Long-term distribution, mobility and plant availability of compost-derived heavy metals in a landfill covering soil

Long-term distribution, mobility and plant availability of compost-derived heavy metals in a landfill covering soil

S CI EN CE OF T H E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5 a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m w w w. e...

1MB Sizes 0 Downloads 6 Views

S CI EN CE OF T H E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

a v a i l a b l e a t w w w. s c i e n c e d i r e c t . c o m

w w w. e l s e v i e r. c o m / l o c a t e / s c i t o t e n v

Long-term distribution, mobility and plant availability of compost-derived heavy metals in a landfill covering soil D. Businelli, L. Massaccesi, D. Said-Pullicino, G. Gigliotti⁎ Department of Agricultural and Environmental Science, University of Perugia, Borgo XX Giugno 72, 06121 Perugia, Italy

AR TIC LE D ATA

ABSTR ACT

Article history:

The application of municipal waste compost (MWC) and other organic materials may serve

Received 2 July 2008

to enhance soil fertility of earthen materials and mine spoils used in land reclamation

Received in revised form

activities, particularly in the recovery of degraded areas left by exhausted quarries, mines

23 October 2008

and landfill sites among others. The long-term distribution, mobility and phytoavailability

Accepted 23 October 2008

of heavy metals in such anthropogenic soils were studied by collecting soil samples at

Available online 22 November 2008

different depths over a 10 y chronosequence subsequent to amendment of the top layer of a landfill covering soil with a single dose of mechanically-separated MWC. Amendment

Keywords:

resulted in a significant enhancement of the metal loadings in the amended topsoils

Anthropogenic soil

particularly for Cu, Zn and Pb, which were also the predominant metals in the compost

Municipal waste compost

utilised. Although metals were predominantly retained in the compost amended soil

Metal leaching

horizon, with time their vertical distribution resulted in a moderate enrichment of the

Dissolved organic matter

underlying mineral horizons, not directly influenced by compost amendment. This

Chemical fractionation

enrichment generally resulted from the leaching of soluble organo-metal complexes and

Phytoavailability

subsequent adsorption to mineral horizons. However, in the course of the 10-y experimental period, metal concentrations in the underlying horizons generally returned to background concentrations suggesting a potential loss of metals from the soil system. Analysis of the tissues of plants growing spontaneously on the landfill site suggests that metal phytoavailability was limited and generally species-dependent. © 2008 Elsevier B.V. All rights reserved.

1.

Introduction

Compost production from the organic fraction of urban solid waste is an important means of recovering organic matter and plays an important role in modern waste management strategies. Indeed, composting is a useful way of transforming organic wastes into a valuable amendment for soils. The application of organic residues as a source of organic matter is a common practice aimed at improving soil physical, chemical and biochemical properties (Giusquiani et al., 1995; Leifeld et al., 2002). Amendment of natural, arable or anthropogenic, degraded and nutrient-depleted soils, leads to an increase in their fertility through an increase in soil organic matter and

nutrient content, soil porosity, water holding capacity, soil microbial activity, soil microbial biomass, and plant productivity (Leifeld et al., 2002; Clemente et al., 2006; Businelli et al., 2007). Moreover, compost amendment has been also utilised for the reclamation and bioremediation of anthropogenically degraded areas, such as landfill and mining sites, in order to minimize land degradation and benefit from both an economical and landscaping point of view (Clemente et al., 2006; Businelli et al., 2007). At the present time, legislative measures in different countries prohibit the application of mechanically-separated municipal waste compost (MWC) to arable soils and as such, these organic materials have been widely utilised in land

⁎ Corresponding author. Dept. of Agricultural and Environmental Science (DSAA), University of Perugia, Borgo XX Giugno, 72, 06121 Perugia, Italy. Tel.: +39 075 585 6237; fax: +39 075 585 6239. E-mail address: [email protected] (G. Gigliotti). 0048-9697/$ – see front matter © 2008 Elsevier B.V. All rights reserved. doi:10.1016/j.scitotenv.2008.10.052

S CI EN C E OF TH E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

reclamation activities. As a matter of fact, the presence of considerable residual concentrations of heavy metals (including Cd, Cr, Cu, Ni, Pb and Zn) in these organic materials is the main problem associated with their application to soil. These heavy metals can be leached through the soil profile, transported in drainage waters, and may pollute groundwater, or they can accumulate in the upper soil layer and can be toxic to plants and soil microbial biomass (Gigliotti et al., 1996, Planquart et al., 1999; Gove et al., 2001; Kaschl et al., 2002; Ashworth and Alloway, 2004; Clemente et al., 2006). Many studies have shown that several soil variables other than pH, such as texture, organic matter and clay contents, cation exchange capacity and redox potential, may influence the behaviour and bioavailability of heavy metals (Planquart et al., 1999; Gove et al., 2001; Clemente et al., 2006). Moreover, organic matter has important direct and indirect effects on the fractionation of metals in soil, because of their adsorption on solid and particulate soil organic matter and their complexation with dissolved organic matter (DOM; Hoffmann et al., 1998; Wang and Staunton, 2006). DOM has the ability to form stable, soluble complexes with heavy metals (e.g. Cu, Cd, Ni, Pb), and can thus play an important role in the dissolution and translocation of heavy metals in the soil profile. Despite the high adsorption capacity of organic matter added with MWC, little is known as yet about the long-term environmental and ecotoxicological risks associated with the application of MWC for land reclamation activities with respect to heavy metal pollution. Various authors have reported that compost amendment may result in the vertical displacement of mobile metal-organic complexes in the soil solution (Planquart et al., 1999; Kaschl et al., 2002). Moreover, the form of metal present and the change in compost composition and soil conditions over time may control the release of metals from the biosolid matrix in the long-term, after application to soil (Stacey et al., 2001; Clemente et al., 2006). For these reasons, it can be hypothesised that the mobilisation and consequently phytoavailability, of compost-derived heavy metals from amended topsoils along a soil profile may, even in the longterm, be significantly influenced by the biodegradation of compost organic matter and/or release of DOM. This hypothesis was tested by elucidating the fate of heavy metals over a 10-y period, along the soil profile of an anthropogenic landfill covering soil in which the topsoil was amended with MWC.

2.

Materials and methods

2.1.

Site description and soil sampling

The site under study was a controlled landfill located in the municipality of Perugia, Italy (43°16′54″N, 12°25′17″E; mean annual precipitation 720 mm; mean annual temperature 13.7 °C) and managed by Gesenu S.p.A. Since 1993, waste disposal was carried out in horizontal overlapping layers with each depositional layer representing a yearly accumulation of waste materials with its respective covering layer. The external fronts of each depositional layer constituted a step in the landfill surface (Fig. 1). Excavations were made in the external fronts to expose the soil profiles which were characterised by three major horizons. The uppermost layer was an A horizon

1427

composed of a silt loam excavation material amended with compost. The second layer was a C1 horizon composed of an unamended, silt loam excavation soil. Beneath this was a C2 horizon consisting of low-permeability clay loam calcareous material. Chemical properties of the substrates used as landfill covering materials are reported in Table 1, while more detailed site and pedological descriptions have been provided elsewhere (Businelli et al., 2007). To study the effects of a single application of MWC on the long-term mobility of compostderived heavy metals, samples were collected (in 2003) from the depositional layers corresponding to the years 1993, 1994, 1997, and 2001, thus representing a time series of soils 10, 9, 6 and 2 y from compost amendment, respectively. Soil samples were collected directly from each horizon of each soil profile. Replicate soil samples (for a total of three soil samples for each horizon of each depositional layer) were sampled with an auger since further excavation of profiles was deemed to be hazardous for landfill stability. Collected soil samples were airdried, passed through a 2-mm sieve, and stored at room temperature for subsequent chemical analyses.

2.2.

Soil chemical analysis

The pH of soils was measured in a 1:5 water extract. Waterextractable organic matter was extracted from the soil samples with deionised water (solid to water ratio of 1:2 w/w) for 24 h in a horizontal shaker at room temperature. The suspensions were then centrifuged at 10,000 rpm for 10 min and filtered through a 0.45 μm membrane filter. The water extracts were analysed for organic C using Pt-catalysed, high-temperature combustion (680 °C) followed by infrared detection of CO2 (TOC-5000A, Shimadzu Corp., Tokyo, Japan). Before determination of organic C, the inorganic C was removed by adjusting the solution to pH 2 with concentrated H3PO4 and sparging with CO2-free synthetic air at a flow rate of 50 ml min− 1 for 2 min. No flocculation of organic matter was observed on acidification of samples. All values of water-extractable organic C (WEOC) were normalised to the total organic C content of the soils. For the determination of aqua-regia extractable trace metal concentrations, soil samples were digested with HNO3:HCl (1:3 v/v) for 16 h at room temperature followed by 2 h at 130 °C (Zemberyova et al., 2006). The concentrations of Pb, Ni, Cr and Cd were determined by graphite-furnace atomic absorption spectrophotometry (GF-AAS; GFA-EX7, Shimadzu Corp.) with deuterium lamp background correction and using a matrix modifier (Pd (NO3)2, 0.5 mol l− 1 in HNO3) whereas Cu and Zn were determined by flame AAS (AA-6800, Shimadzu Corp.). All analyses were carried out in triplicate on ground dry samples.

2.3.

Chemical fractionation of metals

An experimental approach commonly used for studying the mobility, transport and bioavailability of metals in amended soils is the use of selective sequential extraction procedures. In these techniques, various chemical extractants are sequentially applied to the soil sample, each follow-up treatment having a different and generally more severe chemical action than the previous one. After a critical evaluation of the available literature, the following chemical fractionation procedure, based on the methods proposed by Tessier et al. (1979) and

1428

S CI EN CE OF T H E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

Fig. 1 – Scheme and aerial photograph (insert) of the landfill site showing yearly depositional layers and (○) chronosequence of eternal fronts from which samples were collected.

Quevauviller et al. (1994), was used to extract and quantify metals that are exchangeable (F1), bound to carbonates (F2), bound to Fe–Mn oxides (F3), bound to organic matter (F4) and residual (F5). The sequential extraction procedure was conTable 1 – Chemical properties of the substrates used as landfill covering materials

Texture pH (H2O) CEC a (cmol(+) kg− 1) TOC b (g kg− 1) WEOC c (mg kg− 1) Metals Cu (mg kg− 1) Zn (mg kg− 1) Pb (mg kg− 1) Ni (mg kg− 1) Cr (mg kg− 1) Cd (mg kg− 1)

Excavation soil

Low permeability material

Municipal waste compost

Silt loam 8.2 ± 0.1 7.5 ± 0.2 9.4 ± 0.1 91.0 ± 0.9

Clay loam 7.6 ± 0.1 3.5 ± 0.2 5.3 ± 0.1 28.7 ± 2.5

– 7.6 ± 0.2 – 274 ± 19 20 ± 2 × 103

18.9 ± 0.3 70.3 ± 0.8 52.0 ± 1.8 44.1 ± 3.2 nd d nd

12.4 ± 0.1 66.4 ± 0.9 34.0 ± 1.8 64.7 ± 0.6 33.8 ± 1.7 nd

240 ± 28 647 ± 72 750 ± 105 52 ± 8 81 ± 10 5.0 ± 0.4

Values reported as means ± standard error; metal concentrations in municipal waste compost were determined according to US EPA Method 3050B (1996). a Cation exchange capacity. b Total organic carbon. c Water-extractable organic carbon. d Not detected.

ducted on aliquots of 2 g air-dried and sieved soil in polypropylene centrifuge tubes with screw caps and entailed the following sequentially applied reagents and extraction times: F1: 1 M CH3COONa (1:10, w/v; pH 8.2) for 16 h at room temperature; F2: 1 M CH3COONa (1:10, w/v; adjusted to pH 5.0 with CH3COOH) for 16 h at room temperature; F3: 0.1 M NH2OH·HCl (1:10, w/v; adjusted to pH 2.0 with HNO3) for 16 h at room temperature; F4: 30% H2O2 (adjusted to pH 2.0 with HNO3) at 85 °C, followed by extraction with 1 M CH3COONH4 (1:10, w/v; adjusted to pH 2.0 with HNO3) for 16 h at room temperature; F5: aqua-regia digestion as described above.

2.4.

Plant tissue sampling and analysis

At the time of sampling, the major plant species growing spontaneously on the landfill site included the annual Amaranthus sp., Festuca sp., and Conyza canadensis. Various specimens were randomly collected from depositional layers corresponding to 2, 6 and 10 y after compost application as well as from control, unamended, rural sites. Immediately after harvest, plants were washed with deionised water and separated into roots, stems, and leaves. Plant tissues of different specimens of the same species collected from each depositional layer were pooled together to obtain single representative samples. During

1429

S CI EN C E OF TH E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

sequent to compost addition. These results show that the addition of MWC (pH ≈ 7.6) only slightly influenced the pH of the A horizon, while the pH of the underlying, unamended C1 horizon was similar to that of the original excavation soil (pH ≈ 8.2). Mean aqua-regia extractable metal concentrations for the soils sampled from 2 to 10 y after compost amendment and pertaining to the different soil horizons are reported in Table 2. All metal concentrations in the A horizon were more or less enhanced with respect to the original excavation soil used to cover the landfill (Table 1; evaluated through enrichment factors calculated as the ratio between the concentration of metal in the soil sample to that in the covering material). Over the whole 10-y period, the most abundant metals in this horizon were Cu, Zn and Pb and, to a lesser extent Ni and Cr, whereas Cd was the least abundant. This corresponds to the mean metal concentrations in MWC (Table 1). Most of the metals in the A horizon showed a significant increase in their mean concentration with time after compost application (p b 0.05), except for Ni that showed an initial increase followed by a significant decrease, and Cd. In fact, over the 10-y time period, Cu and Zn experienced a ∼ 3 fold increase while that for Pb was nearly 10 fold. Mean metal concentrations for the C1 horizon (Table 2) generally showed an initial enrichment followed by a decreasing trend with time after compost amendment particularly for Zn, Pb and Cr. Ten years after compost application mean concentrations of Zn, Pb and Ni in the C1 horizon were similar to those obtained for the original excavation soil. According to the calculated enrichment factors (data not shown) soil samples from the C2 horizon were often enriched in metals with respect to both the original low permeability material used in the landfill covering soil (Table 1), as well as the C1 horizon. Even in the C2 horizon, samples generally exhibited an initial relative increase in metal concentration with time, followed by a significant decrease thereafter (p b 0.05), with concentrations of Zn, Pb, Ni and Cr returning to values similar to those obtained for the original low permeability material. However, peak concentrations occurring 6 to 9 y after compost application, were slightly delayed with respect to the C1 horizon.

washing of the roots, rhizospheric soil was collected in a bucket, dried, ground and analysed for aqua-regia extractable metal content as described for the bulk soil samples. Plant tissue samples were dried at 60 °C in a ventilated oven and ground to b0.5 mm. Aliquots (2 g) were subsequently digested with repeated additions of concentrated HNO3 and 30% H2O2 as recommended by the US Environment Protection Agency (1996). The concentrations of Pb, Ni, Cr and Cd in the plant digests were analysed using GF-AAS whereas Cu and Zn were determined by flame AAS. All analyses were carried out in triplicate on ground dry samples. The analytical accuracy of this procedure was confirmed by analysis of a standard reference material (NIST-SRM-1573a; Tomato Leaves). The total amounts of metals recovered were generally within 80% of the total certified concentrations.

2.5.

Statistical analysis

All results of the chemical analysis of soil samples were reported (or plotted) as the average values ± standard error of determinations made on three randomly collected replicate samples. The temporal and spatial variations in the chemical parameters obtained were initially analysed by means of a two-way analysis of variance to evaluate the effects of time after compost application and soil depth. All significant effects were confirmed by means of the least significant difference test (p b 0.05) for comparison of means among or within the appropriate soil horizons. Plant tissue data were also analysed by analysis of variance considering time as the independent variable. Significant differences between means were established by the least significant difference test (p b 0.05).

3.

Results

3.1. Temporal and spatial variations in heavy metal concentrations The values of pH obtained for all soil samples analysed were all included between 7.6 and 8.1. These slightly alkaline pH values were consistent with both soil depth and age sub-

Table 2 – Mean aqua-regia extractable metal concentrations (mg kg− 1) in the soils sampled 2, 6, 9 and 10 y after compost amendment and variation with soil depth Time after compost application (y) Horizon A 2 6 9 10 Horizon C1 2 6 9 10 Horizon C2 2 6 9 10

Cu

Zn

Pb

Ni

Cr

Cd

39.74 40.67 86.65 139.91

c c b a

84.78 150.23 174.93 239.00

d c b a

36.47 52.99 166.70 323.92

d c b a

62.40 70.05 68.88 50.57

c a b d

42.57 39.07 51.26 54.11

b b a a

1.56 1.34 1.10 2.73

b bc c a

34.12 23.33 30.20 26.86

a d b c

103.64 76.54 84.51 77.73

a b ab b

42.63 33.17 32.29 32.80

a b b b

65.93 56.89 68.10 49.06

a b a c

50.29 26.51 41.68 31.69

a d b c

1.69 0.95 0.97 1.88

a b b a

20.48 31.54 32.54 26.26

c a a b

65.49 133.98 88.67 71.42

a a a a

24.53 41.41 34.40 36.82

c a b ab

54.59 70.80 69.05 50.08

b a a c

35.99 36.47 45.35 32.02

b b a c

1.21 1.73 0.71 2.06

c b d a

Different letters indicate significant differences between soil samples within each horizon for the same column (p b 0.05).

1430 S CI EN CE OF T H E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

Fig. 2 – Variations in the distribution of metal fractions with time after compost amendment and soil depth. Dotted reference lines represent metal concentrations in original excavation soil (constituting A and C1 horizons) or low permeability material (constituting C2 horizon).

S CI EN C E OF TH E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

3.2. Temporal and spatial variations in the different heavy metal fractions The distribution of heavy metals among their different chemical forms, evaluated through chemical fractionation, and the percentage contribution of each fraction to the total concentration, may yield important information to elucidate the leaching potential of metals along the soil profile in the long-term. Metal fractionation for the original excavation soil evidenced that the major forms of Cu, Zn and Ni, were represented by the residual fraction (F5; 84%, 93%, and 83% respectively) and to a lesser extent the organic fraction (F4; 16%, 7%, and 17% respectively). Pb was distributed evenly between the residual and organic fractions, while Cr and Cd concentrations were below the limits of detection (b1.0 and 0.1 ppb respectively). Similar results were obtained for the low permeability material where the organically-bound metal fraction (F4) constituted 40%, 10%, 50%, 18% and 8% of the total concentration of Cu, Zn, Pb, Ni and Cr respectively, and Cd was not detected. Fig. 2 shows the results obtained from the chemical fractionation of heavy metals in the landfill covering soil with depth, over the 10-y period subsequent to compost addition. When the sum of the metals extracted in each fraction is compared to the aqua-regia extractable metal concentrations, mean % recoveries were Cu 119, Zn 100, Pb 117, Ni 108, Cr 102 and Cd 97% for soil samples from the A horizon, Cu 101, Zn 97, Pb 109, Ni 105, Cr 95 and Cd 91% for the C1 horizon, and Cu 115, Zn 112, Pb 117, Ni 111, Cr 102 and Cd 100% for the C2 horizon. It is immediately evident that for all the metals analysed, the organically-bound (F4) and residual (F5) forms were the most highly represented fractions. Application of MWC to the superficial layer (A horizon) resulted in an increase in the percentage distribution of Cu, Zn and Cr in the organically-bound fraction (F4) over the 10-y period (Fig. 2). On the contrary, the percentage of total Ni in the F4 fraction remained relatively constant possibly due to the limited contribution of MWC to the concentration of Ni in the A horizon. Although the concentration of Pb in the superficial soil samples was significantly enriched by MWC application (p b 0.05), this horizon experienced a relative decrease in the percentage distribution of this metal in the F4 fraction, suggesting that organically-bound Pb was not the principal form of this metal in MWC. Cd was the only metal that had exchangeable forms (10 to 15% with respect to the total concentration) in the A horizon, particularly in the youngest soils (2 y subsequent to compost addition). Nevertheless, the greatest proportion of the Cd in the superficial soil layer was organically-bound while residual forms were only present in the youngest and oldest soil samples (2 and 10 y after compost application). Among the metals whose concentrations in the C1 horizon were greater with respect to the original excavation soil, Zn showed the lowest percentage distribution of organicallybound forms (F4; b5%). The residual fraction was particularly responsible for the enrichment in Cu, Zn, Ni and Cr for this horizon (Fig. 2). Nonetheless, the absolute values for organically-bound forms of these metals showed an initial enrichment in the youngest soil samples with respect to the original excavation soil, followed by a general decrease over the 10-y

1431

period. Pb which was relatively depleted in this horizon had the highest percentage distribution in the organic fraction (F4; ∼ 45–50%), similar to that obtained for the original excavation soil (∼ 49%). Nevertheless, the absolute concentration of Pb in this fraction tended to decrease with time, suggesting a possible loss of metal from this horizon in an organicallybound form. Even in this horizon, the youngest samples showed the presence of exchangeable forms of Cd, while absolute concentrations of organically-bound metal remained rather constant over time. Chemical fractionation also evidenced an increase in the organically-bound forms (F4) of all metals in the C2 horizon (Fig. 2), particularly during the first 6 y subsequent to compost application. In the later years, organically-bound forms of Cu, Zn, Ni and Cr tended to diminish with time (Fig. 2), resulting in absolute concentrations of metals in this fraction which were significantly depleted with respect to the original low permeability material (p b 0.05). On the other hand, concentrations of Pb in the F4 fraction did not show any significant variations with time. Even residual forms of Zn and Ni exhibited an initial enrichment followed by a decreasing trend to values similar to, or even lower than those originally present in the low permeability material.

3.3.

Water-extractable organic carbon

Considering the high concentration of soluble organic C extracted from MWC (∼20 g kg− 1) with respect to that obtained for the original excavation soil (Table 1), amendment of the landfill covering soil probably resulted in an important input of soluble organic compounds. Fig. 3 shows the variations in the C-normalised concentrations of WEOC in the landfill covering soil with depth along the soil profile over the 10-y period subsequent to compost application. Highest yields of WEOC were obtained for the youngest soil samples from the A horizon (2 y), with concentrations showing a decreasing trend with both time and depth. On the other hand, WEOC concentrations in the C2 horizon showed an increasing trend with time, leading to an increase in WEOC with depth for the oldest soil profile (10 y).

Fig. 3 – Variations in the C-normalised concentration of water-extractable organic C with time after compost amendment and soil depth.

1432 3.4.

S CI EN CE OF T H E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

Bioavailability of heavy metals to plants

Table 4 – Mean total concentrations (mg kg− 1) of metals in the Festuca sp. tissues sampled 2, 6, and 10 y after compost amendment

Throughout the 10-y chronosequence, the concentrations of metals in the rhizospheric soil samples (Tables 3–5) were highly correlated to metal concentrations in the respective superficial layers of the landfill covering soil (Table 2), with r values of 0.929, 0.843 and 0.936 (p b 0.001) for Amaranthus, Festuca and Conyza respectively. As observed for metal concentrations in the amended A horizon, even rhizospheric soils were significantly enriched with respect to the control samples particularly in those metals (Cu, Zn and Pb) most abundant in the MWC. Control rhizospheric soil samples for Festuca (Table 4) were particularly rich in Zn, over and above the metal concentrations that generally characterise uncontaminated sites (17 to 125 ppm; Kabata-Pendias and Pendias, 1992). This was possibly due to some form of contamination and did not allow for proper comparison between the control and landfill site with respect to Zn phytoavailability in Festuca. Tables 3–5 also report the mean total metal concentrations in the tissues of plants sampled from the landfill site. At elevated rhizospheric soil concentrations, Cu concentration in the roots ranged from 8.8 to 69.8 mg kg− 1 and tended to exceed that of control plants, particularly for Amaranthus and Festuca. However, no significant enrichment was observed in the aerial parts of all three plant species (p b 0.05). In both Amaranthus and Festuca (Tables 3 and 4 respectively) an increase in rhizospheric soil Zn concentration corresponded to an increase in Zn content of their roots and, in most cases, also of their leaves. Zn concentrations in the tissues of plants sampled from the landfill site ranged from 59.4 to 349.5 mg kg− 1 in the roots, and 53.3 to 259.3 mg kg− 1 in the leaves. In contrast, a significant enrichment of Zn in the roots or shoot of Conyza was

not observed (Table 5), except for a particularly elevated concentration in the leaves of the plant relative to the 10 y covering layer. Although Pb concentrations in the rhizospheric soil were

Table 3 – Mean total concentrations (mg kg− 1) of metals in the Amaranthus sp. tissues sampled 2, 6, and 10 y after compost amendment

Table 5 – Mean total concentrations (mg kg− 1) of metals in the Conyza canadensis tissues sampled 2, 6, and 10 y after compost amendment

Cu

Zn

Pb

Ni

Cr

Leaves Control 2y 6y 10 y Stems Control 2y 6y 10 y Roots Control 2y 6y 10 y

12.8 5.7 12.7 6.2

Zn a a a a

– 4.8 a 1.7 a –

22.8 13.4 21.4 69.8

Rhizospheric soil Control 32.8 2y 47.3 6y 59.6 10 y 134.9

95.5 103.8 154.1 259.3

Pb c c b a

– 52.5 a 34.3 a –

0.1 2.5 20.4 5.6

Ni b b a b

– 1.2 a 1.4 a –

7.4 11.7 4.0 1.2

ab ab b b

– 9.0 a 3.3 a –

Cr 3.5 4.1 0.3 5.0

ab a b a

– 6.9 a 2.9 a –

Cd 0.9 a nd 2.6 a 0.3 a

– 0.1 a 0.1 a –

b b b a

567.0 173.3 237.3 349.5

a d c b

0.3 11.2 10.0 14.6

c ab b a

8.7 6.0 1.8 4.2

a a a a

3.5 6.1 7.2 22.4

b b b a

0.8 1.3 0.3 1.6

a a a a

c c bc a

859.5 145.8 182.8 618.5

a c b a

65.5 47.2 34.1 316.9

b b b a

23.9 53.0 10.5 17.8

a a a a

37.6 42.9 51.3 59.0

b ab a a

0.2 0.3 0.5 1.4

a a a a

Different letters indicate significant differences between metal concentrations within plant tissue for the same column (p b 0.05).

Cd

Cu

Zn

Pb

Ni

Cr

Cd

a a a a

Leaves Control 2y 6y 10 y

18.1 7.5 12.1 28.2

a a a a

158.8 179.2 110.6 218.6

bc b c a

3.0 5.4 1.1 34.2

b b b a

2.5 10.8 6.6 4.8

b a a ab

2.5 2.7 5.5 7.6

b b ab a

nd 0.4 a 0.3 a 1.0 a

nd 0.8 a 1.3 a 1.7 a

0.2 a 0.3 a nd 0.2 a

Stems Control 2y 6y 10 y

6.6 5.6 4.9 4.0

a a a a

48.6 31.0 29.4 59.5

a ab b a

18.2 2.0 0.7 12.8

a b b a

5.6 2.9 1.3 0.9

a a a a

2.3 1.3 1.4 0.2

a a a a

3.7 1.6 0.7 0.7

a a a a

a a a a

1.7 2.0 5.2 2.5

a a a a

0.3 a 0.4 a nd 0.4 a

Roots Control 2y 6y 10 y

15.6 8.9 11.3 12.3

a a a a

81.6 90.8 59.4 92.5

a a b a

3.3 1.0 5.7 5.1

a a a a

3.9 2.2 3.4 0.3

a a a a

0.1 0.9 3.2 2.6

a a a a

0.8 1.2 0.4 0.5

a a a a

a a a a

35.5 46.5 32.2 30.3

a a a a

0.3 a nd 0.4 a 1.3 a

Rhizospheric soil Control 22.8 c 2y 30.7 b 6y 36.4 b 10 y 84.5 a

125.6 104.4 114.5 224.0

b b b a

16.1 20.8 29.5 189.8

b b b a

17.9 18.9 26.3 35.0

b b a a

23.3 17.0 42.8 43.7

b b a a

nd 0.3 a 0.4 a 0.7 a

Leaves Control 2y 6y 10 y

7.7 5.6 4.6 5.2

a a a a

48.2 80.3 53.3 80.1

b a b a

7.7 14.0 0.3 10.4

a a b a

4.6 0.7 1.8 1.3

a a a a

0.7 2.3 1.3 1.1

Stems Control 2y 6y 10 y

2.6 3.0 1.0 1.0

a a a a

29.5 44.5 37.5 42.0

a a a a

3.2 6.1 0.7 4.5

ab a b a

3.7 2.0 2.4 0.2

a a a a

Roots Control 2y 6y 10 y

7.5 8.8 14.8 20.9

c c b a

47.8 64.8 109.0 127.0

d c b a

8.1 10.9 43.1 213.3

c c b a

6.7 6.8 3.4 2.1

c b bc a

129.9 165.8 158.9 576.3

c b b a

24.3 42.3 19.5 479.6

c b c a

28.1 19.9 29.3 18.6

Rhizospheric soil Control 28.9 2y 60.0 6y 51.0 10 y 170.4

Cu

a a a a

0.4 0.5 0.3 0.3

Different letters indicate significant differences between metal concentrations within plant tissue for the same column (p b 0.05).

Different letters indicate significant differences between metal concentrations within plant tissue for the same column (p b 0.05).

S CI EN C E OF TH E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

generally higher in the landfill with respect to the control sites, metal enrichment in the plant tissues was restricted to the roots of Amaranthus and Festuca samples with concentrations ranging from 10.0 to 213.3 mg kg− 1 and limited translocation to the aerial parts of the plants. Here again Conyza did not generally show any significant enrichment (Table 5). Ni content in the rhizospheric soils of all three plants were relatively depleted with respect to soil samples from the A horizon of the landfill covering soil, and in all three plants analysed, no significant enrichment in Ni was observed in either the roots or shoots with respect to the control samples (p b 0.05). Rhizospheric soil samples from all three plants collected from the landfill site did not generally show any significant differences in Cr and Cd contents with respect to the controls, except for slightly higher Cr concentrations in the samples relative to Festuca and Conyza from the 6 and 10 y landfill covering layers (p b 0.05). Similarly, Cr and Cd concentrations in plant tissues were generally comparable to those obtained for the control. Nevertheless, a slight enrichment in Cr was observed for the roots and leaves of Festuca from the 10 y covering layer and Conyza from the 6 and 10 y layers, which in all cases corresponds to the higher Cr concentration in the rhizospheric soils.

4.

Discussion

It is widely recognised that pH is one of the most important soil properties that determines the solubility of trace metals in soils. This is due to the pH-dependence of the variable charge components of soil, pH-dependent cation selectivity of sorption sites, and hydrolysis and precipitation reactions. The slightly alkaline pH values obtained for the soil samples and the minimal influence of MWC application on pH suggest that compost application should not have greatly influenced the solubility of trace metals. However, considering the particular affinity of metals like Cu, Ni and Cr for DOM (Giusquiani et al., 1992; Planquart et al., 1999; Kaschl et al., 2002; Ashworth and Alloway, 2004), the elevated concentrations of WEOC in the A horizon as a direct result of compost application (Table 1), could have significantly influenced metal mobility in such soil systems. Decreasing WEOC concentrations with time indicate that dissolved organic compounds were subject to mineralisation or potentially mobilised into the underlying mineral horizons with the percolation water. Kaiser et al. (2000) suggest that in strongly structured soils, rapid macropore flow may lead to conservative vertical transfer of DOM due to a lack of interaction with mineral surfaces. This is not the case in this landfill scenario where soil structure, particularly in the earlier years was found to be practically absent, as reported elsewhere (Businelli et al., 2007). Moreover, according to the findings of several authors (Qualls and Haines, 1992; Kaiser et al., 1996; Hoffmann et al., 1998), sorption to the solid phase is the major DOM-eliminating process in deeper soil strata and may thus be responsible for the decreasing concentrations of WEOC with depth in the youngest profile. This implies that, in the soil system under study, the mobilisation of compostderived DOM from the superficial A horizon and sorption to the underlying mineral surfaces could have important implications in the vertical transfer of heavy metals (Hoffmann et al., 1998; Kaschl et al., 2002).

1433

Many investigations on the distribution of metals in relation to depth in the profiles of amended soils have shown that, in the short term, relatively little downward movement of metals occurs below the depth of compost application (Planquart et al., 1999). However, other authors have reported a more pronounced movement of metals within the profile of amended soils. Clemente et al. (2006) have reported that fresh organic wastes, such as unstabilised MWC, are rich in soluble organic compounds which can increase the mobility of metals shortly after the addition of these materials to soil, through the formation of soluble organo-metallic complexes. Moreover, the composition of organic amendments can change with time due to decomposition of organic matter by soil microorganisms, with soluble organo-metal complexes being released into soil solution (Hooda and Alloway, 1994; Kaschl et al., 2002). As expected, the application of compost to the landfill topsoil led to a significant enrichment in metal loadings with respect to the original unamended covering soil, particularly for those metals which were mainly represented in the applied MWC (i.e. Cu, Zn and Pb). Although the metal loadings in the MWC utilised in the landfill covering would generally prohibit its application to rural soils, even the highest metal concentrations observed in the amended A horizon were generally lower than the metal concentrations generally reported for contaminated sites in which organic materials were used for bioremediation purposes (Clemente et al., 2006). Compost application increased the proportion of organically-bound metals in the A horizon, particularly for Cu, Zn and Cr. The significantly higher metal concentrations in the older depositions could hardly have been due to organic matter mineralisation, but rather a result of higher compost doses used in the respective years (Businelli et al., 2007). Whereas metal concentrations in the A horizon were directly influenced by MWC application, their enrichment in the underlying C1 and C2 horizons depended on the mobilisation of metals from the overlying A horizon and the sorption of soluble organo-metal complexes and free metals to the underlying mineral surfaces (Hoffmann et al., 1998; Ashworth and Alloway, 2004). In fact, Kaschl et al. (2002) have shown that over 98% of compostderived Cu and Zn added to a sandy soil column accumulated in the soil, even though they had been added as water-soluble species with a compost extract. The observed enrichment in most metals of the youngest soils samples from the C1 and C2 horizons suggests that compost application resulted in an initial relatively rapid vertical distribution of these metals, possibly as organo-metal complexes. The more pronounced enrichment in the C2 horizon with respect to the overlying C1 horizon, particularly for Cu and Zn (highest enrichment factors), may be explained by a higher adsorption capacity for organo-metal moieties. This is in line with the elevated clay content of the low permeability material constituting this horizon with respect to the silt loam of the C1 horizon (Businelli et al., 2007). In fact, Kaschl et al. (2002) have reported a reduction in the leaching of Cu, Ni, Zn and Cr through compostamended soils with elevated clay content with respect to sandy calcareous soils. Similarly, Gove et al. (2001) reported more rapid leaching of Zn and Cu from lighter textured soils particularly with low organic matter content. The delay in enrichment of the C2 horizon (peak concentrations obtained 6 to 9 y subsequent to compost application) with respect to the overlying C1 horizon,

1434

S CI EN CE OF T H E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

indicates that the mobilisation of metals is not an immediate process but involves various equilibria that control the adsorption and desorption of metals, depends on soil characteristics (e.g. hydraulic conductivity) and climatic conditions (e.g. precipitation, temperature), and also involves biochemical processes responsible for the microbial degradation of autochthonous and compost-derived organic matter. This highlights the importance of long-term experimentation when studying the environmental fate of metals in compost-amended soils on large scales. The subsequent decrease in the metal concentrations with time generally observed for both C1 and C2 horizons, suggests that a portion of the metals initially retained were subsequently further re-mobilised. In fact, the decrease in the percentage distribution of organically-bound Cu, Zn, Ni and Cr with time in the C2 horizon, to absolute concentrations which were depleted with respect to the original low permeability material, seems to suggest that the reduction in metal concentrations was predominantly due to a loss of organo-metal moieties, in particular for Cu and Ni. These results suggest that over the 10 y period after application of MWC to the topsoils, metal mobilisation was primarily influenced by organic matter dynamics, potentially also leading to some extent of metal leaching from the soil profile. The chemical fractionation of Zn evidenced that the slight enrichment observed in soil samples from the C1 and C2 horizons was primarily due to an increase in the content of residual forms. This is attributable to the significant mobility of this metal (Planquart et al., 1999), while the relatively low stability constant for organo-Zn complexes (Ashworth and Alloway, 2004) explains the negligible enrichment in organically-bound forms throughout the soil profile. Gove et al. (2001) have reported that Zn is more susceptible to complexation by fresh biosolid-derived organic matter rather than composted materials, implying that the results obtained in this work cannot exclude an early (b2 y after compost amendment) mobilisation of organically-bound Zn. Among the metals studied in this investigation, Pb is known to be very immobile in soils due to association with soil components (Planquart et al., 1999; Kaschl et al., 2002). In fact, none of the soils samples from the C1 or C2 horizons showed any significant enrichment in Pb with respect to the original excavation soil and low permeability material respectively. Moreover, considering the known affinity of Pb for organic matter ligands (Planquart et al., 1999), the observed depletion in the C1 horizon over the 10-y period seems to suggest that the interaction with compost-derived DOM percolating through this horizon could have led to the mobilisation of Pb originally present in the silt loam covering material. As for Cu and Ni, the affinity of Cr for soluble organic matter may have been responsible for the enrichment in the unamended C1 and C2 horizons. Although the slight increase in Cr concentrations in the C2 horizon for soil samples 6 and 9 y after compost application, is predominantly due to organically-bound forms which could have found their way to the deepest horizon as soluble organo-metal forms, the lack of significant difference between the concentrations of Cr in the oldest samples from the C2 horizon and the original low permeability clay loam, seems to suggest that this metal was not strongly retained by the soil system. Although Cd was relatively enriched throughout the soil profile with respect to the original landfill covering materials,

no reasonable explanation could be provided for the observed variations in its concentrations over the 10-y period. Nevertheless, the significant enrichment in soil samples throughout the profile, as well as the significant presence of exchangeable forms even in the deepest horizon (up to 24% of the total Cd concentrations, Fig. 2), suggests that Cd was rather mobile even though its concentrations were relatively low. Studies on the changes in the bioavailability of trace metals with time have generally shown that heavy metals have their highest bioavailability immediately after application to soil (Rajaie et al., 2006). However, previous soil textural and porosity measurements on the landfill site under investigation have also suggested that, being a relatively young anthropogenic soil, the lack of soil structure may actually be primarily responsible for the inhibition of plant growth particularly in the more recent amended topsoils (Businelli et al., 2007). In fact, at the time of sampling, the youngest fronts were completely void of vegetation while the external fronts of the older depositional layers were dominated by herbaceous species with maximum ground coverage, as well as some shrub species with up to 30% ground coverage. Moreover, metal fractionation results showed that the enrichment in metal loadings of the superficial A horizon was generally accompanied by an increase in organically-bound metals which generally exhibit a low degree of phytoavailability. The relatively fast enrichment of the underlying mineral horizons in metal concentrations suggests that any DOMassociated forms of the metals added to the topsoils were mobilised relatively quickly subsequent to compost application, further reducing the bioavailability potential of metals in the A horizon. These considerations suggest that in such anthropogenic compost amended soils, although topsoils remain relatively enriched in metal concentrations over time, their mobility and bioavailability largely depend on various mechanisms such as organic matter mineralisation, surface adsorption by soil colloids, exchange reactions, chelation, redox reactions, etc. Plant tissue analysis evidenced that metal uptake was mainly restricted to Cu, Zn and Pb, the predominant metals in the compost applied, and generally restricted to the roots. In all cases the three plant species analysed showed different responses to metal uptake and translocation. Over the range of concentrations encountered in the compost amended topsoil, metal phytoavailability was therefore a function of plant species and generally restricted to the underground tissues with limited distribution to upper plant parts. In fact, translocation to the shoot was mainly limited to Zn. Several findings support the high phytoavailability of Zn and its tendency to mobilise throughout the plant in appreciable quantities (Kabata-Pendias and Pendias, 1992; Gigliotti et al., 1996; Planquart et al., 1999). These results are somewhat in contrast with the reduction in the relative transfer of Zn from soils to plants as a result of compost application, reported by Zheljazkov and Warman (2004). Nevertheless, the higher availability of Zn with respect to Cu, agrees well with the findings of Basta and Soal (1999) and Zheljazkov and Warman (2004) that suggest a higher adsorption affinity of Cu to organic compounds with respect to Zn in compost-amended soils. The observed enrichment of root tissues in Pb agrees well with the reported ability of plants to passively take up great amounts of Pb (Kabata-Pendias and Pendias, 1992), resulting in an accumulation of metal in the roots due to low rate of transport to the other parts of the plant.

S CI EN C E OF TH E T OTAL EN V I RO N M EN T 4 0 7 ( 2 0 09 ) 14 2 6–1 43 5

Based on these observations, the spontaneous plants growing on the landfill site are not suitable for metal phytoextraction procedures however, they may be beneficial for phytostabilisation of the metals in the amended topsoil.

5.

Conclusions

The evaluation of long-term distribution, mobility and phytoavailability of heavy metals in a compost amended landfill covering soil may give rise to the following implications: (i) Application of mechanically-separated MWC may be directly responsible for the increase in metal loadings in anthropogenic soils as a direct result of the metal content of the organic materials utilised. However, the greatest proportion of metals applied with the compost is retained in the amended topsoil and the presence of organically-bound forms tends to increase with time. (ii) The significant amounts of compost-derived DOC released from the amended topsoils in the early stages subsequent to compost application, may be responsible for the mobilisation of soluble organo-metal complexes from the amended topsoils and subsequent adsorption to mineral surfaces in the deeper horizons. However, loss of organically-bound forms of particular metals from the underlying mineral horizons, suggests that this same pathway may potentially lead to some extent of metal leaching from the soil system. (iii) In the long-term, compost amendment of topsoil in such anthropogenic soils may significantly support the development of native vegetation with important implications on metal uptake and phytostabilisation. However, since metal phytoavailability is a function of plant species, phytoextraction procedures would generally require growing specialised plant species.

Acknowledgements The authors would like to thank Gesenu S.p.A. for rendering their landfill to complete disposition for the purposes of this research. We also thank R. Calandra and A. Leccese for their assistance with sample collection, and G. Pero for her technical assistance.

REFERENCES Ashworth DJ, Alloway BJ. Soil mobility of sewage sludge-derived dissolved organic matter, copper, nickel and zinc. Environ Pollut 2004;127:137–44. Basta NT, Soal JJ. Bioavailability of heavy metals in strongly acidic soils treated with exceptional quality biosolids. J Environ Qual 1999;28:633–8. Businelli M, Calandra R, Pagliai M, Businelli D, Gigliotti G, Grasselli O, et al. Transformation of a landfill covering amended with municipal waste compost, Perugia, Italy. J Environ Qual 2007;36:254–61. Clemente R, Escolar Á, Bernal MP. Heavy metal fractionation and organic matter mineralisation in contaminated calcareous soil amended with organic materials. Bioresour Technol 2006;97:1894–901.

1435

Gigliotti G, Businelli D, Giusquiani PL. Trace metal uptake and distribution in corn plants grown on a 6-year urban waste compost amended soil. Agric Ecosyst Environ 1996;58:199–206. Giusquiani PL, Gigliotti G, Businelli D. Mobility of heavy metals in urban waste-amended soils. J Environ Qual 1992;21:330–5. Giusquiani PL, Pagliai M, Gigliotti G, Businelli D, Benetti A. Urban waste compost: effects of physical, chemical and biochemical soil properties. J Environ Qual 1995;24:175–82. Gove L, Cooke CM, Nicholson FA, Beck AJ. Movement of water and heavy metals (Zn, Cu, Pb and Ni) through sand and sandy loam amended with biosolids under steady-state hydrological conditions. Bioresource Technology 2001;78:171–9. Hoffmann C, Marschner B, Renger M. Influence of DOM-quality, DOM-quantity and water regime on the transport of selected heavy metals. Phys Chem Earth 1998;23:205–9. Hooda PS, Alloway BJ. Sorption of Cd and Pb by selected temperate and semi-arid soils: effects of sludge application and ageing of sludged soils. Water Air Soil Pollut 1994;74:235–50. Kabata-Pendias A, Pendias H. Trace elements in soils and plants. London: CRC Press; 1992. 365 pp. Kaiser K, Guggenberger G, Zech W. Sorption of DOM and DOM fractions to forest soils. Geoderma 1996;74:281–303. Kaiser K, Guggenberger G, Zech W. Organically bound nutrients in dissolved organic matter fractions in seepage and pore water of weakly developed forest soils. Acta Hydrochim Hydrobiol 2000;28:411–9. Kaschl A, Römheld V, Chen Y. The influence of soluble organic matter from municipal solid waste compost on trace metal leaching in calcareous soils. Sci Tot Environ 2002;291:45–57. Leifeld J, Siebert S, Kögel-Knabner I. Changes in the chemical composition of soil organic matter after application of compost. Eur J Soil Sci 2002;53:299–309. Planquart P, Bonin G, Prone A, Massiani C. Distribution, movement and plant availability of trace metals in soils amended with sewage sludge composts: application to low metal loadings. Sci Tot Environ 1999;241:161–79. Quevauviller P, Rauret G, Muntau H, Ure AM, Rubio R, Lopez-Sanchez JF, et al. Evaluation of a sequential extraction procedure for the determination of extractable trace metal contents in sediments. Fresenius J Anal Chem 1994;349:808–14. Qualls RG, Haines BL. Biodegradability of dissolved organic matter in forest throughfall, soil solution and stream water. Soil Sci Soc Am J 1992;56:578–86. Rajaie M, Karimian N, Maftoun M, Yasrebi J, Assad MT. Chemical forms of cadmium in two calcareous soil textural classes as affected by application of cadmium-enriched compost and incubation time. Geoderma 2006;136:533–41. Stacey S, Merrington G, McLaughlin MJ. The effect of aging biosolids on the availability of cadmium and zinc in soil. Eur J Soil Sci 2001;52:313–21. Tessier A, Campbell PGC, Bisson M. Sequential extraction procedure for the speciation of particulate trace metals. Anal Chem 1979;51:844–51. US Environment Protection Agency. Method 3050B (SW-846): Acid digestion of sediments, sludges and soils, Revision 2, Washington DC; 1996 (Available online: http://www.epa.gov/ sam/pdfs/EPA-3050b.pdf). Wang G, Staunton S. Evolution of water-extractable copper in soil with time as a function of organic matter amendments and aeration. Eur J Soil Sci 2006;57:372–80. Zemberyova M, Bartekova J, Hagarova I. The utilization of modified BCR three-step sequential extraction procedure for the fractionation of Cd, Cr, Cu, Ni, Pb and Zn in soil reference materials of different origins. Talanta 2006;70:973–8. Zheljazkov VD, Warman PR. Phytoavailability and fractionation of copper, manganese, and zinc in soil following application of two composts to four crops. Environ Pollut 2004;131:187–95.