricultural Land Low-Input Systems for Biomass Production

This study deals with approaches for a social-ecological friendly European bioeconomy based on biomass from industrial crops cultivated on marginal agricultural land. The selected crops to be investigated are: Biomass sorghum, camelina, cardoon, castor, crambe, Ethiopian mustard, giant reed, hemp, lupin, miscanthus, pennycress, poplar, reed canary grass, safflower, Siberian elm, switchgrass, tall wheatgrass, wild sugarcane, and willow. The research question focused on the overall crop growth suitability under low-input management. The study assessed: (i) How the growth suitability of industrial crops can be defined under the given natural constraints of European marginal agricultural lands; and (ii) which agricultural practices are required for marginal agricultural land low-input systems (MALLIS). For the growth-suitability analysis, available thresholds and growth requirements of the selected industrial crops were defined. The marginal agricultural land was categorized according to the agro-ecological zone (AEZ) concept in combination with the marginality constraints, so-called ‘marginal agro-ecological zones’ (M-AEZ). It was found that both large marginal agricultural areas and numerous agricultural practices are available for industrial crop cultivation Energies 2019, 12, 3123; doi:10.3390/en12163123 www.mdpi.com/journal/energies Energies 2019, 12, 3123 2 of 25 on European marginal agricultural lands. These results help to further describe the suitability of industrial crops for the development of social-ecologically friendly MALLIS in Europe.


Introduction
In the targeted 'ideal' bioeconomy, the production of biomass will take social, ecological, and health aspects into account [1] to help achieve the sustainable development goals 2015-2030. From the bioeconomy's ambitions and definitions, conclusions can be drawn that the growth of the bioeconomy demands both a reduction of waste and losses and an adequate supply of sustainably grown biomass [2]. However, an increasing biomass production also carries a higher risk of social-ecological threats, such as increased use of fertilizers and pesticides, negative impacts from land-use changes and additional pressure on water resources [3][4][5]. The EU Horizon 2020 project MAGIC (Grant agreement ID: 727698) was established with the ambition of supporting the mitigation of these risks. This study deals with the basic findings of the 'Low-input agricultural practices for industrial crops on marginal land'.
Low-input agriculture ( Figure 1) generally provides a number of promising practices that can help improve the social-ecological sustainability of biomass production while maintaining economic feasibility [6]. Here, a key parameter is the ratio between on-and off-farm inputs. According to Biala et al. (2007) [6], in low-input agriculture, the use of on-farm inputs should be maximized and off-farm inputs minimized. Currently, there are four concrete and real farming system types which follow these low-input agriculture principles (taken from Reference [7]): (i) Integrated farming, (ii) organic farming, (iii) precision farming, and (iv) conservation farming.

Introduction
In the targeted 'ideal' bioeconomy, the production of biomass will take social, ecological, and health aspects into account [1] to help achieve the sustainable development goals 2015-2030. From the bioeconomy's ambitions and definitions, conclusions can be drawn that the growth of the bioeconomy demands both a reduction of waste and losses and an adequate supply of sustainably grown biomass [2]. However, an increasing biomass production also carries a higher risk of socialecological threats, such as increased use of fertilizers and pesticides, negative impacts from land-use changes and additional pressure on water resources [3][4][5]. The EU Horizon 2020 project MAGIC (Grant agreement ID: 727698) was established with the ambition of supporting the mitigation of these risks. This study deals with the basic findings of the 'Low-input agricultural practices for industrial crops on marginal land'.
Low-input agriculture ( Figure 1) generally provides a number of promising practices that can help improve the social-ecological sustainability of biomass production while maintaining economic feasibility [6]. Here, a key parameter is the ratio between on-and off-farm inputs. According to Biala et al. (2007) [6], in low-input agriculture, the use of on-farm inputs should be maximized and off-farm inputs minimized. Currently, there are four concrete and real farming system types which follow these low-input agriculture principles (taken from Reference [7]): (i) Integrated farming, (ii) organic farming, (iii) precision farming, and (iv) conservation farming.  For each of these farming systems, crop selection was found to be highly relevant for efficient use of resources during their cultivation [8][9][10][11]. The resource use efficiency becomes even more relevant for industrial crop cultivation on marginal agricultural lands ( Figure 2). This is because both the yield potential and the resilience of the agro-ecosystems (their robustness against cropping failures) may be lower on marginal agricultural lands compared to fertile agricultural lands [9,[12][13][14][15][16][17]. According to Elbersen et al. [12], marginal agricultural lands can be defined as 'lands having limitations which in aggregate are severe for sustained application of a given use and/or are sensitive to land degradation, as a result of inappropriate human intervention, and/or have lost already part or all of their productive capacity as a result of inappropriate human intervention and also include contaminated and potentially contaminated sites that form a potential risk to humans, water, ecosystems, or other receptors'. The implementation of a low-input approach that can potentially reduce the risk to humans, water, ecosystems or other receptors is mainly dependent on the farming system and requires site-specific consideration [18,19]. For each of these farming systems, crop selection was found to be highly relevant for efficient use of resources during their cultivation [8][9][10][11]. The resource use efficiency becomes even more relevant for industrial crop cultivation on marginal agricultural lands ( Figure 2). This is because both the yield potential and the resilience of the agro-ecosystems (their robustness against cropping failures) may be lower on marginal agricultural lands compared to fertile agricultural lands [9,[12][13][14][15][16][17]. According to Elbersen et al. [12], marginal agricultural lands can be defined as 'lands having limitations which in aggregate are severe for sustained application of a given use and/or are sensitive to land degradation, as a result of inappropriate human intervention, and/or have lost already part or all of their productive capacity as a result of inappropriate human intervention and also include contaminated and potentially contaminated sites that form a potential risk to humans, water, ecosystems, or other receptors'. The implementation of a low-input approach that can potentially reduce the risk to humans, water, ecosystems or other receptors is mainly dependent on the farming system and requires site-specific consideration [18,19].  [20][21][22]. The other parameters either influence (main constraints) or follow on (combined constraints) from the major biophysical constraints, which limit the site-specific plant growth suitability (Table A1). The economic and social-ecological challenges have been added, due to their increasing relevance for modern agricultural systems [23][24][25][26]. These challenges can render a site marginal under both economic and social-ecological aspects, such as environmental protection, biodiversity conservation, infrastructure, markets and landscape appearance.

Material and Methods
For ethical reasons, low-input industrial crop cultivation on marginal agricultural lands is to be preferred in order to reduce competition for agricultural land use with both food crop cultivation and biodiversity conservation [9,[27][28][29][30]. As favorable agricultural lands should primarily be used for food  [20][21][22]. The other parameters either influence (main constraints) or follow on (combined constraints) from the major biophysical constraints, which limit the site-specific plant growth suitability (Table A1). The economic and social-ecological challenges have been added, due to their increasing relevance for modern agricultural systems [23][24][25][26]. These challenges can render a site marginal under both economic and social-ecological aspects, such as environmental protection, biodiversity conservation, infrastructure, markets and landscape appearance.

Material and Methods
For ethical reasons, low-input industrial crop cultivation on marginal agricultural lands is to be preferred in order to reduce competition for agricultural land use with both food crop cultivation and biodiversity conservation [9,[27][28][29][30]. As favorable agricultural lands should primarily be used for food crop cultivation, this study focuses on the use of marginal agricultural lands for low-input industrial crop cultivation. Consequently, it aimed at:

1.
Mapping the major climatic and biophysical constraints across European marginal agricultural lands; 2.
Assessing the growth suitability of pre-selected industrial crops under the prevailing climatic and biophysical constraints; and 3.
The development of social-ecologically friendly marginal agricultural land low-input systems (MALLIS) for industrial crop cultivation.
To address the above-mentioned research objectives, a thorough literature review was conducted using the search engines of SCOPUS (Elsevier, B.V.) and Google Scholar (Google LLC.). The pre-selection of the industrial crops (Table 1) which was based on a multi-criteria analysis (among others, the maturity of knowledge on industrial crops on marginal land and crops' productivity on marginal land) did not form part of this study. Instead, the study deals with the further evaluation of the growth suitability of 19 promising industrial crops (Table 1), and thus how they meet the requirements for successful development of MALLIS. The following sub-sections present the concepts underlying the key elements of this study. These key elements are (i) the identification of marginal agro-ecological zones (M-AEZ), (ii) the determination of the growth suitability of the pre-selected industrial crops in the prevailing M-AEZ, and (iii) the development of MALLIS for industrial crop cultivation.

The Identification of Marginal Agro-Ecological Zones (M-AEZ)
To achieve the first two key elements, mapping was performed as follows: Marginal agricultural lands were mapped [31] according to the biophysical limitations defined and classified by JRC [20][21][22]. The mapping was limited to a so-called 'agricultural mask'. This mask includes all land that was classified in an agricultural land cover class in at least one of the four Corine Land Cover (CLC) versions (1990, 2000, 2006, and 2012). Further details of the methodological approaches are provided in the following sub-sections.

Determination of the Growth Suitability of the Pre-Selected Industrial Crops in the Prevailing M-AEZ
The approach to mapping the growth suitability of the 19 pre-selected crops involves the identification of the minimum and maximum climate and soil requirements per crop. The growth suitability requirements of the selected industrial crops were determined according to the literature [32,33]. They were used to map and calculate both the distribution and size of the crop-specific growth suitability areas across European marginal agricultural land. The thresholds for the suitability parameters were set as the starting point at which the crop can grow and survive. The suitable area is, thus, given as the area where all suitability factors are within the minimum and maximum range. In this mapping assessment, a distinction was made between suitable and unsuitable area per crop. However, no further classification of the suitable area was made, for example, into high to low suitability. For an easier interpretation of the results, the European land surface was divided into the three agro-ecological zones (AEZ): Mediterranean (AEZ1), Atlantic (AEZ2) and Continental and Boreal (AEZ3) (Figure 3, Table 2). Each combination of an AEZ with at least one biophysical constraint (Table A1) refers to as 'M-AEZ' ( Table 2).    The basic crop-specific biophysical growth requirements were compiled according to Ramirez-Almeyda et al. (2017) [32]. Each biophysical parameter was divided into a number of classes. For instance, the parameter "precipitation" was divided into eight classes (in mm a −1 ): 0-100, 100-200, 200-300, 300-400, 400-500, 500-600, 600-800, and 800-1000 (Table A2). Afterwards, the growth suitability of each crop was ranked according to these classes based on available literature. Additionally, the basic climatic growth requirements of the crops were compiled (Table 3).  a Accumulated mean daily temperature equal to or above than the crop-specific base temperature.
When mapping the crop-specific growth suitability areas, we only considered whether a crop could potentially grow. We did not take different yield levels into account. In the constraint-specific ranking, classes 0 and 1 were denoted as not suitable. Therefore, if any of the basic climatic growth requirements are not met or any of the constraint-specific rankings falls within class 0 or 1, the area is designated as 'not suitable'. The result was an overview of the potential growth suitability of the pre-selected industrial crops across European marginal agricultural land. This means that only agricultural areas were considered; woodlands and urban areas were excluded from the mapping of marginal agricultural land.

Definition and Methodology of Marginal Land Low-Input Systems (MALLIS) Development for Industrial Crops
In this sub-section, the definition of best-practice low-input management systems for the pre-selected industrial crops (Table 1) is elaborated. This ties in with current knowledge on best low-input agricultural practices for food crop production on good soils [6]. The concept of best-practice low-input agricultural cropping systems considers management approaches from many categories of agricultural production, including organic, integrated, conservation agriculture and mixed crop-livestock farming [35][36][37]. These all have one constant: Low-input agricultural practices seek to optimize the use of on-farm resources while minimizing off-farm resources [6,35,36]. This leads to a more 'closed' cycle of production (and less external inputs) [37]. Note, that this more closed production cycle requires both more advanced agronomic skills [38,39] and additional links within the value chain, such as application of biochar [40][41][42][43][44][45][46][47][48][49] or phosphate salt recovery from the digestates [50,51]. Therefore, practical guidelines for industrial crops are also under development within the MAGIC project.
Agronomic strategies for the successful application of low-input agricultural practices in a crop management system should be seen as a set of strategies that take into consideration both the interactions between plants, soil, the atmosphere and the efficient use of inputs to enable the highest output with minimal (on-farm and/or off-farm) input supply [6,[52][53][54]. Agronomic strategies for low-input systems may also match good agricultural practices-cultivation practices that address economic, social and environmental sustainability [37] for high-quality food and non-food agricultural products [38,55]. Such practices include the implementation of appropriate crop rotations, pasture management, manure application, soil management that maintains or improves soil organic matter, and other land-use practices, as well as conservation tillage practices [8,37,48].
Diversity in crop rotations is a way to reduce reliance on synthetic chemicals, control weeds and pests, maintain soil fertility and reduce soil erosion, prevent soil-borne diseases, leading to the reduction of off-farm inputs [54]. Reduced soil tillage is a way to reduce soil erosion, improve water buffer capacity, and increase both soil fertility and organic matter [37]. Water management is a major challenge in the Common Agricultural Policy (CAP) and requires the monitoring of soil and crop water status to schedule irrigation efficiently. Fertilizers and agrochemicals should be applied following the good agricultural practices, e.g., to replace only the amount of nutrients that were extracted by harvest [37].
Crop protection should be done in a way that maximizes the biological prevention of pests and diseases, in particular by promoting integrated pest management (IPM) and though appropriate rates and timings of agrochemicals. Preventive crop protection can also be supported by the selection of resistant cultivars and varieties, crop sequences, crop associations (e.g., intercropping), and proper cultural practices [35].
The development of 'marginal agricultural land low-input systems', referred to as 'MALLIS', is based on the following definition: 'MALLIS is defined as a set of low-input practices which are relevant management components to form viable cropping systems on marginal (arable) lands under specific climatic conditions and are sustainable in both socio-economic and environmental terms'. The implementation of MALLIS should enable farmers to cultivate industrial crops on marginal agricultural lands, considering both economic and socio-environmental aspects. Consequently, MALLIS should not only allow for profitable net farm income under the challenging biophysical growth conditions of marginal lands. It also helps to (i) reduce off-farm inputs, such as synthetic fertilizer, pesticides and energy (e.g., for water pumps, fuel, crop harvest machinery, storage, processing, etc.) and (ii) mitigate negative macro-economic externalities (GHG emissions, biodiversity loss, ground-and surface water contamination, soil organic matter loss, erosion, degradation, land-use change), while (iii) ensuring feasible economic benefits at farm level. Therefore, the development of MALLIS considers not only the biophysical constraints, but also socio-economic and ecological demands of the respective areas.
The conceptualization of MALLIS development always begins with the selection of the most promising industrial crop, because all other agricultural practices (tillage, fertilization, weeding, irrigation, etc.) strongly depend on the type and site-specific performance of the crop. This MAEZspecific growth-suitability ranking (and mapping) of the pre-selected industrial crops was based on the crop-suitability rankings. The basic climatical growth suitability thresholds are presented in Table 3. After the identification of suitable crops, the conceptualization of MALLIS for MAEZ was done on a general level (regional scale), since detailed best practice recommendations for the optimized management of agricultural practices very much depend on local conditions (field-to-farm scale) [56][57][58][59][60]. Therefore, the MALLIS for the new field trials to be conducted in the MAGIC project (field-to-farm scale) were developed considering three main MAEZ criteria: • The crop's performance according to site-specific climatic and geographic conditions, especially under given biophysical constraints; • The kind and quality of biomass required in the given infrastructure, processing industries and distribution channels (markets); • The agricultural status of the farm(s), e.g., the techniques, knowledge and resources available to ensure successful cultivation of the crop.

Marginal Agro-Ecological Zones in Europe
As illustrated in Figure 2, there are various biophysical constraints and socio-economic challenges which need to be considered for MALLIS development. Table 2 shows the relevance of the numerous biophysical constraint combinations within each of the three AEZ. According to category 1 ('natural constraints'), the total marginal area across European land surface amounts to 646,833 km 2 ( Table 2)-an area as large as France. However, this marginal agricultural land is widely scattered across Europe ( Figure 4). Furthermore, there were 38 combinations of ≥ 2 constraints identified ( Table 2). Across Europe, the most prevailing constraints are adverse rooting conditions, (155,519 km 2 ), adverse climatic conditions (112,096 km 2 ) and excessive soil wetness (108,081 km 2 ). The total marginal arable land characterized by soil constraints accounts for about 535,000 km 2 . This is about 155,000 km 2 more than reported by Gerwin

The Growth Suitability of the Pre-Selected Industrial Crops in the Prevailing M-AEZ
Potentially suitable industrial crops were identified for virtually all types of marginal agricultural land across Europe (Table A3). Each AEZ appears to have its own best-adapted industrial crops. A closer look at the type of biomass reveals that, for instance, oil crops are more suitable for Mediterranean regions than for the Atlantic region (Table A3). Among the woody lignocellulosic

The Growth Suitability of the Pre-Selected Industrial Crops in the Prevailing M-AEZ
Potentially suitable industrial crops were identified for virtually all types of marginal agricultural land across Europe (Table A3). Each AEZ appears to have its own best-adapted industrial crops. A closer look at the type of biomass reveals that, for instance, oil crops are more suitable for Mediterranean regions than for the Atlantic region (Table A3). Among the woody lignocellulosic crops, Siberian elm outperforms poplar in the Mediterranean region (Table A3). The dominating lignocellulosic crops are tall wheatgrass, followed by reed canary grass and miscanthus (Table A3).

Marginal Agricultural Land Low-Input Systems (MALLIS) for Industrial Crop Cultivation
Sections 3.1 and 3.2 revealed both the major M-AEZ in Europe and the growth suitability of the pre-selected industrial crops. This section explains how MALLIS could be developed using the information on M-AEZ and the crops' growth suitabilities. Furthermore, it discusses which other aspects need to be taken into account for MALLIS development in order to improve both the economic and social-ecological sustainability of the MALLIS in the long term.

Agricultural Measures for MALLIS Development
The potential effects of structured and systematic agricultural measures on agriculture facing biophysical constraints are provided in Tables A4 and A5. Furthermore, the literature review revealed that there are several ways to overcome each of the biophysical constraints. Tables A4 and A5 provide an overview of the suitability of agricultural management options for dealing with the prevailing biophysical constraints on marginal agricultural lands. For example, the use of mulch helps to increase the soil thermal time, and thus increase the yield level in regions affected by water limitations and low temperatures [62].

Environmental Threats and Social Requirements
MALLIS implementations at a regional scale should also take both environmental threats and social requirements into consideration. Marginal agricultural lands could be characterized as fragile environments being highly susceptible to any types of external disturbance and input [6,12,63]. Key measures that can be highly recommended for the improvement of resilience include (i) the selection of low-demanding industrial crops (reduces the amount of fertilizers, and thus the risk of nutrient-leaching) [27], (ii) the development of heterogeneous landscape concepts (many small fields rather than only a few large fields) [64][65][66][67], and (iii) the implementation of agricultural diversification measures (intercropping, crop rotations, wildflower strips) [35,68,69]. Consequently, the assessment of the environmental performance of MALLIS should not be exclusively based on the global warming potential, but also on a number of other environmental impact categories, such as human toxicity threats, marine ecotoxicity, freshwater eutrophication and freshwater ecotoxicity, biodiversity and soil quality, pollution [70], and use of resources, e.g., water resources [71]. However, to enable a long-term sustainable implementation of MALLIS, besides the environmental impact categories, the social demands and the economic and market aspects must also be taken into account. The potential and viability of agricultural investments have to take into account land and labor costs, inputs, such as mechanical equipment costs, and income (which is linked with the market opportunities) [72]. The socioeconomic impacts can be measured via quantitative and qualitative parameters [73]. Moreover, aspects related to technological viability should also be taken into consideration. The yield loss associated with cultivation on marginal agricultural land may lead to higher contents of nutrients, such as nitrogen and potassium in the biomass, which may complicate further processing of the biomass [74]. Generally, this means that the prevailing structures of the existing agricultural systems [75], the farm typology [76], and the behavior patterns of the rural communities [24,77] require specific bottom-up research structures, such as the Integrated Renewable Energy Potential Assessment (IREPA) [78]. This would enable a better adaptation of MALLIS to the farm diversity [76,77] and the local community.
Finally, this could potentially have a positive influence on the overall public acceptance of the MALLIS [79].

Biodiversity Conservation
Another aspect worthy of discussion is the ecosystem functionality [80] of the pre-selected industrial crops in terms of biodiversity conservation. Concerning the soil ecological functions fulfilled by pedofauna, recent works on the following of bioenergy crops establishment on marginal contaminated soils showed that belowground fauna was stimulated [81]. Higher densities and diversity of soil invertebrates were found under miscanthus compared to annual cropping systems [82], as well as the positive effect on microbial diversity [83]. These crops were specifically selected as representative of those that deliver the most important crop-based biomass resources for current biomass industries. However, the recent (alarming) decrease in pollinator abundances across Central European landscapes [23,25] may induce changes in the priorities for crop selection, and thus the development of MALLIS in the future. For example, pollinator-supporting traits, such as nectar provision and high resistance to pests and diseases could become more important than economic traits, such as biomass yield and biomass quality if public awareness of this topic continues to increase [84]. There are a number of reports on alternative pollinator-supporting industrial crops, such as perennial wild plants [85][86][87][88][89], cup plant (Silphium perfoliatum L.) [90][91][92], sida (Sida hermaphrodita L.) [93][94][95], and amaranth (Amaranthus hypochondriacus L.) [96][97][98]. However, many of the pre-selected industrial crops are also expected to have positive effects on pollinators. These include camelina [99][100][101], crambe [100,102], safflower [103,104], lupin [105,106], cardoon [107,108] and willow [63,109,110]. In addition, the suitability of the MALLIS for habitat networking in combination with other highly diverse cropping systems, such as species-rich meadows [111] should be investigated to improve the overall efficiency of the MALLIS for biodiversity conservation. Also, marginal land can anchor rich biodiversity components (plants with high significance for locals, e.g., for medicinal or food purposes), and change of land use should take this element into account [112].

Explanatory Setup of a MALLIS on a Shallow Stony Soil
This section provides an example on how MALLIS could be implemented on a marginal agricultural site characterized by two biophysical constraints [21]: (i) Shallow soil (<35 cm topsoil depth); and (ii) stoniness (≥15% of topsoil volume is coarse material, rock outcrop or boulder). Due to these constraints, both the rooting conditions and the soil fertility are lower than in deep soils. It is economically not feasible to grow food crops under these conditions, and thus, the cultivation of certain industrial crops would not compete with food security on sites like this. However, not all industrial crops are able to grow well under these conditions either. Thus, the identification of a best-adapted industrial crop is the first step in developing a site-specifically suitable MALLIS. In this case, perennial crops, such as miscanthus and switchgrass are found to be suitable because (i) they do not require soil tillage and sowing each year compared to annual crops which helps both increasing soil fertility [113,114] and reducing erosion [115] in the long term, (ii) they can manage to root deep enough despite shallow soil, because their root systems are stronger and more developed than those of annual crops, and (iii) the climatic conditions meet the crop-specific growth requirements. In this case, the perennial C4-grass miscanthus (Miscanthus × giganteus Greef et Deuter) was chosen ( Figure 5), due to its low demanding nature and high biomass yield potential under challenging conditions [116]. This is part of ongoing research on the cultivation of miscanthus on marginal agricultural lands in MAGIC [117]. In the EU-funded project 'GRACE' (Grant agreement ID: 745012), it is also investigated how the cultivation of miscanthus on marginal agricultural lands can be optimized [118].
Preliminary results of a field trial in southwest Germany indicate that miscanthus can establish well ( Figure 5) under the given conditions [119]. The dry matter yield (DMY) averages 13 Mg ha −1 a −1 from the second year onwards [119]. This is a medium DMY level compared with miscanthus grown on good soil [116,120,121]. However, it should be mentioned that miscanthus requires very low nitrogen (N) fertilization [122], especially when harvested for combustion in winter [60,123]. This is because miscanthus has very efficient nutrient-recycling when harvested in winter [79,124]. The low demand for nitrogen fertilization renders a key low-input factor [6,32] of this MALLIS, due to an improvement of the on-farm/off-farm-ratio in favor of the on-farm inputs. Furthermore, low N fertilization levels help improve the ecosystem services of miscanthus cultivation, such as groundwater protection, environmental protection [26,80,120], while maintaining the soil nitrogen balance [125]. Overall, both the improved ecosystem services and low production costs justify the medium DMY level of miscanthus at comparable marginal agricultural sites (shallow soil, stoniness, etc.). Consequently, MALLIS must be developed under careful consideration of the given site-specific conditions [57]. Therefore, the major development steps are (i) the identification of the growth conditions and the biophysical constraints [20,20,22], (ii) the selection of best-adapted crops, and (iii) the conceptualization of best-adapted site-specific low-input agricultural practices.

Recommendations and Outlook
The results of the suitability mapping are in line with the available literature [17,56,61,121,[126][127][128]. Uncertainties were identified within the basic climatic requirements, because in some cases the distribution does not meet the expectations. This could be caused by the wide genetic variation within both perennial crop species, such as switchgrass and miscanthus, and annual crop species, such as camelina and safflower. To improve the representability, the basic climatic growth requirements should either include ranges (minimum-maximum) for each parameter per crop or different genotypes for each crop. For instance, there is a wide genetic variation among miscanthus genotypes with regard to their heat and cold tolerance [129][130][131]. For some annual industrial crops, such as camelina and safflower, winter-annual genotypes are also available [132][133][134][135]. It would very likely further increase the potential growth suitability of the pre-selected industrial crops to take these genetic variations into account. Nevertheless, this study provides valuable first insights into the potential distribution of growth suitability, contributing to an improved crop selection for the development of MALLIS across Europe.
The results of this study indicate that there are large areas potentially available for industrial crop cultivation. This is in line with available literature [17,56,61,[126][127]. In many cases there are ≥2 crops suitable for the same area (Table A3). Thus, careful consideration should be given to the selection of crops or their most favorable combination according to the site conditions [136]. For an

Recommendations and Outlook
The results of the suitability mapping are in line with the available literature [17,56,61,121,[126][127][128]. Uncertainties were identified within the basic climatic requirements, because in some cases the distribution does not meet the expectations. This could be caused by the wide genetic variation within both perennial crop species, such as switchgrass and miscanthus, and annual crop species, such as camelina and safflower. To improve the representability, the basic climatic growth requirements should either include ranges (minimum-maximum) for each parameter per crop or different genotypes for each crop. For instance, there is a wide genetic variation among miscanthus genotypes with regard to their heat and cold tolerance [129][130][131]. For some annual industrial crops, such as camelina and safflower, winter-annual genotypes are also available [132][133][134][135]. It would very likely further increase the potential growth suitability of the pre-selected industrial crops to take these genetic variations into account. Nevertheless, this study provides valuable first insights into the potential distribution of growth suitability, contributing to an improved crop selection for the development of MALLIS across Europe.
The results of this study indicate that there are large areas potentially available for industrial crop cultivation. This is in line with available literature [17,56,61,126,127]. In many cases there are ≥2 crops suitable for the same area (Table A3). Thus, careful consideration should be given to the selection of crops or their most favorable combination according to the site conditions [136]. For an adequate crop selection, site-specific conditions other than the growth suitability should also be considered, such as the local social-ecological needs and the distance to the markets. For instance, if a site is prone to erosion, a perennial cropping system would be preferable to an annual (rotational) cropping system [115,[137][138][139]. This could help ensure a more sustainable biomass production from both an environmental and economic point of view in the long term [140]. It would reduce the risk of further degradation through erosion, and thus help maintain or even improve the resilience of the given agroecosystem [14,141,142].
In this study, the growth suitability of the crops did not include yield and quality levels. This means that potential differences in yield or biomass quality between suitable industrial crops for the various types of marginal land across Europe remain unclear. Furthermore, the study did not cover macroeconomic aspects, such as infrastructure and market accessibility, which also play a vital role in the determination of the best site-specific crop selections across European marginal agricultural lands. In some cases, the suitability of an industrial crop also depends on the local conditions of the farms. For example, either the technical equipment or the know-how may impede an optimal MALLIS implementation. However, this study contributes to the ongoing research into how biomass for a growing bioeconomy can be provided in low-input systems, as the growth suitability of the crops forms the basis for the successful development of MALLIS. The site-specific growth suitability presented here are also available in the form of a decision support system [136]. This aims at enabling the selection of suitable case study regions for further optimization of site-specific MALLIS for industrial crop cultivation. In addition, the missing links mentioned above, including detailed information on the best crop-and site-specific harvesting technology and guidelines for farmers are also explored in the EU Horizon 2020 project MAGIC (GA 727698) [117]. As climate-change-forced shifts in the distribution of both marginal agricultural land and growth suitability of the industrial crops are to be expected [126,[143][144][145], they are also under investigation [58,146]. This could help to better prepare European agriculture for the projected severe effects of climate change [143,144,147].

Conclusions
This study introduces the concept of marginal agricultural land low-input systems (MALLIS) for industrial crop cultivation. MALLIS are defined as sets of agricultural low-input practices to form viable cropping systems on marginal agricultural lands under specific climatic conditions. These sets of practices are intended to be holistically sustainable in both social-ecological and economic terms. The study identified the climatic and geophysical constraints on biomass production and the ability of 19 industrial crops to cope with these limitations. Overall, the industrial crops showed high suitability for low-input cultivation on marginal agricultural lands across Europe. However, further investigations of MALLIS are required to investigate their social-ecological sustainability and climate change effects on marginal agricultural lands. Acknowledgments: The authors are thankful to Nicole Gaudet for proof-reading the manuscript.

Conflicts of Interest:
The authors declare no conflict of interest. The funders had no role in the design of the study; in the collection, analyses, or interpretation of data; in the writing of the manuscript, or in the decision to publish the results.
Appendix A Table A1. Overview of the three categories of marginality constraints as classified within this deliverable. Category 1 was adapted from Reference [20]. Categories 2 and 3 were developed based on the literature review.

Constraint Category
Factor Category Thresholds/Specifications Lands which are currently productive, but will be transformed into marginal lands in the long term if not managed properly (also, lack of know-how or lack of awareness from farmers/government).    lifecycle  1  1  1  1  1  0  1  1  2  0  0  1  1  Perennial lifecycle  2  1  3  2  2  1  2  2  3  1  1  2  1  Table A5. Suitability ranking of selected components of agricultural management systems for competing with the prevailing biophysical constraints on arable marginal lands (from −3 = strong negative effect to +3 = strong positive effect).