The concentration and accumulation of metals in plant tissues depend on their availability in the soil solution, and the concentration of metals in the root tissues and in the aerial part increase when their concentration in the soil solution increases. Tolerant species generally accumulate higher concentrations of heavy metals in the root than in the aerial part (Verkleij, et al. 1989) [6] . This indicates that plants growing under these conditions cannot avoid uptake of metals, but limit their translocation. Tolerant species can be characterized according to their relative capacity to absorb, translocate and concentrate metals in the plant, being considered accumulators, indicators and excluders, according to the relative concentrations of metals in the roots and in the aerial part (Baker, 1981) [7] .

The term heavy metal refers to any chemical element that has metallic properties, atomic number greater than 20 and density greater than 5.0 g·cm⁻³ (Kabata-Pendias, 2010) [8] , (Alloway, 2013) [9] . Certain metal with these characteristics can be classified as heavy metal, and does not signify that it is necessarily toxic, and some such as Fe, Mn, Cu and Zn are plant nutrients. Some heavy metals are essential and beneficial, when in adequate concentrations, for the development of plants, when in high concentrations exert toxic effects (Alloway, 2013) [9] .

The known or postulated mechanisms to explain the tolerance of plants to heavy metals are grouped into three main categories with several subdivisions (Kabata-Pendias and Pendias, 1984) [10] : 1) limitation in absorption: a) selective exclusion of the metal in the absorption process, b) metal excretion, and c) root excretion of compounds that decrease element availability, 2) compartmentalization: a) metal retention in the root, vessels or both, b) metal immobilization in the cell wall, and c) immobilization of the metal in the vacuole, and 3) biochemical detoxification: a) production by the plant of sequestering and inactivating compounds of toxic metals, and b) tolerance of enzyme systems activated by metals.

Phytotoxic causes of the effects of excessive Cu(II) have been reported in research in recent years. Cu(II), toxic in many non-tolerant plants, may be associated with disturbance in mitosis (Jiang, et al. 2000) [11] , inhibition of root elongation and damage to root epidermal cell membranes (Ouzounidou, et al. 1995) [12] .

Cr(VI) has been evaluated to be slightly more toxic than Cr(III) to a bush bean variety (Wallace, et al. 1977) [13] . The Cr(III) absorbed by plants is poorly translocated to other parts of the plant (Adriano, 1986) [14] . Pratt (1966) [15] , apud Adriano (1986) [14] , observed that high concentrations of chromium in plant tissues can cause symptoms of toxicity, starting at approximately 5.0 mg·kg⁻¹ for barley, oats and citrus plants and up to 175 mg·kg⁻¹ for tobacco.

Some plants have the characteristic of accumulating extremely high amounts of some heavy metals without suffering adverse consequences. There are cases in which determined plants serve as indicators of the presence of minerals (Cu, Fe, Mn, Pb, Zn and even Ag and Au). It is not known why such plants act as accumulators. There seem to be genetic factors involved as some related species often have this ability, as shown in Table 1.

Plants that have more than 1000 mg·kg⁻¹ of heavy metal in their dry matter are called hyperaccumulators, such as A. bertoloni and A. musale that have 4000

Source: Kabata-Pendias and Pendias, (1984) [10] ; Bereket, (1997) [17] ; Okiemen, (1991) [18] .

mg·kg⁻¹ of Ni in leaves and seeds, and 2500 mg·kg⁻¹ in other tissues (Mishra, 1974) [16] .

The tolerance to excess heavy metals is highly heritable and persists in seeds produced when tolerant plants are cultivated in uncontaminated soil. The “tolerance” character is generally dominant, depending on the element and species, one or more genes are responsible for tolerance (Gerloff, et al. 1983) [19] .

The accumulation of metals in plants will depend on factors such as: the species, variety and organ or part studied (Maeda, et al. 1990) [20] . Cereals, grasses and legumes tend to accumulate less metal than leafy fast-growing plants such as spinach and lettuce (Lake, 1987) [21] .

The plants generally retain most of the heavy metals in their roots. The mobile portion is generally concentrated in the vegetative tissue, and little is translocated to plant reserve organs (Latterel, et al. 1978) [22] . Cd(II) and Zn(II), for example, are intensively translocated to the shoot, whereas Cu(II), Cr(III) and Pb(II) are strongly retained in the root, while Ni(II) it is equally distributed by the plant (Matthews, 1984) [23] , (Lake, 1987) [21] .

The resistance of plants to heavy metals occurs through mechanisms that include immobilization of metal ions in roots and cell wall. The ability to administer, that is, tolerance, in turn, is based on the sequestration of metal ions in vacuoles, bonds with appropriate ligands, such as organic acids, proteins and peptides, and on the presence of enzymes (Garbisu, et al. 2001) [24] .

There are at least two aspects of practical interest: obtaining genotypes capable of growing and producing in soils with high levels of heavy metals and obtaining genotypes in which the toxic metal is not concentrated in the edible part of the animal, including man. The accumulation of heavy metals up to toxic levels for animals and humans in forage and food is an important aspect to be considered. In this aspect, as described by Ortega (1981) [25] , indicating that plants are more resistant to high amounts of heavy metals than animals.

Therefore, to know the contamination, in terms of the effects on plants and the food chain, it is necessary to determine the phytoavailable concentrations of these metals (Leschber, et al. 1985) [26] . In Table 2, the contents of the metals Cd(II), Cr(III) and Pb(II) are presented.

There are studies in the literature, as described by (Alloway, et al. 1993) [27] , which prove a strong relation between soil and environmental factors and the availability of metals “trace elements” for plants. Table 3 shows the concentration considered normal and phytotoxic in soils and plants of heavy metals.

The toxicity of an element must be monitored and verified by the growth rate, productivity, visible symptoms and concentration in the plant tissue (Beckett, 1991) [31] . Some toxicity symptoms caused by heavy metals in plants, are shown in Table 4.

The distribution of toxic elements in the soil profile is variable, due to differences

*Total concentrations of elements considered excessive from a phytotoxic point of view.

CNS―Normal soil content, CCTS―Critical total soil concentration, CNP―Normal plant content (Bowen, 1979)¹ [28] : A―Toxicity (Kabata-Pendias and Pendias, 1992)² [29] and B―Values where the reduction of 10% plant growth (Mcnichol, et al. 1985)³ [30] .

Source: (Kabata-Pendias and Pendias, 1985) [32] .

in the retention capacity of the components, due to the different layers present according to the soil class (Berrow, et al. 1980) [33] . As for the origin, the metals present in the soil can be divided into lithogenic and anthropogenic.

The lithogenic fraction comes from geological sources, such as rock residue or is released during the soil weathering process. The natural content of toxic metallic elements in the soil varies greatly with weathering and the chemical composition of the source material. Although the presence of toxic elements, especially heavy metals, be generalized in soils under natural conditions, human activities, that is, anthropogenic action, somehow end up adding to the soil materials containing these elements, which can reach very high concentrations that compromise the quality of the ecosystem. The main anthropogenic sources of metals in the soil are mining, metal processing, the application of pesticides and fertilizers, urban or industrial sewage sludge, reuse water, etc.

Table 5 presents the total and soluble contents of some toxic chemical elements found in the surface layer of soils.

The availability of chemical elements in soils for plants depends on the process of desorption from the surface of organic and inorganic colloidal particles from the soil to the soil solution, which are physically bound, i.e., electrostatic or chemically bound, which are stronger links.

As described by Hogg, et al. (1993) [47] , verified that the desorption process depends not only on the total labile metal content of the soil, but also on the soil pH, temperature, element concentration and soil/solution contact time. Metals bound to organic matter are rapidly adsorbed, while desorption is slower. Thus, the release tends to be slow or incomplete due to hysteresis, and the inner sphere complexes require greater activation energy in the desorption process (McBride, 1989) [48] .

Regarding the charges derived from complexes, it can be reported that in addition to the structural and proton charges, the charge density of inner sphere complexes can also be defined, which are represented by ionic pairs that have a short-distance bond between the ion and the particle, without interposition of water molecules, that is, it produces specific adsorption effects. The charge density of external sphere complexes, which are represented by ionic pairs having at least one hydration sphere between the ion and the particle, that is, it produces non-specific adsorption effects (Langmuir, 1979) [49] .

The complexity of the nature of soil processes and soil-plant relationships is probably one of the biggest reasons for the existence of many methodologies to assess nutrients and heavy metals available to the plant. Most methods are based on establishing a significant correlation between amounts of metals extracted from the soil and contained in plants (Krishnamurti, et al. 2000) [50] .

The extractors currently used can be grouped into various types according to their properties and mode of action. Chemical extractors are classified as follows: 1) saline CaCl₂, KCl, 2) organic EDTA and DTPA; and, 3) acids, such as 0.1 mol·L⁻¹ HCl solution, Mehlich-1 and Mehlich-3 acid solution.

The Saline extractors have the ability to extract forms that occur in the soil solution and are weakly adsorbed, while acids extract weakly adsorbed metal contents and part of the chemically adsorbed ones in soil organic and inorganic colloids, and organic ones extract metals that are associated with organic forms and carbonates (Sposito, et al. 1982) [51] , (Haddad, et al. 1993) [52] .

Therefore, the objectives of this study are to evaluate the phytoavailability of Cu(II) and Cr(III) in four soils of a basalt-derived toposequence, contaminated with salts of CuSO₄·5H₂O and Cr₄(SO₄)₅·(OH)₂, using corn as an indicator plant, and comparing the efficiency of different metal extraction methods, as follows: 1) distilled and deionized water, 2) Mehlich-1, 3) CaCl₂ 0.01 mol·L⁻¹ and 4) KCl 1.0 mol·L⁻¹, in evaluating the phytoavailability of Cu(II) and Cr(III) for corn plant.

The experiment was carried out in a greenhouse, at the Department of Agronomy, State University of Maringá (UEM), using samples collected from the 0 - 0.2 m layer of four soils of a basalt-derived toposequence in the region of Maringá-PR.

Samples of LATOSSOLO Dystroferric Red (LVdf), NITOSSOLO Red Eutroferric (NVef), CHERNOSSOLO Ferric Clayey (MTf) and VERTISSOLO Hydromorphic Ortic (VGo) were used, as described by (EMBRAPA, 1999) [53] . Since each soil was conditioned in pots with a capacity of 2.5 kg, and treated with salts of CuSO₄·5H₂O and Cr₄(SO₄)₅·(OH)₂ at doses of 0.0, 25.0, 50.0, 100.0, 200.0, 400.0 and 800.0 mg·kg⁻¹, without drainage and three replicates per treatment.

Due to its high acidity, the LVdf samples were corrected with CaCO₃ PA, increasing the base saturation to the cultivation levels in 70%. The amount of CaCO₃ used was 1.16 g·kg⁻¹ of soil, being incubated for 10 days in plastic bags at a temperature of 45˚C and another 14 days at room temperature, that is, before the application of Cu(II) and salts Cr(III).

The samples of the four soils, LVdf, NVef, MTf and VGo, after application of copper and chromium salts, CuSO₄·5H₂O PA and Cr₄(SO₄)₅·(OH)₂ PA, respectively, were subjected to wetting and drying during 45 days, in order to obtain conditions similar to the field. Afterwards, the samples were ground and sieved in meshes of 2.0 mm of mesh opening and conditioned in the respective vases.

The corn variety planted is a hybrid of Dow AgroScience CO32, where four plants were conditioned per vase, leaving two plants per vase, at the time of thinning, that is, one week after planting.

The amounts of distilled water added, initially, corresponded to 60% of the soil water retention capacity. The irrigation of the vases was subsequently monitored by weighing them, replacing the consumed and evaporated water when necessary. The vases were rotated weekly, in order to minimize the effect of environmental variations to which they were exposed.

Several applications of N, P and K were made to better evaluate the symptoms of deficiency, referring to doses of copper and chromium.

At 14 and 21 days of germination, NPK fertilization diluted in water was applied, in the proportion of 10 cm³ of the solution per vase, and the salts in the forms of K₂PO₄ 0.04 mol·L⁻¹ and NH₄NO₃ 0.125 mol·L⁻¹.

At 28 and 35 days after germination, NK fertilization diluted in water was applied, in the proportion of 10 cm³ of solution per vase, and the salts were in the form of KCl 0.1 mol·L⁻¹ and NH₄NO₃ 0.125 mol·L⁻¹.

At 35 days after germination, 10 cm³ of 0.25 mol·L⁻¹ K₂SO₄ solution were applied to the vases for sulfur supplementation.

At 42 days after germination, 10 cm³ of 0.25 mol·L⁻¹ (NH₄)₂SO₄ solution were applied to the vases for maintenance fertilization.

On the 51st day of planting, the aerial part of the corn was harvested (leaves + stalks), being washed and dried in an oven at 65˚C, with air circulation for 72 hours. Subsequently, the material was weighed and ground to evaluate the content of the respective metals Cu(II) and Cr(III), via nitro-perchloric digestion (Sarruge, et al. 1974) [54] .

Considering that some heavy metals often present in extremely low concentrations, 500 mg of plant material were digested with 10 cm³ of nitro-perchloric solution in a 6:1 ratio and recovery of the extract in a volume of 50 cm³ for later analysis.

The roots were also removed from each pot by dismantling the soil and sieved, being washed and dried in an oven at 65˚C, with air circulation for 72 hours, and then ground to evaluate the content of the metals Cu(II) and Cr(III), via nitro-perchloric digestion (Sarruge, et al. 1974) [54] .

The analytical determination of metals in plant tissue was performed by atomic absorption spectrometry (EAA), model GBC932 AA equipment, using air acetylene flame for Cu(II) and acetylene-nitrous oxide for Cr(III) and wavelength 327.4 nm, 1.0 nm slit for Cu and 528.2 nm, 0.5nm slit for Cr.

The fundamental principle of the atomic absorption spectrometry analysis method involves measuring the absorption of the intensity of electromagnetic radiation, coming from a source of primary radiation, produced by gaseous atoms in the ground state. Currently there are options in the equipment that offers different atomizers that can be flame and graphite furnace.

The availability of metals present in the soils was evaluated in function of absorption by maize plants and harvest on the 51st day for soil classes NVef and MTf, intermediate to toposequence, that is, one more weathered and another less weathered.

After corn harvest, the contents of Cu(II) and Cr(III) phytoavailable in the samples (control, 100, 400 and 800 mg·kg⁻¹) were evaluated. The analytical determination of metals in all soil samples was performed by EAA, model GBC932 AA, using air-acetylene flame for Cu and acetylene-nitrous oxide for Cr and wavelength of e 327.4 nm, slit 1.0 nm for Cu and 528.2 nm, 0.5 nm slit for Cr

To evaluate the availability of Cu²⁺ and Cr³⁺ ions in soils, the following extractors were used: 1) distilled and deionized water, 2) Mehlich-1, 3) CaCl₂ 0.01 mol·L⁻¹ and 4) KCl 1.0 mol. L⁻¹. Duplicate determinations were carried out in samples of NVef and MTf soils, representing a more developed soil class, that is, more weathered and another less developed, respectively.

1) Distilled and Deionized Water

5.0 g of soil were stirred with 50 cm³ of distilled and deionized water in a 250 cm³ Erlenmeyer flask for 1 hour at 160 rpm and rest for an equal period. Afterwards, the supernatant was centrifuged at 2000 rpm, for 10 minutes, collecting a 25 cm³ aliquot, kept in a dark flask, for later evaluation of Cu²⁺ and Cr³⁺ ions.

2) Mehlich-1

5.0 g of soil were stirred with 50 cm³ of Mehlich-1 solution in a 250 cm³ Erlenmeyer flask for 1 hour at 160 rpm and rest for the same period. Afterwards, the supernatant was centrifuged at 2000 rpm for 10 minutes, collecting a 25 cm³ aliquot, kept in a dark flask, for later evaluation of Cu²⁺ and Cr³⁺ ions.

3) CaCl₂ 0.01 mol·L⁻¹

5.0 g of soil were stirred with 50 cm³ of CaCl₂ 0.01 mol·L⁻¹ solution in a 250 cm³ Erlenmeyer flask for 1 hour at 160 rpm and rest for the same period. Afterwards, the supernatant was centrifuged at 2000 rpm for 10 minutes, collecting a 25 cm³ aliquot, kept in a dark flask, for later evaluation of Cu²⁺ and Cr³⁺ ions.

4) KCl 1.0 mol·L⁻¹

5.0 g of soil were stirred with 50 cm³ of KCl 1.0 mol·L⁻¹ solution in a 250 cm³ Erlenmeyer flask for 5 minutes at 160 rpm and rest for 12 hours. Afterwards, the supernatant was centrifuged at 2000 rpm for 10 minutes, collecting a 25 cm³ aliquot, kept in a dark flask, for later evaluation of Cu²⁺ and Cr³⁺ ions.

5) Sulfuric Digestion

The medium-sulfuric digestion was carried out in air-dried fine earth, in duplicates in the NVef and MTf classes, respectively. After corn harvest on the 51st day of planting in the samples (control, 100, 400 and 800 mg·kg⁻¹) respectively.

0.5 g of soil was added in 100 cm³ test tubes, together with 20 cm³ of 1:1 H₂SO₄ solution. The material boils for 30 minutes. Afterwards, 1 cm³ of 65% HNO₃ was added, until complete digestion of the organic matter. Afterwards, the samples were heated for another hour at a temperature of 250˚C. Then the solution was transferred to a stock flask, increasing the final volume to 50 cm³.

The Cu²⁺ and Cr³⁺ ions were determined by EAA, as described by (EMBRAPA, 1997) [55] .

The determination of pH was carried out directly in the soil-solution suspension, after carrying out each procedure using the different extractors listed above, with the solutions of Cu²⁺ and Cr³⁺ ions. According to methodology (EMBRAPA, 1997) [55] .

The toposequence soils with the chemical element Cu(II) present at defined concentrations in the experiment, were favorable for dry matter production up to levels of 25 mg·kg⁻¹ for MTf class and 50 mg·kg⁻¹ for VGo. Classes LVdf and NVef, showed loss of production from the first dose, indicating that possibly the soils have high levels of pre-existing Cu(II), whose added values become phytotoxic.

The VGo class was the one that produced more shoot dry matter of corn, in treatments with doses of Cu(II) and Cr(III), due to its own fertility characteristics.

Cr(III) was less phytotoxic at the same doses used for Cu(II), and a corn plant tolerance at levels of approximately 400 mg·kg⁻¹, except for the NVef class, observed at 200 mg·kg⁻¹, indicating metal retention in the roots of maize plants, which is in agreement with Matthews (1984) [23] and Lake (1987) [21] .

Source: (Peternele, et al. 2014) [56] .

Doses above 400 mg·kg⁻¹ of Cu(II) and Cr(III) strongly affect the development, and consequently the production of dry mass of corn, indicating that the levels of toxicity were reached to the point of plant damage. Except for Cr(III) in LVdf and VGo classes with low reduction in dry mass production for the dose of 800 mg·kg⁻¹.

Table 7 shows the amount of corn dry matter produced in different soils and doses of Cu(II) and Cr(III) metals.

It can also be seen in Table 7, the effect of different doses of Cu(II) was highly harmful to the development of the maize plant, in the four soils of the basalt-derived toposequence compared to Cr(III), where the experiments with the chemical element Cr(III), for the same soils showed to be more tolerant.

It can be seen in Table 8 and Table 9, that both Cu(II) and Cr(III) were absorbed by the corn plant, as a function of increasing doses of the respective metals, verifying that Cu(II) was more absorbed and accumulated in the root system and translocated in smaller amounts to other parts of the plant, the same not happening with Cr(III), where there was a smaller absorption by the root system that was retained in it, more specifically in the vacuole, as described by (Garbisu, et al. 2001) [24] .

The greater absorption of Cu(II) by the maize plant verified for the LVdf class, which also produced the smallest amount of dry matter, can be explained by the effect of the liming that this soil underwent before planting, favoring greater absorption.

It can also be observed that the average contents of Cu(II) found in the aerial part of the corn plant for the LVdf, NVef and MTf classes are within the range considered adequate of 6 - 20 mg·kg⁻¹, as per described by (Raij, et al. 1996) [57] , exception to these values were observed in the VGo class in the control treatments, that is, 0.0 dose of Cu(II), whose concentration was from (3.56 mg·kg⁻¹ to

Values represent the average of 3 replicates. (*) Copper outline and (**) Chrome outline.

Aerial part (leaf + culm): represents the average of 3 repetitions. Roots part: represents the average of 2 repetitions.

Aerial part: represents the average of 3 repetitions; Root part: represents the average of 2 repetitions; ND = not detectable in the sensitivity limit of the EAA.

4.56 mg·kg⁻¹) up to the dose of 100 mg·kg⁻¹ of the metal and increasing for the other doses. In general, increasing values of Cu(II) content are observed for all classes of soils in the toposequence as a function of the applied dose, both for the aerial part and for the root system in the corn crop. This can be explained by the difference in the mineralogy of each class and its physicochemical attributes between the classes, where the metal is available in a different way for plants, considering that the organic matter content for the studied soil classes it is between 14.08 and 16.69 g·dm⁻³, that is, values relatively close, as described by Peternele, et al. (2014) [56] .

Table 9 shows the absence of Cr(III) in the aerial part of corn for all soil classes and rates, which was also verified by Anjos and Mattiazzo (2001) [58] .

It can also be observed that Cr(III) presented low mobility in the soil classes of the toposequence, indicating low absorption by the maize plant and still located in the roots, as a natural defense mechanism by the plant, which was also verified by Garbisu, et al. (2001) [24] .

In Table 9, it can also be observed that the soil class that made less Cr(III) available in solution was VGo, indicating that the chemical, physical and mineralogical characteristics of each soil have a strong influence on the retention capacity of these metals.

The evaluation of the phytoavailability of metals to maize plants was evaluated for Cu(II) and Cr(III), for intermediate soil classes of the basalt-derived toposequence, that is, a more developed class (NVef) and a less developed (MTf).

The amounts of Cu(II) and Cr(III), obtained by the extractors of the samples treated with increasing doses of these metals, after harvesting on the 51st day of planting, are shown in Table 10. Considering the amount of metals added to the soil, in form of salts, it is possible to estimate that the average amount of Cu(II) and Cr(III) extracted by the different extractors in the respective treatments, knowing that the control, that is, the soil without addition of metal presents an initial concentration of the respective metals Cu(II) and Cr(III).

It is observed in Table 10 that the total content of metal ions present in the two soil classes can be determined from the sulfur attack experiment, that is, the values obtained in each dose of the experiment compared to the control. It can also be observed that the extractors used in the extraction of the respective Cu(II) and Cr(III) ions for the two soil classes evaluated presented results that do not correspond to a recovery of added metals in the form of salts, that is, considering the total original content of the soil plus that added via salinization in that treatment at different doses. Indicating that part of the metal went to the corn plant and most of it was chemically and physically fixed in clay minerals, whose mineralogy is characteristic of each class. Also observed by Warman, et al. (2000) [59] .

Physical and chemical processes cause metals to be in soluble form, fixed by minerals in the soil, precipitated with other components, in biomass and complexed with some components of organic matter. Thus, a given metal present in the soil solution has its balance related to clay particles, iron oxyhydroxide,

Values represent the average of 2 repetitions. ND―not detected.

aluminum and manganese, in addition to soluble chelators, as described by (Warman, et al. 2000) [59] .

The greater amount of metals extracted in NVef and MTf soils is mainly due to the richness of metals in the source material of these soils using the extractor from digestion with sulfuric acid.

Comparing the two soils of the NVef and MTf experiment, with all extractors, except sulfur attack, it is verified that the extraction with Mehlich-1 removed higher contents of Cu(II) and Cr(III) in treatments at different doses.

Cr(III) was not detected in the extracts from the treatments of the samples of two evaluated soil classes, that is, NVef and MTf using the extractors KCl 1 mol·L⁻¹ solution, deionized distilled water and CaCl₂ solution 0.01 mol·L⁻¹, being detected at low levels when extracted with Mehlich-1, which highlights the importance of soil type in evaluating the efficiency of each extractor.

The pH values of the suspensions for the different extractors are also shown in Table 11. Knowing that most extractors acidify the soil, reproducing the behavior

Values represent the average of 2 repetitions.

of plants in order to absorb the nutrient. It can be observed that there is practically no variation in the pH result of the extractions using the Mehlich-1 extractor.

Table 11 also shows that the difference in pH between the soils also occurred in relation to the metals Cu(II) and Cr(III) for the different extractors, except for the Mehlich-1 extractor. Clearly indicating that the reaction of the soil with the extractor is an important factor in the availability of metals. Thus, it seems convenient that the extractor has the ability to discriminate the effect of pH on these metals availability in different soils.

Most Brazilian soils are acidic, in this condition, the results are not very conclusive in predicting phytoavailability, also observed by Anjos & Mattiazzo, (2001) [58] . The low correlations are due to the fact that the extractors used do not simulate the reactions that occur in the rhizosphere, as descibed by Berton, (2000) [60] .