Bioenergy, Food Production and Biodiversity – An Unlikely Alliance?

The global demand for bioenergy is set to increase to a point where it may supply up to one-third of global primary energy by 2050 (IEA, 2012). Current cultivation of biofuel and bioenergy crops has attracted considerable criticism due to their encroachment into areas traditionally occupied by food crops and natural and seminatural ecosystems. Among the most notorious of these bioenergy crops are sugarcane, jotropha, ethanol maize, and palm oil, crops closely associated with biodiversity and habitat loss, water deficit, and perhaps most ironically, given their potential for greenhouse gas emissions reduction, negative impacts on carbon storage and sequestration (Fargione et al., 2008; Martinelli & Filoso, 2008; Koh et al., 2009; Romijn, 2011). In the medium-term bioenergy demand is likely to be met by so-called second-generation (2G) lignocellulose crops, principally perennial grasses and woody trees (Somerville et al., 2010). Many of the problems associated with established first-generation biofuel crops, which are often also food crops, could be avoided by cultivating 2G bioenergy crops on existing farmland. However, this option is often undesirable due to increasing food demand and the potential displacement of food crops (Dauber et al., 2010; Godfray et al., 2010; Gelfand et al., 2013). This conflict between land use objectives, the so-called ‘food, energy, environment trilemma’ (Tilman et al., 2009), raises concerns that increased demand for bioenergy crops will displace food production and/or cause further destruction of natural and seminatural (so-called marginal) ecosystems. The problems of potential biodiversity loss and land competition with food production remain central to the debate surrounding the possible contribution that bioenergy crops could make towards meeting renewable energy and greenhouse gas emissions targets (Nonhebel, 2012). Policy makers are left struggling with the problems that have dogged bioenergy cultivation for decades; where should 2G bioenergy crops be planted and what proportion of land should they occupy? Such questions raise doubts as to the viability of bioenergy as a major future energy source; a major recent report into bioenergy crop viability in the United States concluded that economic and environmental uncertainty will strongly limit future deployment of 2G crops (Committee on Economic & Environmental Impacts of Increasing Biofuels Production, 2011). However, this assertion relies on the old agricultural paradigm of large-scale cultivation of monocultures on existing agricultural land. Indeed, current policy advice largely ignores the potential to better manage bioenergy crops to reduce their impacts on food, biodiversity and ecosystem services and for strategic deployment of 2G bioenergy crops within the landscape so that they mitigate environmental damage and enable a more balanced use of limited land resources. We propose two main strategies for managing bioenergy crops that could reduce their impact on food production and the environment, and at the same time may even boost biodiversity and food production via increases in ecosystem service provision (Table 1). The first of these is that existing bioenergy crops are managed to facilitate biodiversity and support ecosystem service providers (ESPs). When properly managed using well-designed agri-environment schemes, wildlife-friendly farming practices can be successful in sustaining biodiversity (Ausden, 2007; Carvell et al., 2007, 2011; Whittingham, 2011), but this approach is not widely applied to bioenergy crops at present. Our second proposal is that new bioenergy plantations are located in a more strategic way that considers landscape context and is sensitive to how they affect biodiversity and ecosystem services (Table 2). This involves planning the spatial arrangement of bioenergy plantations so that they interact positively with other landscape units. It also requires an understanding of how the impact of bioenergy crops on biodiversity and food security varies depending upon their biological and physical environmental context. We suggest that the strategic planting of bioenergy crops, relative to other land use packages, could mitigate environmental damage caused by other land uses, benefit ecosystem service provision and in some cases boost arable productivity. Bioenergy crops could therefore make up a component of multifunctional landscapes in which limited land resources are used to meet the requirements and aspirations of a diverse community of users. To illustrate these points further, we focus on three geographically distinct cases where these general principles could be applied. Our choice of these case studies is based upon literature availability, the extent (current or potential) of bioenergy cultivation and the degree of conflict with other land uses within these regions. Land availability and competition between bioenergy cultivation and other land uses have received considerable attention elsewhere (e.g. Haughton et al., 2009; Dauber et al., 2010, 2012; Godfray et al., 2010; Gelfand et al., 2013) and we do not discuss this issue at length here. Instead, we focus on how a policy of targeted integration of bioenergy with conventional crops and other landscape features might facilitate biodiversity and ESPs. *High inputs of biocides and fertilizer *Absence of field margins and other landscape features *Regular tillage *High nutrient run-off to surrounding ecosystems *High agricultural productivity but often unsustainable *Soil degradation *Low carbon storage *High GHG emissions *Low inputs of biocides and fertilizer *Maintenance and planting of landscape features *Encouragement of within-crop biodiversity *Intermediate nutrient run-off to surrounding ecosystems *Lower agricultural productivity but sustainable *Intermediate carbon storage *Intermediate GHG emissions *High inputs of biocides and fertilizer *Absence of field margins and other landscape features *Displacement of food crops *Bioenergy food crops *Removal of natural and seminatural ecosystems *Nutrient run-off *Disease carrying organisms *Low carbon storage *Very high GHG emissions *Novel landscape features and management practices that encourage biodiversity *2G nonfood perennial bioenergy crops *Greater aesthetic appeal *Very low GHG emissions *High carbon storage *Low nutrient run-off *Large-scale monocultures planted for industrial convenience with little consideration of environmental impact *Geographically separate systems *Conversion of natural and seminatural habitat *Absence of habitat corridors *Lack of habitat heterogeneity *Low local and landscape-scale diversity and ESP abundance *Alteration of local hydrology *Low GHG emissions *Low aesthetic appeal *Soil degradation *Nutrient run-off *Side by side planting of complementary land packages *Avoidance of bioenergy planting in environmentally sensitive areas *Native crop species planted within bioenergy areas *Bioenergy crops form habitat corridors in food crops and buffer strips alongside water bodies *Higher biodiversity of bioenergy patches supplements low diversity arable patches *Landscape-scale habitat heterogeneity and diversity *High ESP abundance in both crop types. *Multifunctional landscape in which ecosystem services are maximized e.g. planting of water demanding crops on wetter areas only *Provision of pollinator and natural enemy control in arable crops by neighbouring bioenergy crop ESPs. *Higher aesthetic appeal *Reduced soil degradation *Reduced nutrient run-off Our first example is more environmentally sensitive management of sugarcane (Saccharum spp.) plantations in Brazil. Presently, these provide around 17.5% of the nation's energy production (Ministry of Mines & Energy, Brazil, 2013) and are managed intensively, thus imposing a range of negative environmental impacts including atmospheric pollution from the burning of sugarcane residues, soil degradation and nutrient run-off into aquatic systems (Martinelli & Filoso, 2008). These impacts could be minimized without loss of income and productivity by using altered field and landscape management practices; for example, by applying nitrogen fertilizer in organic form, instead of more environmentally mobile nitrates that sugarcane makes little use of (Robinson et al., 2012). Furthermore, the use of buffer strips in sugar cane plantations could minimize soil erosion and the eutrophication of riparian habitats (Barling & Moore, 1994), while simultaneously boosting biodiversity. Such strips might be placed along field margins and the edges of water bodies and could contain a range of species that provide multiple ecosystem services; e.g. native tree crop species such as candeia (Eremanthus erythropappus), a source of essential oils (Gries et al., 2011), and Copaifera spp., which produce copaiba resins with diverse medicinal uses (Veiga & Pinto, 2002). In addition to providing economically valuable products these strips might also provide habitat for the natural enemies of crop pest species, e.g. the sugarcane borer moth (Diatraea saccharalis) (Oliviera et al., 2012) and predators of the disease carrying rodents that have proliferated under sugarcane expansion (Verdade et al., 2012). Such benefits may outweigh those of sugarcane displacement. Further economic and social benefits could be realized in the increased diversification and aesthetic appeal of the landscape. A similar proposal has been made for the incorporation of agroforestry strips between high conservation value areas and intensive oil palm plantations, and it is expected that such areas would benefit both biodiversity and rural communities (Koh et al., 2009). However, the scientific evidence base to support the planning of such multifunctional landscapes remains underdeveloped for most bioenergy crops. Our second example focuses upon the strategic planting of 2G bioenergy crops into European arable landscapes. It is now widely established that 2G bioenergy crops can promote the abundance of numerous taxonomic groups in comparison with alternative, adjacent land uses such as arable, improved grassland or fallow. Landscape-scale diversity might also be increased as different species are found in bioenergy plantations and cropland (Dauber et al., 2010; Rowe et al., 2011, 2013; Stanley & Stout, 2013). Any measure that increases biodiversity and the abundance of wildlife in this intensively managed agri-environment is welcome, but we suggest that this higher abundance of organisms could also boost ecosystem service delivery both within the bioenergy crop and in the surrounding farmland. This may provide benefits to food production that may outweigh the cost of arable crop displacement. For example, the enhanced arthropod predator diversity and abundance that is found within bioenergy crop plantations (Rowe et al., 2013) can play an important role in herbivore pest control within the crop (Björkman et al., 2004). If bioenergy plantations were situated next to arable crops then the potential ‘spillover’ of spiders, beetles, hoverflies and various wasps could see significant boosts to this service in these surrounding land units. The relative increase in the abundance and diversity of hymenoptera and lepidoptera assemblages observed in and around bioenergy crop plantations (Haughton et al., 2009; Rowe et al., 2013) may also benefit crop pollination in adjacent farmland, an ecosystem service estimated to be worth >US$ 190 billion globally (Gallai et al., 2009). Similarly, weed seed predation by granivorous birds could be enhanced by the presence of bioenergy crops as they provide nesting and roosting sites for many bird species (Campbell et al., 2012). Crop pest butterflies are also less abundant in the field margins of bioenergy crops than in arable crop margins (Haughton et al., 2009). All these studies indicate that strategic bioenergy planting could simultaneously benefit biodiversity and food production. Indeed, where appropriate, bioenergy planting could even be integrated into agri-environment schemes, whereby payments are used to encourage environmentally friendly practices. In addition, agri-environment elements could become integrated into policies that encourage bioenergy crop planning, e.g. the UK's Energy Crops Scheme (Natural England, 2013). An example would be the deployment of short rotation coppice willow plantations as habitat corridors. These might complement existing features such as hedgerows and patches of seminatural habitat in providing the connectivity demanded by contemporary biodiversity policy (Lawton et al., 2010). Our third example centres upon the wide-scale concern that 2G bioenergy crops, in particular Miscanthus and jatropha, damage hydrological resources. Recent reports suggest detrimental effects on water supply following large-scale cultivation of perennial energy grasses such as Miscanthus, with water use in the Midwestern United States increased more than 50% compared with maize (VanLoocke et al., 2010; Le et al., 2011). The water-use footprint of 13 biofuel/energy crops was estimated by Gerbens-Leenes et al. (2009) and showed that Jatropha (a 2G crop) used more water than all of the 1G crops studied, including five times the water used by ethanol maize. However, these concerns rely on modelled data or inventories; blunt tools with which to determine future policy as there is very little experimental validation of model assumptions. These models also assume uniform cultivation across landscapes. This need not happen and plantations could be managed and sited to more effectively use limited water resources. Indeed, when spatial water use and variation in crop cover are incorporated into hydrodynamic models, it is possible to identify less sensitive areas for Miscanthus cultivation and reduce predicted hydrological impacts (Gerbens-Leenes et al., 2009). Experimental verification and wider identification of low impact areas, with regards to water and other land resources should enable prescriptions for the hydrologically and environmentally sustainable cultivation of Miscanthus and other bioenergy crops to be developed. At present these relationships remain hypothetical; studies of the environmental impact of bioenergy crops have focused largely on the response of biodiversity (Dauber et al., 2010; Fletcher et al., 2011; Wiens et al., 2011) and greenhouse gas emissions (Fargione et al., 2008;. Romijn, 2011; Gelfand et al., 2013). By contrast, most other ecosystem processes and services have received very little attention (but see Rowe et al., 2013). While the potential benefits of more strategic bioenergy crop planting and management are considerable, several ecological and economic factors could limit the success of such schemes. Perhaps foremost among these is the potentially low cost-effectiveness of small and disparate bioenergy crops. The costs of managing and harvesting diffuse bioenergy crop plantations, and transporting the products, may well preclude their establishment (Dauber et al., 2012). Also, if biodiversity is to be maintained in such plantations then small-scale rotational management, as opposed to the clear felling of large areas, may also be required (Hanowski et al., 2003). Such practices are common in traditionally managed hazel coppice woodland in the United Kingdom (Ausden, 2007), but their economic viability in bioenergy plantations remains unknown. The availability of suitable land may also limit the cost-effectiveness of mixed bioenergy food production landscapes (Dauber et al., 2012). It is clear that the spillover benefits of bioenergy crop strips, as well as the price for bioenergy crop products would need to be considerable for bioenergy cultivation to displace food production in productive regions. At the other end of the spectrum, in less productive regions, the extent of natural and seminatural habitats is often greater and these habitats will typically contain far greater biodiversity than bioenergy plantations (Christian et al., 1997; Hanowski et al., 2003; Ausden, 2007). These habitats may also provide ecosystem services to surrounding crops more effectively than bioenergy crops, thus limiting the benefits of bioenergy cultivation to the farmer. Nevertheless, there are land use types, even within generally productive landscapes, that may be suitable. These include abandoned agricultural lands with low biodiversity value, unproductive agricultural land and degraded sites colonized by invasive species, although all of these could also be targeted for agricultural improvement or ecological restoration (Haughton et al., 2009; Tilman et al., 2009; Nackley et al., 2013). A third problem is that the edge and connectivity effects of bioenergy crops that we propose may not be manifested and could even be negative. Instead of spilling over into neighbouring crops, many ESPs may remain in perennial bioenergy crops and seldom enter the arable crop (e.g. Christian et al., 1997; Hanowski et al., 2003). Some species of seminatural habitats may respond the bioenergy plantations similarly, remaining within isolated patches of habitat and not using bioenergy crops as habitat corridors. Contact effects between bioenergy and arable crops might also be negative. For example, the emission of isoprene by short rotation poplar forests could potentially result in ozone formation with crop damage and human health impacts (Ashworth et al., 2013). The planting of poplar into areas of high potential ozone formation should therefore be avoided, thus limiting the potential coverage of these crops. With respect to the more sustainable and wildlife-friendly management of bioenergy crops, challenges are likely to be similar to those for food crops; payments and/or landscape-scale crop benefits must be sufficient for farmers to adopt sustainable practices. Management practices also need to be coordinated at a landscape scale to benefit organisms with large ranges and limited dispersal. Such topics have been discussed elsewhere (e.g. Whittingham, 2011 and related papers) and so we do not elaborate upon them here. To see if the potential benefits of the strategic cultivation and management of bioenergy crops can be realized, new research is required. Such research would contribute to the current drive to develop ‘sustainable intensification’ strategies that foster synergies between land use packages and attempt to reduce the trade-offs between the delivery of multiple ecosystem services within our limited land resources (Garnett et al., 2013). While the examples presented here illustrate our general points, they also demonstrate that the trade-offs and synergies between bioenergy cultivation and the provision of other ecosystem services vary depending upon their geographical setting, as will the socially desired balance and prioritization of multiple land uses. Novel landscape management strategies will therefore need to be developed at local, national and global scales (Garnett et al., 2013), but unfortunately few mechanisms are currently in place to ensure that land use policy and practice is developed in an evidence-based way. Such evidence is essential as most of the environmental damage caused by bioenergy cultivation has been a consequence of policy advancing ahead of a well-formulated scientific evidence base. For example, the original European Renewable Transport Fuel Obligation (2009/28/EC) mandated a target of 10% renewable energy in the transport sector by 2020. This policy was implemented with limited understanding of global land use implications and minimal sustainability criteria in place. As a result, it inadvertently encouraged widespread planting of food crops such as palm oil for biofuel production, the displacement of food crops and the destruction of natural habitats (Gilbert, 2012). The recent 2012 directive (com 595) addresses this issue and promotes the use of nonfood crops and residues with minimal indirect Land Use Change (iLUC). However, this is an unrefined approach that relies on biophysical and global economic modelling, and it has rarely been tested at the landscape scale. We propose directed research that identifies new strategies for sustainable landscape-scale 2G planting in which the cost to other ecosystem services is reduced or removed. This would likely require a combination of observational field measurements, large-scale manipulation experiments and spatially explicit simulation models. Once identified, such strategies can be implemented using policy mechanisms such as subsidy, law and certification of produce. If performed strategically, bioenergy crop cultivation and management need not be to the detriment of the environment, and could even boost biodiversity, ecosystem services, and by consequence increase crop yield. The research that is required to tell us if this is possible must bear eventual policy instruments in mind as these are likely to be as much a barrier to successful deployment as technical plausibility. The scientific and policy challenges in developing the resulting multifunctional landscapes are considerable. Nevertheless, the importance of such work is paramount given the prevailing societal desire to produce large quantities of food and energy in agroecosystems that are underpinned by a stable and biodiverse network of ecosystem service providers. Helio Lima Neto provided valuable comments on the article. Research on 2G bioenergy crops in the laboratory of GT is funded by NERC, the Energy and and