To integrate or to segregate food crop and energy crop cultivation at the landscape scale? Perspectives on biodiversity conservation in agriculture in Europe

A decrease in available arable land per capita due to a rapid population growth [1], a growing number of animals in agriculture that need to be fed and actions for climate change mitigation and adaptation [2, 3] are aggravating the competition for land resources. These demands will result in conversion of additional land to agricultural use and/or further agricultural intensification on existing productive land [4, 5]. In this context, the potential impacts of bioenergy production on food security have been highly controversial (e.g. recent food vs. energy debate). Bioenergy expansion has exerted significant land-use pressures, intensifying the competition for land, water and other natural resources with food production, which resulted in various concerns over environmental and social sustainability [6, 7]. On the other hand, energy crop production can bring opportunities for increased agricultural production and productivity, rural development, agricultural diversification and climate change mitigation [8–10].

Past research on bioenergy expansion mainly focused on markets and feedstock distribution [9, 10] or bioenergy potentials and land availability on national and global scales (e.g. [11–13]). Accordingly, studies looking into the impacts of bioenergy production on biodiversity and/or ecosystem services were mainly undertaken from national or global perspectives (e.g. [14, 15]). A number of spatially explicit modelling studies have also emerged to evaluate the biodiversity responses from energy crop production at regional, national and pan-national scales, maize in particular [16–19], some of which considered land-use change scenarios involving energy crop cultivation on marginal land of high nature value [20, 21]. In addition, many field-scale or empirical studies with a focus on a particular energy crop have been conducted to compare biodiversity impacts with those of row crops or grasslands [22–24]. However, a small but increasing number of studies have also tackled the biodiversity issues associated with energy crop production on a farm and landscape scale. These studies included the spatial and temporal aspects of bioenergy-driven land-use change and/or adaptations of the whole farming system (e.g. [25–30]).

The questions of where and to what extent energy crops should be cultivated in agricultural landscapes to maintain and/or enhance biodiversity and ecosystem service provision cannot be easily answered. There is a discrepancy between the findings of coarse-scale/modelling studies (mostly indicating negative impacts) and fine-scale field-based studies (mostly suggesting positive effects in comparison with traditional agricultural crops). The biodiversity effects of energy crop cultivation can also significantly differ from climatic regions and crop types [31] and specific regional circumstances, such as landscape pattern and structure [24, 32]. Therefore, farming system-based approaches are more relevant to answer this question. The potential impacts on local biodiversity from energy crop cultivation needs to be considered in the context of severe biodiversity decline on agricultural land [33, 34]. Thus, we suggest linking the discussion on biodiversity impacts from energy crop production with those of agricultural land-management strategies for biodiversity conservation, namely ecological intensification [35] and the land-sparing vs. land-sharing debate [36]. Addressing both issues may be possible through the right selection of energy crop [6] and the optimal agricultural practices [37]. Therefore, we focus on the spatial-temporal scale of a farming or a production system because the solutions for food and energy security, biodiversity and ecosystem services must be considered altogether. To define the research scope for this paper, our scenarios and discussions only focus on direct land-use change, while we fully acknowledge the importance of challenges resulting from indirect land-use change (iLUC) [38]. The latter would have to be considered in a future discussion. The aim of this paper is to explore and stimulate a discussion about innovative agricultural production systems, which can meet multiple goals at the same time—production of food, fodder and biomass feedstock for energy and industry uses, biodiversity conservation and promotion of ecosystem services. In this paper, we describe nine conceptual land-use scenarios, which both segregate and integrate energy crop cultivation from/with food and fodder production systems on land having different levels of productivity. Potential impacts of those scenarios are discussed on the basis of yield-biodiversity relationships [39]. We also discuss opportunities and barriers for these scenarios in relation to the biodiversity outcomes based on the findings from past and ongoing studies. Finally, we suggest alternative scenarios that integrate food and bioenergy systems, involving first-generation energy crops (i.e. usable as food/fodder or energy feedstock) and dedicated energy crops (i.e. only usable as energy feedstock), taking account of the productivity levels of the farmland.

Integration or segregation, or in other terms “land sparing” or “land sharing”, are two land-management strategies for biodiversity conservation worldwide. The debate originated from their efficiency in different ecosystems, landscapes and biogeographic regions [36, 40, 41]. These two strategies can be easily adapted to the question on where to cultivate energy crops (e.g. [42]) to avoid negative land-use change (LUC) effects, including the impacts on biodiversity and ecosystems. The concepts of land sparing and land sharing can be applied to any energy crop production system on a range of spatial scales (sensu [43]). There are two types of land-sparing approaches. The first approach restricts energy crop production to less productive, degraded, marginal or abandoned agricultural land to minimise the LUC effects including iLUC [40, 44–46]. The second approach cultivates energy crops on land that has been freed up for other purposes through (sustainable) agricultural intensification, resulting in higher yields per hectare (ha) [47, 48]. In contrast, a land-sharing approach for bioenergy involves combining systems of food and energy production on a field or a set of fields (CFE). There are two possible systems: (i) both energy and food crops are included in the crop rotation cycles or (ii) energy crops are planted as structural elements in landscapes such as “energy hedgerows” in agroforestry systems (e.g. [26]). The integration of both crops can be implemented on both land of high- and low-productivity levels. Trials have found CFE to be considerably less resource demanding and more amenable to sustainable production than conventional food-cropping systems [49]. In the developed countries, CFE systems are also expected to improve overall ecosystem functions [50]. Land sharing of food and energy crops could thus be a means of supporting sustainable intensification of agriculture [51].

For a comparative discussion, we have sketched a diagram to visualise those alternatives of integrating or segregating food and energy cropping at coarse spatial scales (Fig. 1). To redefine the idea of restricting energy crop cultivation to less productive land, we have adopted some definitions by Shortall [46]. However, the definitions and characterizations of land productivity were currently highly ambiguous. We have therefore opted for a pragmatic classification into (i) high-grade productive agricultural land which is environmentally and economically suitable for all food crops and (ii) lower grade land which is economically marginal for food production (i.e. areas where cost-efficient production is not possible due to a combination of given conditions including low soil fertility, adverse climate, impediments to cultivation techniques, agriculture policies and macroeconomic and legal conditions). The “marginality” of the land is determined with respect to the particular economic opportunities offered by land-use choices [52]. Areas with extremely poor soils, harsh weather environments and severely degraded areas are classified into (iii) land unsuited for food production. Nevertheless, the notion of this land category still contains technical and economic assumptions that production of energy crops is possible and economically feasible. Owing to advanced plant-breeding and genetic modification technologies, future energy crops are likely to be more resilient, requiring fewer nutrients and water [46]. Our final land category of (iv) “natural” land includes all types of land which are reserved for nature conservation. Those areas must be completely excluded from any agricultural activities (Fig. 1), although the use of biomass waste and forest residues from such areas for energy purposes is possible (e.g. [53]). We are also aware that the classification of land categories can vary across countries and regions. For example, remnant natural land hardly exists in rural landscapes in Europe, which are mostly dominated by cultural landscapes. Thus, current European nature conservation policy has more of a focus on species-rich farmland, such as designated Natura2000 sites and high nature value (HNV) farmlands [54]. These areas are classified as economically marginal land in our classification, not into this “natural” land category. On the other hand, “rewilded” areas are included in the natural land category [55].

We have adopted a coarse yield-biodiversity relationship [40] to characterise the four land categories described above (Fig. 1). The level of biodiversity in the low-grade land unsuitable for food production may be much lower than assumed in Fig. 1, in particular where the unfavourable conditions result from soil degradation or contamination. For such areas, restoration or phytoremediation are important strategies. However, this is beyond the scope of the paper.

The “pre-bioenergy” scenario (Fig. 1) presents the distribution of dominant land cover for the three land categories. Food crops are mainly produced on the most productive agricultural land. Crop yields on that land are high, owing to high soil fertility, high external inputs and intensive production systems. Accordingly, levels of biodiversity on that land are very low (Fig. 1; [33]). As land becomes more marginal, crop yields decline, options for intensification of management become economically restricted and therefore production systems become less intensive. Permanent grassland is becoming the dominant cover on such land. In some regions with economically marginal land, low-intensity farmland, sometimes with traditional farming practices, prevails [56]. Levels of farmland biodiversity on such land are high in comparison to those on productive agricultural land [54, 56]. On the low-grade land unsuitable for food production, high proportions of abandoned agricultural land occur due to mainly economic reasons, and high proportions of land not utilised for agriculture also occur due to environmental reasons. Extensive pastoral systems may exist on such land. Within the coarse yield-biodiversity relationship applied here, we expect to find high levels of biodiversity on such land. The highest levels of biodiversity are expected on “natural” land which is exempted from agricultural production (Fig. 1).

The “segregation of food and energy crops” scenario (Fig. 1) follows the “food-first” directive, postulating that food crops should be produced on the most productive land and energy crops should only be produced on land unsuitable for food production. In comparison to the “pre-bioenergy” scenario, this scenario involves conversion of only the most economically marginal lands (e.g. [57]), such as extensive agriculture, grazing, hay production and abandoned agricultural lands. Environmental impacts of restricting energy crops to these lands were discussed in our previous papers [45, 58]. From the perspective of biodiversity conservation in agriculture, this LUC may not result in positive biodiversity outcomes; thus, the scenarios should be carefully scrutinised. This is highly relevant to the European context. First, this scenario is unlikely to result in any improvement for biodiversity in the existing intensive farming areas, especially when the productive land is used for only food production and no steps are taken to support farmland biodiversity. Secondly, these types of land are generally characterised by comparatively high levels of farmland biodiversity, especially in Europe. Depending on the biodiversity values of the energy crops chosen, the biodiversity values of the land category converted and the functional integration (i.e. supplementing/complementing habitat functions) of the novel crops into the landscapes, the outcome may be positive, neutral or negative. Therefore, thorough regional impact assessments are necessary. Lastly, HNV farming systems in Europe which may exist in those lands are often threatened by non-sustainable socio-economic conditions [59]. The integration of energy crops can help to retain the HNV farming system and the associated species depending on the extensive and low-impact agricultural lands. The diversification also results in the opportunity to increase farmers’ income and a stabilisation of the rural communities without harming the ecological component of those socio-ecological systems (sensu [59]). Permanent grassland with extensive grazing is an important type of land use on economically marginal land and land unsuitable for food production.

The “integration of food and energy crops” scenario (Fig. 1) combines the ideas that (i) a diversification of food production systems through energy crops does not undermine food security [50] and (ii) sufficient and sustainable cultivation of energy crops on marginal land is not achievable because of low yields, low feedstock quality, higher land take and the associated environmental and social issues [45]. In comparison to the segregation scenario, LUC occurs predominantly on the productive agricultural land and also on the economically marginal land. In the latter case, the boundaries of economic marginality shift due to a presumably higher profitability of combined food and energy systems in comparison to food production only. Due to the higher yields on the productive land, land unsuitable for food production is less affected by the expansion of energy crop production. A possible biodiversity outcome of this scenario depends on how the integration is undertaken. Hence, more detailed scenarios of the respective options at the farm to landscape scale are presented and discussed in the following section. In general, this scenario has the potential to halt or even reverse the biodiversity loss from intensive agriculture—if sustainable intensification is understood as ecological intensification [35]. For instance, a strategic integration of perennial crops is suggested as a means for the restoration of agroecosystems [60]. However, the benefit of such changes may mostly affect functional components of biodiversity and more common species as ecosystem service providers and not so much affect species of nature conservation concern. For the permanent grassland in the economically marginal land class affected in this scenario, considerations similar to the ones made for the segregation scenario would apply. In this context, integration of dedicated energy crops—mostly woody perennial crops, such as a variety of eucalypts—into extensive grazing have been also suggested in Australia to remedy natural resource management problems (e.g. salinity, biodiversity loss, water quality degradation), to increase farmers’ profits through product diversification [61] and to explore carbon-offset opportunities [62].

To reconcile agricultural yields for food and/or energy crops and conservation of farmland biodiversity at the landscape level, choosing between pure land-sharing or land-sparing strategies may lead to suboptimal outcomes in many cases. This is because “optimal” landscapes will likely have features of both strategies [63]. Moreover, there is a wide variation in definitions of what constitutes the spared land, ranging from natural habitats, grazed grasslands to field boundaries [43]. Likewise, there is a range of ideas about the spatial scale at which land sparing is applicable [40, 64]. Ekroos et al. [43] suggested a multiple-scale land-sparing approach where the respective scale at which land-sparing options were applied corresponds to the most appropriate aspects of biodiversity conservation.

For our scenarios, the sparing of habitats within landscape mosaics is of relevance. Many common species including ecosystem service providers (e.g. biocontrol agents and pollinators) move tens to thousands of metres during a day or season to use multiple habitats that offer complementary or supplementary resources [65, 66]. To this effect, a field of a low-input dedicated energy crops may function as a “spared” habitat in providing resources to farmland species, comparable to a spared semi-natural habitat. Accordingly, including an energy crop into a crop rotation cycle or in double-cropping systems may function as a sharing type of measure for supporting biodiversity.

Looking at the options for segregating or integrating food and energy crops at the farm to landscape scale (i.e. a spatial cluster of farms), pure energy farms would represent a pure sparing solution and combined food and energy farms a sharing solution. In reality, farms specialising only on the production of energy crops are uncommon in an agricultural context. Therefore, we focus our scenario development for the integration of both food and energy crop production. Such integration could be achieved by converting both productive and economically marginal agricultural lands, albeit in each case different energy crops are involved, including either intensive crops (i.e. first-generation energy crops) or dedicated energy crops (e.g. grass and woody perennial crops), which are often less intensive in management [67]. For the latter, cultivation on land unsuitable for food production may also be viable. We therefore include scenarios with first-generation energy crops only (A1, B1; Fig. 2), spatio-temporal combinations of first-generation energy crops and dedicated energy crops (A2, B2; Fig. 2) and dedicated energy crops only (perennial crops, A3, B3, C1–3; Fig. 2). The options for an integration of food and energy crops therefore differ in the spatial and temporal combinations of the respective crops, in particular. Use of waste materials, slurry or manure for bioenergy is not explicitly taken into account here but might be an integral part of the respective bioenergy chains.

The scenarios presented in Fig. 2 are not exhaustive, and several more options for integrating food and energy crops may exist. The detailed description of nine scenarios with a range of production intensities and systems (A1–C3) enables us to identify the risks and the opportunities associated with biodiversity conservation within the respective scenarios. Possible impacts of land-use change induced by energy crop integration across scenarios are not taken into consideration. All scenarios (Fig. 2) were realised in a farming context or at least tested in experimental setups. Their characteristics are presented below together with examples of their implementations.

By comparing scenarios of segregated and integrated food and energy cropping at coarse spatial scales with yield-biodiversity relationships across land productivity classes (Fig. 1), we could explore possible risks and opportunities for biodiversity conservation across different bioenergy deployment strategies. We reasoned that a strict segregation of food and energy crops with a restriction of energy crop cultivation on economically marginal land and land unsuitable for food crop production is unlikely to mitigate the current situation of biodiversity loss originating from intensive agricultural systems on productive land, especially in the European context. If no alternative measures for biodiversity conservation were developed within intensive agriculture, we would share the concern of Ekroos et al. [43], who argued that segregation strategies could increase the existing negative effects on biodiversity, and even erode ecosystem services that could underpin sustainable, high-yielding farming systems based on ecological intensification [35]. Our analysis demonstrated that most of the LUC in the sparing strategy would occur within the economically marginal land category. Further regional/local investigations are certainly required to establish whether energy crops on this type of land would (i) replace HNV farmland and/or habitats of nature conservation value [20], potentially resulting in habitat loss and biodiversity decline, and (ii) sustain economically threatened socio-ecological HNV farming systems [59] and thereby support conservation. Strohbach et al. [54] suggested that if the social component underlying a HNV farming system has ceased to exist or is not sustainable anymore, new HNV farming systems based on different techniques and diversification of farm income would have to be developed. Integration of energy crops into those farming systems may be one possible “transformation strategy” (sensu [59]). However, an introduction of novel crops will most likely alter farmland species communities and may conflict with existing nature conservation targets.

The scenario of integrating food and energy crops at a coarse scale (Fig. 1) resulted in LUC on productive land. Integration of energy crops into intensive and industrialised production systems may facilitate transformation towards agroecological cropping practices [94] and support sustainable intensification [50]. More evidence is required to answer questions over whether such integration will (i) increase the provision of ecosystem services relevant for agriculture, (ii) promote endangered farmland species and (iii) facilitate migration of non-farmland species through the agricultural matrix. Those functions were so far only connected to semi-natural habitats [41]. Exemplary studies have already given evidence that the integration strategy could work (e.g. [24, 41, 72, 73, 91]). Nevertheless, many open questions remain to be investigated that include the amount of land that needs to be converted to energy crops and the landscape conditions under which ecosystem service provisions vary over time and interactions between different sustainable intensification strategies [43].

The development of scenarios on integrated and segregated food and energy cropping was intended to stimulate and broaden the discussion on the development of more sustainable, especially more biodiversity-friendly, farming systems that contribute to energy as well as food security, within the context of current debates on land sparing and sharing, and ecological intensification. We argue that energy crop production (as well as other industrial crops) should not be rationalised away from the food security debate especially in Europe where land resource is highly limited. Thus, further evaluations are required for both land-sparing and land-sharing scenarios. We propose that it is important to consider how the wide range of energy crops, with their specific traits and management requirements, may play a role when devising novel and more sustainable agricultural production systems. Such integration may result in either positive or negative outcomes, depending on how well it is designed.