Effects of nutrient supply and soil cd con

Effects of nutrient supply and soil cd con

Biomass and Bioenergy 23 (2002) 415 – 426 E ects of nutrient supply and soil cadmium concentration on cadmium removal by willow Erika Klang-Westina; ∗ , Kurth Perttub a Department b Department of Soil Sciences, Swedish University of Agricultural Sciences, P.O Box 7014, SE-750 07 Uppsala, Sweden of Short Rotation Forestry, Swedish University of Agricultural Sciences, P.O Box 7016, SE-750 07 Uppsala, Sweden Received November 2001; received in revised form 10 May 2002; accepted 23 May 2002 Abstract This investigation studied the e ect of an increased biomass production as a result of fertilization and an elevated Cd concentration in the topsoil on concentration and amount of Cd in two clones of Salix (81090 and 78183) The experiment was conducted over a three year period using 200-dm3 lysimeters ÿlled with clay soil A liquid fertilizer containing all essential macro- and micronutrients in balanced proportions by weight was applied at two rates according to growth The lower rate corresponded to 0, 20 and 20 kg N ha−1 during years 1, and 3, respectively, while the higher rate was 30, 60 and 60 kg N ha−1 for the same period The Cd levels in the topsoil were an initial content of 0:3 mg Cd (kg dw soil)−1 and 0:6 mg Cd (kg dw soil)−1 after addition of CdSO4 Biomass production increased signiÿcantly due to fertilization In general, this increase in biomass resulted in a higher Cd amount in the stem However, the magnitude was small and only statistically signiÿcant in some cases, mainly because increased biomass also resulted in a lowered Cd concentration due to an e ect of biological dilution Addition of Cd to the topsoil resulted in higher Cd concentrations and total Cd amounts (concentration × biomass) in the Salix plants In most cases the increase in total stem Cd amount was 40 –80% of the increase in soil Cd concentration, although a directly proportional increase was observed occasionally Clone 81090 had higher concentrations and total amounts of Cd in the stems than clone 78183, while clone 78183 produced more stem biomass The leaves had the highest Cd concentrations, but the total amounts of Cd were largest in the stems ? 2002 Elsevier Science Ltd All rights reserved Keywords: Biomass production; Fertilization; Salix; Clone; Cd Introduction During the 20th Century, arable land in Sweden has been subjected to anthropogenic input of Cd, mainly via phosphorus fertilizers and deposition [1] Calculations by Andersson [2] indicate a 33% increase ∗ Corresponding author Tel.: +46-(0)18-672888; fax: +46(0)18-672795 E-mail address: erika.klang.westin@mv.slu.se (E Klang-Westin) in average Cd concentrations in the topsoil between 1900 and 1990, based on the levels around 1900 Furthermore, Eriksson [3] found that –10% of an annual harvest of Swedish winter wheat had Cd concentrations near or above the limit value for cereals (0:1 mg kg−1 ) proposed by the CODEX committee [4] Elevated Cd concentrations were also found in spring wheat, potatoes and carrots [5] In the past decade plantations of willow, consisting mainly of Salix viminalis L and S dasyclados 0961-9534/02/$ - see front matter ? 2002 Elsevier Science Ltd All rights reserved PII: S - ( ) 0 - 416 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 Wimm., have been established on arable land in Sweden The stem biomass produced has mainly been used as a biofuel in municipal district heating plants Today around 16,000 are cropped with Salix, which corresponds to approximately 0.5% of the total agricultural land in Sweden Several studies have shown that Salix accumulates high levels of Cd [6 –10] Therefore, the role of Salix as a potential phytoextractor to remove Cd from moderately contaminated soils at stem harvest has been discussed In relation to other species known to accumulate Cd, Salix can be deÿned as a high accumulator rather than a hyperaccumulator of Cd According to the deÿnition by Baker et al [11] hyperaccumulators accumulate ¿ 0:01% Cd in leaf dry mass and may have the metal evenly distributed throughout the plant Examples of hyperaccumulators of Cd are Thlaspi caerulescens and Alyssum murale within the Brassicaceae family In contrast to these more e cient species, Salix has a high biomass production, making it possible to have a proÿtable production of biofuel (see above) at the same time as the soil is being restored A di erence in Cd accumulation (uptake and translocation) between genotypes of Salix has also been demonstrated [7,12] The choice of clone will therefore also be of importance for the phytoextraction e ect of a Salix stand The mechanisms that regulate the Cd uptake in plants are still not known Plant uptake of Cd at low solution concentrations has been reported in reviews by Grant et al [13] and Greger [14] to be either passive, metabolic or partially metabolic and partially passive and may be in competition with the uptake system for essential trace elements An incorporation of Cd into the stems of Salix in direct proportion to the biomass production would imply that Cd uptake is dependent on the uxes of water and mineral nutrients through the plant Even if the uptake of Cd is not directly proportional to biomass production, the Cd incorporated into the stems will still increase with increased stem yield as long as the increase in biomass is larger than the decrease in Cd concentration Factors such as temperature, light and ow of water through the plant in uence growth and may therefore also a ect the Cd uptake However, Perttu et al [15] concluded that the factors mentioned not a ect the Cd uptake in Salix The lack of knowledge regarding the factors that determine the Cd concentration and mechanisms behind Cd uptake in Salix makes it di cult to predict the e ect of di erent management practices, e.g fertilisation, on the removal of Cd at stem harvest This study was undertaken to investigate the effect of di erent nutrient supplies and soil Cd concentrations on Cd concentrations in stems, leaves and roots in two di erent clones of Salix The hypothesis was that an increased biomass production induced by fertilisation would increase the Cd content in the stems and hence the removal of Cd at stem harvest Materials and methods The experimental area is situated in Uppsala in ◦ ◦ the east-central Sweden (lat 59 49 N, long 17 40 E, 15 m a.s.) The experiment was carried out in closed lysimeters made from plastic containers (volume approx 200 dm3 and depth 0:9 m) [16] Soil columns consisting of approximately 0:25 m topsoil and 0:50 m subsoil were built up in the lysimeters (Table 1) using an arable clay soil (Eutric Cambisol) collected in the vicinity of Uppsala A drainage pipe covered with sand (approx 0:15 m) was put in the bottom of each lysimeter At the beginning of May 1997, one unrooted cutting (weight 13 ± 0:5 g) of willow (Salix viminalis L or Salix dasyclados Wimm.) was planted in each lysimeter The lysimeters were covered with a lid, with a hole for the shoots The experiment was conducted over three growing seasons, from 1997 to 1999 Treatments consisted of two clones (81090 of Salix dasyclados and 78183 of Salix viminalis), two nutrient levels (based on the N-supply), two soil Cd concentrations and two harvest occasions (2 and years old) Nutrient level (N1) corresponded to application of 0, 20 and 20 kg N ha−1 during years 1, and 3, respectively Corresponding amounts of N for nutrient level (N2) were 30, 60 and 60 kg N ha−1 Table pH, total Cd concentration (7M HNO3 ), exchangeable Cd (NH4 NO3 ) and carbon content in topsoil and subsoil used in the lysimeters Soil type pH Total Cd ( g kg−1 ) Exchangeable Cd ( g kg−1 ) C (%) Topsoil Subsoil 6.7 6.9 296 231 4.9 3.5 1.1 1.0 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 Nutrien level Accumulated irrigation (mm) 420 Nutrient level Season 1997 Season 1997 Season 1998 Season 1998 Season 1999 350 417 Season 1999 280 210 140 70 Accumulated fertilisation (kg N ha-1) 60 50 40 30 20 10 M J J A S O M J J A S O Fig Accumulated irrigation and fertilization for nutrient levels N1 and N2 during the growing seasons 1997 (solid line), 1998 (dotted line), 1999 (dashed line) The other macronutrients were supplied in relation to N (see below) The two soil Cd concentrations in the topsoil were 0:3 mg Cd (kg DW)−1 , which was the concentration of the parent material and 0:6 mg Cd (kg DW)−1 The increased Cd concentration was achieved by adding 0:3 mg Cd (kg DW)−1 in the form of CdSO4 to the topsoil The subsoil had the same Cd concentration in both treatments (0:23 mg Cd kg dw−1 ) Each treatment had two replicates The lysimeters were installed in the ground in groups of (32 lysimeters in total) Plants of Salix were grown around the lysimeters to simulate a stand structure and to eliminate edge e ects The spacing between plants inside and outside the lysimeters was such that the area for each plant was 0:5 m2 For practical reasons, the four combinations of clones and nutrient levels were distributed so that clone and nutrient level were the same for all experimental units within each group Soil Cd treatments were randomized within each group of lysimeters Irrigation was performed daily from late May until the beginning of October each year with a computerized drip irrigation system (Fig 1) The plants received the same amount of water irrespective of nutrient level, in order to reduce the number of experimental factors and also to simulate ÿeld conditions A liquid fertilizer (Blomstra, Cederroth International) containing all essential macro- and micronutrients in the following proportions (by weight); 100 N, 20 P, 84 K, Ca, Mg, S, 0.3 Fe, 0.4 Mn, 0.2 B, 0.06 Zn, 0.03 Cu and 0.0008 Mo was applied with the drip irrigation system according to a sigmoid growth curve (cf [17]) (Fig 1) Climate data were obtained from the Ultuna meteorological station close to the ÿeld The soil water potential in the lysimeters was controlled by TDR measurements during the season Excess water collected in the drainage pipe at the bottom of each lysimeter was occasionally pumped out Samples of this water were taken for Cd analyses and they showed that almost no Cd had leached 418 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 out In order to prevent the roots penetrating into the sand at the bottom, a nylon straining-cloth (80 m, Bewatex AB) enclosed the topsoil and subsoil In November 1998, when the plants were years old (2-Au98), the ÿrst harvest was carried out in half of the lysimeters In the lysimeters without added Cd, the roots were destructively sampled During the winter 1998–1999, the remaining plants were unfortunately severely damaged by ÿeld-mice and these also had to be harvested in early spring 1999 (2-Sp99) The plants coppiced in spring were allowed to resprout and were harvested again in autumn 1999 (1-Au99) The roots were destructively sampled after harvest in all lysimeters harvested in autumn 1999 The shoots were cut at about cm from where they were attached to the cutting (stem base) During the second and third growing seasons, shed leaves were collected (from middle of July until all leaves had fallen) from the plants later being harvested in the same year A net was mounted surrounding each plant and held open above the canopy by a sti circular wire To record plant nutrient status, mature leaves that had not yet abscised were also collected from every plant at the beginning of September in the second and third growing season In order to cover the concentration gradient along the shoot, these leaves were taken randomly within three sections of the aboveground stool (unit of roots and shoots originating from the same cutting) The levels separating each section were set by dividing the tallest shoot into three Each section was analyzed separately The collected shed leaves and the non-abscised leaves sampled for nutrient analysis were dried ◦ (70 C), weighed and ground on a Thomas–Wiley laboratory mill (mesh size mm) and on a Retsch knife mill (mesh size 0:2 mm), respectively Subsamples from the leaf material were wet ashed (heating ◦ block 150 C) in a mixture of 10 ml conc HNO3 and ml conc HClO4 The acids were evaporated until 0.5 ml of perchloric residue remained, then this was diluted with H2 O to a ÿnal volume of 35 ml The extracts were analyzed for Cd on a JY-70 Plus ICP Emission Spectrometer In addition, the extracts from the non-abscised leaves were analyzed for the macro elements P, K, Ca and Mg on ICP (see above) and subsamples from the same leaf material were also analyzed for N (Carlo Erba NA 1500 elemental analyzer) Stems, roots and cuttings were oven dried ◦ at 70 C to constant weight, weighed and ground on a Thomas-Wiley laboratory mill (mesh size mm) and analyzed for Cd as described above The roots were clipped before grinding Soil samples were air ◦ dried (30 –40 C), ground to pass through a mm sieve and analyzed for total Cd, exchangeable Cd, organic carbon (C) and pH Total Cd (Cd–HNO3 ) was analyzed after extraction with M nitric acid ◦ (110 C, 2h) [18] Exchangeable Cd was estimated by extracting the soils with 1:0 M NH4 NO3 (Cd– NH4 NO3 ) Total carbon content was analyzed on an elemental analyzer (LECO CHN-932) and pH was measured in H2 O (soil:water ratio 1:5) Water samples were ÿltered (0:2 m) and 1% by volume of conc HNO3 was added for conservation Analyses of Cd on the water samples and soil extracts were performed by means of atomic absorption spectrophotometry using the graphite furnace technique (Zeeman 4110 ZL) Because of the unplanned harvest in spring of the third growing season, each harvest was treated separately in a 3-factorial design Statistical analyses (ANOVA) were performed with the programme Systat 10.0 (SPSS Inc) Results 3.1 Weather conditions The summer of 1997 was warmer than normal for Swedish conditions (Fig 2) In spite of this, the potential evaporation (Penman) did not exceed precipitation by very much and the accumulated precipitation during May–September was quite high (Fig 2) The growing season in 1998 was cooler than normal and accumulated precipitation was again quite high (Fig 2) During the growing season of 1999, the temperature was once again higher than normal for the area concerned However, the accumulated precipitation was very low and well below accumulated potential evaporation 3.2 Plant nutrient status In the second growing season (1998), the leaf N concentration was around 22 mg N (g DW)−1 and N E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 40 1997 40 Acc precip 330 Acc pot evapo 416 1997 Mean air temp 14.5 30 20 20 10 10 0 40 1998 30 Acc precip 329 Acc pot evapo 338 1998 Mean air temp 12.5 40 30 20 20 10 10 0 40 1999 30 Acc precip 163 Acc pot evapo 447 1999 Mean air temp 14.5 40 Daily mean temperature (°C) Daily precipitation and pot evapo.transp (mm) 30 419 30 20 20 10 10 0 M J J A S O M J J A S O Fig Daily precipitation (solid line), potential evaporation (dotted line) and mean air temperature for the growing seasons 1997–1999 In each graph, values for accumulated precipitation (Acc precip.), accumulated potential evaporation (Acc pot.evapo.) and mean air temperature (Mean air temp.) for each growing season (May–October) are given Table Means of treatment e ects ± SD for N, P, K, Ca, Mg concentrations in non-abscised leaves sampled and analyzed from three sections within the shoot in autumn 1998 and autumn 1999 Treatment N (mg g−1 ) P K Ca (mg g−1 ) Mg 1998 1999 1998 1999 1998 1999 1998 1999 1998 1999 Nutrient level Soil Cd conc Clone 81090 78183 22 ± 3a 22 ± 3a 22 ± 3a 22 ± 2a 19 ± 1a 24 ± 1b 19 ± 1a 23 ± 4b 22 ± 4a 19 ± 2a 20 ± 5a 22 ± 2a 5:6 ± 1:1a 4:1 ± 1:0b 4:8 ± 1:5a 4:9 ± 1:1a 4:5 ± 0:9a 5:2 ± 1:5a 8:1 ± 0:6a 6:3 ± 1:4b 6:6 ± 1:6a 7:8 ± 0:9b 7:0 ± 1:3a 7:4 ± 1:6a 16 ± 1a 15 ± 1a 15 ± 1a 15 ± 1a 15 ± 1a 16 ± 1a 17 ± 2a 16 ± 1a 17 ± 2a 16 ± 1a 17 ± 2a 15 ± 1b 15 ± 4a 13 ± 3b 14 ± 3a 14 ± 4a 17 ± 1a 11 ± 2b 20 ± 4a 18 ± 4b 18 ± 4a 19 ± 4a 22 ± 2a 15 ± 2b 1:9 ± 0:3a 1:7 ± 0:2b 1:9 ± 0:3a 1:9 ± 0:3a 2:1 ± 0:2a 1:6 ± 0:2b 3:3 ± 0:3a 2:7 ± 0:4b 2:9 ± 0:5a 3:1 ± 0:3a 3:2 ± 0:3a 2:8 ± 0:4b Means within columns followed by di erent letters are di erent at p 0:05 when comparing levels within the same treatment and harvest occasion concentrations were not signiÿcantly (p ¡ 0:05) affected by nutrient level (Table 2) During the third growing season (1999) when the plants had resprouted after coppicing, the plants at the high nutrient level (N2) had signiÿcantly (p ¡ 0:05) higher leaf N concentrations (23 mg N (g DW)−1 ) than the plants at the low nutrient level (N1) (19 mg N (g DW)−1 ) Leaf concentrations of the macronutrients P, Ca and Mg were signiÿcantly higher at nutrient level N1 compared to nutrient level N2 (p ¡ 0:05), independent of sampling occasion, while leaf concentration of K was not in uenced by nutrient level (Table 2) Clone 81090 had higher leaf concentrations of Ca and Mg than clone 78183, a trend which was also true for K during the third growing season Leaf N concentration was highest for clone 78183 during the second growing season, while leaf concentration of P did not di er between clones Soil Cd concentration did not have any pronounced e ects on leaf nutrient concentration 420 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 Stem biomass (g plant -1 ) 750 Clone 78183 Clone 81090 900 (a) 600 450 300 150 Stem Cd amount (mg plant -1 ) Stem Cd concentration (mg kg -1 ) 20 (b) 15 10 3.0 2.5 (c) 2.0 1.5 1.0 0.5 0.0 N1N2 Cd0 N1N2 Cd1 2-Au98 N1N2 Cd0 N1N2 Cd1 N1 N2 Cd0 2-Sp99 N1N2 Cd1 1-Au99 N1N2 Cd0 N1N2 Cd1 2-Au98 N1N2 Cd0 N1N2 Cd1 2-Sp99 N1N2 Cd0 N1N2 Cd1 1-Au99 Fig Mean values ± SD (n = 2) for stem biomass (DW), concentration (dry weight basis) and total amount of Cd in the stem for each clone (81090 and 78183), nutrient level (N1 and N2) and soil Cd concentration (Cd0 and Cd1) The plants are 2-years old, harvested in autumn 1998 (2-Au98) and spring 1999 (2-Sp99), and 1-year old sprouts coppiced in autumn 1999 (1-Au99) 3.3 E ect of nutrient level and soil Cd concentration on biomass and Cd content of stems Stem biomass production signiÿcantly (p ¡ 0:05) increased with a higher nutrient supply, independent of harvest occasion (Fig 3a) The increase in mean stem biomass between nutrient levels and amounted to 60 –80% (Tables 3a–3c) Figure 3b also shows that stem Cd concentration was a ected by nutrient level In the stems harvested in autumn 1998 (2-Au98) and spring 1999 (2-Sp99), the Cd concentration was signiÿcantly (p ¡ 0:05) higher at the low nutrient level (N1) than at the high nutrient level (N2) (Tables 3a and 3b) The same tendency was found in the resprouting yr old stems (1-Au99), but it was not statistically signiÿcant (Fig 3b and Table 3c) The total amount (concentration × biomass) of Cd in stems tended to be slightly higher at the higher nutrient level (N2) than at the low nutrient level (N1), but the difference was only statistically signiÿcant (p ¡ 0:05) in the case of the resprouting 1-year old stems (1-Au99) (Fig 3c and Tables 3a–3c) The explanation for the weak e ect on amounts of Cd in the stems is the opposing and very consistent relationship between stem biomass production and stem Cd concentration (Fig 3a and b) The e ects of nutrient levels N1 and E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 421 Table 3a Stem biomass, Cd concentration, Cd amount in stems harvested in autumn 1998 (2-Au98) as in uenced by clone, nutrient level and soil Cd concentration Analysis of variance Source df Biomass Cd-conc Cd-amount p-values ( = 0:05) Main e ects Nutrient level (N) Soil Cd concentration (Cd) Clone Interactive e ects Clone*N Clone*Cd N*Cd Clone*N*Cd Error Corrected total 1 0.000 0.168 0.000 0.000 0.000 0.000 0.087 0.000 0.000 1 1 15 0.263 0.027 0.094 0.010 0.002 0.075 0.255 0.381 0.098 0.251 0.061 0.626 0.126 0.743 0.002 (g DW) 264 ± 77 410 ± 137 259 ± 76 415 ± 131 360 ± 171 314 ± 81 (mg kg DW−1 ) 6:1 ± 1:6 3:1 ± 1:2 5:3 ± 2:2 3:9 ± 1:8 3:6 ± 1:8 5:6 ± 1:0 (mg) 1:6 ± 0:4 1:1 ± 0:3 1:3 ± 0:3 1:4 ± 0:5 1:1 ± 0:2 1:6 ± 0:4 Means of main e ects ± SD Clone 81090 Clone 78183 N1 N2 Cd0 Cd1 Data for biomass and Cd amount were log-transformed prior to the statistical analyses N2 on stem biomass, stem Cd concentration and total stem Cd amount for each soil Cd concentration (Cd0 and Cd1) and for both clones in Fig 3, remains the same regardless of soil Cd concentration Increased levels of Cd in the soil did not in uence stem biomass production (Tables 3a–3c) However, higher soil Cd concentrations signiÿcantly raised the concentration and total amount of Cd in the stems (Tables 3a–3c) The concentration was 1.4 –2.2 times higher at soil Cd concentration compared to soil Cd concentration (Fig 3b) The corresponding total Cd amount in the stems at Cd1 was 1.3–2.1 times the Cd0 level (Fig 3c) In the resprouting 1-year-old stems, the increase in stem Cd concentration tended to be somewhat higher than the increase in total Cd amount 3.4 Di erences in stem growth and Cd content between clones Clone 78183 produced signiÿcantly more stem biomass than clone 81090 when harvested after two growing seasons (2-Au98 and 2-Sp99) (Tables 3a and 3b) In the resprouting plants harvested in autumn 1999 (1-Au99) there were no di erences in stem biomass between clones Clone 81090 had higher stem Cd concentrations than clone 78183, independent of harvest occasion The total amount of Cd in the stems was also larger in clone 81090, with the exception of the stems harvested in spring 1999 Clone 81090 yielded a higher root biomass and had a larger amount of Cd in the roots than clone 78183, independent of harvest occasion (Fig 4) 3.5 Comparison between plant compartments The treatment e ects on biomass, Cd concentration and amounts of Cd in leaves and roots followed more or less the same pattern as those in the stems (Fig 4) A comparison of Cd concentrations in the different plant compartments showed that the leaves had the highest concentrations independent of harvest occasion, nutrient level, clone and soil Cd concentration 422 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 Table 3b Stem biomass and Cd concentration and total Cd amount in stems harvested in spring 1998 (2-Sp99) as in uenced by clone, nutrient level and Cd concentration in the soil Analysis of variance Source df Biomass Cd-conc Cd-amount p-values ( = 0:05) Main e ects Nutrient level (N) Soil Cd conc (Cd) Clone 1 0.001 0.409 0.023 0.011 0.005 0.024 0.891 0.001 0.187 Interactive e ects Clone*N Clone*Cd N*Cd Clone*N*Cd Error Corrected total 1 1 15 0.724 0.507 0.818 0.849 0.010 0.899 0.774 0.331 0.778 0.020 0.499 0.489 0.095 0.470 0.095 (g DW) 237 ± 49 423 ± 147 353 ± 181 307 ± 99 276 ± 95 384 ± 168 (mg kg DW−1 ) 7:1 ± 3:6 4:1 ± 1:5 3:8 ± 1:4 7:3 ± 3:4 6:7 ± 3:4 4:5 ± 2:4 (mg) 1:6 ± 0:6 1:6 ± 0:4 1:2 ± 0:2 2:0 ± 0:4 1:7 ± 0:6 1:5 ± 0:5 Means of main e ects ± SD N1 N2 Cd0 Cd1 Clone 81090 Clone 78183 Data for biomass and Cd concentration werelog-transformed prior to the statistical analyses (Fig 4) There were no major di erences in Cd concentration between the other plant parts (stems, roots and cuttings) In the 2-year old plants, the amount of Cd was largest in the stems and lowest in the cuttings, while leaves and roots were intermediate in this respect In the resprouting 1-year old plants, Cd amounts in the roots were relatively larger in comparison to the amounts in the stem For clone 81090, which had a larger root biomass than clone 78183, this meant that the total root Cd amount was larger than that of the stem Discussion An enhanced nutrient supply resulted in signiÿcantly higher stem biomass production In general, this gave rise to lower stem Cd concentrations compared to when the nutrient supply, and hence biomass production, was lower This e ect of enhanced growth on stem Cd concentrations is commonly referred to as a biological dilution e ect Biological dilution has also been reported for heavy metals in other plant species investigated elsewhere For example Singh et al [19] saw a suppressed Cd uptake in lettuce when the N level was high (¿ 150 mg N kg−1 added to the soil) which could partly be explained by a dilution e ect Furthermore, Jones et al [20] observed a substantial increase in the concentration of lead in the shoots of plants whose growth rate was slow due to a nutrient deÿciency E ects of dilution on the concentration of Cd have also been demonstrated in ryegrass plants growing at di erent rates due to the age of the plants [21] The opposing and very consistent trends in stem biomass production and stem Cd concentration resulted in insigniÿcant or small positive e ects on the total amount of Cd in the stems (Fig 3c) This indicates that the incorporation of Cd into the stems is governed by processes which are independent of E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 423 Table 3c Stem biomass and concentration and total amount of Cd in stems harvested in autumn 1999 (1-Au99) as in uenced by clone, nutrient level and Cd concentration in the soil Analysis of variance Source df Biomass Cd-conc Cd-amount p-values ( = 0:05) Main e ects Nutrient level Cd concentration in soil Clone Interactive e ects Clone*N Clone*Cd N*Cd Clone*N*Cd Error Corrected total 1 0.016 0.118 0.922 0.097 0.000 0.019 0.011 0.000 0.003 1 1 15 0.585 0.778 0.638 0.946 0.014 0.598 0.415 0.246 0.460 1.060 0.821 0.449 0.087 0.249 0.004 (g DW) 152 ± 22 239 ± 81 221 ± 86 170 ± 50 190 ± 53 201 ± 92 (mg kg DW−1 ) 5:7 ± 2:4 4:8 ± 2:1 3:5 ± 0:9 7:0 ± 1:7 6:0 ± 2:4 4:5 ± 2:0 (mg) 0:8 ± 0:3 1:0 ± 0:3 0:7 ± 0:2 1:1 ± 0:2 1:1 ± 0:3 0:8 ± 0:2 Means of main e ects ± SD N1 N2 Cd0 Cd1 Clone 81090 Clone 78183 Data for biomass and Cd amount were log-transformed prior to the statistical analyses biomass production Plants of Salix whose growth rate is slow because of nutritional constraints are therefore likely to have elevated concentrations of Cd This seems to be valid also for P, Ca and Mg, but not for N and K In this context it should be mentioned that the leaf concentrations of macronutrients in the present study were in almost the same range as those presented in some other investigations for young, fertilized, high-yielding stands sampled in midsummer [22] Stem biomass production is, however, not only determined by the supply of nutrients Lindroth and Cienciala [23] concluded that water availability is a critical factor for growth of Salix in Sweden During some periods of the growing season, water availability will probably be the most limiting factor for growth It is therefore likely that the plants will grow more slowly than can be expected from available N during some periods of growth In the present investigation the plants received the same amount of water irrespective of nutrient level, in order to reduce the number of experimental factors and also to simulate ÿeld conditions As pointed out earlier, when comparing the amount of Cd in the stems at the two nutrient levels the same pattern could be distinguished regardless of harvest occasion However, in the 2-year old plants (2-Au98 and 2-Sp99) the e ect of nutrient level on total stem Cd amount was insigniÿcant This was not the case in the coppiced 1-year old plants (1-Au99), where the amount of Cd was signiÿcantly higher at the higher nutrient level compared to the lower nutrient level Similar results were observed in a study with Salix conducted in a climate chamber for two growing seasons and where the plants were supplied with N in accordance with growth at two di erent rates [15] In contrast to the present study, the di erence between the high and low nutrient levels was more pronounced and the plants were kept well watered at both nutrient levels, while the growth medium was solely a clay mineral (vermiculite) The somewhat di ering results between the 1- and 2-year old plants described in the 424 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 Clone 81090 Cd concentation (mg kg-1) 30 20 20 10 10 Cd amount (mg plant-1) Clone 78183 30 Leaves Stem Cutting Root 3.5 3.5 2.5 2.5 1.5 1.5 L S C R 0.5 -0.5 0.5 -0.5 -1.5 -1.5 -2.5 -2.5 -3.5 N1 N2 Cd0 N1 N2 Cd1 2-Au98 -3.5 N1 N2 Cd0 N1 N2 Cd1 2-Au99 N1 N2 Cd0 N1 N2 Cd1 2-Au98 N1 N2 Cd0 N1 N2 Cd1 2-Au99 Fig Mean values (n = 2) for Cd concentration (dry weight basis) and total Cd amount per plant in leaves (L), stems (S), cuttings (C) and roots (R) for each clone (81090 and 78183), nutrient level (N1 and N2) and soil Cd concentration (Cd0 and Cd1) The plants are 2-years old harvested in autumn 1998 (2-Au98) and 1-year old sprouts coppiced in autumn 1999 (1-Au99) For the plants harvested in autumn 1998 (2-Au98), data for the cuttings and roots are missing present investigation might be a consequence of coppicing According to Bollmark [24], coppicing may, for example, result in changes in growth rate for different plant parts depending on nutrient level and also changes in mobilization and translocation of nutrients and carbohydrates, which in turn may a ect the concentration and amount of Cd in the stems Other explanations for the di erences between the 2- and 1-year old plants in this study might be changed weather conditions between the second and the third growing season, but also the higher water availability in the 1-year old plants as they received the same amount of water as before coppicing An increased soil Cd concentration in the topsoil in the current investigation increased the Cd concentration and the total Cd amount in the plants, demonstrating the ability of Salix to take up more Cd from more contaminated soils In some cases, the increase in total stem Cd amount tended to be almost directly proportional to the increase in soil Cd concentration However, more often the increase in stem Cd amount was between 40 and 80% of the increase in Cd content of the topsoil The reason that the increase in stem Cd concentration is less than the increase in the topsoil Cd concentration may be that a signiÿcant proportion of Cd in the stems is taken up from the subsoil Clone 81090 had higher stem Cd concentrations compared to clone 78183 The reason might be clone speciÿc or a consequence of the higher stem biomass production of clone 78183 As pointed out earlier, various clones di er in their ability to take up Cd and to transport Cd up to the shoot [7] Clones also di er in their distribution of Cd between stems and leaves, which has been seen by Perttu et al [15] The choice of clone will therefore be important for the removal of Cd at stem harvest Both clones in this investigation had an intermediate transport of Cd up to the shoot In general, the leaves had higher Cd concentrations than the stems, a trend also recorded by Riddel-Black [25] However, the total amount of Cd is larger in the stems, at least if harvest is performed after two or more growing seasons Conclusions • Increased fertilization in this experiment consistently resulted in increased biomass production, E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 and generally in a higher total Cd content in the stems However, the magnitude of the increase in total Cd was small and only statistically signiÿcant in some cases, mainly because increased biomass also resulted in lower Cd concentration Thus, if Salix is used as a phytoextractor of Cd, the possibilities for signiÿcantly increasing removal rate by increased biomass production would seem to be restricted On the other hand, if a low Cd concentration in Salix biofuel is desirable, the prospects of achieving that through increased biomass production are good • Addition of Cd to the topsoil resulted in higher Cd concentrations and Cd amounts in the Salix plants Thus, the e ciency of Salix as a phytoextractor may increase with the degree of pollution of the soil • Clones di ered in concentrations and total amounts of Cd in the stems This indicates that choice of clone may be a better way to increase the phytoextraction e ect of Salix than increasing biomass production by fertilization • The amount of Cd was higher in stems than in leaves Unless a signiÿcant amount of Cd is taken up from the subsoil, this means that more Cd is taken out from the topsoil by stem harvest than is recirculated back by litterfall [2] [3] [4] [5] [6] [7] [8] [9] [10] Acknowledgements The authors cordially thank Richard Childs, Eva-Marie Fryk, Christina Segerqvist, Eira Casenberg, Gunilla Lundberg and Gunilla Hallberg for their help in starting and managing the experiment, sampling, sample preparation and analyses We also want to thank Dr Jan Eriksson, Dr Par Aronsson, Dr Maria Greger and Dr Anders Goransson for valuable discussions and comments on the manuscript This research was ÿnancially supported by Vattenfall AB, The Federation of Swedish Farmers and The Swedish National Energy Administration References [1] Eriksson J, Soderstrom M, Andersson A Cadmium contents in the plough layer of Swedish agricultural soils [11] [12] [13] [14] [15] [16] 425 Naturvardsverket Report 4450, 1995 30p [in Swedish with English summary] Andersson A Trace elements in agricultural soils- uxes, balances and background values Swedish Environmental Protection Agency Report 4077, 1992 Eriksson J A ÿeld study on factors in uencing Cd levels in soils and in grains of oats and winter wheat Water, Air, and Soil Pollution 1990;53:69–81 FAO=WHO Report of the Eighth Session of the Codex Committee on Cereals, Pulses and Legumes held in Washington DC, 26 –30 October 1992 Joint FAO=WHO Food Standards Programme, Codex Alimentarius Commission, 20th Session, Geneva 28 June–July, 1993 p Oborn I, Jansson G, Johnsson L A ÿeld study on the in uence of soil pH on trace element levels in spring wheat (Triticum aestivum), potatoes (Solanum tuberosum) and carrots (Daucus carota) Water Air and Soil Pollution 1995;85(2):835–40 Brieger G, Wells JR, Hunter RD Content in y ash ecosystem Water, Air, and Soil Pollution 1992;63:87–103 Landberg T, Greger M Can heavy metal tolerant clones of Salix be used as vegetation ÿlters on heavy metal contaminated land? In: Aronsson P, Perttu K, editors Willow vegetationÿlters for municipal wastewaters and sludges Report 50 Swedish Univ of Agric Sci., Dept Ecol Envir Res., Sect Short Rotation forestry, 1994 p 133– 44 Ostman G Kadmium i Salix Report Vattenfall utveckling AB, Projekt bioenergi, 1994 22pp [in Swedish with English summary] Felix H Field trials for in situ decontamination of heavy metal polluted soils using crops of metal-accumulating plants Z Planzenernahr Bodenk 1997;160:525–9 Greger M, Landberg T Use of willow in phytoextraction International Journal of Phytoremediation 1999;1:115–23 Baker AJM, McGrath SP, Sidoli CMD, Reeves RD The possibility of in situ heavy metal decontamination of polluted soils using crops of metal-accumulating plants Resources, Conservation and Recycling 1994;11:41–9 Landberg T, Greger M Di erences in uptake and tolerance to heavy metals in Salix from unpolluted and polluted areas Applied Geochemistry 1996;11:175–80 Grant CA, Buckley WT, Bailey LD, Selles F Cadmium accumulation in crops Canadian Journal of Plant Science 1998;78:1–17 Greger M Metal availability and bioconcentration in plants In: Prasad MNV, Hegemeyer J, editors Heavy metal stress in plants Heidelberg: Springer, 1999 p 1–27 Perttu K, Eriksson J, Greger M, Goransson A, Blomback K, Klang-Westin E, Landberg T Slutrapport for projektet Forrad och oden av kadmium i systemet mark-Salix Published by The Swedish National Energy Administration, in press [in Swedish] Rytter R-M Fine root production and carbon and nitrogen allocation in basket willows Dissertation Silvestria 39, Swedish University of Agricultural Sciences, Department of Short-rotation Forestry, 1997 426 E Klang-Westin, K Perttu / Biomass and Bioenergy 23 (2002) 415 – 426 [17] Rytter L, Ericsson T Leaf nutrient analysis in Salix viminalis (L.) energy forest stands growing on agricultural land Z Planzenernahr Bodenk 1993;156:349–56 [18] Swedish Standars Institute Soil analysis-determination of trace elements in soils-Extraction with nitric acid SS 028311, 1997 [in Swedish] [19] Singh JP, Bharat S, Karwasra SPS Yield and uptake response of lettuce to cadmium as in uenced by nitrogen application Fertilizer Research 1988;18(1):49–56 [20] Jones LHP, Jarvis SC, Cowling DW Lead uptake from soils by perennial ryegrass and its relation to the supply of an essential element (sulphur) Plant and Soil 1973;38:605–19 [21] Jarvis SC, Jones LHP Uptake and transport of cadmium by perennial ryegrass from owing solution culture with a constant concentration of cadmium Plant and soil 1978;49:333–42 [22] Ericsson T, Rytter L, Linder S Nutritional dynamics and requirements of short-rotation forests In: Mitchell CP, Ford-Robertson T, Hinckley T, Sennerby-Forsse L, editors Ecophysiology of short rotation forest crops London and New York: Elsevier Science Publishers, 1992 p 35–65 [23] Lindroth A, Cienciala E Water use e ciency of short-rotation Salix viminalis at leaf, tree and stand scales Tree physiology 1996;16:257–62 [24] Bollmark L, Sennerby-Forsse, Ericsson T Accumulation and mobilisation of nutrient reserves in Salix viminalis Dissertation Silvestria 155, Swedish University of Agricultural Sciences, Department of Short-rotation Forestry, 2000 [25] Riddel-Black D Heavy metal uptake by fast growing willow species In: Aronsson P, Perttu K, editors Willow vegetationÿlters for municipal wastewaters and sludges Report 50 Swedish Univ of Agric Sci., Dept Ecol Envir Res., Sect Short Rotation forestry 1994 p 145 –51  ... total stem Cd amount for each soil Cd concentration (Cd0 and Cd1 ) and for both clones in Fig 3, remains the same regardless of soil Cd concentration Increased levels of Cd in the soil did not... biomass (DW), concentration (dry weight basis) and total amount of Cd in the stem for each clone (81090 and 78183), nutrient level (N1 and N2) and soil Cd concentration (Cd0 and Cd1 ) The plants... (c) 2.0 1.5 1.0 0.5 0.0 N1N2 Cd0 N1N2 Cd1 2-Au98 N1N2 Cd0 N1N2 Cd1 N1 N2 Cd0 2-Sp99 N1N2 Cd1 1-Au99 N1N2 Cd0 N1N2 Cd1 2-Au98 N1N2 Cd0 N1N2 Cd1 2-Sp99 N1N2 Cd0 N1N2 Cd1 1-Au99 Fig Mean values ±