COMPARISON OF H/AL STOICHIOMETRY OF MINERAL AND ORGANIC SOILS IN BRAZIL 1071

Exchangeable Al has been used as a criterion for the calculation of lime requirement in several Brazilian States. However, the laboratory method with extraction by a 1 mol L-1 KCl solution followed by indirect alkaline titration is not accurate for some Brazilian soils, mainly in the case of soils with high organic matter content. The objective of this study was therefore to evaluate the stoichiometry of H+/Al3+ in KCl soil extracts. The results suggested that organically complexed Al is the main contributor to exchangeable acidity in soils enriched with organic matter. Liming recommendations for organic soils based exclusively on exchangeable Al determined by the NaOH titration method should therefore be revised.


INTRODUCTION
The determination of Al is a permanent concern of soil fertility management.Soil chemists are constantly testing methods that take into account the effective soil exchange capacity, organic matter content, texture, and the pH-dependent charges of various clays minerals (Menzies, 2003).
The titrimetric determination of Al 3+ in 1 mol L -1 KCl soil extract using bromothymol blue as indicator has been widely used in Brazil (Raij et al., 1994;Bernardi et al., 2002).However, in fact, this procedure determines H + in the KCl extract.This is recognized as the exchangeable acidity (Sposito, 1989).In this context, there is therefore no differentiation between the "exchangeable H +" and H + developed by Al 3+ hydrolysis.
From very acid mineral soils with neutral salt solutions, such as KCl, more Al 3+ than H + is extracted (Veith, 1977;Bloom et al., 2005), while in the organic soil horizons, the quantity of extractable H + typically exceeds Al 3+ (Bloom et al., 2005).But Sparks (2003) observed that much of the H + that appears to be exchangeable in organic soils may be due to Al 3+ hydrolysis in the organic matter resulting in H + .
An important tool for the differentiation between H + and Al 3+ is based on the use of the titration/backtitration procedure (Bertsch & Bloom, 1996).This method was first described by Yuan (1959).However, Coscione et al. (1998) showed the lower sensitivity of exchangeable Al 3+ determined by back-titration.Al determination by spectrophotometric methods, such as ICP-OES, is doubtlessly more sensitive and reliable (Bertsch & Bloom, 1996).The purpose of this study was therefore the evaluation of H and Al in 1 mol L -1 KCl extracts to determine whether the theoretical value of 3 of Al hydrolysis predominates in the exchangeable acidity.

MATERIALS AND METHODS
The method for exchangeable Al extraction of most Brazilian soil laboratories is based on the Al exchange from the surface of soil colloids using a non-buffered 1 mol L -1 KCl solution (Embrapa, 1997;Bernardi et al., 2002).A soil:solution mixture, ratio 1:10, was vigorously shaken for 5 min and left to stand overnight.Then a 25 mL aliquot of the supernatant was sampled and Al was indirectly quantified by titration with 0.025 mol L -1 NaOH using Bromothymol Blue as indicator.In this case, each mol of NaOH neutralizes 1 mol H + .Thus, the number of H + mmoles in the 25 mL aliquot is (A x 0.025), where A is the volume (mL) of NaOH consumed in titration.Considering that the initial volume of KCl was 100 mL and the soil mass was 10 g, the result could be expressed as (A x 10) mmol kg -1 or A cmol kg -1 of H + .For further discussions in the text, this type of mass concentration unit will be used, since it expresses the real titration, i.e., H + rather than Al 3+ .However, to clarify the use of titration in Al 3+ routine analysis, the full calculation will be shown here.To express the result in terms of Al 3+ , it must be assumed that each Al 3+ produces 3 H + after complete hydrolysis.Thus, the result would be expressed as (A x 10)/ 3 mmol kg -1 or (A/3) cmol kg -1 of Al 3+ .Since the mmol Al 3+ is equivalent to 3 mmol c Al 3+ , the final result will be expressed as (A x 10) mmol c kg -1 or A cmol c kg -1 of Al 3+ .
For a direct quantification of Al in the KCl extract, another aliquot of the supernatant was analyzed using an inductively coupled plasma atomic emission spectrometer (Perkin Elmer Optima 3000).The operating frequency was 50-60 MHz at a power of 1,500 W with a torch argon flow rate of 15 L min -1 .The selected Al atomic line was at 308.211 nm.A nitrogen gas purge flow rate of 5.0 L min -1 was used throughout the analysis.The sample flow rate was 2.0 L min -1 .In this case, the result was expressed as B mg kg -1 of Al 3+ , considering that the initial volume of KCl was 100 mL and the soil mass was 10 g.Since the mmol of Al is, approximately, 27 mg, the result could be expressed as (B/27) mmol kg -1 or (B/270) cmol kg -1 of Al 3+ .For further discussions in the text, this kind of mass concentration unit will be used, since it expresses the real mass concentration of the element.However, to clarify the use of spectrophotometric determinations in Al 3+ routine analysis, we will show the full calculation.Since the mmol Al 3+ is equivalent to 3 mmol c Al 3+ , the final result will be expressed as (B/9) mmol c kg -1 or (B/90) cmol c kg -1 of Al 3+ .
In the case of soils with high organic matter content, Al extraction by CuCl 2 0.2 mol L -1 was performed according to Bertsch & Bloom (1996).The Al determination followed the ICP-OES procedure described above.

RESULTS AND DISCUSSION
In a first approach, it is possible to arrange the soil samples in two groups (Tables 2 and 3).The first (Table 2) consisted of 18 horizons of mineral soils where the main dominant clay mineralogy is kaolinite and/or iron/aluminum oxy-hydroxides.In this group, the mean H/Al ratio was around 3.4.In addition to the potential titration errors in a routine aluminum analyses, the presence of soluble Mn 2+ and Fe 3+ in KCl soil extract (Table 2) could represent a small contribution of H + through the hydrolysis of these metals, which would explain a positive variation from the expected 3.0 ratio.Our results corroborate the findings of McLean (1976), Veith (1977) and Bloom et al. (2005) that most of the protons measured in the salt extract of mineral soils, i.e., the exchangeable acidity, may come from the hydrolysis of exchangeable Al 3+ .
The second group (Table 3) was composed of organic horizons with a H/Al ratio of around 0.4.In this case, the lower H + concentration suggested that the "exchangeable H + " pool was not the major part of exchangeable acidity in the studied organic soils.This may be explained by the findings of Ross et al. (2008).They showed that the quantity of exchangeable H + may be insignificant at a pH above 4.5.The mean pH of the studied organic soils was around 4.9 (Table 3).
Another fact that corroborates the insignificant contribution of exchangeable H + to exchangeable acidity is related to the stoichiometry of the organic-Table 1. Soil classification and localization of 37 selected soil samples (1) SiBCS: Brazilian Soil Classification System (Embrapa, 2006); Soil Taxonomy (Soil Survey Staff, 1999).

Table 2. Some chemical analyses of 18 mineral soil samples
(1) Extracted by 1 mol L -1 KCL and determined by ICP-OES. (2)Extracted by 1 mol L -1 KCL and determined by acid-based titration.
Table 3.Some chemical analyses of 19 organic soil samples studied (1) Extracted by 1 mol L -1 KCL and determined by ICP-OES. (2)Extracted by 1 mol L -1 KCL and determined by acid-based titration.
bound Al.The high values of CuCl 2-extracted Al compared with those obtained by KCl (Table 3) are generally related to organic-bound Al (Shuman, 1990;Soon, 1993;Barra et al., 2001).In this case, there is a relative preference for hydroxyl-Al ions over Al 3+  (Schnitzer & Skinner, 1963; Bloom et al., 1979).Hargrove & Thomas (1982) showed that as the Al content of the organic matter increased, the effective valence of the adsorbed Al varied from 2.12 to 0.93.Nissinen et al. (1999) and Simonsson (2000) reported binding stoichiometry values ranging from 1.1 and 2.4 in literature.Therefore, the forced hydrolysis of organic-bound Al by alkaline titration should generate lower H/Al ratio values.
The results obtained in the mineral soils showed that most of the soil exchangeable acidity is primarily due to Al 3+ .However, in the case of the studied organic soils, the contribution of hydroxyl Al bound to the organic matter is probably the main pool of exchangeable acidity.The contribution of exchangeable H + does not seem to be relevant for the studied soils.Considering that most authors indicate that the toxic aluminum is Al 3+ and the monomeric Al-hydroxy species (Bloom et al., 2005), all liming recommendations for organic soils based exclusively on exchangeable Al determined by NaOH-titration should therefore be revised.These findings corroborate the studies of Pereira et al. (2005) and Ebeling et al. (2008).They pointed out the need of new analytical routine methods to properly study and monitor the management of Histosols in Brazil.Besides, the relationship between exchangeable acidity determined by acid-based titration and Al determined directly by spectrophotometry (AAS or ICP) could be a useful index to characterize organic horizons.