**4.2. Availability**

**Norm of Abrangency As Cd Co Cr Cu Hg Mo Ni Pb Se Zn** 

CETESB† São Paulo State 3.5 <0.5 13 40 35 0.05 <4 13 17 0.25 60

†CETESB [90].

**Table 7.** Limits for heavy metals in soils from São Paulo State and Brazil

‡Resolution # 420/2009 of CONAMA [91].

**Limits (mg kg–1)**

886 Environmental Risk Assessment of Soil Contamination

Availability refers to the amount of a chemical (e.g. heavy metal) in the soil that would be available to be absorbed by plants or other biological receptors (e.g. microorganisms). The availability to plants is called phytoavailability, which is a specific term for this class of organisms. In recent years, the availability has also been called bioavailability. Although bioavailability not yet has a clear and accepted definition [93], it has been used to characterize the availability of heavy metals in the environment [94]. However, the uncertainty about the meaning of bioavailability restricts the expansion of its use and thus contributes to the maintenance of the term availability.

The availability of a heavy metal is given by its available concentration. Quantification of the available concentration is performed by extracting the fraction of the metal that is sufficiently soluble to be absorbed by plants. The extraction of this fraction is performed using chemical extractants that try to simulate the potential for uptake of an element of the soil by the roots of plants. There are several extractants which can be used to extract heavy metals from the soil. The extractants used to assess the availability of heavy metals added to Brazilian soils by the application of sewage sludge to land are usually the same ones used in the evaluation of cationic micronutrients, especially Mehlich 1 and DTPA as the most common. Other extrac‐ tants, as 0.1 mol L–1 HCl and Mehlich 3, have also been used in this assessing. The concentration of a metal in the soil can only be considered available if it is closely and positively correlated with the concentration of the metal in plant tissues. In terms of assessing the contamination, the available concentration can be correlated to the concentration of the metal in raw or processed agricultural product. When there is no correlation or the correlation is poor, it is more appropriate to use the term extractable concentration.

Knowledge of available concentrations is essential for assessing the environmental impact of sewage sludge application to land, since such concentrations represent the amount of heavy metals that could be leached to reach groundwater or could be absorbed by plants and transmitted to the food chain levels until reaching the man. Assessment of the availability of heavy metals in soils amended with sewage sludge was carried out in some field experiments conducted in Brazil. The main results are summarized below.

Oliveira and Mattiazzo [68] found different effects of sewage sludge rates (Table 5) on the availability of some heavy metals evaluated. The sludge was applied for two consecutive years to an Oxisol cultivated with sugarcane, and concentrations of available Cd, Cr, Cu, Ni, Pb and Zn by extractants 0.1 mol L–1 HCl, Mehlich 3 and DTPA were determined in soil samples collected in 0-0.20 m depth one year after each application of the waste. Increase in rates of sewage sludge consistently increased concentrations of Cu and Zn extracted with the three extractants in both evaluations. There was increase in concentration of available Ni only at the second assessment and for sewage sludge treatments. The concentrations of Cd, Cr and Pb were below the LOD of the analytical method (AAS). Variations in concentrations of DTPA-Cu and DTPA-Zn between the control (without application of sewage sludge) and the maximum rate of the waste (110 Mg ha–1) were from 0.70 to 10.70 mg kg–1 and 0.62 to 19.12 mg kg–1, respectively. The higher values of Cu and Zn are considered high and very high, respec‐ tively, for soils of São Paulo state (Table 8), assuming a similarity between the values expressed in mg kg–1 and mg dm–3. However, the authors did not report any negative consequence of these high concentrations to the crop. Positive correlations between the availability of Cu and Zn in soil assessed with three extractants and the concentrations of these metals in different plant components (leaf+1, stalk and juice) were significant when included data from treatments with sewage sludge (three rates), treatment with mineral fertilization and control. However, when included only the sludge treatments, there were few significant correlations, indicating generally low efficiency of extractants to assess the availability of Cu and Zn in soil amended with sewage sludge. The exceptions were significant correlations between 0.1 mol L–1 HCl-Zn and juice-Zn in the two years of evaluation and between Zn extracted by the three extractants and stalk-Zn and juice-Zn only in the second year.


**Table 8.** Interpretation limits for concentrations of cationic micronutrients extracted by DTPA pH 7.3 in soils from São Paulo State, Brazil

In another study with sugarcane, Nogueira et al. [78] applied sewage sludge rates up to 10.8 Mg ha–1 to an Ultisol to supply 100 % of N required by the crop, and after 360 and 720 days they assessed the availability of heavy metals in the layer of 0-0.20 m depth using DTPA. The concentrations of Cd, Cu, Ni, Pb and Zn available increased with increasing sludge rates in the two evaluation periods, reaching maximum values of 0.112, 2.64, 0.47, 2.09 and 7.61, respectively. Concentrations of Cu and Zn are considered very high (Table 8), while the concentrations of Cd, Ni and Pb are above the normal range for soils of São Paulo state (Table 9), indicating contamination. Nevertheless, there was no toxicity in the plant. The extractant used was efficient only for Cd and Zn, since their available concentrations in the soil were highly correlated with the concentrations in leaf with top visible dewlap, stalk and juice. For concentrations of As, Cr and Se, there was no effect of sewage sludge application. All metals intentionally evaluated in this study were detected, probably by the use of inductively coupled plasma mass spectrometry (ICP-MS), which had LODs very low.


Source: Adapted from Abreu et al. [93].

to an Oxisol cultivated with sugarcane, and concentrations of available Cd, Cr, Cu, Ni, Pb and Zn by extractants 0.1 mol L–1 HCl, Mehlich 3 and DTPA were determined in soil samples collected in 0-0.20 m depth one year after each application of the waste. Increase in rates of sewage sludge consistently increased concentrations of Cu and Zn extracted with the three extractants in both evaluations. There was increase in concentration of available Ni only at the second assessment and for sewage sludge treatments. The concentrations of Cd, Cr and Pb were below the LOD of the analytical method (AAS). Variations in concentrations of DTPA-Cu and DTPA-Zn between the control (without application of sewage sludge) and the maximum rate of the waste (110 Mg ha–1) were from 0.70 to 10.70 mg kg–1 and 0.62 to 19.12 mg kg–1, respectively. The higher values of Cu and Zn are considered high and very high, respec‐ tively, for soils of São Paulo state (Table 8), assuming a similarity between the values expressed in mg kg–1 and mg dm–3. However, the authors did not report any negative consequence of these high concentrations to the crop. Positive correlations between the availability of Cu and Zn in soil assessed with three extractants and the concentrations of these metals in different plant components (leaf+1, stalk and juice) were significant when included data from treatments with sewage sludge (three rates), treatment with mineral fertilization and control. However, when included only the sludge treatments, there were few significant correlations, indicating generally low efficiency of extractants to assess the availability of Cu and Zn in soil amended with sewage sludge. The exceptions were significant correlations between 0.1 mol L–1 HCl-Zn and juice-Zn in the two years of evaluation and between Zn extracted by the three extractants

> **Limits Cu Fe Mn Zn** \_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_ mg dm–3 \_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_\_

Low 0.0-0.2 0-4 0.0-1.2 0.0-0.5 Medium 0.3-0.8 5-12 1.3-5.0 0.6-1.2 High 0.9-1.5 13-24 5.1-9.0 1.3-2.3 Very high 1.6-15 25-60 10-50 2.4-15 Toxicity >130

**Table 8.** Interpretation limits for concentrations of cationic micronutrients extracted by DTPA pH 7.3 in soils from São

In another study with sugarcane, Nogueira et al. [78] applied sewage sludge rates up to 10.8 Mg ha–1 to an Ultisol to supply 100 % of N required by the crop, and after 360 and 720 days they assessed the availability of heavy metals in the layer of 0-0.20 m depth using DTPA. The concentrations of Cd, Cu, Ni, Pb and Zn available increased with increasing sludge rates in the two evaluation periods, reaching maximum values of 0.112, 2.64, 0.47, 2.09 and 7.61, respectively. Concentrations of Cu and Zn are considered very high (Table 8), while the concentrations of Cd, Ni and Pb are above the normal range for soils of São Paulo state (Table

and stalk-Zn and juice-Zn only in the second year.

888 Environmental Risk Assessment of Soil Contamination

**Interpretation**

Paulo State, Brazil

Source: Adapted from Abreu et al. [93].

**Table 9.** Limits for heavy metals extracted by DTPA pH 7.3 in soils from São Paulo State and other states in Brazil†.

The availability of heavy metals in soils amended with sewage sludge has also been evaluated for maize crop. Martins et al. [81] found that sewage sludge rates up to 80 Mg ha–1 increased the concentrations of extractable Cu, Ni and Zn by DTPA and Mehlich 3 in a clayey Oxisol with or without lime after a maize cultivation. However, the effect of liming on extractability of these metals was different between the two extractants. Liming reduced the concentrations of extractable Ni and Zn by DTPA, but did not change concentrations of Cu. In the case of Mehlich 3, concentrations of extractable Cu and Zn increased with liming, which was unex‐ pected, because it is well known the fact that the increase in soil pH due to liming generally decreases the availability of cationic micronutrients. Positive correlations between soil Cu or Zn extracted by DTPA and Mehlich 3 and concentrations of Cu and Zn in leaf+4 and shoots of maize were observed, suggesting that extractants were effective in assessing the availability of these micronutrients. In the case of Ni, it has not been possible to establish these relationships because its concentration in plant tissues was below LOD of the analytical method (AAS).

Silva et al. [88] applied sewage sludges from Franca and Barueri municipalities (São Paulo state) in three consecutive crops of maize to a clayey Oxisol. For both sludges, rates were to supply up to eight times the amount of N required by the crop, reaching maximum values of 30 and 64 Mg ha–1 for sludges of Franca and Barueri, respectively. Soil samples were taken from the layer 0-0.20 m depth after each waste application and before each maize sowing for assessing availability of Cu, Mn, Ni, Pb and Zn by DTPA and Mehlich 1. The concentrations of Cu, Ni and Zn extracted by both extractants increased in response to the application of the two sludges in the three maize crops. For Mn, the extractability varied with sludge, crop and extractant. In extraction with Mehlich 1, the concentration of Mn increased in the three crops for Franca sewage sludge and only in the first crop for Barueri sludge. The concentration of Mn extracted by DTPA increased for both sludges, but only in first and second crops. There were no significant changes in the extractability of Pb measured by Mehlich 1 for both sludges in all crops. In contrast, the concentration of Pb extracted by DTPA increased in all crops for Barueri sludge and in the second crop for Franca sludge. In all cases in which concentrations of extractable metals increased, increases were higher for Barueri sludge, probably because it had the highest concentrations of metals and was applied in higher rates. Regardless of sewage sludge, increases in concentrations of extractable metals by both extractants were lower in the third than in the second crop, suggesting that successive applications of sewage sludge reduced the solubility of such elements. The significance of the correlations between the concentrations of Cu, Mn, Ni, Pb and Zn in soil and leaf below the ear and grains varied with sewage sludge and crop for both extractants, indicating that they were not consistently effective in assessing the availability of these metals to maize. The element that had more cases of significance was Zn.

In long term experiments conducted by Oliveira et al. [69], sewage sludge rates up to 10 Mg ha–1 year–1 were applied to two Oxisols (Typic Haplorthox and Typic Eutrorthox) for five years and in each year the soils were cultivated with maize. The concentrations of Cu, Ni and Zn extractable by Mehlich 1 in layer of 0-0.20 m depth increased in both soils after the fifth year of sludge application. The concentration of extractable Pb in this layer increased only in Typic Haplorthox. In layer of 0.20-0.40 m depth, there was an increase in extractability of Ni in both soils, whereas concentrations of extractable Cu, Pb and Zn were increased only in Typic Haplorthox. There was no change in concentration of extractable Mn in any of the layers of both soils. Although the concentrations of extractable Cu, Ni, Pb and Zn increased with sludge application, these increases were not adequately correlated with the accumulation of these metals in the shoot and grains of maize plant, indicating that the Mehlich 1 was ineffective in assessing the availability of such heavy metals in both soils. Although the extractant has been inefficient to evaluate the availability, increased extractability in layer of 0.20-0.40 m depth of Typic Haplorthox suggests that Cu, Ni, Pb and Zn were leached due application of sewage sludge.

The results presented in Galdos et al. [77] also suggest the occurrence of leaching, but not for all metals evaluated. The authors applied sewage sludge to a clayey Oxisol to supply up to double of N requirement of maize by two successive crops. The maximum rates were 21.6 Mg ha–1 in first crop and 20.5 Mg ha–1 in second crop. The sludge was incorporated into soil layer of 0-0.10 m depth. The concentrations of extractable Cu, Ni and Zn by DTPA in soil layers of 0-0.05, 0.05-0.10 e 0.10-0.20 m depth were evaluated 267 days after first application and 179 days after second application. The applications of sewage sludge increased the concentrations of Cu and Ni in the layers 0-0.05, 0.05-0.10 m only in the second evaluation, but did not change concentrations in 0.10-0.20 m layer in any of the evaluations, indicating no leaching, probably because the concentrations of these metals in sludge were low (284.1 and 864.8 mg kg–1 Cu and 41.8 and 35.5 mg kg–1 Ni). In contrast, Zn concentrations increased in all layers, including the layer of 0.10-0.20 m in both evaluations, suggesting that Zn was leached to the layer below the incorporation layer of sewage sludge. The leaching may have occurred because of high concentrations of Zn in sludge (11,364.8 and 1,738.1 mg kg–1). Increased mobility of heavy metals in soils amended with sewage sludge is very concerning because the metals leached can reach groundwater, contaminating it.

### **4.3. Chemical fractions**

in all crops. In contrast, the concentration of Pb extracted by DTPA increased in all crops for Barueri sludge and in the second crop for Franca sludge. In all cases in which concentrations of extractable metals increased, increases were higher for Barueri sludge, probably because it had the highest concentrations of metals and was applied in higher rates. Regardless of sewage sludge, increases in concentrations of extractable metals by both extractants were lower in the third than in the second crop, suggesting that successive applications of sewage sludge reduced the solubility of such elements. The significance of the correlations between the concentrations of Cu, Mn, Ni, Pb and Zn in soil and leaf below the ear and grains varied with sewage sludge and crop for both extractants, indicating that they were not consistently effective in assessing the availability of these metals to maize. The element that had more cases

In long term experiments conducted by Oliveira et al. [69], sewage sludge rates up to 10 Mg ha–1 year–1 were applied to two Oxisols (Typic Haplorthox and Typic Eutrorthox) for five years and in each year the soils were cultivated with maize. The concentrations of Cu, Ni and Zn extractable by Mehlich 1 in layer of 0-0.20 m depth increased in both soils after the fifth year of sludge application. The concentration of extractable Pb in this layer increased only in Typic Haplorthox. In layer of 0.20-0.40 m depth, there was an increase in extractability of Ni in both soils, whereas concentrations of extractable Cu, Pb and Zn were increased only in Typic Haplorthox. There was no change in concentration of extractable Mn in any of the layers of both soils. Although the concentrations of extractable Cu, Ni, Pb and Zn increased with sludge application, these increases were not adequately correlated with the accumulation of these metals in the shoot and grains of maize plant, indicating that the Mehlich 1 was ineffective in assessing the availability of such heavy metals in both soils. Although the extractant has been inefficient to evaluate the availability, increased extractability in layer of 0.20-0.40 m depth of Typic Haplorthox suggests that Cu, Ni, Pb and Zn were leached due application of sewage

The results presented in Galdos et al. [77] also suggest the occurrence of leaching, but not for all metals evaluated. The authors applied sewage sludge to a clayey Oxisol to supply up to double of N requirement of maize by two successive crops. The maximum rates were 21.6 Mg ha–1 in first crop and 20.5 Mg ha–1 in second crop. The sludge was incorporated into soil layer of 0-0.10 m depth. The concentrations of extractable Cu, Ni and Zn by DTPA in soil layers of 0-0.05, 0.05-0.10 e 0.10-0.20 m depth were evaluated 267 days after first application and 179 days after second application. The applications of sewage sludge increased the concentrations of Cu and Ni in the layers 0-0.05, 0.05-0.10 m only in the second evaluation, but did not change concentrations in 0.10-0.20 m layer in any of the evaluations, indicating no leaching, probably because the concentrations of these metals in sludge were low (284.1 and 864.8 mg kg–1 Cu and 41.8 and 35.5 mg kg–1 Ni). In contrast, Zn concentrations increased in all layers, including the layer of 0.10-0.20 m in both evaluations, suggesting that Zn was leached to the layer below the incorporation layer of sewage sludge. The leaching may have occurred because of high concentrations of Zn in sludge (11,364.8 and 1,738.1 mg kg–1). Increased mobility of heavy metals in soils amended with sewage sludge is very concerning because the metals leached

of significance was Zn.

890 Environmental Risk Assessment of Soil Contamination

sludge.

can reach groundwater, contaminating it.

In Brazil, field studies about fractionation of heavy metals in soils amended with sewage sludge are scarce. Among the few studies that exist, one of Nogueira et al. [94] can be considered comprehensive because they evaluated Cd, Pb and Zn in different soil fractions [exchangeable (Exch), organic matter (OM), amorphous Fe oxide (AFeO), crystalline Fe oxide (CFeO ), and residual (Res)] and in soil organic matter (SOM) chemical fractions [fulvic acid (FA), humic acid (HA) and humin (Hum)] after nine annual applications of sewage sludge rates up to 20 Mg ha–1 to a clayey Oxisol cultivated with maize. The Cd concentration in the fraction Res was not altered by application of sewage sludge, and concentrations in other fractions were below the LOD of the analytical method (AAS). Similarly, there was no effect of sludge on concen‐ tration of Pb in fractions AFeO, CFeO and Res, and the concentrations in other fractions were below the LOD. In contrast, the sludge application increased the concentration of Zn in all fractions, mainly in AFeO and CFeO fractions, indicating that a considerable part of the added Zn was adsorbed on these oxides. In the fractions of SOM, Cd and Pb concentrations were not changed in response to sewage sludge application. While the concentration of Zn in Hum fraction was not changed, the concentrations in FA and HA fractions increased by sludge application, indicating that, in SOM, the added Zn is bound to fractions less stable and thus can become more easily available in the environment.

In another study, the redistribution of Ni among humic fractions of a medium-textured Oxisol cultivated with maize and amended with sewage sludge rates up to 20 Mg ha–1 year–1 for six years was assessed by Melo et al. [95]. The authors observed that the application of sewage sludge increased more the proportion of Ni in the humin fraction (Figure 1). This means that the sewage sludge redistributed added Ni preferably to more stable fraction of SOM. As the humin fraction is insoluble in acid and alkaline medium, Ni associated with it can be consid‐ ered unavailable to plants and not directly subject to leaching, which favors its accumulation in soil layer where sewage sludge is incorporated.

#### **4.4. Speciation**

In this work, speciation refers to the separation of heavy metals in different chemical forms and possible oxidation states (e.g., chemical species) in the soil solution. Such metals can be separated basically into free ions, complexes and ion pairs [11, 96]. The participation of each form in the total concentration of a metal in the soil solution can be estimated by using the following general procedures: (i) extraction of soil solution, (ii) determining total concentration of cations, anions and organic compounds in solution and (iii) calculation of the activity of the metals of interest, based on determined concentrations, using specific software for this purpose. Speciation is the basis of the free ion activity model (FIAM) and also other models, which relate the responses of plants to different chemical species of metals present in the soil solution [96].

Plants uptake heavy metals preferably as free ions. Thus, if these ions are increased in the soil solution, so they can be absorbed in excess, and excessive uptake of these metals can cause phytotoxicity and also contamination of the food chain. Indeed, uptake and toxicity of metals in plants generally correlate better with the activity of free ions [11, 96]. On the other hand,

7

Source: Adapted from Melo et al. [95].

Figure 1 **Figure 1.** Nickel associated to soil organic matter chemical fractions in medium-textured Oxisol amended with sewage sludge applied annually during six years and cultivated with maize. Means within each fraction followed by the same letter are not significantly different according to Tukey test (*p* < 0.05). †2.5 Mg ha–1 in 1st, 2nd, and 3rd years and 20 Mg ha–1 in 4th, 5th and 6th years.

other chemical species that are not immediately relevant to the uptake of plants can also cause environmental impact. Heavy metals bound to dissolved organic matter, forming organometal complexes in the soil solution, may have increased their mobility in the soil profile, as demonstrated in experiments with undisturbed soil columns [17], favoring the leaching of these potentially toxic metals, which increases the risk of groundwater contamination

Addition of organic materials to the soil is able to change the availability and mobility of heavy metals in the soil solution, which can minimize some environmental impacts and maximize others. Organic materials added to the soil can decrease the activity of free ions [97] and, consequently, reduce the availability of metals to plants, minimizing the risk of phytotoxicity and contamination of the food chain. In contrast, the addition of organic materials can increase the activity of organo-metal complexes [97] and thus increasing mobility of heavy metals, which become more readily leachable, maximizing the risk of groundwater contamination. Sewage sludge, as an organic material, can also generate these contrasting environmental effects.

As occurs for fractionation, field studies about speciation of heavy metals in soils amended with sewage sludge are also rare in Brazil. The only study found with these characteristics was the Silva's MS thesis [98]. In this work, sewage sludge rates up to 20 Mg ha–1 year–1 were applied for seven years to a clayey Oxisol cultivated with maize. The author evaluated chemical species of heavy metals in the soil solution from the layer 0-0.20 m depth at 22, 26, 37 and 43 months after the seventh application of sludge. Considering all evaluation times, Cd, Cr, Mo, Ni and Pb concentrations in the soil solution were generally below the LOD of the analytical method used (Inductively Coupled Plasma-Optical Emission Spectroscopy–ICP-OES). On the other hand, Cu and Zn were consistently detected. The data of speciation for Cu and Zn show that application of sewage sludge generally decreased the proportion of these metals as free ions (Cu2+ and Zn2+) and increased their proportion as organo-metal complexes [dissolved organic carbon-Cu (DOC-Cu) and DOC-Zn]. The observed decrease in the proportion of Cu and Zn free ions in the soil solution indicates that sewage sludge can restrict the availability and consequently uptake of theses metals by plants, reducing the risk of phytotoxicity and contamination of the food chain. In contrast, increased proportions of COD-Cu and COD-Zn in the soil solution indicate that the sludge can intensify the leaching of Cu and Zn, increasing the risk of groundwater contamination. These results suggest that the risk of environmental contamination by application of sewage sludge on land should be better assessed considering the intensity these two contrasting effects.

### **4.5. Risks of contamination**

other chemical species that are not immediately relevant to the uptake of plants can also cause environmental impact. Heavy metals bound to dissolved organic matter, forming organometal complexes in the soil solution, may have increased their mobility in the soil profile, as demonstrated in experiments with undisturbed soil columns [17], favoring the leaching of

**Figure 1.** Nickel associated to soil organic matter chemical fractions in medium-textured Oxisol amended with sewage sludge applied annually during six years and cultivated with maize. Means within each fraction followed by the same letter are not significantly different according to Tukey test (*p* < 0.05). †2.5 Mg ha–1 in 1st, 2nd, and 3rd years and 20

0 5 10 2.5 + 20†

bc <sup>a</sup>

a

b ab a

Annual sewage sludge rate (Mg ha-1)

c

7

b

Addition of organic materials to the soil is able to change the availability and mobility of heavy metals in the soil solution, which can minimize some environmental impacts and maximize others. Organic materials added to the soil can decrease the activity of free ions [97] and, consequently, reduce the availability of metals to plants, minimizing the risk of phytotoxicity and contamination of the food chain. In contrast, the addition of organic materials can increase the activity of organo-metal complexes [97] and thus increasing mobility of heavy metals, which become more readily leachable, maximizing the risk of groundwater contamination. Sewage sludge, as an organic material, can also generate these contrasting environmental

As occurs for fractionation, field studies about speciation of heavy metals in soils amended with sewage sludge are also rare in Brazil. The only study found with these characteristics was the Silva's MS thesis [98]. In this work, sewage sludge rates up to 20 Mg ha–1 year–1 were applied

these potentially toxic metals, which increases the risk of groundwater contamination

Page 34, Line 20 – Replace "heavy metal" by "heavy metals".

892 Environmental Risk Assessment of Soil Contamination

Fulvic acid Humic acid Humin

<sup>c</sup> bc

d

ab

0

Source: Adapted from Melo et al. [95].

Mg ha–1 in 4th, 5th and 6th years.

3

6

9

Ni in organic matter fractions (mg kg-1)

12

15

effects.

Figure 1

Sewage sludge is an organic material containing heavy metals. Therefore, its application to land generates risks of soil contamination by heavy metals. Although it is difficult to quantify this risk, due to the complex interaction among the factors that determine it, it is possible to mention some steps that can minimize it. Knowledge of these measures is relevant, since they allow the adoption of a more appropriate management for the disposal of sewage sludge on land.

As noted above, research about the impact of sewage sludge agricultural use on soil contam‐ ination by heavy metals has advanced considerably in recent years in Brazil. Field experiments using the total concentration as an indicator of heavy metal contamination have shown interesting results. Sludge rates above 65 Mg ha–1 year–1 can quickly contaminate (approxi‐ mately two years) the soil by Cu, Ni and Zn. Once contaminated with Cu and Zn, soil may remain in this condition for several years, even without additional application of sewage sludge. Zn may be leached in soils amended with relatively high rates of sludge (> 30 Mg ha– 1 year–1). In contrast, rates as low as 10 Mg ha–1 year–1, applied for five years, have low potential for soil contamination by Cu, Ni, Pb and Zn. Single application of sludge rate of this magnitude is unlikely to contaminate soil for As and may be insufficient to cause contamination by Cd. But as the limits for Cd in soils are very low (Table 7), any increase in its total concentration is concerning. Therefore, maximum rates for annual increase of Cd total concentration in soil could be adopted as an optional security criterion to restrict application of sewage sludge on land even though there is no legal restriction to its application.

Studies reviewed in this work also show the usefulness of available concentration to assess soils contaminated by heavy metals. Sewage sludge rates relatively low (~ 10 Mg ha–1 year–1) can contaminate soil by Cd and Zn due to the excessive increase in available concentrations of these metals by DTPA. Similarly, soil contamination by Cu and Zn, indicated by increase in available concentrations by DTPA and Mehlich 3, may occur by applying high sludge rates. Liming reduces the availability of Zn and hence the potential for contamination of the soil, but only DTPA is able of detecting this reduction. Mehlich 1 extractant, which is widely used in routine soil analysis (soil testing) in Brazil, seems to be inefficient to assess whether a soil is contaminated by heavy metals. Likewise, Mehlich 3 and DTPA are not always efficient. The alternative extractant 0.1 mol L–1 HCl has also shown poor performance. As the factors associated to inefficiency of the extractants are variables (e.g., type of heavy metal, soil class, type of crop, cropping sequence and origin of sewage sludge) and they may act in combination, the available concentration can not yet be considered a consistent and relatively safe indicator for assessing the risk of heavy metals contamination in soils amended with sewage sludge.

Nevertheless, extractable concentrations by DTPA and Mehlich 1 may be useful to monitor the leaching of heavy metals in soils amended with sewage sludge. Relatively low sludge rates (~ 10 Mg ha–1 year–1) applied for a long time (5 years) stimulate the leaching of Cu, Ni and Pb, as indicated by increase in extractable concentrations by Mehlich 1 in depth. High sludge rates (~ 20 Mg ha–1 year–1) with high concentration of Zn in its composition (> 1,300 mg kg–1) have the potential to leach Zn from the superficial layers of the soil in a short time (2 years), as suggested by the increase in extractable concentrations by DTPA in depth. Thus, the extractable concentration, used in the monitoring of leaching, can help in assessing the risk of groundwater contamination by these heavy metals added to the soil by sewage sludge application to land.

The results of fractionation of heavy metals, albeit very limited, have shown opposite trends for Zn and Ni. Zn added to the soil by the sludge is preferably bound to Fe and Al oxides. In this form, its solubility is limited and thus the risk of being excessively absorbed by plants or leached is reduced. On the other hand, added Zn which binds to SOM is most often associated with fulvic and humic acids. As these humic substances are poorly stable, its transformation tends to release Zn retained in soil organic matrix, thereby increasing the risk of contamination. In contrast, added Ni which binds to SOM is mainly in humin fraction. Since this fraction is very stable, it restricts the solubility of Ni, minimizing the risk of contamination of crops and groundwater.

Speciation of heavy metals in soils amended with sewage sludge is still incipient. We found only one work on this topic, which is a MS thesis. The results of this thesis, however, have shown significant trends. Applications of sewage sludge rates as low as 20 Mg ha–1 year–1 for seven years declined forms of Cu and Zn (Cu2+ and Zn2+) that are preferentially absorbed by plants, but increased forms of these metals (Cu-DOC and DOC -Zn) that are easily leached. This means that the sewage sludge, when applied in low rates over a long period, decreases the risk of phytotoxicity and contamination of the food chain, but increases the risk of groundwater contamination.

Simultaneous use of different measures to evaluate heavy metals in soils can increase security in the risk assessment of environmental contamination due to the application of sewage sludge to land. From this perspective, total concentration, available concentration, chemical fractions and chemical species of metals should be assessed together to better characterize the dynamics of possible contamination. These different measures represent a gradient of solubility with immediate and potential impact on plant uptake and leaching of heavy metals in soils. Thus, the environmental risk posed by the application of sewage sludge could be predicted for short and long term. In addition, these integrated measures could also be useful for the development of mathematical models to predict the availability of heavy metals as free ions using easily measurable input variables (e.g., total metal concentration, pH and SOM) [96]. Such models would be particularly advantageous in predicting the activity of free ions in response to changes in soil properties by the application of sewage sludge [96]. We did not find studies in Brazil using this integrated approach. The studies summarized in this work enable only superficial considerations, because the data are from independent experiments, and only for Zn, since it was the only metal with results for all measures. Total concentration is shown to be a good measure of soil contamination by Zn, but only at high rates of sewage sludge. Under these conditions, the leaching of metal can be found by the total concentration. For relatively low rates of sludge, however, this measure suggests only low potential for contamination. In both cases, it is not possible to establish a close and reliable relationship between total concentration and toxic effects. Nevertheless, availability and chemical fractions indicate environmental contamination even with application of relatively low rates of sludge. Specia‐ tion, in turn, suggests that the leaching of Zn may be more relevant than its excessive uptake by plants, which increases the risk of groundwater contamination to the detriment of phyto‐ toxicity and contamination of the food chain.
