Obtaining biofuels derived from plantations and forests heavily relies on including parts of trees, such as crowns, that have relatively high concentrations of nutrients (Manley and Richardson 1995; Perry 1998; Sims and Riddell-Black 1998; Pare et al. 2002; Rytter 2002). In the relatively young trees that characterize plantations, nutri­ent concentrations are, moreover, higher than in older trees (Rytter 2002). The use of feedstocks with high nutrient levels adds to losses of nutrients associated with common consequences of tree-harvesting practices such as erosion, increased leach­ing of nutrients and lowered rates of N-fixation by leguminous understory plants (Hamilton 1997; Heilman and Norby 1998; Richardson et al. 1999; Bernhardt et al.

2003) .

Overall losses of nutrients may well have an impact on future productivity. For instance, studies of whole tree harvesting, with branches, tops and needles used as biofuels, as it is currently practiced in Sweden, show deficits in base cations (K, Mg and Ca) (Akselsson et al. 2007). Akselsson et al. (2007) suggest that compensatory fertilization with K, Mg and Ca is necessary to keep forestry sustainable. Studies in French forests show that the budget for Ca and probably Mg is negative, as­suming a 60-year rotation time and a conservative scenario of biomass harvesting (Ranger and Turpault 1999). In the southern USA, P deficits have been noted (Pit­man 2006). In Scandinavian forests, thinning involving whole tree removal has been found to cause significant reduction in stand volume increment linked with nutrient loss (Nord-Larsen 2002). And in tropical dry forests, repeated harvesting may well lead to reduced primary production due to a reduced capture of P from air (Lawrence et al. 2007). Keeping forest soil concentrations of nutrients in a steady state while removing feedstocks for biofuel production with relatively high concentrations of nutrients may, in the absence of nutrient amendments, force the application of long rotations or even an end to harvesting. In Sweden, harvesting trees from nutrient — poor soil is in fact discouraged (Manley and Richardson 1995).

There is also the option of the recycling of nutrients to soils. The extent to which this can be done depends on the use of biomass. For instance, fixed N is almost totally lost during combustion (Sander and Andren 1997). On the other hand, it may be expected that fixed N can, to a large extent, be conserved in the production of methane and ethanol from biomass. In the context of ethanol production from switchgrass, Anex et al. (2007) have proposed a process that may recover about 78% of the fixed N input, which then may be recycled.

Nutrient elements other than N tend to be largely conserved in (fly and bottom) ash during burning and can be retrieved when proper controls are in place. In power plants, biomass is often co-fired with coal, and this will lead to ashes that are often considered unacceptable for nutrient application in forests or on arable soils (Reijn — ders 2005). When only biomass is burned, this may be different. It has been pro­posed to return ashes, especially for their base cations, to forest soils after burning forest-derivedbiomass (Hanell and Magnusson 2005; Pitman 2006; Ozolineius et al.

2007) . Pettersson et al. (2008) have suggested extracting phosphate from ashes for reuse as a nutrient. Also, digestate remaining after anaerobic conversion of biomass to methane, and fermentation residue remaining after converting lignocellulosic biomass into ethanol, may be returned to soils (e. g. Zwart et al. 2007; Reijnders

2006).

As yet, however, nutrient recycling is very limited. Ashes from burning biofuels are not usually returned to soils used for biofuel production, but largely diverted to other destinations, such as landfills (Reijnders 2005; Saikku et al. 2007). There may also be complications in returning nutrients. Most studies have focussed on the recycling of ash, and especially the recycling of base cations such as Ca2+, Mg2+ and K+. Additions of such ashes to forests on mineral soils have shown disappoint­ing effects, which have been linked to N deficiency (Augusto et al. 2008). It has been found that the chemistry of base cations in ashes tends to be different from the original chemistry in soils, and so are local concentrations of base cations in the soil after applications of ash. It has been argued that such differences may be limited by keeping temperatures between 600 and 9000 C during burning and using granulated ashes (Pitman 2006). Furthermore, it has been advised to apply ash to adult stands and not to seedlings (Augusto et al. 2008).

Experience with return of wood ash for its base cations to forest soil shows sub­stantial side effects, for instance, on the levels of aluminium in soil solution and an increased soil emission of CO2 (Maljanen et al. 2006; Ring et al. 2006), sup­pression of denitrification (Odlare and Pell 2009) and especially in the case of high fixed N presence in soils, increased leaching of nitrate (Pitman 2006). Reductions of Mn levels in biomass have been associated with wood ash recycling (Augusto et al.

2008) . Moreover, hazardous elements, such as lead and cadmium, and hazardous organics, such as polycyclic aromatic hydrocarbons, may be present in ashes in sub­stantial amounts. It has been found that even in the apparent absence of substantial anthropogenic contamination, levels of heavy metals in combustion ashes may be remarkably high (Reimann et al. 2008). For example, combustion ashes from South Norwegian birch and spruce wood ashes contained up to 1.3% lead and 203 mgkg~1 cadmium (Reimann et al. 2008). Johansson and van Bavel (2003) and Enell et al.

(2008) looked at the presence of polycyclic aromatic hydrocarbons (PAHs) in wood ash and found that the concentration thereof in a substantial number of cases ex­ceeded the standard applicable in Swedish forests of 2 mgkg-1 of 16 PAHs.

Thus, high concentrations of inorganic and/or organic contaminants may repre­sent a barrier to the sustainable recycling of nutrients. This may require input con­trols for burners (e. g. excluding wood with unacceptable levels of heavy metals), facilities for burning that minimize the formation of hazardous compounds such as polycyclic aromatic hydrocarbons and chlorinated dioxins and/or treatment of ashes to eliminate hazardous compounds. As pointed out in Sect. 3.2, in the case of wastes from lignocellulosic ethanol production, phenolic compounds, ionic compo­sition and pH should be controlled, and the flow of heavy metals should be limited. All in all, it is unlikely that recycling of nutrients after biofuel processing and use can or will be as efficient as nutrient recycling in natural systems.

Regarding plantations, processes such as natural weathering and symbiotic N-fix — ation that are important in providing undisturbed forests with nutrients may well be less productive (Perry 1998). Intensive site preparation, common in plantations, in­volving burning biomass may negatively affect productivity as it leads to the volatil — isationof nutrient N (Perry 1998). Because root systems in short rotation plantations may be less extensive than in undisturbed forests, the leakage of nutrients may well increase (Ong and Leakey 1999). Harvesting practices on plantations tend to in­crease denitrification and leaching, which may lead to increased nutrient deficits (Hamilton 1997; Heilman and Norby 1998). It is likely that erosion on plantations will exceed the value range 0.004-0.05 Mgha-1year_1 that is found in undisturbed forests. Still, by judiciously planting and harvesting trees, it may be possible to keep erosion rates below the level of 1 Mgha-1year_1 (Pimentel et al. 1997b).

Overall, as rates of biomass harvesting also tend to be much higher on plantations than in forests, large deficits in nutrients are to be expected. For instance, regarding aspen-for-fuel plantations, Rytter (2002) estimated the yearly deficit per hectare of N at 30 kg, for P at 4 kg, for Ca at 30 kg, for Mg at 4 kg and for S at 2.5 kg. Indeed, the productivity of growing short rotation trees strongly depends on external nutrient inputs (Adegbidi et al. 2001). Remarkably, current human activity, especially in in­dustrialized countries, has led to increased environmental fluxes of wasted nutrients such as S, N and P, which reach soils via air and/or water (Smil 1991; Kvarnstrom and Nilsson 1999). For instance, in North America, unintended N depositions on soils may be up to 53 hgha^year-1, and in Europe up to 115 kg N ha-1year-1 (Heilman and Norby 1998). It is probable that the unintentional addition of nutri­ents to soils has been important to maintaining productivity in the absence of inten­tional nutrient amendments. For aspen plantations in Southern Sweden, it has, for instance, been calculated that deposition exceeds the yearly S deficit (2.5 kgha-1) and may cover a substantial part of the N deficit. In forests in Southern Sweden, there may still be accumulation of N with current biofuel harvesting practices (Ak — selsson et al. 2007). However, such unintended additions of nutrients to soils are not designed to fit all actual deficits in nutrients. For instance, though S deficits are com­pensated for in Nordic spruce forests, nutrient deficiencies for P, K and B still occur (Rytter 2002). Unintended additions of nutrients to soils may also cause new prob­lems, especially when unintended nutrient additions are relatively high. They may well contribute to long-lasting eutrophication and acidification and deterioration of ground water quality (Galloway et al. 2008).