Studies regarding annual crop production may provide clues about the factors nec­essary for a type of land-based biomass production that has high productivity and may be maintained indefinitely. Syers et al. (1997),Vance (2000) and Lal (2001a, b)

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

have surveyed such studies and did show that one important factor in maintain­ing high productivity is soil conservation and the maintenance of high levels of organic matter in the upper layer of the soil. Loss of soil due to erosion ultimately leads to a strong decline in crop productivity (Lal 2001). Soil erosion is a major problem in annual crops, including the production of major current biofuel feed­stocks such as corn, sugarcane and soybeans (Mahadevan2008; Smeets etal. 2008), but also may be a problem in plantations and forests (Worrell and Hampson 1997; Perry 1998).

Soil organic matter is an important reserve for plant nutrients such as nitrogen (N) and phosphate (P). It improves soil structure and water-holding capacity (Kahle et al.

2002) and limits erosion (Troeh et al. 1999). Soil organic matter is also involved in weathering that extends the availability of nutrients (McBride 1994). Depletion of organic matter in soils ultimately results in a decrease in yields (Syers et al. 1997; Perry 1998). In many areas of the world, arable land currently shows a net loss of soil organic matter, if compared with its virgin or natural status (Cole et al. 1997; Ogle et al. 2005). Levels of soil organic carbon that are under long-term cultivation with annual crops tend to be lower than under native vegetation. On average, soil carbon levels under arable soils with a long history of cultivation tend to be approxi­mately 18% lower under temperate dry conditions, approximately 30% lower under temperate moist or tropical dry conditions, and approximately 42% lower in tropi­cal moist climates if compared with soils under native vegetation (Ogle et al. 2005). These reductions of carbon levels in soils have contributed to increased levels of greenhouse gases in the atmosphere. In many soils in tropical and subtropical ar­eas, especially in Asia and sub-Saharan Africa, due to systematic excessive residue removal, soil organic carbon pools have decreased to levels that are conducive to soil degradation. Such soils also show substantially reduced production levels (Lal 2008).

Current agricultural practice often leads to further losses of soil carbon from arable soils. Net losses of soil carbon have been documented for the European Union (Vleeshouwers and Verhagen 2002), Eastern Canada (Gregorich et al. 2005), China (Li et al. 2003; Tang et al. 2006; Wright 2006), Nepal (Matthews and Pilbeam 2005; Shrestha et al. 2006), Brazil (Zinn et al. 2005; Jantalia et al. 2007), Sudan (Ardo and Olsson 2003), the southern Ethiopian highlands (Lemenih and Itanna 2004) and in West Africa (Ayanlaja et al. 1991; Ouattara et al. 2006; Bationo et al. 2007; Lufafa et al. 2008). From peaty arable soils losses may be especially high when there is deep drainage and intensive mechanical soil disturbance (Freibauer et al. 2004). Net carbon losses varying from 6 Mg in northern Norway up to 15MgCha-1year-1 in the tropics have been reported (Granlund et al. 2006; Reijnders and Huijbregts 2007, 2008).

It has emerged that low crop yields, the absence of cover crops, low additions of crop residues, low additions of manure, mechanical tillage and high temperatures enhance the loss of carbon from arable soils (Vleeshouwers and Verhagen 2002; Pretty et al. 2002; Freibauer et al. 2004; Lal and Pimentel 2007; Valentin et al. 2008). There is evidence that excessive reliance of synthetic nitrogen fertilizers may be conducive to carbon loss (Kahn et al. 2007; Triberti et al. 2008). It has also been

found that depending on the nature of the soil, climate, crops and crop rotation, crop residues management and tillage system, a partial (20-50%) removal of crop residues from the field reduces the pool of soil organic carbon, can exacerbate soil erosion hazard and negatively impact future yields of crops (Wilhelm et al. 2004; Lal 2005; Blanco-Canqui and Lal 2007; Lal 2008; Varvel et al. 2008). Models with parameters based on empirical data have been developed to estimate the impact of residue removal on soil organic carbon (e. g. Saffih-Hdadi and Mary 2008).

Sustainable use of soil and soil organic matter should be such that levels of soil organic matter do not decrease and that soil loss (erosion) should not exceed addition to topsoil stocks by natural processes. The latter may add between 0.004-0.5 mm of topsoil per year (Cannell and Hawes 1994). A variety of measures has been pro­posed to reduce erosion in annual cropping. These come under the umbrella of the term conservation tillage (Cannell and Hawes 1994; Lal 1997, 2001, 2008). They include the reduction of tillage (preferably to no-till or zero tillage), the use of cover crops and nitrogen-fixing legumes, intercropping, contour farming, increased return of harvest residues or residue mulches to the soil, use of manure, direct seeding, cor­recting effects of soil compaction due to vehicles and catching soil subject to water erosion on sloping soils by terracing or barriers (Gumbs 1993; Cannell and Hawes 1994; Lal 1997, 2001, 2008; Lal et al. 2007; Mills and Fey 2003). High inputs of carbon (residues, manure, compost) in soils that are subject to tillage are conducive to maintaining soil organic carbon levels (Jenkinson et al. 1990; Grace et al. 2006; Reijneveld et al. 2009).

In forests, limitations to harvesting and the use of heavy machinery on erodible soils and judicious planting may be necessary to prevent soil losses from exceeding additions to topsoil stocks (Pimentel et al. 1997b; Worrell and Hampson 1997). To maintain soil organic matter levels, in intensively managed forests and on planta­tions, intensive site preparation involving burning should be avoided, as this leads to volatilisation of soil carbon and prospective soil carbon (Perry 1998; Bauhus et al.

2002) . Also, limitations to removal of harvest residues from forests may be neces­sary to maintain levels of soil organic matter (Worrell and Hampson 1997).

Soils to which crop residues are returned tend to store more soil organic carbon (and nitrogen) than plots where residues are taken away (Vance 2000; Mendham et al. 2002; Dolan et al. 2006; Epron et al. 2006). In this respect, there are two phe­nomena with opposite effects for C4 and C3 plants. Residues from plants that have C4 photosynthesis seem less effective in contributing to soil carbon than the same amounts of residues from plants with C3 photosynthesis (Wynn and Bird 2007). On the other hand, C4 crops rather often generate relatively large amounts of below — and aboveground biomass, if compared with C3 crops (Wright et al. 2001; Wilhelm et al. 2004). Vleeshouwers and Verhagen (2002) estimate that adding to arable soils the cereal straw that is currently taken away may, on average, increase European soil carbon levels by 0.15Mgha~1year~1. Other studies have shown that full return of crop residues to arable soils in temperate climates may increase soil carbon lev­els by up to 0.7Mg C ha-1year-1 (Webb et al. 2003; Smith 2004; Freibauer et al. 2004; Rees et al. 2005). For maintaining soil carbon stocks in tropical soils, return­ing residues, application of other organic matter such as manure, shrub prunings and household composts, cover crops and fallows have been advocated (Lal and Bruce 1999; Bationo and Buerkert 2001; Nandwa 2001; Lufafa et al. 2008).

A changing climate will impact soil carbon stocks. In part, this impact is de­pendent on crop productivity. There is only limited empirical evidence about likely future crop productivity. Kim et al. (2007) have studied the C4 crop corn and con­cluded that under elevated CO2 concentration, productivity may remain unchanged. It has also been suggested that under elevated CO2 concentrations in the atmosphere, productivity of C3 plants may increase and that this may enhance soil carbon se­questration (Marhan et al. 2008), but increased temperature also leads to increased respiration in soils, and soil carbon dynamics may be impacted by changes in precip­itation (Marhan et al. 2008). Overall effects may vary for different climate regions. In the European context, Vleeshouwers and Verhagen (2002) estimate that an in­crease in average temperature of 1 °C caused by an increase of CO2 concentration may, ceteris paribus, lead to an average net loss of soil organic carbon of about 0.04 Mgha~1 year-1.

In view of carbon losses, increased use of agricultural residues has been advo­cated in order to maintain (or restore) proper levels of soil organic carbon and ensure agroecosystem sustainability (Lal 1997; Duiker and Lal 1999; Lal 2001,2008; Oue — draogo et al. 2006; Chivenge et al. 2007; Lal and Pimentel 2007). Adding lignocellu — lose to arable soil is more useful in this respect than more easily degradable carbon compounds such as sugars or starches (Sartori et al. 2006). Available evidence is limited but suggests that approximately 4-24% of carbon contained in residues of crops may be converted to refractory soil organic carbon in agricultural soils (Lal 1997; Follett et al. 2005; Razafimbelo et al. 2006; Triberti et al. 2008).

Crop residues contain cellulose in a matrix of lignin and hemicellulose. Lignin is, together with compounds such as cutins, suberins and tannins, largely respon­sible for humus formation in arable soils (Kirk 1971; Rasse et al. 2005) and in such soils is a major contributor to refractory soil organic carbon (Loveland and Webb 2003; Chapman and McCartney 2005). There is evidence that among the compo­nents of lignocellulose (lignin, hemicellulose and cellulose) in arable soils, lignin is by far the most refractory component (Melillo et al. 1989; Spaccini et al. 2000; Quenea et al. 2006). Thus, lignin is more suitable for carbon sequestration in arable soils than hemicellulose. For this reason, removal of residues with high concentra­tion of lignin (such as nut shells) may be expected to be more negative to arable soil carbon stocks than residues with a lower lignin level, such as wheat or rice straw.

Still, the presence of carbon compounds, which are more easily degraded than lignin (with a half life less than 1 year), in arable soil is also important for soil fertil­ity and stability (Spaccini et al. 2000; Loveland and Webb 2003). The carbohydrates hemicellulose and cellulose in harvest residue belong to this category (Spaccini et al. 2000).

Against this background, systematic removal of all crop residues for biofuel pro­duction is not a good idea (Lal 2008; Reijnders 2008; Saffih-Hdadi and Mary 2008). Limitations on crop residue removal will have an upward effect on energy input into, and costs of residue collection for, biofuel production (Higgins et al. 2007). To the extent that crop residues are removed, there is furthermore a case for selecting crop residues for the production of the transport biofuel ethanol that have relatively high concentrations of hemicellulose and cellulose susceptible to conversion into ethanol. In the case of corn stover, this fraction consists of cobs, leaves and husks (Crofcheck and Montross 2004). The crop residue fraction that is relatively rich in lignin may be expected to be a relatively efficient contributor to refractory carbon in arable soils, but also contains a substantial amount of carbohydrates that are more easily degradable and contribute to soil fertility and stability. Thus, it may be that, for example, the scope of residue removal for ethanol production may be widened by selecting residues on the basis of their relative suitability for ethanol production and for the formation of refractory soil carbon, respectively.

Another option is to consider a return of processing ‘wastes’ of crop residues that are relatively rich in lignin. In generating ethanol from crop residues by enzymatic conversion (see Chap. 1), a residue emerges that is rich in lignin and also contains unreacted cellulose and hemicellulose (Mosier et al. 2005). It would seem worth­while to consider applying this residue to arable soils. Such an application would serve the presence of refractory carbon in arable soil, while it may also contribute to the presence of more rapidly degradable carbohydrates. In doing so, one should limit or prevent undesirable side effects of adding this processing residue. A matter to consider in this respect is the accumulation potential of the residue for phenolic carbon compounds. Such accumulation may occur under anaerobic conditions, and this may have a negative effect on soil fertility (Olk et al. 2006). Ionic composition and pH of the processing residue are subject to limitation when use of the residue is to be sustainable (Mahmoudkhani et al. 2007). Also, one should be aware that lignin binds heavy metals such as cadmium much better than cellulose does (Basso et al. 2005). Thus, provisions should be in place to limit the flow of heavy metals to soils when the fraction that is rich in lignin is applied. If the processing residue has acceptable quality, it may well be that the amount of crop residue that can be removed from the field without a negative impact on soil characteristics can be in­creased. The quantitative and qualitative aspects of applying processing residues to arable soils would seem to merit further research. Finally, it should be noted that the refractory character of lignin in the arable soils studies cannot be generalized to all soils. There is, for instance, evidence that in lowland tree plantations in Costa Rica, litter decay increases with increasing lignin content (Raich et al. 2007).