Zah et al. (2007) have studied cumulative energy demand associated with methane production from a variety of wastes using allocation on the basis of price and a zero value for the waste itself. Thus, the calculation of energy demand and emissions linked to methane production from wastes was restricted to the waste-to-wheel stages of the life cycle. Comparison was with natural gas. The wastes considered were: sewage sludge, ‘biowaste’, manure and manure plus co-substrates. Cumula­tive fossil energy demand for methane from these wastes was typically in the order of approximately 45% of the fossil reference. The outcomes of the study of Zah et al. (2007) are more favourable to transport biofuels made from wastes than to transport biofuels made from food crops. Zwart et al. (2006) made a more detailed study of the conversion of manure from cattle and swine into biogas (methane) in the Netherlands and concluded that the fossil fuel input energetically roughly equalled the biogas output. One should keep in mind that these outcomes are based on the assumption that life cycle impacts up to the waste can be neglected. When wastes change into secondary resources, fetching a price, or when the seed-to-wheel allo­cation is based on mass or energy, this would raise cumulative fossil energy demand of transport biofuels made from residues (cf. Reijnders and Huijbregts 2005).

The seed-to-wheel emissions of the transport biofuels considered here are substan­tial. The most important current transport biofuels (bioethanol from starch and sugar crops and biodiesel from edible vegetable oil crops) are often not a substantial im­provement over current fossil transport fuels or do even worse. Thus, the question arises as to what the possibilities are for reducing this impact. Much of the impact tends to come from the production of feedstock for transport biofuel production. So the possibilities for the reduction of environmental impact associated with this stage of the transport biofuel life cycles will be considered here.

Increasing soil carbon stocks while growing feedstocks may reduce the emission of biogenic carbonaceous greenhouse gases. Reduced tillage, the use of cover crops and/or fallows, including nitrogen fixers, and the return of residues and application of other organic matter, such as manure and household composts, may contribute to such an increase in C stocks (Nandwa 2001; Bationo and Buerkert 2001; D^az — Zorita et al. 2002; Cowie et al. 2006; Reijnders and Huijbregts 2007,2008b). The net emissions of greenhouse gases due to changes in ecosystem carbon stocks linked to land use change following from expansion of feedstock cropping may be lowered or even reversed by growing feedstocks on soils with low aboveground carbon stocks (Germer and Sauerborn 2007). Also, increases in yield achieved at relatively low inputs of fossil fuels and improved efficiencies in converting feedstock to biofuel may reduce net greenhouse gas emissions (Gibbs et al. 2008).

In the case of CH4 emission due to anaerobic conversion linked to the process­ing of biomass, capture of CH4 and application thereof in energy generation will help (Reijnders and Huijbregts 2008a). When biofuels are burned in power plants, there is the option of CO2 sequestration in abandoned gas and oil fields or aquifers, which may lead to net biogenic C sequestration (Mathews 2008). The emission of N2O may be reduced by a better N efficiency of agriculture. Precision agriculture and subsoil irrigation techniques may be conducive to a better N efficiency (Reijn — ders and Huijbregts 2008b). Kim and Dale (2008b) have shown that in conventional corn cropping, there is an environmentally optimal N application rate which en­hances profitability to farmers. Reduction of greenhouse gas emissions linked to the production of synthetic N fertilizer may be possible by including N-fixing crops.

Thomsen and Haugaard-Nielsen (2008) have, for instance, proposed wheat under­sown with clover grass for the production of biomass to be subj ected to simultaneous saccharification and fermentation.

As to the emissions linked to fossil fuels, over time there have been significant re­ductions in the fossil fuel inputs into the production of the currently most important transport biofuel — ethanol due to efficiency gains (Hill et al. 2006; Macedo et al. 2008) — and a further significant reduction linked to efficiency gains is expected (Macedo et al. 2008).

Also, in feedstock processing, there is scope for the replacement of fossil fuels by agricultural residues, especially in the case of high-yielding crops (Reijnders and Huijbregts 2008a; Reijnders 2008). For instance, in Thailand, coal is an important fuel in the conversion of sugar cane molasses into ethanol (Nguyen et al. 2008), and coal can be replaced by residues of sugar cane or oil palm fruit processing. In the case of ethanol from sugar cane and within limitations linked to the need for maintenance of soil C stocks, the possibility exists to additionally produce electricity from bagasse (a sugar cane residue) for use elsewhere (Macedo et al. 2008).

In the case of sugar cane production, emissions of a variety of pollutants can be reduced when cane burning is replaced by mechanical harvesting (Macedo et al. 2008; Machado et al. 2008). Improving the nutrient efficiency of biofuel cropping (for instance by precision agriculture) may reduce nutrient emissions from arable land, and there is also scope for reduced pesticide emissions (e. g. Muilerman 2008).

1.1 Introduction

Production and use of transport biofuels have a history of considerable length. The prototype of the Otto motor, which currently powers gasoline cars, was developed for burning ethanol and sponsored by a sugar factory. The Ford Model T (Tin Lizzy) did run on ethanol. In the early twentieth century, ethanol-fuelled cars were praised because they experienced less wear and tear, were quieter and produced a less smoky exhaust than gasoline-fuelled cars (Dimitri and Effland 2007). Also in the early twentieth century, a significant part of train locomotives in Germany were powered by ethanol (Antoni et al. 2007). In the same country, ethanol from potato starch was used in gasoline as an anti-knocking additive between 1925 and 1945 (Antoni et al. 2007). In the 1930s, ethanol produced from starch or sugar made something of a comeback as road transport fuel in the Midwestern states of the USA, because agricultural prices were very depressed (Solomon et al. 2007). Also in the 1930s, the Brazilian government stimulated gasoline blends with 5% bioethanol.

Early demonstrations of the diesel motor around 1900 in Paris and St Peters­burg were with a variety of plant — and animal-derived oils. These were thought especially interesting for use in tropical and subtropical countries, where the rela­tively high viscosity of such oils, if compared with fossil diesel, is less of a problem than in colder countries (Knothe 2001). The first patent on making fatty acid esters (biodiesel) was awarded in 1937 and applied in 1938 to powering buses in Belgium (Knothe 2001). During the Second World War, vegetable oils re-emerged as fuels for diesel motors in countries like Brazil, Argentina and China (Knothe 2001). In Japan, soybean oil was used to power ships, pine root oil was used as a high-octane motor fuel and biogenic butanol was used in airplanes (Tsutsui 2003). The Japanese navy conducted extensive research on the production of diesel fuel from coconut oil, birch bark, orange peel and pine needles (Tsutsui 2003). Also, during the Second World War, substitutes for mineral-oil-based gasoline and kerosene were produced in China by the catalytic cracking of vegetable oils (Knothe 2001). Furthermore,

L. Reijnders, M. A.J. Huibregts, Biofuls for Road Transport © Springer 2009

thermal destruction of wood was used for producing road transport fuel during the World Wars in Europe (Reed and Lerner 1973).

The post-World War II re-emergence of transport biofuel use dates from the 1973 hike in petroleum prices, or the ‘first oil crisis’. Tax reductions, subsidies, support for research and development, obligations to fuel providers and artificially high fuel economy ratings for flex fuel cars, which are suitable for high percentages of biofuel in transport fuel, were important government instruments used in this re-emergence (Demirba§ 2007; Szklo et al. 2007; Tyner 2007; Wiesenthal et al. 2008). By now, large sums of money are involved in such support. It has been estimated that in 2006, about US$11 billion was spent on public support measures by the USA, Canada and the European Union (OECD 2008).

Due to the first oil crisis of 1973, Brazil decided to reduce its dependence on the import of mineral oil by establishing a National Alcohol Program to supply ve­hicles. This program started in 1975, using sugar cane as a feedstock. A second program stimulating the use of ethanol began in the USA in 1978, using mainly corn and to a much lesser extent sorghum as feedstocks (Wheals et al. 1999; Wang et al. 2008a). In the USA, arguments for subsidizing the production of bioethanol since 1978 have included energy security, supporting farm prices and incomes and improvement of air quality (Tyner 2007). Several Canadian provinces started out using 5-10% ethanol-gasoline mixtures in the 1980s (Szklo et al. 2007). The ‘re­discovery’ of biodiesel occurred in the 1980s. Biodiesel initiatives were announced in 1981 in South Africa and in 1982 in Germany, New Zealand and Austria (Kor — bitz 1999). In Europe, substantial production of biodiesel started from about 1987 and in the USA from the 1990s (Knothe 2001). The relatively large production of biofuels in countries such as Germany, France, Italy, Austria and Spain had much to do with an interest in the development of new agricultural markets (Di Lucia and Nilsson 2007). Geopolitical worries about the supply of crude mineral oil and price rises affecting this dominating feedstock for current transport fuels furthered a rapid increase in biofuel production in the twenty-first century, especially after 2004 (Heimanand Solomon 2007).

The production of conventional mineral oil is likely to peak in the coming decades (GAO 2007; Bentley et al. 2007; Kaufmann and Shiers 2008). An adequate supply thereof may therefore become increasingly expensive and difficult. This has led to calls to — in the words of former US president G. W. Bush — kick the oil ‘ad­diction’ (Bush 2006). Timeliness of a transition to alternative fuels has been stressed (Kaufmann and Shiers 2008). ‘Home-grown’ biofuels, especially, have been argued to be suitable for energy security (Tyner 2007). There is also much concern about the pollution originating in the burning of fossil fuels. Recently, the effects thereof on climate have become important on the international political agenda. This, in turn, has led to increasing calls to reduce the emission of greenhouse gases, such as CO2. Such calls extend to transportation because worldwide transport accounts for about 22% of the total use of primary energy and is overwhelmingly mineral oil based (de la Rue du Can and Price 2008). For instance, regarding the USA, min­eral oil accounted in 2006 for about 97.8% of total transport energy use (Heiman and Solomon 2007). Worldwide, the consumption of petroleum products represents 94% of energy use in the transportation sector (de la Rue du Can and Price 2008), whereas in 2004, about 60% of all mineral oil was used for transportation (Quadrelli and Peterson 2007). Proponents of biofuels have argued that replacement of mineral oil by biofuels is a good way to reduce greenhouse gas emissions.

It has furthermore been stated that the potential for replacing fossil transport fuels with biofuels is very substantial indeed. de Vries et al. (2007) have suggested that by 2050, up to 300 EJ (= 300 x 1018 J) of liquid biofuels may be produced worldwide. An even higher estimate for liquid biofuel production by 2050 (455 EJ) has been proposed by Moreira (2006). Such amounts can in all probability cover demand for transport fuels in 2050, as the 2007 primary energy consumption for transport amounted to about 100 EJ (de la Rue du Can and Price 2008). Use of transport fuels by means of transport (‘end use’) was probably in the 85-90 EJ range, with the remaining amount used for winning, refining and distribution (Colella et al. 2005; EUCAR et al. 2007; Winebrake et al. 2007). The potential importance of biofuels in replacing fossil transport fuels is by now much stressed by the Brazilian government. In Brazil, ethanol from sugar cane is currently a substantial transport biofuel. In 2004, its share in energy for road transport was near 14% and in 2007 about 20% (OECD 2008). In 2006, 70% of the new cars sold in Brazil were ‘flex cars’, able to run on either 100% ethanol or a fossil fuel-ethanol blend (Quadrelli and Peterson 2007). The claims about the benefits and potential of transport biofuels have, however, been contested. And the resulting debate has been much fuelled by the high food prices in 2008, which have been partially linked to increasing transport biofuel production (OECD-FAO 2007).

This book will give a seed-to-wheel perspective on biofuels for road transport and will deal with a number of environmental issues that have emerged in the current biofuel debate. This first chapter is introductory and structured as follows: firstly, Sects. 1.2-1.6 will deal with the physical basis and the variety of biofuels and the ways to produce and apply them in transport. Secondly, in Sect. 1.7, developments in production volume, costs and prices will be discussed. Thereafter, in Sect. 1.8, the debate on the pros and cons of transport biofuels that has emerged will be briefly surveyed, and the rest of the book will be outlined.

Table 2.4, finally, shows overall estimated conversion efficiencies for solar irradia­tion to car kilometres, corrected for the input of fossil fuels, which are calculated by

TSCEXi = SCEx ■ СЕх,,

where TSCEx, i is the transport solar energy conversion efficiency (%) and CEx, the efficiency drive train of transport option i derived from biofuel type x (%).

According to the estimates in Table 2.4 regarding seed-to-wheel solar energy conversion efficiency, ethanol from sugar cane outperforms ethanol from European wheat by about a factor of five to ten, and biodiesel from European rapeseed by about a factor of two to three. Electrical traction from lignocellulosic biomass, how-

ever, in turn outperforms ethanol from sugar cane by roughly a factor of two to four. The relatively high efficiency of using biomass for electricity production has also been noted by other authors, such as Zhang et al. (2007). All biomass-based automotive power is, however, far less efficient than electricity from solar cells that is stored for use in electrical traction. This way of powering motor cars is roughly at least two orders of magnitude better than ethanol from Brazilian sugar cane and three orders of magnitude better than ethanol from European wheat. In calculating the values for Table 2.4, it has been assumed that solar cells and the plug-in facil­ity for cars are in the same region. When distances are large or conversion to H2 is necessary for long distance transport, the efficiency will be lower than indicated in Table 2.4 because of transport-linked losses. For instance, an estimate has been made regarding the life cycle emission of greenhouse gases linked to electrolysis powered by concentrated solar power (CSP) in the Sahara, liquefaction of H2 and transport to, and distribution of, hydrogen in Western Europe. In such a case, a re­duction of the life cycle efficiency by somewhat less than 10% has been found (Ros et al. 2009). Such a reduction applied to electricity from solar cells (last row of Table 2.4) would reduce the overall efficiency in the last column to approximately 3.5-9.4%.

A lesson from this chapter is that conversions lead to substantial reductions in solar conversion efficiency. In Chap. 1, quite a number of proposals have been sum­marized that rely on such conversions. Examples are: the conversion of methane (from the anaerobic conversion of biomass) to methanol, the conversion of lipids and ethanol to hydrocarbons or H2 and the conversion of methanol to hydrocarbons. As the starting products may in principle be used directly as transport biofuels, there is good reason to be sceptical about such sequential conversions in view of the neg­ative impact that they have on the overall solar energy conversion efficiency.

The data presented in this chapter allow for estimates of the ability of biofuels to energetically displace fossil fuels. It appears that in this respect, palm oil and ethanol from sugar cane do much better, especially when processing is powered by harvest residues, than rapeseed oil or ethanol from corn or wheat, as produced in industrial­ized countries. It should be noted, though, that the ability to energetically displace fossil fuels may be at variance with their ability to do so in the economy. The lat­ter is strongly impacted by prices and government policy. An interesting illustration thereof concerns the use of corn-derived ethanol in US gasoline, which has mainly been by E10 fuels, containing 10% ethanol and 90% conventional gasoline. The use of E10 fuels has been stimulated by a federal excise tax which in recent years led to E10 gasoline being cheaper than conventional gasoline (Tyner 2008; Vedenov and Wetzstein 2008), which in turn had an upward effect on the overall consump­tion of gasoline, thereby partly negating the downward effect of ethanol use on the consumption of conventional gasoline (Vedenov and Wetzstein 2008).

The data in this chapter also allow for estimates of land requirements linked to a large-scale displacement of fossil transport fuels by biofuels. This may be illus­trated by the following back-of-the-envelope calculation. As explained in Chap. 1, mineral oil is the dominating fossil fuel for powering transport, and about 60% of all crude oil is used for this transport. Let us suppose that all mineral oil that is currently used as an input in worldwide transport were to be replaced by vegetable oil. Corrected for the difference in lower heating value between crude oil and veg­etable oil (see Table 1.2) and the cumulative fossil fuel input into vegetable oil (estimated here at 40% of the energetic value of vegetable oil), this would require an increase of vegetable oil production by about a factor of 37.5. Part of this in­crease may be met with the increase of yields per hectare. Estimates made for the 23 most important food crops suggest that such an increase may range from 0.63- 1.76%year-1 for developing countries and from 0.59-0.79%year-1 for developed countries up to 2050 (Balmford et al. 2005), to a large extent by intensification of cropping (Tilman et al. 2001). Using intermediate values, this would allow for an increase in yield by a factor of approximately 1.75 for developing countries and by a factor of approximately 1.42 for developed countries between 2000 and 2050 (Balmford et al. 2005), far below the factor of 37.5 needed to displace all mineral oil by vegetable oil. Moreover, it may well be that the average productivity of addi­tional land is lower than that of land currently in use. Thus, even if yield increases in the future would be much larger than currently estimated, there would seem no way around large additional land requirements linked to large-scale displacement of fossil fuels by biofuels. Current policy targets are estimated to require between 55 and 166 million ha (Mha) (Renewable Fuels Agency 2008).

Moreover, expanding transport may well lead to even larger land claims in the future. Gurgel et al. (2007) studied an expansion of the production of cellulosic biofuel to supply up to 368 EJ in 2100. This, according to their scenario, would require about 2.5 x 103 Mha, an amount greater than any other land cover category. For comparison: worldwide, current cropland is about 1.6 x 103 Mha, and the land area that is currently considered fit for additional cropland is estimated at between 400 and approximately 1.2 x 103 Mha (Renewable Fuels Agency 2008).

Non-energy Natural Resource Demand

3.1 Introduction

Several stocks of natural resources are highly important to biomass-for-energy, in­cluding transport biofuel, production. These are: soil and soil organic matter, nutri­ents and water. In this chapter, these will be discussed in turn as to their current sta­tus. We will also discuss the sustainable use of such resources. Sustainable is a term that has by now many meanings. Here, the term will be used in its original meaning in the modern environmental debate, linked to a steady state economy (Daly 1973; Gliessman 1989; Hueting and Reijnders 1998; Reijnders 2006). Thus, sustainable use of biomass is defined in this chapter as a type of use that can be continued indefinitely while maintaining the supply or availability of natural resources. Sus­tainability leads to limitations regarding the use of renewable resources which are characterized by large additions to the stock of these resources and concerning the use of resources that are geochemically scarce and formed in slow geological pro­cesses (‘virtually non-renewable resources’). To allow for indefinite use, the usage of renewables should not exceed addition to stock, and resource quality should be maintained (Reijnders 2000). As to geochemically scarce virtually non-renewable resources, such as phosphate ore for which no substitution seems possible, the way to define sustainability is that wastes irretrievably lost should not substantially ex­ceed the small addition to the stock by geological processes (Goodland and Daly 1996; Reijnders 2006). This requirement corresponds with a large reduction in cur­rent wastage.

Searchinger et al. (2008) have shown by careful modelling of land use change that a large and fast expansion of bioethanol production from corn in the USA is coun­terproductive in tackling climate change. This is largely related to the land use change necessary for displaced food and feed production, which leads to a large desequestration of C. It is likely that a similar conclusion applies to biofuel produc­tion in Europe. Regarding Brazil, it has been shown that soybean-based biodiesel is counterproductive in tackling climate change (Reijnders and Huijbregts 2008a), and the same holds for palm-oil-based transport biofuels for which tropical forests are cleared (Danielsen et al. 2008; Fargione et al. 2008; Reijnders and Huijbregts 2008a).

As explained in Chap. 3 and as also pointed out by Fargione et al. (2008), any strategy aiming at expanding transport biofuel production by converting native wooded ecosystems to cropland is likely to be counterproductive in tackling climate change in the coming decades. When abandoned or fallow land over a number of decades has developed into a secondary forest (e. g. Grau et al. 2003), the same will probably apply. In the case that abandoned or fallow agricultural land is peat land, further C losses from soil will be greater than C gains to be made by reductions in fossil fuel use associated with biofuel use, possibly for many centuries (Reijn — ders and Huijbregts 2007; Danielsen et al. 2008; Fargione et al. 2008; Reijnders and

Huijbregts 2008a). Moreover, restoration of peat bogs may lead to the additional sequestration of C (Tuittila et al. 1999; Dukes 2003).

This leaves a limited number of terrestrial transport biofuel options conducive to tackling climate change. It would seem that from the point of view of effectiveness in limiting climate change, the burning of biofuels in power plants for use in electric traction is to be preferred, as the solar energy conversion efficiency is relatively high and net greenhouse gas emissions relatively low, as indicated in Chaps. 2 and 4.

The biofuel options that can help in tackling climate change are the following:

1. Linking biofuel use to the man-made sequestration of C in soils. This may be done as soil organic matter (Blaine Metting et al. 2001; Read 2008), as ‘biochar’ (Lehman et al. 2006) or as CO2. Examples of the latter option are: burning biofuels in power plants for electrical traction and sequestration of CO2 in de­pleted oil and gas fields or aquifers (Haszeldine 2006; Mathews 2008). Simi­larly, CO2 generated during fermentation or anaerobic conversion of biomass (into compounds such as ethanol, methane and hydrogen) might be captured or sequestered.

2. Limited use of residues of forestry and agriculture as feedstocks for biofuel production. Limitations in part stem from sustainability requirements. The scope of residue removal is, as pointed out in Chap. 3, limited by the need to maintain soil carbon and nutrient stocks, because otherwise future productivity will be impaired. Moreover, some conversions of agricultural residues do not appear to make energetic sense, such as, for instance, the conversion of swine and bovine manure into methane in NW Europe (cf. Chap. 2).

3. Producing biofuels in areas with currently relatively little aboveground biomass, as also discussed in Sect. 6.3, for instance within the framework of reclaiming deserts and reclamation of saline soils and on abandoned lands (Lal and Bruce 1999; Banetjee et al. 2006; Germer and Sauerborn 2007; Danielsen et al. 2008; Lal 2008).

However, from the point of view of mitigating climate change, alternative uses of land with currently little aboveground biomass merit consideration. In the case of abandoned and fallow agricultural mineral soils which may support secondary forests, transport biofuel production — as, for instance, proposed by Cunha da Costa (2004) for deforested areas in the Brazilian Amazon — may be compared to C se­questration linked to re-growth of forest. For instance, in the case of Saccharum spontaneum grasslands in Panama that preclude forest regeneration, it has been found that low-cost management options exist for restoring forest cover (Hooper et al. 2005). And also in other contexts, feasible ways have been developed to convert degraded agricultural land into secondary forests (Vieira and Scariot 2006; Cummings et al. 2007). In doing so, high levels of C sequestration may be achieved. Steininger (2000) found in the Brazilian Amazon a biomass accumulation of, on average, 9.1Mgha~1year~1 over a 12-year period, whereas Zarin et al. (2001), studying re-growth of Amazonian forests, found for a 20-year period an average sequestration of about 6 Mg biomass ha^year-1. These values are well above the amount of C that can be displaced by, for example, growing soybeans for biodiesel production (Reijnders and Huijbregts 2008b). Righelato and Spracklen (2007) esti­mated that as to tackling climate change during the coming decades, the gains per hectare of reforestation would be higher than those from most biofuels. Expansion of secondary forests is now a substantial development in Latin America, Africa and Southeast Asia (Lambin et al. 2003) and is known to have been successful in coun­tries like Puerto Rico, Bhutan, Vietnam, Gambia and Cuba (Chazdon 2008). All in all, an estimated 96 x 106 ha of abandoned agricultural land has been reforested (Field etal. 2008).

Another alternative that merits consideration in cases where forests may be re­established is agroforestry, which has been advocated as more in line with the press­ing needs of local populations than plantations (De Foresta and Michon 1996), and has now been successfully applied in a number of countries, including the Philippines, Peru, Indonesia, China and Vietnam (Chazdon 2008). Agroforestry allows for substantial sequestration of C. For instance, the cacao agroforests of humid Africa sequester up to about 62% of the C of primary forests (Duguma et al. 2001), and rates of C sequestration in agroforests varying from 2-9 Mg C ha-1year-1 have been found (Pandey 2002). Of course, comparisons with alterna­tives such as secondary forests and agroforestry are only meaningful when there is not a fixed mandate for biofuel production. When there is such a fixed man­date, growth of crops for transport biofuel production will move elsewhere, and may, for instance, be associated with cutting down virgin forests (a phenomenon called leakage, discussed in Sect. 1.4). Also, as stated in Sect. 1.4, one should re­alize that expansion of secondary forests and agroforestry as such do not provide fuels.

The use of a variety of photosynthetic organisms has been proposed that directly or indirectly biocatalyse the splitting of water into H2 and O2 (Melis and Happe 2001; Hallenbeck and Benemann 2002; Nath and Das 2004; Hahn et al. 2007; Hankamer et al. 2007). The production of hydrogen from wastewaters or carbohydrates by H2- producing bacteria has also been proposed (Van Ginkel et al. 2005; Rupprecht et al. 2006; Wongtanet et al. 2007; Jones 2008). The latter approach to generating H2 so far has a poor conversion efficiency (Jones 2008).

Melis and Melnicki (2006) have suggested the combined production of hydrogen by H2-producing bacteria and photosynthetic algae, and Westermann et al. (2007) have proposed biorefineries producing ethanol and hydrogen. Furthermore, there have been proposals to electrolytically generate hydrogen from wastes in the pres­ence of microbes (Stams et al. 2006; Dumas et al. 2008).

The methods to produce H2 with the help of micro-organisms would often re­quire closed systems (Rupprecht et al. 2006). This is necessary for the capture and removal of H2, which may inhibit H2 production, maintain anaerobic conditions and be highly conducive to limiting infection by unwanted organisms, but limits the scope for large-scale production.

N2O emissions may originate on several occasions along the biofuel life cycle (Reijnders and Huijbregts 2005). The production of N fertilizers is often accom­panied by the emission of N2O. NOx emissions linked to burning fossil fuels may be deposited as N compounds in soils, and there, they may be partly converted by microorganisms into N2O. N inputs in cropping are also partly converted into N2O. It is usually assumed that the latter process is, directly and indirectly, responsible for most of the N2O emission linked to the transport biofuel life cycle. The actual quantity of the N2O emission, given a specified input of N and soil, is, however, the subject of a lively debate (Mosier et al. 1998; Crutzen et al. 2007). According to the estimates of Crutzen et al. (2007), 3-5% of the N input into growing biofuel crops will be converted into N2O. Mosier et al. (1998) have presented data suggesting that direct N2O emissions from agricultural fields associated with biofuel cropping may be about 1.25% of added fixed nitrogen. In addition, they argue that fixed nitrogen lost from agricultural fields may also be subject to microbial conversion to N2O (estimated at 2.5% of fixed N lost).

Local conditions may have a significant impact on actual N2O emissions. The presence of soil moisture matters (Rebelo de Mira and Kroeze 2006; Guo and Zhou 2007; Scheer et al. 2008). Higher temperatures tend to be conducive to higher emis­sions of N2O (Ding et al. 2007; Scheer et al. 2008). So are higher levels of soil organic carbon (Guo and Zhou 2007; Liu et al. 2007). In some soils, nitrate is rel­atively favourable to N2O production, but in other soils, it is rather ammonia (Guo and Zhou 2007; Liu et al. 2007; Scheer et al. 2008). Moreover, it may be noted that N2O emissions may change when the climate changes. Temperature and precipi­tation are significant determinants of N2O emissions, and as temperature and pre­cipitation are expected to change when the climate changes, N2O emissions from biofuel cropping may be different in the future from what they are now (Novoa and Tejeda 2006). In view of variability, there is a case for a rather wide range for the conversion of fixed N into N2O: 1.5-5%.

Fossil fuel inputs in producing microalgae tend to be high. When microalgae are grown in bioreactors, outputs are unlikely to energetically outperform inputs (Wijf- fels 2008; Reijnders 2008). A claim has been made for ultrahigh bioproductivity from algae in thin channel ultradense culture bioreactors indirectly irradiated by the sun (Gordon and Polle 2007). The cultures are irradiated with pulsed light emitting diodes, powered by photovoltaic cells. The efficiency of converting solar radiation into biomass is probably below 0.2%, and the corresponding energetic yield is likely to be exceeded by fossil fuel inputs (Wijffels 2008).

As to producing microalgal biofuels in open ponds, it is a remarkable aspect of several recent publications strongly advocating algal transport biofuels (e. g. Chisti 2007; Huntley and Redalje 2007; Chisti 2008a; Dismukes et al. 2008) that inputs of fossil fuels are not addressed. Two less recent studies are available that looked at energy inputs and outputs in open pond cultures of microalgae. They did not take account of all inputs, though. For instance, fossil fuel input into the handling and clean-up of discharges from ponds (which will probably be necessary in view of the extreme pH and/or salt concentrations and high nutrient levels in algal ponds) was considered by neither of the studies. Sawayama et al. (1999) studied operational life cycle energy inputs in growing and processing Dunaliella tertiolecta to sup­ply bio-oil. Processing was by thermal liquefaction (also Yang et al. 2004). Oper­ational energy inputs (fossil fuels) exceeded energetic output by 56% when microal­gal yield was 15 Mg ha-1 year-1. Hirano et al. (1998) studied Spirulina production and processing to supply methanol (via synthesis gas). Here the assumed yield was approximately 110 Mg ha-1 year-1. Both fossil fuel inputs in infrastructure and op­eration were considered. The energetic output exceeded the life cycle fossil fuel input by 10%. At more realistic estimates of Spirulina yield, which are in the order of 10-30 Mgha-1 year-1 (Vonshak and Richmond 1988; Jimenez etal. 2003), fossil fuel inputs would have exceeded energetic outputs. Chisti (2008b) has argued that the energetic inputs used in the studies of Hirano et al. (1998) and Sawayama et al. (1999) are ‘grossly overestimated’. However, even at Chisti’s (2008b) estimate, the fossil fuel input energetically would equal an output of approximately 30 Mg dry weight algal biomass ha-1 year-1, which is at the upper end of the range for the commercial production of Spirulina (Jimenez et al. 2003).

Though experimentally, yields have been demonstrated that may energetically exceed fossil fuel inputs (Hirano et al. 1998; Chisti 2008b), it is far from certain that such yields can be achieved in actual commercial practice. Large differences between experimental yields and average commercial yields are also common in the production of terrestrial crops, as will be explained in Sect. 2.4.1.

A ‘high yield’ has furthermore been claimed for oil from Haematococcus plu — vialis produced by a combination of a closed bioreactor and 1.3 days in a pond (Huntley and Redalje 2007). This yield probably corresponds with a photosynthetic efficiency in producing biomass of just over 1% and a photosynthetic efficiency in producing algal oil of roughly 0.6% (Vasudevan and Briggs 2008). No data have been published about the cumulative energetic inputs in this type of culture, but from the above, it would seem unlikely that the energetic value of algal oil would much exceed the cumulative energy input into the infrastructural and operational inputs.

Studies regarding algal production of H2 suggest that the cumulative energy de­mand for algal H2 production is probably of the same order of magnitude as the energetic output, when the solar energy conversion efficiency does not exceed 1% (Burgess and Fernandez-Velasco 2007).

On the other hand, it may be that the yield of microalgae grown in water satu­rated by CO2 from power stations may exceed fossil fuel inputs when there is no allocation of the fossil fuel input into electricity production to these algae. However, whether this application will actually become operational is unclear, as algal perfor­mance has so far been disappointing, and sequestration of CO2 in abandoned gas and oil fields and aquifers has a higher efficiency (Benemann et al. 2003; Vunjak- Novakovic et al. 2005; Odeh and Cockerill 2008).

The emergence of some saltwater and freshwater macroalgae and macrophytes as pests offers scope for their conversion into transport biofuels. Only for one of the macrophytes (water hyacinth) are data available about the overall energy efficiency of conversion into ethanol. These data suggest a negative energy balance (Gunnars — son and Petersen 2007).

5.1 Introduction

In 2007, there was a lively discussion about a plan to replace the Mabira Forest Reserve in Uganda with a sugar cane plantation for the production of the trans­port biofuel ethanol. This forest reserve is home to almost 300 bird species (among which is the very rare Nahan’s Francolin) and supports 75 endemic species. In Oc­tober 2007, the Ugandan government announced that the plan had been scrapped, because the income from conserving the Mabira Forest would dwarf the profits from bioethanol production (Williams 2007). In the case of the Mabira Forest, much of this income is derived from ecotourism (Williams 2007), with additional revenues coming from harvesting timber, making charcoal and the collection of fuelwood (Naidoo and Adamowicz 2005). There may also be other sources of income from such forests, such as the collection of food, ornamental plants and organisms that have medicinal value (Brown and Rosendo 2000; Shanley and Luz 2003; Brennan etal. 2005;Mutimukuruetal. 2006). In being a target for ecotourism and providing natural resources, living nature may be said to provide ecosystem services that have monetary value.

Decisions about clearing nature to make way for biofuel production may also turn out differently. For instance, in 2007, the replacement of the high ecological value Tanoe Swamp Forest (Ivory Coast) by an oil palm plantation for biofuel production was also started (Sielhorst et al. 2008).

Direct replacement of nature by growing biofuels is only a part of the conse­quences of the expansion of biofuel production. There are also indirect effects of this expansion. These follow from the relative inelasticity of the demand for food (von Braun 2007; Searchinger et al. 2008). When cropping for biofuels replaces cropping for food or feed, food or feed crops largely have to be grown somewhere else. When the expansion of cropping for biofuel production is small, it may be pos­sible that the extra production of food and feed can be accommodated on existing agricultural soils, due to increasing productivity of agriculture. But when there is a fast and major expansion, this is not possible, and food and feed production may

L. Reijnders, M. A.J. Huibregts, Biofuls for Road Transport © Springer 2009

have to expand in areas that were so far left to nature (Searchinger et al. 2008). When the expansion of biomass for biofuel production is on highly productive soils, the effect can be relatively large, as it may well be that part of the expansion of food or feed production has to take place on soils with lower productivity, which will lead to relatively large land claims.

Replacement of living nature by agriculture directly or indirectly related to the expansion of biofuel production is a significant matter in the current debate about transport biofuels. And financial considerations are important to the outcome of the conflicts between nature and (agri)culture. However, financial interests are not the only matters that are at stake. It has been argued that natural species have an in­trinsic value, which requires protection against extinction. This type of argument has led to laws such as the (US) Endangered Species Act, which aims at protec­tion of endangered species. Secondly, ecosystems provide non-monetary ecosystem services to humankind conducive to a benign biological, chemical and physical en­vironment and to socio-cultural fulfilment (Daily 1997; Moberg and Folke 1999; Daily 2000; Batabyal et al. 2003; D^az et al. 2006; Brauman et al. 2007; Marrs et al. 2007; Wallace 2007). The non-monetary ecosystem services include beneficial im­pacts on water quality and quantity, climate, soil retention, pest and disease control and pollination. A more extended list is in Table 5.1.

Table 5.1 Non-monetary ecosystem services to mankind (Daily 1997; Diaz et al. 2006; Lelieveld et al. 2008)

Non-monetary ecosystem service to mankind

— Cleansing of air and water

— Contribution to preservation of soil fertility and stability

— Regulation of water quantity and quality available to humans, crops, animal husbandry and domestic animals

— Pollination of plants important to humans

— Pest and disease control in agriculture

— Resistance to invasive organisms that have negative impacts

— Climate regulation

— Protection against natural hazards (floods, fires, storms)

— Contribution to productivity and stability of plant production important to humans

— Prevention of leakage of nutrients and metals from soil to surface and ground water

Ecosystem services have been linked to biodiversity (Daily 1997, 2000), under­stood here as the diversity of species present in an ecosystem. There are a large num­ber of empirical studies that underpin this link (Salonius 1981; Naeem et al. 1995; Walker 1995; vanderHeijdenetal. 1999; Schlapfer and Schmid 1999; Duarte 2000; Emmerson et al. 2001; Engelhardt and Ritchie 2001; Hector et al. 2001; Lyons and Schwarz 2001; Loreau et al. 2001; Tilman et al. 2001b; Duffy 2002; Emmerling et al. 2002; Symstadet al. 2003; Tilman etal. 1996; Armsworth et al. 2004; Heems — bergen et al. 2004; Reusch et al. 2005; Balvanera et al. 2006; Cardinale et al. 2006; Worm et al. 2006; Brussaard et al. 2007; D^az et al. 2007; Fargione et al. 2007;

Turner et al. 2007; Flombaum and Sala 2008; Fornara and Tilman 2008; Ptacnik et al. 2008; Weigelt et al. 2008). From the studies done so far, it appears that there may be considerable differences between species in terms of what they contribute to ecosystem services (Cardinale et al. 2006; Jordan et al. 2006). On one hand, there are ‘keystone’ species that appear to have a large impact on such services. Keystone species have key functions in ecosystems. For instance, in sea grass communities, there are engineering species which, by changing the environment, facilitate the presence of species that would otherwise be absent (Duarte 2000). In arid envi­ronments, nurse plants such as Cercidium microphyllum and Carnegiea gigantea (saguaro cactus) have been identified that promote the establishment and survival of other species, including a variety of trees and shrubs (Withgott 2000; Drezner 2006,

2007) . And in the US Great Basin, sagebrush serves as a nurse plant for pinyon pine (Withgott 2000).

When keystone species disappear, the loss of ecosystem services may be dispro — portional. On the other hand, there are species which are in one or more respects rather similar to others in what they do in ecosystems. The loss of such a species may give rise to a less-than-proportionate loss of ecosystem services. The latter is reflected in studies that suggest that halving the number of plant species, on average, leads to a reduction in primary production of about 10-20% (Tilman et al. 1996). Still, there is also the possibility that cumulative biotic changes which at first ap­pear to have little effect may give rise to a sudden collapse of ecosystem services (Scheffer et al. 2001; Folke et al. 2004; Balmford and Bond 2005).

Such a collapse and the disproportionate effect of the loss of keystone species ex­emplify the possibility that the relationship between biodiversity and ecosystem ser­vices may show non-linearities (Lovelock 1988; Scheffer et al. 2001; Strange 2007). There may also be other causes for non-linearity. For instance, it has been found that there may be synergistic relations between invasive species (Grosholz 2005). And there may be interactions between losses of biodiversity and other human interven­tions that may give rise to non-linear effects. Malhi et al. (2008) discussed such a possibility in the context of deforestation and fire use in Amazonia. Here, there is a synergism between forest fragmentation and fire. Once burnt, a forest becomes more vulnerable to further burns and loses many primary forest species. Malhi et al. (2008) suggest that a tipping point may be reached when gasses establish in the forest understory, providing a source of fuel for repeated burns.

Monetary valuation of non-monetary ecosystem services is inherently problem­atic. One cannot buy them on markets. Provided that nature is there, they are freely provided. Without the ecosystem services of living nature, we would not even exist. The latter may be argued to suggest an infinite value, the former a zero value. And there have also been estimates in between (Costanza et al. 1997, 2007; Sukhdev

2008) . Given the problematical character of monetary valuation of non-monetary services, the following overview considers non-monetary ecosystem services with­out attributing monetary values.

In Chap. 2, it was pointed out that increased yields per hectare, partly linked to intensification of agriculture, are expected to contribute to the displacement of fossil fuels by biofuels, but that large-scale displacement of fossil fuels by biofuels requires large areas of land. Current policy targets for 2020 would require the use of between 55-166 million hectares (Mha) for biofuel cropping (Renewable Fuels Agency 2008). One option is to use surplus and degraded or abandoned and fallow agricultural land for this purpose (e. g. Hoogwijk et al. 2003). An estimate suggest­ing that by 2050, up to 300 EJyear~1 of liquid biofuels can be produced worldwide indeed assumes that 80% of the land area needed for that purpose will be aban­doned land (de Vries et al. 2007). In this case, there will probably be a large effect on biodiversity (Huston and Marland 2003; Marland and Obersteiner 2008), and biodiversity on abandoned and fallow land is linked to ecosystem services (Borner et al. 2007; Williams et al. 2008). It is likely that abandoned and fallow land will have a lower productivity than good-quality land. As a result, when fixed amounts of transport biofuels have to be produced, as mandated under current regulations in the USA, Canada, Brazil, India and the European Union, larger areas of land will be needed for biofuel production.

It is likely that abandoned and fallow land rather often harbours substantial biodi­versity (vanNoordwijk 2002; Zechmeister et al. 2003; Karlowski 2006; Bowen et al. 2007; Royal Society 2008). Especially after long periods of abandonment, biodiver­sity may be much increased (Fournier and Planchon 1998; van Noordwijk 2002; Williams et al. 2008) and may approach biodiversity in undisturbed ecosystems. However, there are also abandoned and fallow lands with relatively low biodiversity. Examples are the Imperata cylindrica and Saccharum spontaneum dominated grass­lands in formerly forested areas (Hooper et al. 2005; Germer and Sauerborn 2007). Though there are parts of such grasslands of great ecological importance (Peet et al. 1999), often they are not (MacDonald 2004). There are hundreds of mega-hectares of Imperata cylindrica grasslands, mainly in Africa and Asia (MacDonald 2004). In Southeastern Asia, these grasslands cover an estimated 25-35 Mha (Garrity et al. 1996; Otsamo 2000). Such grasslands are currently used for feeding livestock and thatching material (MacDonald 2004) but can also be used for biofuel production by harvesting the grasses as lignocellulosic feedstock for biofuel production or by the cultivation of, for example, short rotation woody crops that may serve as feedstock. In practice so far, the use of abandoned agricultural land for biofuel production has been very limited. For instance, expansion of Brazilian sugar cane production for the biofuel ethanol has largely been in the Cerrado region, a hotspot for biodiversity (Klink and Machado 2005; Koh 2007), and it has been suggested that further expan­sion may mainly take place on current pastureland (Goldemberg 2008; Goldemberg et al. 2008). Similarly, in Malaysia and Indonesia, there has been large-scale con­version of tropical forest into plantations that produce palm oil, notwithstanding the presence of large areas of degraded land in these countries (Germer and Sauerborn 2007).

There have been earlier proposals and attempts to exploit Imperata cylindrica grasslands for the production of wood and lignocellulosic feedstocks for the pulp and paper industry (Potter 1996; Lamb 1998; Otsamo 2000). These have met with some success (Marjokorpi and Otsamo 2006). In Southeast Asia, an estimated 2 Mha of former Imperata cylindrica grassland is now converted into Acacia mangium plantations, serving the supply of wood and lignocellulosic feedstock for the pulp and paper industry (Yamashita et al. 2008). However, attempts to convert Imperata cylindrica grasslands has had limited success because of limited support by local people who made use of those grasslands, e. g. for keeping livestock, and/or felt such plantations at variance with their pressing needs (Potter 1996; Otsamo 2000; Marjokorpi and Otsamo 2006), which is illustrated with a quote by one of those affected: ‘The trees are healthy, but the people are sick at heart’ (Potter 1996). And the production of lignocellulosic feedstock is not the only option for the conversion of Imperata cylindrica grasslands. It has, for instance, been suggested that, where possible, replacement of Imperata cylindrica grasslands by agroforests may be more in line with the needs of local populations (De Foresta and Michon 1996).

Assuming ‘business as usual’, a strong future expansion of transport biofuel pro­duction is expected to cause large-scale replacement of nature (Germer and Sauer — born 2007; Gurgel et al. 2007; Johansson and Azar 2007; Sivaram 2007; Christers — son 2008). At a regional scale, this seems to be confirmed by studies about a future expansion of biofuel production. For instance, in studies regarding the perspectives for ‘sustainable’ modern biomass production in Asian countries such as China, In­dia, Sri Lanka, Malaysia and Thailand, production of biofuels means to a consid­erable extent conversion of forests into plantations (Bhattacharya et al. 2003). And use of marginal lands for biofuel production in Southwestern China is doubted as much of this land is on sloping land that is prone to serious erosion (Naylor et al.

2007) . In Africa, wetlands of high ecological value are increasingly considered for biofuel crops such as sugar cane and oil palm (Sielhorst et al. 2008). Moreover, the land claims associated with the expansion of transport biofuel production should be considered against the background of increasing food production.

Currently on land, about one fifth of net primary production, or somewhat more, is appropriated by humankind (Imhoff et al. 2004; Haberl et al. 2007), and about 38% of land is in agricultural use (FAO 2007). Tilman et al. (2001a) have esti­mated that the area needed for expansion of agriculture for food production until 2050 may be about twice as large as the area of surplus and degraded agricul­tural land identified by Hoogwijk et al. (2003). And the total amount of arable land where wheat, maize, oilseeds and sugar is grown for food and feed purposes is pro­jected to grow between 2000 and 2020, while assuming substantial improvements in yield per hectare (Eickhout et al. 2008). Growing additional crops for ethanol or biodiesel production or establishing plantations of trees will, assuming business as usual, only add to the conversion of nature into ‘culture’, and thus to loss of habitats for living nature (Koh 2007; del Carmen Vera-Diaz et al. 2008; Searchinger et al.

2008) .

Replacement is not the only impact that an expanded production of transport biofuels may have on living nature. It is likely (e. g. Goldemberg 2008) that expan­sion of biofuel production will partly result in intensification of agriculture, which is often associated with the increased use of inputs such as nutrients, pesticides and (irrigation) water and increased drainage (Tilman et al. 2001a; Tscharntke et al. 2005; Liira et al. 2008). This will have side effects that affect biodiversity. It is also possible to harvest nature for biomass that may serve as the basis for biofuel pro­duction. For a specified amount of biofuel, this may well affect a larger area than

for a similar amount of biofuels from crops. The removal of biofuels from forests and other natural ecosystems may impact biodiversity. A case in point is the use of forestry residues, which may negatively affect a variety of species (Norden et al. 2004; Rudolphi and Gustafsson 2005). Natural habitats may also be invaded by species planted for biofuel production (Zedler and Kercher 2004; Lavi et al. 2005; Raghu et al. 2006; Nash 2007).

Effects of replacement of nature by agriculture on biodiversity and ecosystem services will be discussed in Sect. 5.3. The effects on biodiversity and ecosystem services of cropping and harvesting practices and of invasive species used in biofuel cropping will be considered in Sects. 5.4 and 5.5, respectively. But first, in Sect. 5.2, we will go briefly into the impact of biodiversity loss on natural resources which have monetary value.

Biofuels are ultimately based on the ability of photosynthetic organisms to use solar irradiation for the conversion of CO2 into glucose (C6H12O6) and subsequently into biomass; the overall reaction for the conversion into glucose usually being:

6CO2 + 6H2O ^ C6H12O6 + 6O2

Some photosynthetic bacteria may not produce oxygen but give off elemental sul­phur.

In practice, only part of incident solar radiation is captured by plants. And of the solar irradiation captured by plants, only a part (approximately 43-45% of radia­tion in the visible part of the spectrum for land plants) is photosynthetically active (Sinclair and Muchow 1999; Vasudevan and Briggs 2008).

The synthesis of glucose is powered by light reactions generating NADPH, ATP and O2. Thereafter, the reactions can proceed in the dark. In these reactions, collec­tively known as the Calvin cycle, ATP, NADPH and CO2 are converted into glucose, NADP+, ADP and phosphate.

The first enzyme of the Calvin cycle is ribulose bisphosphate carboxylase. As ribulose bisphosphate carboxylase is sensitive to oxygen, photorespiration is impor­tant to protect the enzyme. When CO2 levels in the atmosphere increase, protec­tion by photorespiration can be reduced. At the present atmospheric concentration of CO2, in most plants, photorespiration leads to the release of up to about 50% of the CO2 originally fixed by photosynthesis. These plants are called C3 plants. This name is linked to the first product of photosynthesis that contains 3 C atoms, 3-phosphoglyceric acid. All large trees are C3 plants (Heaton et al. 2008). More re­cently in the evolution of terrestrial plants, a retrofit to the Calvin cycle has emerged that reduces the need for photorespiration. The plants having such a retrofit are called C4 plants. This name is again linked to the first product(s) of photosynthesis that are organic acids with 4 C atoms. Examples of C3 terrestrial plants relevant to biofuels are wheat, rapeseed, soybean, sunflower, eucalyptus, sugar beet, potato, poplar, coconut, cassava, cotton and Jatropha, while examples of C4 plants are sugar cane, corn (maize), switchgrass, sorghum, millet, and Miscanthus.

Natural C4 species tend to be better adapted to relatively warm climates than C3 species. However, breeding and selection have changed the temperature response in a number of C3 and C4 species. Thus, there are now C3 species that do optimally in relatively warm climates (e. g. cotton) and C4 species, such as corn varieties which have been well adapted to temperate climates (El Bassam 1998). The reduced need for photorespiration in C4 species is reflected in a higher maximum theoretical effi­ciency in the conversion of solar irradiation into biomass.

For C4 plants on land at the present concentration of CO2, the maximum theoret­ical efficiency is estimated at 5.5-6.7% and for C3 plants on land at 3.3-4.6% (Hall 1982; El Bassam 1998; Kheshgi et al. 2000; Heaton et al. 2008). For algae, a theo­retical efficiency varying between 5.5 and 11.6% has been suggested (Heaton et al. 2008; Vasudevan and Briggs 2008). Actual efficiencies in commercial cultivation are much lower, as will be discussed in Chap. 2. Most transport biofuels are derived from photosynthetic organisms, though there is also a limited supply of biofuels derived from animals (based on, for example, yellow grease and animal meal).