EVALUATION OF SIKORA INSTEAD OF SMP BUFFER TO ESTIMATE THE POTENTIAL ACIDITY OF BRAZILIAN SOILS

Despite the efficiency of the Shoemaker, McLean, Pratt (SMP) buffer method in estimating soil acidity, the presence of p-nitrophenol and potassium chromate in the solution, both hazardous substances, has caused increasing environmental concerns. The purpose of this study was to test Sikora method (Sikora, 2006) as an alternative to the adapted SMP buffer method, generally used to estimate potential acidity of Southern Brazilian soils. For the test, 21 soils in the South and Cerrado regions of Brazil were sampled. (1) The potential acidity values of these soils range from 35.95 to 4.02 cmolc kg-1 of soil, reflecting a wide acidity variation. The Sikora buffer does not mimic the adapted SMP buffer used in Southern Brazil, since the former has a low ability to distinguish soils with different acidity from each other, probably due to the higher buffer capacity than of the adapted SMP solution.


INTRODUCTION
The SMP buffer, developed by Shoemaker et al. (1961), is one of most commonly used methods to esrtimate soil potential acidity in Brazil, due to its simplicity, efficiency and quickness. SMP solution contains four chemical substances that act as weak bases to buffer the pH (triethanolamine, p-nitrophenol, potassium chromate and calcium acetate), along with calcium chloride, to control the ionic strength. Around five decades ago, the SMP method was adapted to estimate potential acidity in the States of Rio Grande do Sul and Santa Catarina, in southern Brazil (CQFSRS/SC, 2004). The formulation of the adapted SMP buffer (Tedesco et al., 1995) differs slightly from the original SMP buffer proposed by Shoemaker et al. (1961), since the former contains half the amount of calcium acetate of the original solution. The procedure to determine the pH-SMP is analogous to a singlepoint, potentiometric titrimetry, where the SMP solution acts as a base of known concentration and the soil as an acid. Thus, when the soil sample is added to the buffer solution, the pH in the soil suspension decreases due to the neutralization reaction between the soil potential acidity (H + Al) and buffer bases. The final pH value of the soil-waterbuffer suspension is called pH-SMP. However, for a quantitative estimate, the measured pH-SMP must be calibrated against the real potential acidity of the soil (Kaminski et al., 2002). This method is used to estimate the lime requirement of acid soils as well as the potential acidity to calculate the cation exchange capacity at pH = 7.0 (Kaminski et al., 2007).
For very acid soils that contain high amounts of organic matter, the SMP solution is not an ideal method, since frequently the pH-SMP values are lower than the calibrated limit of most mineral soils. This suggests that the buffer capacity of the SMP solution is insufficient to counterbalance the high acidity of these soils, which is due to the several acid functional groups of the organic matter. Another problem is the composition of the SMP solution, which contains two toxic substances: p-nitrophenol and potassium chromate, in which hexavalent chromium is present. Upon conclusion of a pH SMP analysis, these compounds remain in the suspension residues. In most soil laboratories in Brazil, these residues are not adequately treated; instead, they are usually disposed in the environment. Therefore, to avoid environmental damage caused by these hazardous chemicals, it would be better to replace these two substances in the SMP solution, if possible, maintaining their efficiency as buffer.
In the literature, solutions have been described with similar behavior to some of the most commonly used buffers, in which the toxic substances were substituted by species without health or environmental risks, based on the similar buffer capacity of substitute and substituted species (Huluka, 2005;Sikora, 2006;Sikora & Moore, 2008;Wolf et al., 2008). In relation to the SMP buffer, Vaughan (2004) suggested that the compounds imidazole or ethylenediamine could replace p-nitrophenol, while nitric or succinic acids could substitute the chromate salt.
However, since the buffer pH values of SMP and these new buffers are not the same, a lime calibration study (or conversion to SMP buffer pH) is required to establish lime recommendations. Sikora (2006) addressed this problem a little differently, proposing a solution that would mimic the SMP buffer behavior, without requiring a new lime calibration study for soils where this method is already used. This author chose imidazole (pKa = 6.95) to replace p-nitrophenol (pKa = 6.85), due to the similarity of the pKa values. Based on the same criteria, 2-(N-morpholino) ethanesulfonic acid monohydrate (MES) (pKa= 6.10) was selected to replace chromate (pKa=5.81). The functional acidic group for imidazole and MES is a heterocyclic N atom. MES is a zwitterion (from the German zwitter, meaning hybrid), because the molecule has a positive charge in the protonated N atom and a negative one in the sulfonic group. The linearity of response of the new buffer to soil acidity is ensured by the similarity between pKa values of the replacing substances and those of the hazardous chemicals to be substituted. Besides, both MES and imidazole are stable, relatively inexpensive, readily available on the market and neither of the substances is known as hazardous.
Sikora buffer was closely correlated with SMP values for USA soils and, therefore, highly efficient in estimating their soil lime requirement. However, the efficiency for Brazilian soils has not been evaluated so far. The purpose of this study was to test Sikora as an alternative to the SMP buffer for potential acidity determination in soils in the South and Cerrado region of Brazil.

Soils
Twenty one soil samples were collected from the 0-20 cm layer in different physiographic regions of Brazil: eleven from lowlands and seven from highlands of the State of Rio Grande do Sul and three from the Cerrado region. All samples were taken from native areas that had never been limed. After collection, the sample were air-dried, sieved through 2 mm mesh, plant debris manually removed, and chemical and physical analysis applied. Each sample was dispersed in sodium hydroxide solution before determining the particle size distribution; sand was separated by sieving and silt and clay, by sedimentation (Embrapa, 1997). The organic matter content was determined by wet digestion in a digester heating block (Yeomans & Bremner, 1988). Water pH and SMP pH were measured by a pH meter equipped with a combined pH electrode, consisting of a glass electrode and a silver/silver chloride reference electrode. The exchangeable cations Ca 2+ , Mg 2+ and Al 3+ were extracted from the soil using a 1 mol L -1 KCl solution. Al 3+ was determined by titration with a 0.0125 mol L -1 NaOH solution, while Ca 2+ and Mg 2+ were analyzed by atomic absorption spectroscopy (Embrapa, 1997). The results of physical and chemical analysis are given in table 1.

Lime incubation study
Soil samples (1.00 kg) in three replications were moistened to 90 % of the field capacity and then treated with increasing doses of CaCO 3 , equivalent to 0, 50, 75, 100, 125, 150 % of the potential acidity estimated by the SMP method to a target pH of 6.5, before being stored in plastic bags and incubated for 130 days. Weekly, the plastic bags were opened and the soil was revolved. After incubation, soils samples were airdried, ground, sieved through 2 mm mesh and subjected to pH analysis (in water and SMP solution). The pH achieved in water by the incubated soil samples were plotted as a function of the amount of CaCO 3 added. From this data, the potential acidity (measured H + Al) at pH 7.0 was estimated for each soil as the lime requirement to raise the soil pH to a target level of 7.0.  Tedesco et al. (1995).

Buffer pH measurements
The pH measurements were carried out with a pH meter equipped with a combined pH electrode. Before measurements, the pH meter was calibrated with two buffer solutions at pH 4.00 and 7.00. All measurements were made at room temperature, between 21 and 27 o C.
The pH-SMP and pH-Sikora were determined in the unlimed and limed soils (50 % of the potential acidity estimated by SMP method to attain pH 6.5). A slurry of 10 cm 3 of soil and 10 mL of water were agitated with a glass rod and left to stand for 30 min, before determining the pH, i.e. the soil-water pH. Then 5 mL of SMP buffer was added, the mixture stirred with a glass rod, left to stand for 20 min and then stirred again, before determining the new pH value (pH-SMP) (Tedesco et al., 1995). Measurements of Sikora buffer pH are performed similarly, except for the buffer: water: soil proportion, which was 2:1:1.

Statistical analysis
Linear regression was applied to data of pH-Sikora versus pH-SMP of soils with natural potential acidity and soils with potential acidity influenced by liming, separately, considering the SMP-pH as independent variable.
Exponential regression of potential acidity (estimated by wet incubation) was performed as a function of the soil buffer pH measured using SMP-RS/SC buffer (or Sikora buffer) of the limed and unlimed soils.

RESULTS AND DISCUSSION
The values of potential acidity observed within the range of 35.95 to 4.02 cmol c kg -1 of soil reflected a wide acidity variation (Table 1). This can be attributed to the influence of the acidity factors present in these soils, such as variable contents of organic matter and exchangeable aluminum (Kaminski, 1974). The highest organic matter content (82-11 g kg -1 ) and the highest amplitude of variation were observed in the lowland soils of Rio Grande do Sul, where different kinds of soils are found. The organic matter content of the highland soils of Rio Grande do Sul is similar (all > 42 g kg -1 ). The variation in exchangeable Al 3+ contents in the studied soils was wide (0-5 cmol c kg -1 ). The Cerrado soils contained low values of exchangeable Al 3+ (~ 0.2 cmol c kg -1 ). The data (Table 1) indicate that the samples were taken from a representative group of soils with a broad range of physical and Table 2

. Comparison of concentrations used in the soil suspension of the original Sikora buffer components and Sikora buffer in this study
(1) Concentration in the soil suspension. chemical characteristics, mainly with a wide variation in soil acidity levels.
When the buffer solutions (SMP or Sikora) get in contact with a soil sample, a pH reduction is observed, which is related with the potential acidity content of the soil. The obtained values were within the ranges 4.3 to 6.8 for pH-SMP and 4.9 to 7.1 for pH-Sikora (Table 1). A linear relation between pH-Sikora and pH-SMP values of limed and unlimed soils was obtained (Figure 1a), with correlation factor of 0.99, angular coefficient of 0.846 and linear coefficient of 1.350; in other words, the buffer capacity of Sikora was higher than that of the adjusted SMP solution used in this study. These results differ from those obtained with the Sikora buffer for several soils in the USA (Sikora, 2006). This author also obtained a straight line, but with an angular coefficient very close to 1.0 and linear coefficient around zero. However, it is worth mentioning that Sikora used the SMP buffer in its original composition, as described by Shoemaker et al. (1961). The SMP solution used in this study is slightly different, containing half of the calcium acetate amount of the original SMP buffer. Since the anion acetate is the base of acetic acid, its buffering action is higher for soils with pH values near 4.5 (pKa of acetic acid with ionic strength of 0.5 mol L -1 , Martell & Smith, 1977). Therefore, the lower amount of acetate in the SMP buffer used is responsible for the lower buffering capacity when in contact with more acid soils, and is the cause of the lower angular coefficient of the straight line, compared to the analogous result of Sikora (2006) for soils of USA. Figure 1a also shows that a linear relation between pH-SMP and pH-Sikora can be obtained based on only the original soils (not treated with calcium carbonate), where the correlation factor is 0.98, angular coefficient 0.869 and linear coefficient 1.229. On the other hand, considering only the soils after correction of 50 % of potential acidity (limed soils), another linear relation between pH-SMP and pH-Sikora was obtained, with a correlation factor of 0.99, angular coefficient of 0.825 and linear coefficient of 1.465. However, in both cases the linear and the angular coefficients of these straight lines are far from 0.0 and 1.0, respectively, which would be expected if the Sikora buffer mimicked the SMP perfectly. Figure 1b presents the relation between the real potential acidity (H + Al) and the values of pH-SMP and pH-Sikora determined for these soils, showing that in both cases the real H + Al decreases exponentially as a function of pH, but the decrease curve is more accentuated for pH-Sikora than for pH-SMP. In this figure, for both SMP and Sikora buffers, the soil samples with lower real potential acidity (due to liming, mostly) correspond to points located close to each other in the narrow range of higher pH values. However, this effect was much more pronounced in the case of Sikora than the SMP buffer. This behavior in not desirable, since it will increase laboratorial errors in the routine soil acidity estimation performed with Sikora buffer, because a slight variation in the measured Sikora pH will correspond to a great variation in the estimated value of H + Al. Soils with H + Al values > 12 cmol c kg -1 are particularly important, since the Sikora solution has a lower ability to distinguish these soils from each other. Therefore, a higher probability of laboratory error is expected in the estimate of potential acidity of the studied soils with the Sikora than the SMP buffer. To be able to replace the SMP method, a solution with the same bases as Sikora buffer should be used in other concentrations to reduce the buffer capacity and consequently improve the ability to distinguish Brazilian soils with different values of real potential acidity.

CONCLUSION
The Sikora buffer does not reproduce the same soilbuffer pH as the SMP solution normally used to estimate the acidity of soils in Southern Brazil (Rio Grande do Sul and Santa Catarina states).