J Sci Food Agric 1997, 73, 446È454 Plant Availability of Heavy Metals in Soils Previously Amended with Heavy Applications of Sewage Sludge P S Hooda* Department of Geography, Queen Mary and WestÐeld College, University of London, Mile End Road, London E1 4NS, UK D McNulty Biomathematics and Statistics, Scotland, Hannah Research Institute, Ayr KA6 5HL, UK B J Alloway Department of Soil Science, University of Reading, PO Box 233, Reading RG6 6DW, UK and M N Aitken Department of Environmental Science, SAC, Auchincruive, Ayr KA6 5HW, UK (Received 25 July 1996 ; accepted 2 October 1996) Abstract : Plant uptake is one of the major pathways by which sludge-borne potentially toxic metals enter the food chain. This study examined the accumulation of Cd, Cu, Ni, Pb and Zn in wheat, carrots and spinach grown on soils from 13 sites previously amended with sewage sludge. Winter wheat, carrots and spinach were grown consecutively under Ðeld like conditions. The results showed that plant availability of heavy metals di†ered widely among the crop species. The accumulation of Cd, Ni and Zn in the plants showed the greatest increases compared to their background levels. The Cu and Pb accumulation in the plants grown on sludge-amended soils showed only small increases compared to those grown on uncontaminated soils. Multiple regression analysis of various soil properties showed that the surest way to control the accumulation of metals in food plants is by controlling their concentrations in the soils. Furthermore, soils with a non-acidic pH and a clayey texture tended to achieve better control of metal accumulation in food plants compared to those with an acidic reaction and a coarse texture. Metal concentrations in the plants generally correlated well with those extracted from soils in 0É005 M DTPA, 0É05 M EDTA-(Na) , 1 M 2 NH NO and 0É05 M CaCl . The EDTA, however, proved to be a more reliable 4 3 2 and consistent test in predicting the accumulation of metals in the plants. The results also showed that liming soils to pH 7 e†ectively reduced the metal contents in carrots and spinach, but liming to pH 6É5 had little e†ect on metal concentrations in wheat grain. Key words : plant availability, sludge-amended soils, metal contents, wheat, carrot, spinach, soil extractants, pH, soil type and liming. * To whom correspondence should be addressed at : Institute of Environmental and Biological Sciences, Lancaster University, Lancaster, LA1 4YQ, UK. 446 J Sci Food Agric 0022-5142/97/$09.00 ( 1997 SCI. Printed in Great Britain Plant availability of heavy metals in soils INTRODUCTION Currently some 42% of the total sewage sludge produced in the UK is diposed of by application to agricultural land (DoE 1993). With increasing environmental concern, which has outlawed sea disposal by 1998 (CEC 1991), sewage sludge utilisation in agriculture is likely to increase in future years. Sewage sludge contains considerable amounts of N and P. In addition, it may also contain high concentrations of potentially toxic elements (PTEs), particularly heavy metals (eg Cd, Cu, Ni, Pb, Zn). While sewage sludge recycling to agricultural land is generally recognised as a fertiliser resource (Coker et al 1987), the contamination of soils with sludge-borne heavy metals continues to be an area of concern because of their persistence in the soils and their increased uptake by crop plants many years after sludge applications have ceased (McGrath 1987). Plant uptake is one of the major pathways by which sludge-borne PTEs enter the food chain (Chaney 1990). Although a large body of research exists on factors controlling the uptake of these elements by crop plants (see the review by Alloway and Jackson 1991), the conclusions drawn have tended to be based on studies conducted on soils recently treated with sludge. After the terminal sludge application, the soil will gradually establish a new biochemical equilibrium due to the decomposition of sludge-added organic matter and, in many cases, acidiÐcation of the soil. Consequently, the plant availability of sludge-borne elements may change after the biodegradation of sewage sludge in soils (Robertson et al 1982 ; Hooda and Alloway 1993). Liming soils o†ers a means of minimising the risk of food chain contamination by reducing the plant uptake of sludge-borne heavy metals (Williams et al 1987 ; Jackson and Alloway 1991 ; Smith 1994). Several studies, however, have indicated that liming may not always have a signiÐcant e†ect, and that the e†ectiveness of liming could also vary depending on the soil, metal, pH value of the limed soils and crop species (Hemphill et al 1982 ; Pepper et al 1983 ; Kuo et al 1985 ; Eriksson 1989). Furthermore, the e†ect of liming on plant metal uptake from recently sludge-applied soils may also be confounded due to a possible progressive acidiÐcation of the soil during the period of sludge biodegradation. Research on the behaviour of metals in soils which have stabilised after sludge applications is therefore needed to understand the long-term e†ects of sludge disposal on agricultural land. This paper reports studies conducted on 13 sludge-amended soils which had equilibrated in the Ðeld for several years after sludge applications. The speciÐc objectives of the studies were to : (i) compare the uptake behaviour of Cd, Cu, Ni, Pb and Zn by wheat, carrots and spinach ; 447 (ii) examine the relation of various soil properties to metal contents in the plants ; (iii) test the suitability of four soil extractants in predicting metal uptake by crop plants ; and (iv) study the e†ect of liming on plant metal contents. MATERIALS AND METHODS Samples of 13 soils, c 50 kg, previously amended with large quantities of sewage sludge over many years were collected from various sites in the UK. In addition, similar samples of nine uncontaminated soils were also included for measuring the background levels of metals in the food crops. All soil samples were passed through a coarse sieve (\5 mm) and divided into two equal halves before being put into polyethylene tubs. The tubs were kept in the Ðeld in a rural location near Brentwood in Essex (UK). Winter wheat (T riticum aestivum L), carrots (Daucus carota L) and spinach (Spinacia oleracea L) were grown consecutively between 1989 and 1991. All crops received a pre-sowing basal dressing of an inorganic fertiliser supplying 60, 26 and 50 kg N, P and K ha~1, respectively. Of the 13 sludge-amended soils, nine (soil numbers 1È9, Table 1) were acidic in reaction, and one replicate of these acidic soils was limed to pH 7 in 1987. Winter wheat was grown on soils which had been limed two years previously, and the pH of these soils averaged 6É5 when measured after harvesting the wheat crop. Soils which showed a decrease in pH values were re-limed to pH 7 again before sowing of carrots and spinach. All crops were harvested at maturity. The edible parts of wheat (grain), carrot (roots) and spinach (leaves) were thoroughly washed with de-ionised water before being dried in an oven at 65¡C. The samples of plant material were Ðnely ground in a centrifugal ball mill before being digested in concentrated AristaR grade HNO follow3 ing the procedure described in Jackson and Alloway (1991). Soil particle size fractions were quantitatively determined by the pipette method (Day 1965). Soil pH values were measured in water (Avery and Bascomb 1974). The contents of free iron oxides in soils were extracted with sodium dithionate (Avery and Bascomb 1974), and hydroxylamine hydrochloride was used for the extraction of hydrous manganese oxides (Chao 1972). Soil organic matter and cation exchange capacity were measured using the standard procedures described in Hesse (1971). The “plant availableÏ metal concentrations in soils were extracted using 0É005 M DTPA (Lindsay and Norvell 1978), 0É05 M EDTA-(Na) (Anon 1986), 1 M 2 NH NO and 0É05 M CaCl (Alloway and Morgan 4 3 2 P S Hooda et al 448 TABLE 1 Heavy metal contents and selected physico-chemical characteristics of soilsa Soil number Metal concentration (mg kg~1) Cd 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 4É35 10É40 0É71 12É30 5É07 7É30 8É07 3É35 1É57 2É45 2É40 7É40 3É82 0É49 0É73 0É49 0É49 0É52 0É40 0É30 0É46 0É46 Cu Ni 86É4 65É2 199É8 121É4 148É0 49É5 164É9 45É2 72É5 13É1 119É2 44É3 884É0 103É7 41É1 51É6 187É4 53É2 210É5 42É0 68É4 32É8 102É2 40É8 85É0 11É2 30É8 21É8 29É9 21É9 15É6 18É5 19É4 28É1 126É5 39É2 24É2 30É2 14É5 4É8 16É7 8É5 16É5 29É2 Pb Clay (%) OM (%) pH CEC (cmol kg~1) c 31É0 29É8 46É1 44É2 18É0 33É3 26É6 28É0 81É0 45É3 41É5 20É5 19É5 40É9 30É5 28É5 60É5 20É0 48É7 20É3 28É6 32É8 5É6 5É6 4É3 17É5 3É3 5É9 13É8 4É5 11É0 22É5 5É0 4É1 4É6 38É2 5É7 4É5 11É5 18É1 4É7 4É1 3É8 4É0 6É65 6É15 5É11 5É54 6É35 5É50 6É09 5É81 6É40 6É96 7É29 7É04 7É30 6É60 6É04 7É37 7É21 4É42 6É90 7É42 4É70 4É20 17É9 18É7 29É6 41É9 19É4 35É1 42É3 20É0 45É8 48É2 35É5 18É3 22É4 126É2 21É8 22É6 41É8 47É3 35É2 21É5 23É4 21É8 Zn 85É9 221É0 177É5 497É5 902É5 322É5 200É9 485É3 149É7 150É0 107É6 291É6 813É7 1125 63É5 346É0 200É8 462É5 450É5 370É0 42É0 170É0 92É6 260É0 85É0 202É5 75É2 82É3 79É5 128É7 22É3 37É3 44É1 75É8 268É5 158É0 32É5 126É2 36É0 95É2 45É0 182É3 23É5 124É6 FFeO HMnO CaCO L ast sludge 3 (%) (mg kg~1) (%) application 1É55 1É24 1É15 0É57 1É04 2É16 1É12 1É30 3É11 2É66 2É08 4É70 0É46 1É27 1É20 2É09 0É80 2É45 2É21 0É46 1É27 3É07 444É8 315É2 195É6 75É9 128É8 95É6 139É9 91É7 70É4 292É0 95É6 598É7 292É9 86É2 581É9 357É8 370É1 108É8 73É7 329É8 83É2 266É7 1É5 1É9 2É7 41É5 2É8 2É5 5É9 1É9 3É2 4É9 2É9 3É0 64É0 65É2 2É0 26É0 24É9 1É2 2É4 8É2 1É5 1É0 1985 1985 1984 1984 1982 1979 1987 1978 1988 1984 1985 1979 1976 NCSS NCSS NCSS NCSS NCSS NCSS NCSS NCSS NCSS a Abbreviations : OM, organic matter ; CEC, cation exchange capacity ; HMnO, hydrous manganese oxides ; FFeO, free Fe oxides ; NCSS, no history of contamination with sewage sludge. 1985). For total metal concentrations, soil samples were digested in a mixture of HClO and HF. The metal con4 centrations in food and soil samples were determined using Ñame atomic absorption spectrometry or electrothermal atomisation atomic absorption spectrometry for very low metal concentrations. RESULTS AND DISCUSSION The concentrations of metals in many of the soils used in this study are higher than the maximum permissible values under the current European Union regulations (CEC 1986). For this reason the data from this study may not be suitable for considerations of dietary metal intake scenarios. However, sludge-amended soils which have equilibrated in the Ðeld for many years provide a valuable means of studying the soil-plant relationships of PTEs. The soils used in these experiments di†ered widely in physico-chemical properties, metal composition and history of sludge application (Table 1). The results from this study were therefore expected to be of wider applicability and more useful than those stemming from similar studies conducted on a few freshly sludge-amended soils. Metal concentrations in wheat, carrots and spinach The concentration of metals in the dry matter of wheat grain, carrot roots and spinach leaves produced on the sewage sludge-amended and uncontaminated soils are summarised in Table 2. While the concentrations of Cd and Ni in wheat grain grown on the sludge-amended soils were much greater than their concentrations in those produced on the uncontaminated soils, Pb and Zn concentrations in wheat grown on the sludgeamended soils were only marginally higher (Table 2). The observed mean concentrations of 0É32 mg kg~1 for Pb and 47É6 mg kg~1 for Zn in wheat grain from sludge-amended soils, however are common for wheat crops grown on soils not contaminated with sewage sludges (Vigerust and Selmer-Olsen 1986 ; Lubben and Sauerbeck 1991). The mean Cu concentration (3É90 mg kg~1) in wheat grain produced on the sludgeamended soils appeared to be rather lower than its background level (4É17 mg kg~1). The mean concentrations of Cd, Ni and Zn in carrots from the sludge-amended soils were very much higher than those grown on the uncontaminated soils (Table 2). The concentrations of Cu and Pb in carrots produced on the sludge-amended soils were also elevated compared to their background levels, but the increases were Plant availability of heavy metals in soils 449 TABLE 2 Contents of Cd, Cu, Ni, Pb and Zn (mg kg~1 DW) in wheat, carrots and spinach produced on uncontaminated and sludgeamended soils, and their soil-to-plant transfer coefficients Element Uncontaminated soils (n \ 9) Sludge contaminated soils (n \ 13) Mean T ransfer ratioa Mean T ransfer ratioa W heat grain Cadmium Copper Nickel Lead Zinc 0É24 4É17 0É58 0É23 47É62 0É451 0É189 0É051 0É006 0É412 0É68 3É90 2É15 0É32 58É35 0É112 0É031 0É034 0É002 0É133 Carrot roots Cadmium Copper Nickel Lead Zinc 0É63 5É18 2É17 0É33 25É48 1É204 0É249 0É185 0É009 0É278 1É71 7É23 5É28 0É48 41É74 0É350 0É058 0É118 0É004 0É132 0É94 9É48 4É76 0É82 206É0 2É092 1É806 0É462 0É020 2É297 12É76 16É91 9É46 0É95 455É5 1É991 0É156 0É178 0É008 1É216 Spinach leaves Cadmium Copper Nickel Lead Zinc a Transfer ratio \ plant metal concentration/soil metal concentration. relatively small compared to those observed for Cd, Ni and Zn (Table 2). The observed average concentrations of 7É23 mg kg~1 for Cu and 0É48 mg kg~1 for Pb in carrots grown on the sludge-amended soils, however are within the ranges reported for carrots grown on soils not contaminated with sewage sludges (Keefer et al 1986). Spinach leaves grown on the sludge-amended soils accumulated substantially higher metal contents than those grown on the uncontaminated soils. Spinach grown on the sludge-amended soils, on an average, absorbed more than 13-times as much Cd, more than twice as much Zn and Ni and nearly twice as much Cu compared to their background levels in plants on the uncontaminated soils (Table 2). The concentration of Pb in spinach leaves produced on sludge-amended soils showed only a small increase compared to its background level (Table 2). Spinach tissues accumulated the highest concentrations of all Ðve metals, while wheat grain accumulated the lowest concentrations of Cd, Cu, Ni and Pb among the three crops ; the lowest Zn concentration was in carrot roots (Table 2). The sequence of metal concentrations in the plant tissues from sludge-amended soils was Zn [ Cu [ Ni [ Cd [ Pb for wheat grain and carrot roots, and Zn [ Cu [ Cd [ Ni [ Pb for spinach leaves. The concentrations of metals in wheat, carrots and spinach grown on sludge-amended soils in this study clearly showed that metal uptake is plant-species dependent. The Ðndings are consistent with those reported in the literature (Schauer et al 1980 ; Keefer et al 1986 ; Vigerust and Selmer-Olsen 1986 ; Lubben and Sauerbeck 1991 ; Sauerbeck 1991 ; Smith 1994). The metal uptake data also showed that soil-to-plant transfer of metals varied widely among wheat, carrots and spinach. Soil-to-plant metal transfer ratios (ratio of the concentration of a metal in plants to its concentration in the soil) were much higher for spinach than for wheat and carrots, and this was consistent with all Ðve metals (Table 2). This implies that, for a given level of metals in soils, spinach leaves accumulate far greater amounts of metals than wheat grain and carrot roots. Overall, the amounts of Cd, Ni and Zn in the food plants showed the greatest increases compared to their background levels. The concentrations of Cu and Pb in the plants grown on the sludge-amended soils were also elevated, but to a much less extent. This is primarily because of their very low soil-to-plant transfer ratios compared to those of Cd, Ni and Zn (Table 2). Plant uptake is one of the major pathways by which metals from sludge-amended soils enter the food chain. The food-chain plants might absorb enough amounts of heavy metals to become a potential health hazard to consumers. In this context, the results showed that Cd, Ni and Zn pose the greatest threat among the metals studied because the levels of Cu and Pb were either relatively low or did not show any appreciable increase compared to their background concentrations in the food crops. Soil factors a†ecting metal accumulation in plants The inÑuence of various soil properties on metal concentrations in wheat, carrots and spinach was evaluated initially by stepwise multiple regression analyses and involved all the soil variables mentioned in Table 1. Soil total metal concentration was found to be the principal factor controlling metal contents in the plants, and was the only factor which consistently appeared in all models. The inclusion of other soil variables, such as pH, cation exchange capacity (CEC), organic matter, iron and manganese oxides and texture (clay), in the models gave results which were inconsistent and somewhat ambiguous. It is well known that many soil descriptors (eg CEC and texture or CEC and organic matter) are highly correlated and this may be one of the reasons why the multiple stepwise regression gave inconsistent results. To overcome the problems associated with correlated descriptors the soil variables were mapped onto an uncorrelated set of predictors formed by principal component analysis (PCA) of the soil variables. The stepwise regression was repeated using the P S Hooda et al 450 new uncorrelated variables and the regression equation back transformed to the original variables. The back transformed regression equations were then used to identify a small set of independent soil descriptors for inclusion into Ðnal equations. All calculations were performed using Genstat 5 (release 3.1 copyright ; Lawes Agricultural Trust, Rothamsted Experimental Station, UK). The PCA results indicated that total metal, pH and clay content were inÑuential in most stepwise regression equations. Consequently it was decided that the Ðnal regression equation ought to consist of three soil descriptors ie total metal concentration, pH and clay fraction. The multiple regression equations describing the relationships between metal concentrations in plants and these three soil variables are given in Table 3. The results show that the relationship between metal uptake in plants and the soil variables in the regression model is strongly inÑuenced by metal and plant species. For example, this model accounts for 62, 91 and 67%, respectively, of the variation in Cd uptake data for carrots, spinach and wheat (Table 3). Regression coefficients with relatively small standard errors of estimate (SE), ie with a high t-ratio (coefficient/ SE) indicate that a particular variable has a strong e†ect on the uptake of a metal in a plant. The coefficients for total metal concentration generally had the lowest SE values (data not given) among the three soil variables in the model. This means that the concentration of a metal in soils is the principal factor for predicting its concentration in carrots, spinach and wheat. The relatively small regression coefficients (Table 3) with somewhat large SE values for clay fraction, as compared with pH, suggest that the soil clay fraction does not predict metal accumulation in plants as well as pH. Obviously the pH is more important than clay content in regulating the plant availability of these metals in the soils used in this study. The e†ect of organic matter, CEC, free Fe oxides and hydrous Mn oxides on metal accumulation in the plants was less clear. This may be partly due to the large variability in metal concentrations in these soils (Table 1). Had the concentrations of the metals been similar across all soils, these soil variables would possibly have also a†ected metal accumulation in the plants in a more clear manner. These results however are consistent with a study conducted under similar conditions involving soils from 13 long-term Ðeld experiments (Sauerbeck 1991). The author found that organic matter and texture are less important than the metal concentrations in soils and pH in regulating the availability of Cd, Cu, Ni and Zn to several crop plant species. From the above discussion it could be concluded that although metal accumulation in plants is strongly a†ected by their concentrations in soils, metal availability to plants varies with soil type. Plants grown on soils with pH values in the neutral range and a clayey texture tend to accumulate less metals than those grown on soils with an acidic pH and a coarse texture. This means disposal of sewage sludge on clayey textured soils with either a neutral to alkaline pH or combined with lime treatment might help achieve better control of metal accumulation in food plants grown on sludgeamended soils. TABLE 3 Multiple regression equationsa for plant metal content (mg kg~1) in relation to total soil metal concentration (mg kg~1), pH and soil clay fraction (%) Equation Carrots Cd \ 3É44 Cu \ 11É06 Ni \ 5É20 Pb \ 1É164 Zn \ 43É50 ]0É284 ]0É018 ]0É218 ]0É009 ]0É068 Spinach Cd \ 29É43 Cu \ 50É20 Ni \ 6É80 Pb \ 0É601 Zn \ 829 W heat Cd \ 0É990 Cu \ 6É65 Ni \ 0É344 Pb \ 0É712 Zn \ 208 R2 P-value [0É328 pH [1É179 pH [0É450 pH [0É123 pH [3É80 pH [0É017 clay ]0É018 clay [0É097 clay ]0É0004 clay ]0É027 clay 0É622 0É836 0É838 0É582 0É725 0É008 \0É001 \0É001 0É012 0É002 ]1É919 soilÈCd ]0É014 soilÈCu ]0É185 soilÈNi ]0É0016 soilÈPb ]0É342 soilÈZn [3É93 pH [4É83 pH [1É00 pH ]0É011 pH [64É10 pH [0É087 clay [0É133 clay ]0É028 clay ]0É0023 clay [4É82 clay 0É910 0É626 0É723 0É911 0É383 \0É001 0É007 0É002 \0É001 0É045 ]0É183 soilÈCd ]0É002 soil [ Cu ]0É082 soilÈNi ]0É0008 soilÈPb ]0É019 soilÈZn [0É156 pH [0É503 pH [0É051 pH [0É047 pH [24É86 pH [0É003 clay [0É021 clay [0É029 clay [0É0059 clay [0É024 clay 0É672 0É664 0É761 0É759 0É595 0É012 0É013 0É004 0É004 0É025 soilÈCd soilÈCu soilÈNi soilÈPb soilÈZn a Based on data from 13 sludge-amended soils. Plant availability of heavy metals in soils 451 Soil extractants and metal concentrations in plants The concentrations of Cd, Cu, Ni, Pb and Zn in the edible parts of carrots, wheat and spinach were regressed against the concentrations of trace elements extracted with the four chemical reagents (0É005 M DTPA, 0É05 M EDTA-(Na) , 1 M NH NO and 0É05 M 2 4 3 CaCl ). Cadmium extracted with all the four extractants 2 correlated signiÐcantly with Cd contents in each of the crops. However, Cd concentrations in spinach were best predicted by the DTPA and those in carrots and wheat by the EDTA (Table 4). Similarly, Cu contents in the plant tissues were signiÐcantly related to soil Cu concentrations extracted by all the four reagents but DTPA for carrots, NH NO for spinach and CaCl for wheat 4 3 2 gave the best relationships among the four extractants employed (Table 4). For Ni concentrations in plant tissues, all the four soil extractants gave highly signiÐcant relationships but the concentrations were best predicted by the CaCl for carrots, EDTA for spinach and 2 NH NO for wheat (Table 4). 4 3 The Pb concentrations in the plants were best predicted by the EDTA test (Table 4). Soil Pb concentrations extracted in CaCl and NH NO were not found 2 4 3 to be related to Pb contents in the crop plants. This is partly due to the fact that the Pb concentrations extracted by these neutral salt solutions were frequently near the detection limit (0É05 mg dm~3) of the Ñame atomic absorption spectrometer (FAAS) used. It would appear, therefore, that the NH NO and CaCl extract4 3 2 ants may well be suitable for predicting Pb availability to plants if a more sensitive analytical method became available. The Zn concentrations in both spinach and wheat were best predicted by those extracted from the soils with the CaCl solution, only in carrots was the Zn 2 concentration better related to the EDTA-extractable Zn (Table 4). In general, the ability of the extractants to predict plant-available metals depended on the crop species, the metal and extractant used. Overall, the results show that 0É05 M EDTA is a reliable test for predicting metal availability to carrots, spinach and wheat from sludgeamended soils. This was because it was the only extractant which consistently correlated with all Ðve elements in each of the crops and proved to be the best extractant in many cases (Table 4). EDTA and DTPA are the most commonly used soil tests for assessing plantavailable metals, and, because of their greater extracting strength, are likely to be more sensitive to reduced metal solubilities at higher soil pH or to soils low in metal contents. For a speciÐc metal or crop, however 0É05 M CaCl or 1 M NH NO may provide a better 2 4 3 prediction than EDTA or DTPA. For example, the Ni concentrations in carrots and wheat were better predict- TABLE 4 Linear regression equations of the form y \ a ] bx where y represents metal concentration in the edible plant tissues (mg kg~1) and x is the soil metal concentration (mg kg~1) extracted by one of the four extractants used y Cadmium Carrot Spinach Wheat Copper Carrot Spinach Wheat a 0É653 2É41 0É069 5É23 12É67 2É40 SE b SE 0É297 1É91 0É125 0É469 3É92 0É224 0É071 0É672 0É029 0É790 1É81 0É187 0É052 1É93 1É21 x R2 P EDTA DTPA EDTA 0É781 0É734 0É855 \0É001 \0É001 \0É001 0É008 0É543 0É218 DTPA NH NO 4 3 CaCl 2 0É769 0É493 0É752 \0É001 0É004 \0É001 Nickel Carrot Spinach Wheat 1É50 4É52 0É453 0É770 1É37 0É137 3É48 0É301 1É52 0É307 0É047 0É080 CaCl 2 EDTA NH NO 4 3 0É914 0É763 0É973 \0É001 \0É001 \0É001 L ead Carrot Spinach Wheat 0É327 0É698 0É179 0É097 0É089 0É061 0É002 0É004 0É002 0É0008 0É0007 0É0005 EDTA EDTA EDTA 0É433 0É711 0É653 0É009 \0É001 0É002 0É130 16É86 2É60 0É026 5É59 0É494 EDTA CaCl 2 CaCl 2 0É663 0É403 0É728 \0É001 0É012 \0É001 Zinc Carrot Spinach Wheat 26É62 299É10 37É78 5É04 56É50 5É42 a Based on data from 13 sludge-amended soils. 452 ed by 0É05 M CaCl and 1 M NH NO , respectively, 2 4 3 compared to the EDTA and DTPA (Table 4). Similarly, soil Ni concentrations extracted by neutral salts such as CaCl (0É05 M), MgCl (1 M) and NaNO (1 M) were 2 2 3 found to be better correlated with Ni contents in 13 crop species compared to other extractants (EDTA, DTPA, NH OAc, CuCl ) (Sauerbeck and Hein 1991). 4 2 Indeed, metals extracted with a weaker extractant such as 0É05 M CaCl or water would represent the most real 2 plant-available metals in soils. These extractants, however, have the disadvantage of extracting metals in very small concentrations which are often too low for FAAS, and are difficult to determine by ETAAS because of the chloride matrix interference and possible contamination problems (Ure 1990). As a result, stronger extractants such as DTPA and EDTA have been most widely used (Alloway and Jackson 1991). In this study, 0É05 M EDTA-(Na) was found to be a better and 2 P S Hooda et al consistent predictor of plant-available metals among those tested. E†ect of liming on metal contents in the plant tissues Figure 1(A) shows the e†ect of liming on soil pH values. The application of lime to soils signiÐcantly increased soil pH values. The e†ects of liming on metal concentrations in wheat, carrots and spinach are summarised in Figs 1(B)È(F). Liming the soils to pH 7 signiÐcantly reduced the concentrations of all Ðve metals in carrot roots (Figs 1(B)È(F)). The metal concentrations in carrots grown on the limed soils were lower, on average, by 39% for Cd and Ni, 44% for Zn, 48% for Pb and 28% for Cu compared to their concentrations in plants grown on the unlimed soils. Liming the soils to pH 7 also decreased metal contents in spinach leaves. As with Fig 1. E†ects of liming on soil pH values and accumulation of metals in wheat grain, carrot roots and spinach leaves. The error bars represent the least signiÐcant di†erence (LSD, P O 0É05) between the two sets of data. Plant availability of heavy metals in soils carrots, the metal concentrations in spinach leaves produced on the limed soils were reduced by 39% for Cd, 27% for Cu, and more than 40% for Ni and Zn (Figs 1(B)È(F)). The Pb concentration in spinach leaves, however showed a relatively small reduction of 19% due to liming (Fig 1(E)). Winter wheat was grown on soils which had been limed two years previously, and the pH of these soils averaged 6É5 when measured after harvesting the wheat crop. This pH was still statistically higher than that of unlimed soils (Fig 1(A)), but had no consequential e†ect on the contents of Cd, Cu, Ni and Zn in wheat grain (Figs 1(B)È(F)). In contrast, however the wheat grain produced on the limed soils had signiÐcantly lower contents of Pb than those on the unlimed soils (Fig 1(E)). Liming sludge-amended soils to pH P 6É5 is often recommended to minimise plant uptake of PTEs. The reduction in metal concentrations in carrots and spinach due to liming the soils is consistent with the Ðndings reported in the current literature (Eriksson 1989 ; Jackson and Alloway 1991 ; Lubben and Sauerbeck 1991 ; Smith 1994), but the liming had no e†ect on metal contents in wheat grain, except for Pb. The results indicate that while liming soils to pH 7É0 was e†ective in reducing metal accumulation in carrots and spinach, liming to pH 6É5 had little e†ect on metal accumulation in wheat grain. This infers that either pH 6É5 was not high enough to have a substantial e†ect on metal uptake or that metal accumulation in wheat grain was independent of soil pH. Wheat straw samples were not analysed for their metal contents in the present study. As a result, it is difficult to say whether liming to pH 6É5 was not sufficient enough to have a signiÐcant reduction in wheat metal contents or metal translocation from wheat vegetative parts to grain was independent of soil pH. Results from a recent study, however suggested that metal contents in reproductive plant parts, such as grain, were independent of soil pH. For example, while liming sludge-amended soils over a pH range of 3É9È7É6 substantially reduced Cd concentrations in potatoes, ryegrass and oat straw, it had no e†ect on Cd contents in oat grain (Smith 1994). The results from the present investigation together with other recent studies (Smith 1994) indicate that liming is less likely to control metal accumulation in the seeds of cereal crops such as wheat and oat grain. The Council of the European Communities has Ðxed mandatory concentrations for PTEs in soils receiving sewage sludge (CEC 1986). The maximum permissible metal concentrations in the 1986 EC Directive on sludge regulation (CEC 1986) apply to soils with pH in the range of 6É0È7É0. Where sewage sludge is applied to soils with pH values \6É0, Member States are required to reduce the maximum permissible concentrations accordingly to counterbalance the increased availability of metals to crop plants. In the UK, lower metal limits apply for Cu, Ni and Zn where sludge is applied to soils 453 with pH values \6É0 compared to pH band 6É0È7É0 (Statutory Instrument 1989). However, the maximum concentration of Cd (3 mg kg~1) and Pb (300 mg kg~1) in the EC Directive of 1986 are permitted in sludgeamended soils with a pH value as low as 5É0 (Statutory Instrument 1989), without taking account of their potentially increased bioavailability at lower pH values (\6É0). The results in this study clearly demonstrate an increased uptake of heavy metals at lower soil pH values. This is particularly signiÐcant with regard to Cd and Pb which potentially can accumulate in humans following ingestion of contaminated foods. The current UK Regulations for both these metals take no account of the soil pH and are at the highest level of the maximum range permitted in the European Union regulations (CEC 1986). It would, therefore, be prudent to base maximum Cd and Pb loading values on the soil pH (as is the case with Cu, Zn and Ni). REFERENCES Alloway B J, Jackson A P 1991 Behaviour of trace metals in sludge-amended soils. Sci T otal Environ 100 151È176. Alloway B J, Morgan H 1985 The behaviour of Cd, Ni and Pb in polluted soils In : Contaminated Soils, eds Assinnk J W & van den Brink W J. Martinus Nijho† Publishers, Dordrecht, pp 101È113. Anon 1986 T he Analysis of Agricultural Materials. Reference Book No 27, Ministry of Agriculture Fisheries and Food. HMSO, London. Avery B W, Bascomb C L (Eds) 1974 Soil Survey Methods. Soil Survey Technical Monograph No 6, Harpenden : Soil Survey of England and Wales. CEC 1986 Council Directive of 12th June 1986 on the protection of the environment and in particular of the soils, when sewage sludge is used in agriculture. Official J European Communities, Directive 86/278/EEC, No L 181/6-12. CEC 1991 Urban Waste Water Treatment Directive. Official J European Communities, Directive 91/271/EEC, No L 375/18. Chaney R L 1990 Public health and sludge utilisation. Biocycle 30 68È73. Chao T T 1972 Selective dissolution of Mn oxides from soils with acidiÐed hydroxylamine hydrochloride. Soil Sci Soc Am Proce 36 764È768. Coker E G, Hall J E, Carlton-Smith C H, Davis R D 1987 Field investigations into the manurial value of lagoonmatured digested sewage sludge. J Agric Sci 109 467È478. Day P R 1965 Particle fractionation and particle size analysis. In : Methods of Soil Analysis, ed Black C A. American Society of Agronomy, Madison, Wisconsin, pp 545È567. DoE 1993 Sludge Uses in 1990È91, UK Report to the Commission under Directive 86/275/EEC. DoE, HMSO, London. Eriksson J E 1989 The inÑuence of pH, soil type and time on adsorption and uptake by plants of Cd added to soils. W ater Air Soil Pollut 48 317È335. Hemphill D D Jr, Jacobson T L, Martin L W, Kiemnec G L, Hanson D, Volk V V 1982 Sweet corn response to application of three sewage sludges. J Environ Qual 11 191È196. Hesse P R 1971 A T extbook of Soil Chemical Analysis, John Murray, London. 454 Hooda P S, Alloway B J 1993 E†ects of time and temperature on the bioavailability of Cd and Pb from sludge-amended soils. J Soil Sci 44 97È110. Jackson A P, Alloway B J 1991 Bioavailability of cadmium to lettuce and cabbage in soils previously treated with sewage sludge. Plant and Soil 132 179È186. Keefer R F, Singh R N, Horvath D J 1986 Chemical composition of vegetables grown on agricultural soils amended with sewage sludge. J Environ Qual 15 146È152. Kuo S, Jellum E J, Baker A S 1985 E†ects of soil type, liming and sludge application on zinc and cadmium availability to Swiss chard. Soil Sci 139 122È130. Lindsay W L, Norvell W A 1978 Development of DTPA test for zinc, iron, manganese and copper. Soil Sci Soc Am J 42 421È428. Lubben S, Sauerbeck D R 1991 The uptake and distribution of heavy metals by spring wheat. W ater Air Soil Pollut 56È57 239È247. McGrath S P 1987 Long-term studies of metal transfer following the application of sewage sludge. In : Pollutant T ransport and Fate in Ecosystems, eds Coughtrey P J, Martin M H & Unsworth M H. Blackwell ScientiÐc, Oxford, pp 301È317. Pepper I L, Bezdicek D F, Baker A S, Sims J M 1983 Silage corn uptake of sludge applied zinc and cadmium as a†ected by soil pH. J Environ Qual 12 270È275. Robertson W K, Lutrick M C, Yuan T L 1982 Heavy applications of liquid-digested sludge on three Ultisols : 1. Soil fertility. J Environ Qual 11 450È454. P S Hooda et al Sauerbeck D R 1991 Plant element and soil properties governing uptake and availability of heavy metals derived from sewage sludge. W ater Air Soil Pollut 57È58 227È237. Sauerbeck D R, Hein A 1991 The nickel uptake from di†erent soils and its prediction by chemical extractions. W ater Air Soil Pollut 57È58 861È871. Schauer P S, Wright W R, Pelcha J 1980 Sludge-borne heavy metal availability and uptake by vegetable crops under Ðeld conditions. J Environ Qual 9 69È73. Statutory Instrument 1989 United Kingdom Statutory Instrument No 1263. The Sludge (Use in Agriculture) Regulations. HMSO, London. Smith S R 1994 E†ect of soil pH on availability of metals in sewage sludge-treated soils. II. Cadmium uptake by crops and implications for human dietary intake. Environ Pollut 86 5È13. Ure A M 1990 Methods of analysis for heavy metals in soils. In : Heavy Metals in Soils, ed Alloway B J. Blackie, Glasgow, pp 40È80. Vigerust E, Selmer-Olsen A R 1986 Basis for metal limits relevant to sludge utilisation. In : Factors InÑuencing Sludge Utilisation Practices in Europe, eds Davis R D, Haeni H & LÏHermite P. Elsevier Applied Science Publishers, London, pp 26È41. Williams D E, Vlamis J, Pukite J, Corey J E 1987 Metal movement in sludge-amended soils : a nine year study. Soil Sci 143 124È131.