PERGAMON Biomass and Bioenergy 16 (1999) 137±154

Environmental e€ects of energy crop cultivation in SwedenÐI: Identi®cation and quanti®cation

PaÊ lBoÈ rjesson *

Department of Environmental and Energy Systems Studies, Lund University, Gerdagatan 13, SE-223 62 Lund, Sweden

Abstract

This paper presents an analysis of how energy crop cultivations in Sweden, consisting of short-rotation forest (Salix) and energy grass (reed canary grass), can be located and managed to maximise environmental bene®ts. The overall conclusion is that substantial environmental bene®ts, ranging from global to site-speci®c, could be achieved when traditional annual food crops produced with current agriculture practices are replaced by dedicated perennial energy crops. The emission of greenhouse gases could be reduced by reduced carbon dioxide emissions from organic soils, by reduced nitrous oxide emissions caused by the use of fertilisers and through accumulation of soil carbon in mineral soils, which also leads to increased soil fertility. Nutrient leaching could be reduced by using energy crop cultivations as bu€er strips along open streams and wind erosion could be reduced by using Salix plantations as shelter belts. Cultivation of Salix and energy grass can also be used to purify municipal waste, such as waste water, land®ll leachate, and sewage sludge. Furthermore, the content of heavy metals in the soil can be reduced through Salix cultivation. The biodiversity is estimated to be almost unchanged, or slightly increased in open farmland. These environmental bene®ts, which could be achieved on up to 60% of current Swedish arable land and last for 25 years or more, will increase the value of the energy crops. The economic value of these bene®ts is calculated in Part II of the analysis, which is presented in a second paper. # 1999 Elsevier Science Ltd. All rights reserved.

Keywords: Energy crops; Salix; Reed canary grass; Cultivation; Environmental e€ects

1. Introduction cated perennial energy crops on Swedish arable land, negative environmental e€ects from current Replacing fossil fuels with biomass leads to agriculture practices such as erosion, nutrient global environmental bene®ts such as reduced leaching and the emission of greenhouse gases, emissions of greenhouse gases. However, not may be reduced. Similar conclusions have been only the use of biomass could lead to bene®ts for drawn in earlier studies concerning, for example, the environment, but also the production of the perennial energy crop production in the USA biomass. This paper will show that when tra- (see e.g., Tolbert [1]; McLaughlin and Walsh [2]; ditional annual food crops are replaced by dedi- Kort et al. [3]; Grigal and Berguson [4]; Bransby et al. [5]). Energy crop systems can also be used to purify municipal waste, thus reducing negative * E-mail: [email protected] impacts from the community. Furthermore, the

0961-9534/99/$ - see front matter # 1999 Elsevier Science Ltd. All rights reserved. PII: S0961-9534(98)00080-4 138 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 content of heavy metals in the soil could be energy purposes as the climatic conditions are reduced through cultivation of short-rotation for- not suitable for Salix cultivation [9]. This situ- est. These environmental bene®ts could increase ation can, however, change in the future if more the value of the energy crops, thereby a€ecting frost resistance Salix clones will be available. The future market conditions for biomass. energy crop cultivations are assumed to be geo- In this paper, environmental changes are ident- graphical equally distributed on current arable i®ed and quanti®ed when traditional food and land used for perennial forage and annual food forage crops produced with current agriculture crop production [10]. Today, about 40% of practices are replaced by perennial energy crops Swedish arable land is used for the production of on Swedish arable land. This paper, which is perennial forage crops, mainly ley, while about based mainly on a literature review, represents 60% is used for annual crop production. When Part I of the analysis, which is complemented by perennial energy crops replace perennial forage a second part in which an economic valuation of crops (in Sweden normally clover-grass ley), the the environmental changes is carried out (see environmental changes are assumed to be small BoÈ rjesson [6]). Thus, this paper is a background and are therefore neglected, except when the to the second paper, which includes a more inte- impact is particularly caused by the cultivation of grating synthesis. The purpose with the overall Salix, e.g. reduced wind erosion and cadmium study is to analyse how energy crop cultivations removal. In annual crop cultivation, which in in Sweden could be located and managed to Sweden is dominated by grain cultivation, maximise environmental bene®ts. Neither en- ploughing is assumed to be undertaken annually, vironmental e€ects arising from replacing fossil the average dose of chemical nitrogen (N) fertili- fuels by biomass nor environmental bene®ts from ser applied is about 140 kg N/ha yr, and the the reduced use of fossil fuels in perennial energy average dose of chemical pesticides is about 1 kg crop cultivation, in comparison to annual crop active ingredients/ha [7]. Reed canary grass culti- cultivation, are included in this study as such vation, harvested once a year, extends over 10 bene®ts have already been analysed in earlier stu- years, while Salix cultivation, harvested every dies (see e.g., BoÈ rjesson [7, 8]). The magnitude of fourth year, extends over 25 years. The biomass the environmental bene®ts depends on, for yield of Salix is assumed to be, on average, 10 example, geological and geographical conditions, tonne dry matter/ha yr, and of reed canary grass, while population density in¯uences the amount 6.5 tonne dry matter/ha yr [7]. of municipal waste that can be recycled on The environmental impact of energy crop culti- energy crop cultivations. Thus, the environmental vation varies between di€erent regions in Sweden impact is analysed on a regional basis and then due to local conditions. Here, impacts are ana- aggregated to a national level. Since agricultural lysed for each county of Sweden (24 in 1996) and practices, the load of antrophogenic pollutants then aggregated to a national level. and the generation of waste, are changing over Soil type distribution on arable land in Sweden time, the potential impact of energy crop cultiva- is divided into four classes using the American tion also changes over time. The time during soil classi®cation system, regarding di€erences in which environmental advantages could be physical and chemical properties of the soils achieved is therefore estimated. (Table 1). Cultivation methods in agriculture are chan- ging, as well as the load of antrophogenic pollu- 2. Methodology and assumptions tants and the generation of waste. Thus, the time during which environmental impact could be sus- Energy crops included in this analysis are tained is estimated. This estimate is based on short-rotation forest (Salix) and energy grass physical and chemical properties of the soils, en- (reed canary grass). In the north of Sweden, only vironmental goals for future agricultural prac- reed canary grass is assumed to be grown for tices, recycling of waste and reduction of P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 139

Table 1 cultivation, in comparison to annual crop cultiva- Soil type distribution (%) on arable land in Sweden [11±13] tion, is restricted to be included only in the cat- Mineral soils egory ``biodiversity''. No analysis of the reduced risk of, for example, ground water pollution has Clayey Loamy Sandy Organic here been carried out due to lack of data. soils* soils$ soils soils%

55 19 17 9 3. Identi®cation and quanti®cation of * Including sandy clay loam, silty clay loam, and clay loam. environmental e€ects $Dominated by silt. %More than 20% organic matter 3.1. Greenhouse gases antrophogenic emissions, and prognoses of the possibility of attaining these goals. Greenhouse gas emissions from arable land The term ``environmental change'' or ``changed can be reduced in three di€erent ways when environmental impact'' is used here in a broad annual crops are replaced by perennial energy sense, as e€ects on soil, water, air, ¯ora and crops, through (i) accumulation of soil carbon fauna, and the recirculation and purifying of mu- (C) in mineral soils, (ii) reduced carbon dioxide nicipal waste, are included. Environmental (CO2) emissions from organic soils, and (iii) changes when perennial energy crops replace reduced nitrous oxide (N2O) emission caused by annual crops identi®ed in this paper have been the use of fertilisers. Changes in the emissions of divided into six categories: (1) greenhouse gases, other greenhouse gases are estimated to be small (2) nutrient leaching, (3) heavy metals, (4) soil and are therefore neglected. fertility and erosion, (5) municipal waste treat- ment, and (6) biodiversity. 3.1.1. Accumulation of soil carbon in mineral soils The aesthetic impact of short-rotation forest The rate of change in soil carbon depends on on the landscape has not been considered here. the rate at which organic matter is added to the Depending on the nature of the surrounding soil and the rate at which erosion and biological landscape, and the design of the cultivations, oxidation remove the organic matter from the Salix cultivation is assumed to have the possi- soil [16]. The two latter mechanisms are substan- bility of causing both positive and negative tially reduced if soil tillage is reduced or elimi- e€ects on the landscape [14]. Neither is the nated, which is the case when annual crops are impact of Salix on tile drainage considered, replaced by perennial energy crops. Also, the which could lead to expenses. There is a great input of organic matter from litter and roots to risk of damage to the tile drainage when Salix is the soil is high in energy crop production. In an cultivated, due to the deep Salix roots [15]. This established Salix cultivation, 4.5±10 tonne dry risk has not been considered, as about two thirds matter/ha is recirculated yearly, of which about of the total Swedish arable land lacks tile two thirds are litter and one third is roots [17]. drainage [10]. Salix may also be grown in ®elds Only a minor part, however, is converted into with old tile drainage, which has mainly served stable humus containing, on average, about 60% its function. Salix cultivations are assumed to C, as most of the organic residues are decom- extend over 25 years, which is about the same posed within period of a few weeks up to a period of time as the technical life-time of a tile year [16]. drainage system. If energy crops are to be culti- Results from Swedish [17] and German [18] vated in ®elds with new tile drainage systems, ®eld trials shows that a vertical gradient of soil energy grass should be grown in preference to carbon is developed after a few years of Salix Salix. cultivation, with a higher carbon content in the Environmental bene®ts from the reduced use top soil. No signi®cant changes in the total con- of chemical pesticides in perennial energy crop tent of soil carbon have yet been recorded, as 140 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 these changes take place over a long period of ganic soil is drained, a lowering of ground level time [16]. Long-term ®eld trials with ley grass in is also caused by compression and shrinking [21]. Sweden, which could be comparable to Salix The carbon loss is higher in annual crop cultiva- regarding changes in soil carbon [17], show an tion than in perennial crop cultivation due to increased level of humus equivalent to 30±40 more frequent soil tillage. Results from Swedish tonne C/ha, compared with food crops after 30 investigations [21] show the subsidence rates are years [19]. To some extent, this di€erence could greatest when the crops are potatoes and carrots, be explained by changes in cultivation practices, 2±3 cm/yr, followed by grain 1±2, ley 1, and per- as the depth of ploughing is probably greater manent pasture 0.5 cm/yr. Based on these results, now than in the 1960 s, causing a higher dilution a change from cultivation of annual crops to per- of soil carbon in food crop cultivation with fre- ennial energy crops is estimated to reduce the quent ploughing. Therefore, a change from subsidence rate from, on average, 2 to 1 cm/yr. annual to perennial crop production is estimated This is equivalent to about 7 tonne C/ha yr, as to lead to a somewhat lower increase in soil car- the density and carbon content of cultivated or- bon content than the results of ®eld trials pre- ganic soils in Sweden are about 200 kg/m3 and sented above indicate. 35%, respectively [22]. This theoretical calcu- An annual carbon accumulation in mineral lation has, however, not been veri®ed by ®eld soils of 0.5 tonne C/ha is assumed here, over a trials [23]. Salix thrives best on organic soils con- period of about 50 years. After that, the humus sisting of fen peat (Carex). If the organic soil is level is estimated to reach a new, higher, steady composed of bog peat (Spha gnum), however, state of about 1±1.5 percentage points higher Salix is not a suitable crop due to the low pH of than today. This estimate is, however, uncertain the soil. Thus, bog peat soil is more suitable for and must be veri®ed by long-term ®eld trials. the cultivation of reed canary grass.

Changes in soil carbon storage also vary for The reduced CO2 emission from organic soils di€erent soils and locations due, for example, to when cultivating energy crops is assumed to last, soil texture, drainage, base status and climatic on average, 80 years after which the peat layer parameters [4]. The proportion of the total will have been dissipated by biological oxidation. Swedish arable land where energy crop cultiva- A rough estimate is that the thickness of the peat tion could increase the soil carbon level is esti- layer, which can vary greatly both within a ®eld mated to be 55%. This is equivalent to the area and between di€erent ®elds [21, 24], varies from a covered by mineral soils (which amounts to few decimetres up to two metres in Swedish or- about 90% of the total arable land, see Table 1) ganic soils, with an average of about 80 cm [22]. on which annual crops are cultivated today. At The proportion of arable land where energy crop present, about 60% of Swedish arable land is cultivation could decrease the CO2 emission by used for annual crop production [10]. On the about 7 tonne C/ha yr, is estimated to be 5%. remaining 40%, where perennial forage crops are This is equivalent to the area covered by organic cultivated, no increase in the soil carbon level is soils (which amounts to about 9% of the total assumed to occur. arable land, see Table 1) on which annual crops As a comparison, Ranney and Mann [20] esti- are cultivated today. mate that short-rotation plantations in North

America increase the soil carbon inventories 3.1.3. Reduced N2O emission from mineral soils (excluding litter and roots) by about 10 tonne/ha Nitrous oxide emission from arable soils is over 20±50 years. induced by the use of fertilisers. Loss of fertiliser

nitrogen (N) as N2O varies from almost zero up 3.1.2. Reduced CO2 emission from organic soils to 2%, due to di€erences between soils, fertiliser When organic soils are cultivated, soil carbon type, climate and land management is released by oxidation leading to a subsidence practices [25, 26]. Since energy crops have a lower in ground level. In an initial phase when an or- nitrogen demand than food crops, on average, P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 141 about 50 kg N/ha yr [7], a change in land use 3.2.1. Reduced nutrient leaching in general from food to energy production is assumed to Growing perennial energy crops instead of reduce the emission of N2O. Regression analyses annual food crops reduces the risk of water pol- by Bouwman [27] show a statistically weak corre- lution through leaching and runo€, due to lation between N2O emission and input of nitro- reduced input of fertiliser, longer growing season, gen fertiliser. A rough estimate from this soil cover all year round, and a more extensive correlation is, however, that growing perennial root system [30, 31]. Nitrogen (N) leaching from energy crops instead of annual crops on mineral short-rotation forest and ley production, for soils will reduce N2O emission by, on average, example, is estimated to be 30±50% and 75% 0.3 kg N2O-N/ha yr. A similar estimate has been lower, respectively, than from grain made by the National Swedish Environmental production [18, 32, 33]. This bene®t will be great- Protection Board [28]. This reduction is equival- est on coarse-textured sandy soils, as the nitrogen ent to about 40 kg C relative to the Global leakage from these soils is, on average, twice as

Warming Potential (GWP) of CO2 over 100 high as from ®ne-textured clayey soils [33]. The years [29]. average nitrogen leaching from arable land varies Apart from lower input of nitrogen fertilisers, between 13 and 48 kg N/ha yr in southern and the risk of N2O losses in perennial energy crop central Sweden, where the highest losses are for plantations is assumed to be reduced due to the south-west Sweden [34]. low availability of mineral nitrogen in the soil, as It is assumed here that nitrogen leaching is energy crops have a longer growing season and a reduced by, on average, 50% when perennial more extensive root system than annual energy crops are used to replace annual crops, crops [28]. On the other hand, N2O losses could which is equivalent to about 10 kg N/ha yr [34]. be stimulated by the nitrogen-rich litter produced This reduction can be achieved on about 60% of in Salix plantations and if the tile drainage were Swedish arable land (including both mineral and to be damaged by roots, leading to wet organic soils), on which annual crops are culti- soils [17, 18]. vated today. The proportion of the arable land on which energy crop cultivations could decrease the N2O 3.2.2. Reduced nutrient leaching through bu€er emission is estimated to be 55%, which is equiv- strips alent to the mineral soils on which annual crops Perennial energy crops could also be estab- are cultivated today. Changes in N2O emission lished between open streams and food crop culti- from organic soils are not considered here as the vations as vegetation ®lters decreasing leaching input of nitrogen fertiliser is assumed to have a and runo€ [20, 35]. The eciency of nutrient smaller impact on the N2O emission from these retention in these bu€er strips depends on water soils, than from mineral soils. The content of or- ¯ow pathways controlling the transport of nutri- ganic nitrogen, for example, is much higher in or- ents through the landscape, and the composition ganic soils than in mineral soils, and thus the and width of the riparian zones. Nitrogen reten- rate of nitrogen release in organic soils may have tion by biological transformation, e.g., vegetation a greater impact on the N2O emission. uptake and denitri®cation by micro-organisms, is increased by longer water transit times and is mainly con®ned to the upper soil stratum, thus 3.2. Nutrient leaching a€ecting surface water and shallow groundwater [36]. When perennial energy crops replace annual If the ®eld is tile drained, nitrogen retention in crops, the nutrient leaching is reduced. Nutrient bu€er strips will be signi®cantly reduced since leaching from annual crop cultivation could also 80±90% of the nitrogen is transported in subsur- be reduced by using energy crop cultivations as face ¯ows as dissolved nitrate (NO3) [35]. The bu€er strips along open streams. phosphorus (P) retention, on the other hand, is 142 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 only a€ected to a small extent by tile drainage, clay soils with a high degree of tile drainage [10]. as phosphorus transport mainly takes place as This assumption must be investigated further. surface runo€, as more than 70% of the phos- The theoretical maximum area of bu€er strips phorus is normally particle bound. Thus, the (with a width of 50 m) for nitrogen retention, bene®ts of reduced phosphorus losses will be based on an occurrence of open streams of 8 m/ greatest on easily eroded soils, such as clayey and ha in Swedish farmland, is equivalent to 2.6% of silty soils [37]. An increase in width of riparian the current arable land. Arable land with tile zones up to 25 m will increase retention, where drainage, ley production, and nitrogen leaching often more than 70% of the total content of less than 15 kg N/ha yr, is excluded. Nitrogen nitrogen and phosphorus is removed from the leaching from 10±15% of current arable land water ¯ow [35]. A prerequisite for a continuous could be treated in these bu€er strips. The theor- high retention is, however, that the biomass is etical area of bu€er strips for phosphorus reten- harvested regularly leading to an output of nutri- tion on clayey and silty soils where annual crops ents from the bu€er strips, thus avoiding nutrient are cultivated is equivalent to 3.7% of current saturation. arable land. Phosphorus leaching from 15±20% It is assumed here that bu€er strips, with a of Swedish arable land could be treated here. width of 50 m, consisting of energy crops, are However, considering the practical application of established between annual crop cultivations and perennial energy crop cultivations as riparian open streams. Half of the width (25 m) is har- bu€er strips, the potential area would probably vested at a time, leading to a continuous high be lower due to di€erent obstacles not identi®ed uptake of nutrients. The bu€er strips consist here. mainly of Salix, but with a narrow zone of The period of time during which energy crop energy grass closest to the stream, for practical cultivations will signi®cantly reduce nutrient reasons. The nitrogen retention, which is leaching is estimated to 25 years or more. assumed to be 2/3 by root uptake and 1/3 deni- Current measures in agriculture aiming to reduce tri®cation, is estimated to be, on average, 70 kg nutrient leaching are not sucient of attaining N/ha yr, when the nitrogen leaching from stated environmental goals [41]. Also, a structural upstream cultivations of annual crops exceeds 15 change towards a more even geographic distri- kg N/ha yr [38]. There is only a small variation bution of crop production and animal production in the nitrogen retention in absolute terms, as a takes time. According the National Swedish lower load of nutrient on the vegetation zone Environmental Protection Board [41], uneven leads to higher relative retention, and vice versa. geographic distribution of crop production and The phosphorus retention in bu€er strips is esti- animal production is a major reason for the pro- mated to be, on average, 1.5 kg P/ha yr [38]. The blems associated with nutrient leaching and average loss of phosphorus from Swedish arable eutrophication. land is about 0.3 kg P/ha yr [39]. The occurrence of open streams in the farm- 3.3. Heavy metals land a€ects the possibility of establishing riparian bu€er strips. Results from an investigation in The annual accumulation of cadmium (Cd) in southern Sweden show that the occurrence of Swedish arable soils is equivalent to 0.20±0.25% open streams is about 6 m/ha, and the occurrence of the total content of cadmium in the soil, of streams in culverts about 7 m/ha [40]. Since which has increased by around 33% during this there are no data for the rest of Sweden, an aver- century [42]. One result of this cadmium accumu- age value has been assumed in this paper. The lation is that in southern Sweden, the cadmium occurrence of open streams in Swedish farmland content in grain harvested on intensively culti- has been estimated to be, on average, 8 m/ha. vated farmland sometimes exceeds the limit of This is 30% higher than in southern Sweden, 0.1 mg Cd/kg grain, proposed by WHO/FAO [43]. since this region consist of intensively cultivated However, Salix plantations could be used to P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 143 lower the content of heavy metals in arable land, siderably reduced. The atmospheric deposition of as the uptake of heavy metals in Salix shoots is cadmium, as well as the load from phosphorus normally much higher than in other crops. The fertilisers and sewage sludge, is assumed to be uptake of cadmium in Salix shoots, for example, reduced in the future due to national environ- is 35 to 70 times higher than in energy grass and mental programmes in accordance with inter- straw [44]. The cadmium could then be removed national agreements [52, 53]. from the ash through ¯ue gas cleaning when the Salix is combusted [45]. There are, however, sig- ni®cant di€erences in the eciency of cadmium uptake between di€erent Salix species and 3.4. Soil fertility and erosion clones [46±48]. It is mainly the fraction of cadmium in the soil Current agriculture practices employing inten- available to the plant that is reduced by sive soil tillage and a high proportion of annual Salix [49, 50]. The plant-available fraction of cad- crops, has resulted in loss of soil humus and mium in the topsoil normally amounts to 30± decreased soil fertility [27]. Also, the impact on 40% of the total cadmium content, depending on the Swedish agricultural landscape has been con- the pH and contents of clay and organic matter siderable, as both the number of ®elds as well as in the soil. The total content of cadmium in the the occurrence of permanent vegetation zones topsoil of Swedish arable land has been estimated have been drastically reduced since the to be, on average, 600 g/ha [43]. The content of 1930 s [54]. These changes have resulted in cadmium in Salix shoots varies between 0.4 and increased wind and water erosion [55, 56]. An 3.9 mg/kg dry matter [49]. An average cadmium increased proportion of perennial energy crop content of 2.0 mg/kg dry matter is equivalent to cultivations in the agricultural landscape may an uptake of about 20 g Cd/ha yr, when the increase soil fertility and reduce soil erosion. annual increment in biomass is 10 tonne dry mat- ter/ha yr. The annual uptake of cadmium will, however, decrease over time, as the fraction of plant-available cadmium in the soil is reduced. 3.4.1. Increased soil fertility To some extent, cadmium might also be redis- An increased humus level in mineral soils by tributed from the subsoil to the topsoil through 0.5±1 percentage points after about 25 years of deep Salix roots and litter fall, thus reducing the energy crop cultivation (see Section 3.1.1.), is net output of cadmium from the top soil [49]. It estimated to increase food crop harvest by on is estimated that the highest output of cadmium average 5% [57]. This is valid for soils with a from the topsoil through Salix cultivation could humus content of less than about 6%, as physical be achieved on soils with a high content of cad- and chemical properties of the soil are improved mium in the topsoil due to antrophogenic depo- up to this humus level [33]. Increased yields sition, and with a low, natural, content in the through higher levels of humus in mineral soils subsoil [51]. could then theoretically be achieved on about It is here assumed that the annual net output 47% of current Swedish arable land, considering of cadmium from the topsoil through Salix culti- mineral soils with a humus level of less than 6% vation is equivalent to, on average, 1% of the on which annual crops are cultivated [10, 11]. current total content of cadmium. This output of The period of time during which higher yields of about 6 g Cd/ha yr will exceed the yearly input annual crops would be achieved is estimated to of cadmium by 4 to 5 times. The period of time be equal to the period of time during which per- during which Salix plantations could signi®cantly ennial energy crops were cultivated, for example, reduce the content of cadmium in the soil is esti- 25 years considering Salix. After that period, the mated to be about 35 years, after which the humus level may decrease or be maintained, plant-available fraction is assumed to be con- depending on future agriculture practices. 144 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154

3.5. Decreased wind erosion high, above 70%, to prevent water erosion [37]. Results from ®eld trials with short-rotation forest By using Salix plantations as windbreaks in show that the degree of soil cover in matured areas exposed to wind erosion, higher yields in stands may vary from 5% up to 75%, and in surrounding food crop cultivations could be newly harvested stands from 37% up to achieved [58]. Wind erosion reduces crop yields 100% [18]. Therefore, rill erosion and run-o€ through both direct damage, e.g., plant blasting, could occur in Salix plantations with low soil covered plants, and uncovered roots and seeds, cover [38]. Pimentel and Krummel [65] estimate, as well as indirect damage from loss of organic however, that soil erosion in short-rotation for- matter and ®ne soil particles leading to reduced ests and in energy grass cultivations is, on aver- soil fertility [59]. Results from Swedish and age, 7 and 70 times lower, respectively, than in Danish long-term ®eld trials in areas exposed to wheat cultivations. wind erosion show a reduction in wind speed of Up to 20% of Swedish arable land is 30±90% in ®elds with shelter belts less than exposed to rill erosion. On about 5%, rill ero- 100 m apart [60±62]. Compared with ®elds which sion causes soil losses of about 500 kg/ha yr, have shelter belts 150±350 m apart, the loss of which sometimes leads to permanent land clay and silt particles was 200±300 times lower degradation [37]. In this paper, serious rill ero- and the humus level was about one percentage sion is assumed to be prevented on 5% of ara- point higher. ble land consisting of clayey and silty soils, by It is assumed here that shelter belts with a vegetation zones of energy crops (preferably width of 50 m consisting of Salix are estab- energy grass). The vegetation zones, with a lished 100 m apart on ®elds exposed to wind width of 50 m, are located 200 m apart on erosion. One half of the width, or 25 m, is har- slopes, hollows, and other critical areas. Based vested at a time leading to a continuous high on these assumptions, the theoretical area of sheltering e€ect. The height of Salix in mature vegetation zones preventing rill erosion is stands is about 5 m. The yields from intermedi- equivalent to about 1% of current arable land. ate annual crop cultivation is estimated to The erosion, on an area equal to 0.5% of the increase by, on average, 10% [58, 63]. There ®eld size, is assumed to cause permanent land are, however, some uncertainties in this esti- degradation on every third occasion. Direct mate as the shelter e€ect depends on the width crop damage is assumed to occur on every sec- of the shelter belts. The results of the Swedish ond occasion, as rill erosion often occurs and Danish ®eld trials referred to above during winter on bare soil. assume the width of the shelter belts to be less than 25 m. Thus, the estimate made here must 3.6. Municipal waste treatment be veri®ed by practical ®eld trials. The arable land exposed to wind erosion in Sweden, Energy crop cultivation is being tested in large- mainly sandy soils in southern Sweden, is esti- scale trials in Sweden as vegetation ®lters for mu- mated to have increased by 15% since the nicipal waste water, for example, in the munici- 1950 s, and now amounts to 40,000 ha [58, 64]. palities of SvaloÈ v and BromoÈ lla [66]. Energy crop Thus, the theoretical area of Salix shelter belts cultivations can also be utilised for the treatment will be 13,000 ha, equivalent to 0.5% of the of land®ll leachate and sewage sludge. Interest in total area of arable land. vegetation ®lters for municipal waste water treat- ment, as a complement to conventional treatment 3.5.1. Decreased water erosion methods, has increased due to the problem of Perennial energy crop cultivation could also be marine eutrophication [67]. Also, the cost e- used as vegetation zones preventing rill erosion, ciency of puri®cation using vegetation ®lters is particularly on ®elds with clayey and silty soils in normally higher than for conventional hilly areas. The degree of soil cover must be treatment [68, 69]. Another bene®t of waste water P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 145 treatment in vegetation ®lters is that the amount The risk of accumulation of salts, heavy metals of sewage sludge will be reduced [68, 70]. and other components in the soil from waste water irrigation is estimated to be small. The 3.6.1. Waste water treatment reasons for this are that the content of these Around 75±95% of the nitrogen and phos- components is normally low in municipal waste phorus in the waste water could be removed in water, that precipitation in Sweden is much energy crop cultivations, when the waste water higher than evapotranspiration and that waste load is 500±1000 mm /ha yr [70±72]. An increase water irrigation is assumed to occur only during in waste water load to 2000±5000 mm/ha yr, will the summer period. In drier climates, however, decrease the treatment eciency to 10± there might be a risk of an accumulation of tox- 55% [69, 73]. It has been estimated that a waste ius, for example, salts from waste water irriga- water load of about 600 mm/ha yr, containing tion. 125 kg nitrogen, will not cause any long-term en- The cost of waste water treatment in veg- vironmental problems such as nitrogen etation ®lters will increase with increasing size leaching [74]. This amount of waste water will of the village or town, as longer pipes from supply, not only the demand for nitrogen and the populated area to the vegetation ®lters will other macro-nutrients, but also the demand for be needed. A rough estimation is that munici- water, which is often the limiting factor for pal waste water from about 70% of the growth. Nitrogen retention in vegetation ®lters is Swedish population may be treated in veg- also, to some extent, caused by denitri®cation. etation ®lters at a lower cost than if only con- The nutrient content in municipal waste water ventional treatment methods are used [6]. The matches the nutrient demand in energy crop cul- area of energy crop cultivation required for tivation quite well [71]. Waste water irrigation waste water treatment will then be about may increase the biomass yield by 2±3 times, 100,000 ha, which is equivalent to 3.6% of the compared with the case when no fertiliser is total arable land [77]. The period of time applied [75]. Compared with conventional fertili- during which vegetation ®lters will be useful as sation practices, the biomass yield is estimated to a complement to conventional treatment plants increase, on average, by 50% [76]. The nitrogen is assumed to be at least 25 years. This content in the waste water is reduced by 15±20% assumption is based on an estimate by the after a pre-treatment, which is needed for redu- National Swedish Environmental Protection cing the amount of solid particles. Thus, a waste Board [67], bearing in mind that new treatment water load of 600 mm/ha yr will supply around techniques which signi®cantly reduce the 100 kg nitrogen after a pre-treatment. amount of waste water, e.g., urine separation, The combination of waste water treatment in composting/digestion of faeces, or reuse of grey vegetation ®lters during the summer months, and water, will not be extensively implemented conventional phosphorus treatment during the within the near future. This is mainly due to winter months, has the highest cost eciency, as the low turnover of buildings, and technical the cost of storage ponds during the non-growing diculties in changing the sewage system in season is avoided [68]. Enhanced nitrogen existing buildings. removal will then occur only during the summer months, but it is during this season that the 3.6.2. Land®ll leachate treatment aquatic ecosystem is most vulnerable to eutrophi- About 10 of the current 300 sanitary land®lls cation. With this combination of treatment sys- in Sweden employ vegetation ®lters of Salix or tems, waste water from about 60 inhabitants may grass for land®ll leachate treatment [78]. Using be treated on one hectare of energy crop cultiva- on-site biological puri®cation systems, instead of tion. The amount of nitrogen in municipal waste transporting the leachate to a conventional treat- water is, on average, equivalent to about 4 kg N/ ment plant, reduces the risk of contamination of person yr [67]. the sewage sludge and leads to an increased 146 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 possibility of recirculating the sludge on arable Several studies have shown that using sewage land [78]. Energy crops, such as Salix and reed sludge as nitrogen fertiliser in Salix cultivations canary grass, have high evapotranspiration which leads to the same biomass yield as when commer- will reduce the amount of leachate. The amount cial fertilisers are used, that the nutrient and of nitrogen in the leachate, which is normally the metal leaching is insigni®cant, and that the con- most problematic parameter from a pollution tent of heavy metals in the soil is almost point of view concerning old land®lls, can be una€ected [83±87]. Application of sewage sludge reduced by 90% in vegetation ®lters [79, 80]. If would also increase the humus level in the soil. the leachate contains high levels of heavy metals, Extremely high application levels (around 20 these could also be reduced by using Salix (see tonne dry matter/ha yr) could, however, cause Section 3.3.). The content of heavy metals in the nitrogen leaching and an increased content of leachate from Swedish land®lls is, however, nor- cadmium in the soil [87]. mally low [78]. There might be a risk of long- The composition of nutrients in sewage sludge term salt accumulation in the soil as the content is, in contrast to municipal waste water, quite of, for example, chloride is normally much higher di€erent from the plant demand, with an excess in land®ll leachate than in municipal waste of phosphorus compared with nitrogen [71]. water. This risk has, however, not yet been ana- Thus, to ensure the optimum use of the nutrients lysed in ®eld trials. in the sludge, complementary nitrogen fertilisa- The average size of a vegetation ®lter for lea- tion is required [74]. The maximum allowed ap- chate treatment is about 3 hectares per land®ll. plication ratio of sewage sludge in Sweden is 5 The nitrogen content in leachate is, on average, tonne dry matter/ha every ®fth year [82]. This ap- about 150 g N/l [78]. Land®ll leachate produced plication will fully supply the demand for phos- during the winter is stored in ponds. The theor- phorus in energy crop production, and 20±30% etical area of energy crop cultivations as puri®- of the demand for nitrogen [7, 87]. Sewage sludge cation systems for land®ll leachate will then be could be applied to Salix cultivations after plant- around 1000 ha, which is equivalent to about ing, or harvesting, using manure spreaders [88]. 0.04% of current arable land. If old sanitary The theoretical area of energy crop cultivation land®lls are also included, the amount of land that could be fertilised with municipal sewage required for vegetation ®lters will increase. The sludge now being deposited at land®lls is about period of time during which vegetation ®lters will 130,000 ha, or 4.8% of current arable land. The be needed for leachate treatment is as long as the period of time during which a signi®cant amount anaerobic decomposition phase continues, or up of sewage sludge may be used as fertiliser in to 100 years [78]. energy crop cultivation, is here estimated to be 25 years or more. After that, the quality of the sewage sludge may have been improved through 3.6.3. Recirculation of sewage sludge national environmental programmes, and thus be The use of municipal sewage sludge as fertiliser available for recycling in food production [67]. in agriculture is a well established principle in Sweden. However, only about 35% of the sewage 3.7. Biodiversity sludge produced is recycled, mainly on agricul- ture land, while 65% is deposited on land®lls due Replacing annual crops by perennial energy to uncertainties in the amounts of heavy metals crops will a€ect the biodiversity on a genetic, and toxic organic compounds in the sludge [81]. species, and habitat level. The diversity and According to the National Swedish occurrence of soil micro-organisms and soil Environmental Protection Board [82], this sludge fauna, especially decomposers such as earth- may be used in other cultivations than food crop worms, wood lice, harvest men, and carbides, is, cultivation, e.g., energy crop production, until in general, higher in energy forest and grass culti- the sludge quality is established. vations than in annual food crop P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 147 cultivations [17, 18]. This is mainly due to the Large-scale introduction of Salix may increase reduced soil tillage and use of agrochemicals, and the risk of pests and diseases, such as rust fungus to the increased input of litter. The average dose and herbivores. This is, however, not assumed to of chemical pesticides applied in Salix and reed result in an increased use of chemical pesticides, canary grass cultivations is, for example, about as it is dicult, and thereby costly, to carry out 0.2 kg active ingredients/ha yr, while it is about 1 chemical pest control in mature stands [100]. kg in annual food crop cultivation [7]. Alternative methods which can be used are, for Cultivation of Salix (male plants) will also pro- example, breeding of resistant strains, biological mote pollinating insects, such as bees and bum- pest control of herbivores, and a more diverse ble-bees, as Salix ¯owers in early spring when clonal composition which also leads to increased the supply of pollen is limited [89]. About half of yields [101±103]. Another risk associated with the Salix clones sold in Sweden today are male large-scale introduction of Salix cultivation is plants [90]. genetic contamination of surrounding wild native The composition of the ground ¯ora in Salix Salix species through hybridisation with foreign cultivations, which depends on present manage- species [33, 103]. One solution to this problem ment methods and earlier land use, is normally could be the development of sterile clones [20]. dominated by weed species, but rare species Today, mainly native species of Salix are used in could also occur [91]. Results from an investi- Sweden [90]. gation in twelve Salix plantations in southern In this paper, it is estimated that the overall Sweden show an occurrence of 125 species of biodiversity in a larger region is not, or only to a vascular plants, and 18 species of mosses [92]. In small extent, a€ected by a substitution of annual a grain cultivation, the frequency of weed species crops by perennial energy crops. If any change is around 50, of which about 30 occur only occurs, the biodiversity will probably increase, sporadically [93]. Energy crop cultivations may mainly in open farmland. A similar conclusion also improve the biodiversity indirectly in has been drawn by the National Swedish streams through reduced nutrient leaching [94]. Environmental Protection Board [91]. Short-rotation forest will provide a transitional habitat promoting wildlife, specially in open farmland. Several studies show that both short- 4. Conclusions and discussion rotation forests and energy grass cultivations will increase the number of bird species, mainly song- An overall conclusion of this study is that the birds, as well as the abundance of some environmental bene®ts from large-scale introduc- species [95±97]. A proportion of 10 to 20% tion of energy crop production in Sweden could energy forest in open farmland has been esti- be substantial, as the negative environmental mated to be optimal for fauna, especially when impact from current agriculture practices and the harvest is asynchronous in di€erent sub- municipal waste treatment could be signi®cantly areas [96]. This will also improve the diversity of reduced. The carbon dioxide emission from or- the ground ¯ora [92]. Results regarding the ganic soils through biological oxidation of the or- e€ects of short-rotation forest on the occurrence ganic matter, for example, could be signi®cantly of larger mammals, such as moose, deer, and reduced (by about 7 tonne C/ha yr) when annual hare, are not in agreement [33, 98]. However, crops are replaced by perennial energy crops considering a region including several di€erent (Table 2). The greenhouse gas mitigation from types of landscape, an increase in biodiversity in changed land use on mineral soils, through soil open farmland may not lead to an increase in the carbon accumulation and reduced nitrous oxide overall biodiversity in the region. Several of the emission, is around 1/10 and 1/100, respectively, birds and mammals occupying energy crop plan- of that through changed land use on organic tations are often widespread and regionally abun- soils. As a comparison, the reduction of carbon dant in other landscape types [99]. dioxide emissions due to the lower input of fossil 148 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154

Table 2 Summary of environmental changes, including the recirculation and treatment of municipal waste, when perennial energy crops replace annual food crops, the maximum area of energy crop cultivation where a change occurs expressed as per cent of the total Swedish arable land, and the period of time during which the change is estimated to last. When the energy crop cultivations a€ect surrounding cultivations of food crops, this area is given within parentheses, expressed as per cent of the total Swedish arable land

Changed environmental impact Average quantitative impact * Maximum proportion Period of time (years) of total arable land (%)

Accumulation of soil C in mineral 0.5 tonne C/ha yr 55 050 soils

Reduced CO2 emission from 7 tonne C/ha yr 5.0 080 organic soils

Reduced N2O emission from 0.04 tonne C-equivalents/ha yr 55 1 mineral soils Reduced N leaching in general 10 kg N/ha yr 60 > 25 Reduced N leaching through 70 kg N/ha yr 2.6 (10±15) > 25 bu€er strips Reduced P leaching through 1.5 kg P/ha yr 3.7 (15±20) > 25 bu€er strips Cadmium removal 6 g Cd/ha yr 92 035 Increased soil fertility 5% higher yield 47 > 25 Reduced wind erosion 10% higher yield 0.5 (1.5) 1 Reduced water erosion Reduced land degradation and 1.0 (5.0) 1 plant damage on 0.5% of the ®eld area Waste water treatment 100 kg N/ha yr > 25 (sparsely populated areas) 2.5 (populated areas) 0.7 (densily populated areas) 0.4 Land®ll leachate treatment 100 kg N/ha yr 0.04 0100 Recirculation of sewage sludge 1 tonne dry matter/ha yr 4.8 > 25 Biodiversity Unchanged/increased in open 60 1 farmland

* The average quantitative impact presented should be seen as average estimates which could vary signi®cantly due, among other things, to local conditions. A more integrating sensitivity analysis is presented in Part II of the analysis, in the second paper (see BoÈ rjesson [6])

fuels for the production of perennial energy based on the results presented in this paper [105]. crops production than of annual food crops, has Today, the contribution of greenhouse gases been estimated to amount to about 0.2 tonne C/ from agriculture, mainly consisting of carbon ha yr [8]. When the biomass is used to replace dioxide emission from drained and cultivated or- fossil coal in heat and power production, the car- ganic soils, amounts to 15% of the total Swedish bon dioxide mitigation amounts to 5±6 tonne C/ emissions [105]. The Biomass Commission of the ha yr, thus the reduced carbon dioxide emissions Swedish Government [9] has estimated the from changed use of land on organic soils could amount of arable land not needed for food pro- exceed this mitigation [104]. duction for domestic consumption, and thus If perennial energy crops are cultivated on available for energy crop production, to about 30% of the total Swedish arable land, this could 800,000 ha around 2005 (equivalent to about lead to a reduction in the total emissions of 30% of the total Swedish arable land). The greenhouse gases from agriculture by 15±25%, reduced carbon dioxide emission from organic P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 149 soils is, however, caused by the cessation of The environmental bene®ts identi®ed in this annual crop cultivation leading to less frequent paper could be achieved on large areas of arable soil tillage, and not particularly by the cultivation land, up to 60% of the land on which annual of perennial energy crops. Thus, an equivalent re- crops are cultivated today. The occurrence of duction of carbon dioxide emission is assumed to open streams limits the use of energy crop culti- be achieved when the land lies fallow, as when vations as bu€er strips, while the amount of mu- perennial crops are grown. Also, the reduced nicipal waste limits the use of energy crop nitrous oxide emission from mineral soils is cultivations as vegetation ®lters. caused by the reduced input of nitrogen fertili- The environmental advantages could last for a sers, and not particularly by the cultivation of long period, 25 years or more. Reduced wind energy crops. and water erosion will last as long as energy The greatest reduction in nitrogen emission to crops are cultivated. How long soil carbon ac- surface and ground water is achieved when cumulation and cadmium removal will last, energy crop cultivations are used as vegetation depends on the physical and chemical properties ®lters for waste water treatment. A somewhat of the soil, while the thickness of the peat layer lower reduction is achieved when energy crop determines how long the reduction in carbon cultivations are used as riparian bu€er strips, fol- dioxide emissions from organic soils will last. lowed by recirculation of sewage sludge, and The carbon dioxide mitigation through changed replacement of annual crops. When energy crop land use where energy crops have replaced food cultivations are utilised as riparian bu€er strips, crops, will be gradually reduced and then cease. However, this carbon dioxide mitigation will be where possible, nutrient leaching from Swedish achieved again on new cultivation sites when agriculture could be reduced by 15±25% [106]. energy crop cultivations are moved. The possi- Nutrient leaching from arable land amounts to, bility of the future use of energy crop cultivations on average, 30% and 10% of the total Swedish to reduce nutrient leaching and to treat munici- nitrogen and phosphorus loads to water [106]. In pal waste, depends on the development of agri- intensively cultivated agricultural areas, however, cultural practices, the load of pollutants, and the the nutrient leaching could be much higher, up production of waste. to 75% [107, 108]. Through the extensive use of There are several uncertainties in the estimates energy crop cultivations for municipal waste presented in this paper, e.g., concerning carbon water treatment, as a complement to convention- dioxide emission from organic soils, accumu- al treatment methods, the discharge of nitrogen lation of soil carbon in mineral soils, nitrous to water, and the generation of sewage sludge, oxide emission, nutrient removal in bu€er strips, may be reduced by 40%, compared with current cadmium removal, and wind and water erosion. treatment technology in Sweden [109]. Also, The emission of nitrous oxide, for example, may nitrogen equivalent to about 6% of the current vary greatly depending on soil type, type of ferti- use of commercial fertiliser in Sweden could be liser used, climate and land management prac- reused [10, 81]. When sewage sludge now being tices. Thus, local conditions will a€ect the deposited on land ®lls is recirculated on energy magnitude of the reduction of nitrous oxide emis- crop cultivations, nitrogen and phosphorus sions when energy crops replace annual food equivalent to about 2% and 20%, respectively, crops. Concerning the magnitude in the reduction of the current use of commercial fertiliser in of carbon dioxide emissions from organic soils Sweden, could be reused [10, 81]. due to the change in land use and the accumu- The yield increase in surrounding food crop lation of soil carbon in mineral soils, these theor- cultivations is greatest when Salix is used as shel- etical calculations have not been veri®ed by ®eld ter belts preventing wind erosion, followed by trials. The nutrient removal in bu€er strips and energy crop cultivations as vegetation zones pre- the occurrence of water erosion also depends on venting water erosion. local conditions, thus, the variation could be sig- 150 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 ni®cant due to, for example, soil type and how The Swedish National Board for Industrial and hilly the landscape is. How eciently 25±50 m- Technical Development. wide shelter belts of Salix will reduce wind ero- sion, in comparison to 5±10 m-wide, must also to be analysed in ®eld trials. The magnitude of cad- mium removal could vary greatly, depending on References both soil conditions, such as pH and content of clay and organic matter, and the eciency of the [1] Tolbert V. Guest editoral. Biomass and Bioenergy cadmium uptake concerning the speci®c Salix 1998;14:301±6. clone. [2] McLaughlin SB, Walsh ME. Evaluating environmental Another uncertainty concerns the bene®t of the consequences of producing herbaceous crops for bioe- nergy. Biomass and Bioenergy 1998;14:317±24. decreased use of chemical pesticides in the culti- [3] Kort J, Collins M, Ditsch D. A review of soil erosion vation of perennial energy crops as compared to potential associated with biomass crops. Biomass and annual food crop cultivation. As described in Bioenergy 1998;14:351±60. Section 2, no data have been found of, for [4] Grigal DF, Berguson WE. Soil carbon changes associ- example, the risk of water pollution by pesticides ated with short-rotation systems. Biomass and Bioenergy 1998;14:371±8. used in perennial energy crop cultivation, in com- [5] Bransby DI, McLaughlin SB, Parrish DJ. A review of parison with annual food crop cultivation. carbon and nitrogen balances in switchgrass grown for However, such a risk should be lower, as the energy. Biomass and Bioenergy 1998;14:379±84. average dose of chemical pesticides applied in [6] BoÈ rjesson P. Environmental e€ects of energy crop culti- annual food crop cultivation is about ®ve times vation in SwedenÐII: Economic valuation, Biomass and Bioenergy 1999;16:155-170. higher than in Salix and reed canary grass culti- [7] BoÈ rjesson P. Energy analysis of biomass production vations (see Section 3.6.) Thus, the bene®t of a and transportation. Biomass and Bioenergy reduced use of chemical pesticides is probably 1996;11:305±18. underestimated in this analysis, as it is included [8] BoÈ rjesson P. Emissions of CO2 from biomass pro- only in the impact on the biodiversity. duction and transportation in agriculture and forestry. Energy Convers. Mgmt. 1996;37:1235±40. Concerning other environmental impacts ident- [9] Ocial Report of the Swedish Government. BiobraÈ nsle i®ed in this paper, such as waste water and land- foÈ r framtiden: slutbetaÈ nkande av biobraÈ nslekommissio- ®ll leachate treatment, and recirculation of nen (Biomass for the future: ®nal report of the Biomass sewage sludge, the uncertainties in the calcu- Commission). AllmaÈ nna FoÈ rlaget, Stockholm, 1992. lations are estimated to be lower. Thus, consider- [10] Statistics Sweden. Jordbruksstatistisk aÊ rsbok 1995 (Yearbook of agricultural statistics 1995). AllmaÈ nna ing the magnitude of the di€erent environmental FoÈ rlaget, Stockholm, 1995. changes presented in Table 2, these should be [11] Mattsson L. MarkboÈ rdighet och jordart i svensk aÊ ker- seen as average estimates which could vary sig- mark (Soil fertility and soil type of Swedish arable ni®cantly due, among other things, to local con- land). National Swedish Environmental Protecting ditions. A more integrating sensitivity analysis of Board, Stockholm, 1996. [12] Hallgren G, Berglund G. De odlade myrjordarnas the various environmental changes is presented in omfattning och anvaÈ ndning (The extent and use of cul- Part II of the analysis, in the second paper (see tivated peat soils in Sweden). Swedish University of BoÈ rjesson [6]). Agricultural Science, Uppsala, 1962. [13] EkstroÈ mG.AÊ kermarkens matjordstyper (Map of culti- vated soil types) In Atlas of Sweden, Stockholm 1953. [14] Swedish National Board of Industrial Technical Development (NUTEK). Energiskog i landskapetÐraÊ d Acknowledgements och anvisningar foÈ r lokalisering (Short-rotation forest in the landscapeÐguidelines for localisation). I would like to thank L. Gustavsson, B. Stockholm, 1993. [15] Frank K. Energiskogens miljoÈ konsekvenser Johansson and A. Lundborg for reviews and use- (Environmental e€ects of short-rotation forest). ful comments. I also gratefully acknowledge the Department of Agricultural Engineering, Swedish economic support provided by Vattenfall AB and University of Agricultural Science, Uppsala, 1992. P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 151

[16] Reicosky DC, Kemper WD, Langdale GW, Douglas [31] BergstroÈ m L, Johansson R. In¯uence of fertilised short- CL, Jr, Rasmussen PE. Soil organic matter changes rotation forest plantations on nitrogen concentration in resulting from tillage and biomass production. Journal groundwater. Soil Use and Management 1992;8:36±40. of Soil and Water Conservation 1995;50:253±61. [32] Rijtema P, de Vries W. Di€erences in precipitation [17] SjoÈ dahl Svensson K, Granhall U, Andre n O. Soil bio- excess and nitrogen leaching from agricultural lands logical aspects on short-rotation forestry. Swedish and forest plantations. Biomass and Bioenergy National Board of Industrial and Technical 1994;6:103±15. Development (NUTEK), Stockholm, 1994. [33] Andersson R. BiobraÈ nslen fraÊ n jordbruketÐen analys [18] Makeschin F. E€ects of energy forestry on soils. av miljoÈ konsekvenserna (Biofuels from agricultureÐan Biomass and Bioenergy 1994;6:63±80. analysis of the environmental impact). National [19] Thyselius L, Johansson W, Mattsson L, Wallgren B. Swedish Environmental Protection Board, Stockholm, EnergigroÈ dor foÈ r biogas (Energy crops for biogas pro- 1990. duction). Vattenfall, VaÈ llingby, 1992. [34] Johansson H, Ho€man M. Normalutlakning av kvaÈ ve [20] Ranney JW, Mann LK. Environmental considerations fraÊ n svensk aÊ kermark 1985 och 1994 (Nitrogen leaching in energy crop production. Biomass and Bioenergy from Swedish arable land in 1985 and 1994). 1994;6:211±28. Department of Soil Science, Swedish University of [21] Berglund K. YtsaÈ nkning paÊ mosstorvjord (Subsidence Agricultural Sciences, Uppsala, 1996. of cultivated organic soils). Department of Soil Science, [35] Vought L, Dahl J, Lauge Pedersen C, Lacoursie re J. Swedish University of Agricultural Sciences, Uppsala, Nutrient retention in riparian ecotones. Ambio 1989. 1994;23:332±48. [36] Sunblad K, Andersson L. Links between hydrological [22] Berglund K. Personal communication. Department of processes and nutrient ¯uxes. In: Robinson M, editor. Soil Science, Swedish University of Agricultural Flow Regimes from International Experimental and Sciences, Uppsala, 1996. Network Data Vol. III, 1993. [23] Kasimir Klemdtsson AÊ , Klemedtsson L. [37] AlstroÈ m K, Bergman AÊ kerman A. Vattenerosion i MarkanvaÈ ndning paÊ verkar avgivning av vaÈ xthusgas Sydsvensk aÊ kermark (Water erosion of arable land in (Land use and e€ects on emission of greenhouse gases). southern Sweden). Department of Physical Geography, In: Berg B, editor. Markdagen 1996ÐForskningsnytt Lund University, Lund, 1991. om mark. Department of Forest Soil Science, Swedish [38] Leonardsson L. VaÊ tmarker som kvaÈ vefaÈ llorÐsvenska University of Agricultural Sciences, Uppsala, 1996. och internationella erfarenheter (Wetlands as nitrogen [24] McAfee M. YtsaÈ nkning paÊ torvjord (Subsidence of cul- trapsÐSwedish and international experience). National tivated organic soils). Department of Soil Science, Swedish Environmental Protection Board, Stockholm, Swedish University of Agricultural Sciences, Uppsala, 1993. 1985. [39] Statistics Sweden. NaturmiljoÈ n i si€ror (The natural en- [25] Robertsson K. Emission of N2O in SwedenÐnatural vironment in ®gures). OÈ rebro, 1995. and antropogenic sources. Ambio 1991;20:151±5. [40] Krook J. Personal communication. Ekologgruppen, [26] Nevison CD, Esser G, Holland EA. A global model of Landskrona, 1995. changing N2O emissions from natural and perturbed [41] National Swedish Environmental Protection Board. soils. Climatic Change 1996;32:327±78. KvaÈ ve fraÊ n land till hav (Nitrogen from land to sea). [27] Bouwman AF. Exchange of greenhouse gases between Stockholm, 1996. terrestrial ecosystems and the atmosphere. In: [42] Andersson A. Trace elements in agricultural soilsÐ Bouwman AF, editor. Soils and the Greenhouse E€ect. ¯uxes, balances, and background values. National Chichester, Wiley, 1990. Swedish Environmental Protection Board, Stockholm, [28] National Swedish Environmental Protection Board. 1992. VaÈ xthusgasernaÐutslaÈ pp och aÊ tgaÈ rder i ett internatio- [43] Eriksson J, SoÈ derstroÈ m M, Andersson A. nellt perspektiv (Greenhouse gasesÐemission and miti- Kadmiumhalter i matjorden i svensk aÊ kermark gation in an international perspective). Stockholm, (Cadmium content in the topsoil of Swedish arable 1990. land). National Swedish Environmental Protection [29] Intergovernmental Panel of Climate Change (IPCC). Board, Stockholm, 1995. Climate Change 1995. The Science of Climate [44] AÊ byhammar T, Fahlin M, Holmroos S. Kadmium i ChangeÐContribution of Working Group I to the biobraÈ nslesystemet (Cadmium in the biomass based Second Assessment of the IPCC, ch. Chapter 2. energy system). Vattenfall, VaÈ llingby, 1993. Cambridge University Press, Cambridge, 1996. [45] Obernberger I, Biedermann F, Widmann W, Riedl R. [30] Stjernquist I. An integrated environmental analysis of Concentrations of inorganic elements in biomass fuels short-rotation forests as a biomass resource. Biomass and recovery in the di€erent ash fractions. Biomass and and Bioenergy 1994;6:3±10. Bioenergy 1997;12:211±24. 152 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154

[46] Dickinson N, Punshon T, Hodkinson R, Lepp N. vate land owners and municipalities). Lund University, Metal tolerance and accumulation in willows. In: Lund, 1985. Aronsson P, Perttu K, editors. Willow Vegetation [59] Mattsson J, Nihle n T, Olesen F. Vindens skadegoÈ relse Filters for Municipal Wastewaters and Sludges. paÊ aÊ kermark i SkaÊ ne och Danmark (Wind erosion on Department of Ecology and Environmental Research, arable land in Scania and Denmark). Yearbook of Swedish University of Agricultural Sciences, Uppsala, Geography in Sweden, 1983. 1994. [60] JoÈ nsson P. LaÈ planteringars inverkan paÊ vind och ero- [47] Landberg T, Greger M. Can heavy metal tolerant sion (The e€ects of shelter belts on wind and erosion). clones of Salix be used as vegetation ®lters on heavy Manuscript. Department of Physical Geography, Lund metal contaminated land?. In: Aronsson P, Perttu K, University, Lund, 1987. editors. Willow Vegetation Filters for Municipal [61] Nord M. LaÈ e€ekter av vegetationsridaÊ er (Shelter e€ects Wastewaters and Sludges. Department of Ecology and of vegetation belts). National Swedish Institute of Environmental Research, Swedish University of Construction Sciences, GaÈ vle, 1989. Agricultural Sciences, Uppsala, 1994. [62] JoÈ nsson P. In¯uence of shelter on soil sorting by wind [48] Ridell-Black D. Heavy metal uptake by fast growing erosion. Catena 1994;22:35±47. willow species. In: Aronsson P, Perttu K, editors. [63] Olesen F. LñplanteringarÐdyrkningssikkerhed, klimat- Willow Vegetation Filters for Municipal Wastewaters forbedring og landskabspleje (Shelter plantationsÐcul- and Sludges. Department of Ecology and tivation security, climatic improvements and landscape Environmental Research, Swedish University of scenery). Landhusholdningsselskabets Forlag, Agricultural Sciences, Uppsala, 1994. Copenhagen, Denmark, 1979. [49] Eriksson J, Ledin S. E€ekter av laÊ ngvarig Salixodling [64] Ocial Report of the Swedish Government. MarkvaÊ rd paÊ kadmiuminnehaÊ llet i jordenÐen pilotstudie (E€ects och erosionsskydd (Land conservation and erosion con- of long-term Salix cultivation on the total and plant- trol). Stockholm, 1957. available content of cadmium in the soilÐa pilot [65] Pimentel D, Krummel J. Biomass energy and soil ero- study). Vattenfall, VaÈ llingby, 1995. sion: assessment of resource costs. Biomass 1987;14:15± 38. [50] OÈ stman G. Kadmium i Salix (Cadmium in Salix). [66] Perttu K. Personal communication. Department of Vattenfall, VaÈ llingby, 1994. Short-Rotation Forestry, Swedish University of [51] Johansson L. Tungmetaller i traÈ d och energigroÈ dorÐen Agricultural Sciences, Uppsala, 1997. litteraturstudie (Heavy metals in trees and in energy [67] National Swedish Environmental Protection Board. cropsÐa review). Vattenfall, VaÈ llingby, 1995. Vatten, avlopp och miljoÈ (Water, sewage and the en- [52] National Swedish Environmental Protection Board. vironment). Stockholm, 1993. Jordbruk och miljoÈ (Agriculture and the environment). [68] Rosenquist H, Aronsson P, Hasselgren K, Perttu K. Stockholm, 1993. Economics of using municipal wastewater irrigation of [53] National Swedish Environmental Protection Board. willow coppice crops. Biomass and Bioenergy Hur ska Sverige maÊ aÊ r 2020? (The status of the environ- 1997;12:1±8. ment in Sweden in 2020). Stockholm, 1993. [69] Sundblad K, Wittgren HB. Glyceria maxima for waste- [54] JoÈ nsson P. Landscape changes relevant to wind stress water nutrient removal and forage production. 1939±84 in Scania, southern Sweden. In: Proceedings at Biological Wastes 1989;27:29±42. The Fourth International Symposium of Windbreaks [70] Geber U, Tuvesson M. VallvaÈ xters egenskaper som and Agroforestry. July 26-30 1993, Hedeselskabet, producenter av energiÐoch ®berraÊ vara och som biolo- Denmark, 1993. gisk renare av naÈ ringsrika vatten (Characteristics of ley È [55] Jnsson P. Wind erosion on sugar beet ®elds in Scania, crops in the production of energy and ®brous material southern Sweden. Agricultural and Forestry Meterology and for biological puri®cation of nutrient-rich water). 1992;:62, 141±157. Department of Crop Production Science, Swedish [56] Bergman AÊ kerman A, JoÈ nsson P. MarkÐoch vatten- University of Agricultural Sciences, Uppsala, 1993. vaÊ rd i helaÊ kersbygdenÐlaÈ r av historian (Soil and water [71] Perttu K. Biomass production and nutrient removal protection in open farmlandÐlearning from history). from municipal wastes using willow vegetation ®lters. Department of Geography, Lund University, Lund, Journal of Sustainable Forestry 1993;1:57±70. 1994. [72] Kutera J, Soroko M. The use and treatment of waste- [57] Gustavsson LE. MarktaÈ ckningÐe€ekter paÊ olika jord- water in willow and poplar plantations. In: Aronsson P, typer (MulchingÐe€ects on di€erent soil types) Perttu K, editors. Willow Vegetation Filters for Department of Soil Science, Swedish University of Municipal Wastewaters and Sludges. Department of Agricultural Sciences, Uppsala. 1987. Ecology and Environmental Research, Swedish [58] Lewan L, MaÊ sbaÈ ck S, WoÈ rneryd O. LaÈ planteringarÐ University of Agricultural Sciences, Uppsala, 1994. ett gemensamt intresse foÈ r privata markaÈ gare och kom- [73] Kowalik P, Randerson P. Nitrogen and phosphorus muner (Shelter plantationsÐa common interest for pri- removal by willow stands irrigated with municipal P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154 153

wastewaterÐa review of the Polish experience. Biomass [88] SjoÈ svaÈ rd L. Spridningstekniker foÈ r slam i Salixodling and Bioenergy 1994;6:133±40. (Techniques of spreading sewage sludge in Salix planta- [74] Nielsen K. Environmental aspects of using wastewater tions). Vattenfall, VaÈ llingby, 1995. and sludge in energy forest cultivation. Biomass and [89] Stark J. Personal communication. Department of Bioenergy 1994;6:123±32. Entomology, Swedish University of Agriculturl Science, [75] Perttu K, Kowalik PJ. Salix vegetation ®lters for puri®- Uppsala, 1997. cation of waters and soils. Biomass and Bioenergy [90] Larsson S. Personal communication. SvaloÈ f Weibull 1997;12:9±19. AB, SvaloÈ v, 1997. [76] BoÈ rjesson P, Gustavsson L, Christersson L, Linder S. [91] National Swedish Environmental Protection Board. Future production and utilisation of biomass in MarkanvaÈ ndningen och miljoÈ n (Land use and the en- Sweden: potentials and CO mitigation. Biomass and vironment). Stockholm, 1993. 2 [92] Gustavsson L. Plant conservation aspects of energy for- Bioenergy 1997;13:399±412. estryÐa new type of land use in Sweden. Forestry [77] Statistics Sweden. Statistisk aÊ rsbok 1996 (Statistical Ecology and Management 1987;21:141±61. Yearbook of Sweden 1996). Stockholm, 1996. [93] Hallgren E. FoÈ raÈ ndras ograÈ s¯oran och verkan av ett [78] National Swedish Environmental Protection Board. ograÈ smedel med tiden? (Does the occurrence of weeds Lokal lakvattenbehandlingÐerfarenheter fraÊ n42 and the eciency of herbicides change over time?). svenska deponier (Local land®ll leachate treatmentÐex- Department of Crop Production Science, Swedish perience from 42 Swedish land®lls). Stockholm, 1993. University of Agricultural Sciences, Uppsala, 1993. [79] Hasselgren K, Nilsson P. LakvattenbehandlingÐlokala [94] National Swedish Environmental Protection Board. metoder foÈ r behandling av lakvatten fraÊ n avfallsupplag Biologisk maÊ ngfald (Biological diversity in Sweden). (Land®ll leachate treatmentÐmethods for on-site treat- Stockholm, 1993. ment of leachate from land®lls). National Swedish [95] Sage R, Robertson P. Wildlife and game potential of Environmental Protection Board, Stockholm, 1992. short-rotation coppice in the U.K.. Biomass and [80] Hasselgren K. 1994. Land®ll leachate treatment in Bioenergy 1994;6:41±8. energy forest plantations. In: Aronsson P, Perttu K, [96] GoÈ ransson G. Bird fauna of cultivated energy shrub editors. Willow Vegetation Filters for Municipal forests at di€erent heights. Biomass and Bioenergy Wastewaters and Sludges. Department of Ecology and 1994;6:49±53. Environmental Research, Swedish University of [97] Paine L, Peterson T, Undersander D, Rineer K, Bartelt Agricultural Sciences, Uppsala, 1994. G, Temple S, Sample D, Klemme R. Some ecological [81] Statistics Sweden. Avfall och aÊ tervinning i kommunal and socio-economic considerations for biomass energy regi 1994 (Waste and recycling in Swedish municipali- crop production. Biomass and Bioenergy 1996;10:231± ties in 1994). Stockholm, 1995. 42. [82] National Swedish Environmental Protection Board. [98] Christian D. Wintertime use of hybrid Poplar planta- AnvaÈ ndning av avloppsslam i jordbruket (The utilis- tions by deer and medium-sized mammals in the ation of sewage sludge in agriculture). Stockholm, 1996. Midwestern U.S. Biomass and Bioenergy 1997;12:35± [83] Simon M, Zsu€a L, Sennerby-Forsse L, Burgess D. 40. Variation in the response of some North American wil- [99] Christian D, Niemi G, Hanowski J, Collins P. low species and clones to sludge fertilisation. Biomass Perspectives on biomass energy tree plantations and changes in habitat for biological organisms. Biomass and Bioenergy 1991;1:185±92. and Bioenergy 1994;6:31±9. [84] Diedrichs H. Slam och aska i energiodlingarÐen litter- [100] Christersson L, Ramstedt M, Forsberg J. Pests, dis- aturstudie (Sewage sludge and ash in energy planta- eases, and injuries in intensive short-rotation forestry. tionsÐa review). Vattenfall, VaÈ llingby, 1992. In: Mitchell C, Ford-Robertson J, Hinckley T, [85] Nielsen K. 1994. Sludge fertilisation in willow planta- Sennerby-Forsse L, editors. Ecophysiology of Short tions. In: Aronsson P, Perttu K, editors. Willow Rotation Forest Crops. Elsevier, London, 1993. Vegetation Filters for Municipal Wastewaters and [101] Willebrand E, Ledin S, Verwijst T. Willow coppice sys- Sludges. Department of Ecology and Environmental tems in short-rotation forestry: e€ects of plant spacing, Research, Swedish University of Agricultural Sciences, rotation length and clonal composition on biomass pro- Uppsala. duction. Biomass and Bioenergy 1993;4:323±32. [86] Hasselgren K. Kadmiumhalter i Salixodlingar efter [102] Price P, Martinsen G. Biological pest control. Biomass behandling med kommunala restprodukter (The content and Bioenergy 1994;6:93±101. of cadmium in Salix plantations after fertilisation with [103] Ocial Report of the Swedish Government. Den bru- municipal waste products). Vattenfall, VaÈ llingby, 1995. kade maÊ ngfalden: betaÈ nkande av utredningen om jord- [87] Diedrichs H., Bramryd T. Kommunalt slam som goÈ dsel brukets tillgaÊ ngar paÊ vaÈ xtgenetiska resurser (Cultivated i Salixodlingar (Municipal sewage sludge as fertiliser in biodiversity: genetic resources in agriculture). Fritzes Salix plantations). Vattenfall, VaÈ llingby, 1995. FoÈ rlag, Stockholm, 1995. 154 P. BoÈrjesson / Biomass and Bioenergy 16 (1999) 137±154

[104] Gustavsson L, BoÈ rjesson P, Johansson B, Svenningsson of agriculture on surface water and groundwater)

P. Reducing CO2 emissions by substituting biomass for National Swedish Environmental Protection Board, fossil fuels. Energy 1995;20:1097±113. Stockholm, 1993. [105] SkaÈ rby L, Ferm M, Johansson M, Klemedtsson L, [107] Ekologgruppen. VattenvaÊ rdande aÊ tgaÈ rder foÈ r delar av LoÈ vblad G, Phil Karlsson G, Zetterberg L. KaÈ vlingeaÊ ns avrinningsomraÊ de (Water protection in the LuftfoÈ roreningar och vaÈ xthusgaser: jordbrukets roll KaÈ vlinge River water basin). Landskrona, 1992. som mottagare och kaÈ lla (Airborne pollutants and greenhouse gases: agriculture as a source and recipient). [108] LaÈ nsstyrelserna i Kristianstad och MalmoÈ hus laÈ n. Swedish Environmental Research Institute (IVL). VegeaÊ -projektet (The VegeaÊ River project). Kristianstad GoÈ teborg, 1995. och MalmoÈ , 1992. [106] LoÈ fgren S. Jordbrukets inverkan paÊ ytÐoch grundvat- [109] Statistics Sweden. UtslaÈ pp till vatten 1992 (Emissions tenÐtillstaÊ nd, utveckling, orsak och verkan (The e€ects to water in 1992). Stockholm, 1994.