There exist several approaches proposing sustain­ability criteria for bioenergy (e. g. Cramer et al., 2007; EU, 2009; IEA, 2010; RSB, 2011; GEF et al., 2013, where the last two are very detailed). They usually cover some goal for GHG emission reductions, e. g. a 35% reduction of aggregate emissions over some time period with respect to the baseline as suggested in EU regula­tions (50% from 2017 to 60% from 2018 onwards, EU, 2009). While this seems a clear criterion, its assessment is complex. The choice of different default values, soil carbon stock data and land use change definitions, for example, is behind the huge differences in GHG bal­ances between two GHG calculation tools as assessed in Hennecke et al. (2013), one of them being the tool used by the Roundtable of Sustainable Biomaterials (RSB) whose sustainability criteria are discussed below. Other aspects are decisive as well. The choice of the time horizon over which aggregate reductions have to be achieved and the choice of the social discount rate, which influences the relative importance of current and future emissions, also greatly influence the outcome (De Gorter and Tsur, 2009).

The sustainability criteria proposed for biofuel pro­duction that relate to agricultural production and the food system, address land competition, biodiversity, environmental impacts on soil, water and air, and social aspects. Land competition, resp. absence thereof and the related food security are by far the most prominent criteria in the discussion (HLPE, 2013). Potential drivers for land competition are many. First, there is the fact that bioenergy crops need fertile land to achieve economi­cally interesting yields. Biofuels on marginal lands are some option in smallholder and community — self-sufficiency contexts, but for commercial supply of biofuels in significant shares of total global energy de­mand, production on fertile land that potentially is used for food production is necessary (cf. Section Bio­energy Potential on Farm Level). Similarly, bioenergy crop production depends on water availability and nutrient inputs as any other agricultural production sys­tem. Thus, a second potential competition is not only on fertile lands, but also on land with sufficient water avail­ability (Lysen and van Egmond, 2008), in particular in the context of climate change, where water scarcity will become a prevalent problem in many regions (Meehl et al., 2007). Third, relative price differences be­tween bioenergy and food production will be and have been a key driver behind land competition as without further regulation, land will be allocated to the most profitable production. Stated differently, an increasing demand for biofuels leads to higher prices, which trig­gers an increasing supply for it, with corresponding land use (HLPE, 2013). It has to be emphasized that land competition between food and other uses is not new and relative profitability has always been a key driver behind this. As Nhantumbo and Salomao (2010) state, "Competition for higher-value resources existed well before the biofuels campaign was initiated. In this sense, biofuels production per se cannot be blamed for land use conflicts, as the same types of conflicts have occurred in other economic activities. But, in conjunc­tion with other activities like mining, forestry and tourism, biofuels projects further exacerbate competi­tion for land, water and other resources" (p. 4). The key point is thus that biofuel expansion increases the pressure on the already scarce resource of fertile land. In principle, policy measures can be used to mitigate these adverse effects. However, their implementation is often riddled with difficulties and land-use rights pro­tection, enforcement of laws and regulations, etc. have to be carefully considered when establishing potentially promising institutions for sustainable land use. Nhan — tumbo and Salomao (2010) illustrate these challenges for the case of Mozambique and draw a rather pessi­mistic picture.

The land use debate is further complicated by ILUCs (Wicke et al., 2012). Those arise, for example, if biofuel expansion in one region (e. g. sugar cane in southern Brazil) leads to land use change in another region (in this case, deforestation for livestock production in north­ern Brazil). The rationale behind this example is the fact that expanding sugarcane in the South is at the expense of already existing pastures in this region, that then themselves relocate at the expense of other uses such as natural forests (Andrade De Sa et al., 2013). Such ef­fects are very difficult to clearly identify and assess (Wicke et al., 2012). This is also the case in the detailed analysis in Andrade De Sa et al. (2013) who find only very weak significant statistical effects. Nevertheless, there is evidence from many descriptive studies that the potential presence of such mechanisms must not be neglected (PBL, 2010). ILUC is not only relevant for the competition between different land uses but also for the GHG balances of biofuels, as it can have consider­able negative effects on those (PBL, 2010; Faist Emmenegger et al., 2012).

Biodiversity criteria mainly refer to the ban of using forests or protected areas for bioenergy production (e. g. Cramer et al., 2007; EU, 2009) or to being attentive not to use invasive species as bioenergy crops (UNEP, 2010). The use of protected areas can also be seen as a particular aspect of land competition from biofuel pro­duction. As mentioned above, biofuel production com­petes not only with food for land but also with other uses, such as biodiversity protection and also with fiber and biomaterial production that all depend on land availability. Invasive species are seen as a potential danger, due to already existing cases but also due to gen­eral characteristics of biofuel crops that also correlate with invasiveness (e. g. fast growth or tolerance to wide range of soil and climate conditions, UNEP, 2010). Much less prominent in this discussion are the adverse effects of current agricultural production on biodiversity (mainly due to overfertilization with nitro­gen and pesticide use), albeit those are a key driver behind biodiversity losses (Galloway et al. 2008). This is mentioned in Bindraban et al. (2009) and adoption of agricultural practices with low negative effects on biodiversity is a criterion in GEF (2013), but not in EU (2009) or RSB (2011).

Other environmental impacts largely remain rather unspecific in the criteria suggested, although the range of adverse environmental effects of current agricultural production as described above will also realize in bio­energy cropping systems. EU (2009) for example only posits that the production has to meet the Community environmental requirements and in GEF et al. (2013), water contamination is assumed to be no issue if legal requirements are met. The size of the adverse environ­mental effects depends on the types of crops. Grassland or wood products usually perform better than annual crops, for example, WBGU (2009). Somewhat more detailed criteria are usually given in reference to soil — related aspects such as soil fertility and soil organic car­bon contents (see e. g. Cramer et al., 2007; EU, 2009; RSB, 2011; GEF et al., 2013). Regarding soil organic carbon, some sustainability criteria explicitly exclude bioenergy cropping on peatlands and other carbon-rich soils (EU, 2009). On the other hand, some bioenergy crops are judged to be advantageous for soil carbon levels, mainly grassland and forest-based bioenergy. The effect on soil carbon is not that clear for some perennial crops and rather negative for annual crops (WBGU, 2009).

Social sustainability criteria, finally, sometimes tend to be formulated on a very general level. EU (2009), for example, only requires that source countries for bio­energy have "ratified and implemented" (p. 97) a range of conventions referring to labor rights, gender aspects, etc. RSB (2011) and GEF et al. (2013), on the other hand, are quite detailed on the social aspects that cover a range of important criteria for social sustainability in agricul­tural production. RSB (2011) and GEF et al. (2013) also make long-term economic viability of bioenergy projects a criterion for their sustainability assessment. It is not mentioned as a criterion, but bioenergy crops can have some risk-spreading characteristics as they can increase production diversity of a farm and as their demand and price dynamics likely follows different patterns than food or fiber crop demand and prices. Some types of bio­energy crops can also be used for direct on-farm energy provision without much investment needs, such as Jatropha, for example. These energy crops thus have the potential to increase energy access and reduce work­load of women considerable in case they have to collect fuel-wood from far away, which is a common situation in many poor regions in developing countries. Energy crops can also provide specific income sources for women, as many case-studies show (Karlsson and Banda, 2009). However, there is similar evidence of problematic situations from case-studies, and whether a bioenergy project is advantageous for single farmers, the community and women in particular strongly de­pends on the concrete design and institutional context. Further example of positive cases are given in Practical Action Consulting (2009), and some negative cases for example in Ribeiro and Matavel (2009), focusing on Jatropha in Mozambique.

The choice of sustainability criteria for bioenergy thus reflects the classical topics of sustainability criteria, with a focus on environmental aspects and climate change in particular. An additional aspect is land competition, which is covered extensively in the discussion. The focus on environmental criteria is understandable as bioenergy is a climate change mitigation strategy and prime impacts of agricultural production are in the envi­ronment. However, besides GHG emissions and, partly, biodiversity, the assessment of environmental criteria remains rather weak. For a comprehensive sustainability assessment, the topical breadth and depth in analysis must be improved. Generally, bioenergy production has the same impacts as any agricultural production and sustainability in bioenergy production largely links to sustainable agricultural production.

Biomass Use

The assessment of proposed sustainability criteria for bioenergy production shows that the competition for biomass between bioenergy use and for fertilizing sus­tainable agricultural production systems is no topic. It is covered marginally in some other publications on sus­tainable bioenergy, e. g. in Bindraban et al. (2009) or Blonz et al. (2008), although it is of key relevance for sustainable agricultural production. Some publications address this topic as a caveat of agricultural or forestry residues use, as exporting too much of them causes soil carbon losses and soil degradation (e. g. WBGU, 2009). Only Muller (2009) takes up this topic in depth. The export of biomass from the fields for bioenergy use also exports nutrients that have to be replaced by other fertilizers, i. e. mineral fertilizers. The overuse of mineral fertilizers is however a key driver behind many environmental problems of current agricultural production. Sustainable agricultural production systems are based on closed nutrient cycles and organic fertilizers (compost from crop residues, roots and residues that remain on and in the fields, and manure from livestock operations). Those are keys for soil fertility and increased soil organic carbon levels (Lal, 2008; Gat — tinger et al., 2012). The nutrient export becomes particu­larly relevant for second-generation biofuels, where basically the whole plant can be used and no unused res­idues remain, resp. where cellulosic residues from any crops can be utilized (IEA, 2010). This is even suggested as a strategy to mitigate land use competition, as residues come without additional land requirements and feed­stock for second-generation biofuel is claimed to often grow on marginal lands (IEA, 2010, 2011). On marginal lands in particular, high organic matter inputs are key to improve soil fertility, though. Also for bioenergy crops, yields tend to be lower and erratic on marginal lands and economic viability of bioenergy projects is often given on fertile land only (Bindraban et al., 2009). Thus, regarding the biomass competition, the most promising options to avoid land use competition seem particularly problematic.