CORRELATIONS BETWEEN HOT CALCIUM CHLORIDE-EXTRACTED BORON AND CHEMICAL AND PHYSICAL ATTRIBUTES OF SOME BRAZILIAN SOILS

CORRELAÇÕES ENTRE BORO SOLÚVEL EM CLORETO DE CÁLCIO 0,01 mol L-1 A QUENTE E ATRIBUTOS QUÍMICOS E FÍSICOS DE ALGUNS SOLOS BRASILEIROS

CORRELATIONS BETWEEN HOT CALCIUM CHLORIDE-EXTRACTED BORON AND CHEMICAL AND PHYSICAL ATTRIBUTES OF SOME BRAZILIAN SOILS

Luís Reynaldo Ferracciú Alleoni^1,3*; Otávio Antonio de Camargo^2,3; José Maria Aires Silva Valadares²

¹Depto. de Solos e Nutrição de Plantas - ESALQ/USP, C.P. 09 - CEP: 13418-900, Piracicaba, SP.

²Centro de Solos e Recursos Agroambientais - IAC, C.P. 28 - CEP: 13020-902, Campinas, SP.

³ Bolsista do CNPq.

*e-mail: [e-mail]

ABSTRACT: Relationship between soluble boron extracted with a hot CaCl₂ 0.01 mol L^-1 solution and pH; organic carbon; exchangeable cations; cation exchange capacity (CEC); base saturation; total, free and amorphous iron and aluminum oxide contents; clay content; and specific surface area were performed for surface and subsurface samples of five soils. The soils were a Rhodic Hapludox, an Arenic Paleudalf and three Typic Hapludox, all representative soils of the State of São Paulo, Brazil. To quantify the relations between soluble boron and the different soil characteristics, simple linear correlations and multiple regressions, using a stepwise regression program, were performed. Hot-CaCl₂ extractable boron (HCB) was significantly correlated with clay content (r = 0.69^*), specific surface area (r = 0.68^*), CEC (r = 0.63^*) and total aluminum oxides (r = 0.70^*) in all five soils. In addition, there was a correlation between HCB and organic carbon (r = 0.75^*) in the four Oxisols. The correlation coefficient between the product (carbon x clay) and soluble boron contents was also highly significant (r = 0.78^**). Multiple regression analysis showed that total aluminum oxide, as well as exchangeable calcium and aluminum, were correlated with HCB, explaining 85% of the variation. The product (carbon x clay) took into account the effect of textural gradient and showed high positive correlation with hot-CaCl₂ 0.01 mol L^-1extractable boron.

Key words: soluble boron, tropical soils, simple correlation, multiple regression analysis

CORRELAÇÕES ENTRE BORO SOLÚVEL EM CLORETO DE CÁLCIO 0,01 mol L^-1 A QUENTE E ATRIBUTOS QUÍMICOS E FÍSICOS DE ALGUNS SOLOS BRASILEIROS

RESUMO: Teores de boro solúveis em CaCl₂ 0,01 mol L^-1 a quente, de amostras superficiais e subsuperficiais de dois Latossolos argilosos, dois Latossolos de textura média e um Podzólico Vermelho-Amarelo do Estado de São Paulo, Brasil, foram correlacionados com alguns de seus atributos químicos e físicos. As correlações entre boro solúvel e superfície específica (r = 0,68^*), argila (r = 0,69^*), CTC (r = 0,63^*) e alumínio total (r = 0,70^*) dos cinco solos e com o teor de carbono orgânico dos quatro Latossolos (r = 0,75^*) foram estatisticamente significativas. O coeficiente de correlação entre o produto (carbono x argila) e os teores de boro solúvel também foi altamente significativo (r = 0,78^**). Na análise conjunta, alumínio total, cálcio trocável e alumínio trocável foram as variáveis que se correlacionaram significativamente com boro solúvel, explicando 85% da variação. O produto (carbono x argila) levou em conta o efeito do gradiente textural dos solos e apresentou alta correlação positiva com os teores de boro solúvel em CaCl₂ (0,01mol L^-1) a quente.

Palavras-chave: boro solúvel, solos tropicais, correlações simples, regressão múltipla

INTRODUCTION

Boron is an element that occurs in rocks, combined with oxygen, in trigonal coordination, forming borates. In soil, organic matter sequesters most of the boron bioavailable to plants. After mineralization, the element is liberated to the soil solution, where it can be absorbed by plants, lost by leaching or adsorbed to soil colloids (Berger & Pratt, 1963).

Boron content varies a great deal, depending on the types of minerals present in the soil, as well as the various types of organic matter (OM) and clays. Many authors observed that total soil boron content does not constitute an adequate parameter to predict the bioavailability of this nutrient for plants, indicating better correlations between soluble boron in hot water and the concentration of the element available to the plants (Brasil Sobrinho, 1965; Parker &Gardner, 1982). The content of soluble boron in hot water from various soils from the São Paulo State varied from 0.06 to 0.32 mg dm^-3 for the A horizon, from 0.07 to 0.29 mg dm^-3 for the B horizon and from traces to 0.12 mg dm^-3 for the C horizon (Brasil Sobrinho, 1965). These values can be considered within the limits quoted in the literature (Parker & Gardner, 1982; Correa et al., 1985; Jin et al., 1988), although not all researchers used the same extraction technique to obtain solutions containing boron. Better results have been obtained with dilute calcium chloride (from 0.01 to 0.02 mol L^-1), because this salt minimizes problems caused by OM and by clay dispersion (Bataglia & Raij, 1990; Adams et al., 1991).

A variety of soil attributes affects the solubility of boron in soils. Water-soluble boron has been reported to be influenced by soil pH, specific surface area, clay and organic carbon contents (Gupta, 1968). Boron concentrations are usually higher in clayey and loamy soils than in sandy soils (Evans & Sparks, 1983). Soil OM plays an important role in boron availability because it can minimize boron leaching from the soil, and maintains the element in a relatively available form (Jin et al., 1988).

Most of the literature that uses multiple regression analysis shows a significant effect between pH or organic carbon and extracted boron (hot water or a centimolar calcium chloride solution). Organic matter accounted for 69% of the variability of the boron extracted in hot water in thirty-three Egyptian soils at various depths (Elseewi & Elmalky, 1979). A similar effect due to organic matter was obtained in Spanish soils, and Ruiz-Neto et al. (1988) concluded that the organic fraction of the ninety samples had a more significant effect on the soluble boron content than the inorganic fraction, although the clay content also appeared in the group analysis. Elrashidi & O'Connor (1982) observed that none of the simple correlations with oxides were significant, but in a group analysis the total iron and aluminum oxide contents, plus the organic carbon content, represented 98% of the variation of soluble boron content.

The skill of relating attributes to each other is common, mainly when these newly created parameters can be used in group analysis. Miljkovic et al. (1966) tested the correlations between the products of organic carbon and pH (C x pH) or clay content (C x clay) and available boron, and observed an increase in the determination coefficient when compared to the correlations of boron and C, pH or clay, individually.

Following the chemical and physical characterization of four Brazilian Oxisols and one Alfisol, simple and multiple correlations were performed, with the objective of determining which attributes were best correlated with hot CaCl₂ (0.01 mol L^-1)-extracted boron (HCB) contents.

MATERIAL AND METHODS

Surface (0-0.20 m) and subsurface (in the highest expression of B horizon) samples were taken from five representative soils from São Paulo State, Brazil (TABLE 1). In the Oxisols, the depth of the B horizon sampled was between 0.60 to 0.80 m, and in the Alfisol this depth was from 0.75 to 1.00 m.

Chemical analysis, particle size and mineralogical characterization of the samples were performed following the procedures described in Camargo et al. (1986). Hot CaCl₂ (0.01 mol L^-1) soluble boron (HCB) was determined by the azomethine-H method (Bingham, 1982). For specific surface area, the EGME method was used (Cihacek & Bremner, 1979). Instead of sample pre-treatment with phosphorus pentoxide (P₂O₅) to remove water, samples were dried in an oven at 110 ^oC for 24 h (Ratner-Zomar et al., 1983). With the objective of grouping the effects of the carbon and clay contents, the parameter carbon x clay, which is the result of the multiplication of these properties, was calculated. Simple correlations and the multiple regression analysis were obtained using SAS (Statistical Analysis System) analytical program.

RESULTS AND DISCUSSION

A great heterogeneity among chemical and physical attributes of the soil samples was observed. As expected, the Oxisols exhibited great textural homogeneity between the sampled A and B horizons (TABLE 2).

Except for the A horizon of the LE₁, where the soluble boron content was high, the samples exhibited low or low-to-medium soluble boron contents (£ 0.20 mg dm^-3; Raij, 1995), indicating the low soil potential for B adsorption. In three soils (LR, LV and LE₁), the boron contents in the A horizon were higher than those in the B horizon, possibly due to the higher organic carbon content in the surface layer (TABLE 2). This is very important, because the subsoil becomes nutritionally important during dry periods, when boron deficiency becomes prevalent (Baker & Mortensen, 1966).

Among the samples studied, the lowest value of soluble boron was found in the PV surface horizon, possibly linked to the low soil clay content (TABLE 2). The low quantity of boron in this layer may be attributed, in part, to the higher losses of this element due to leaching. In coarse-textured soils, Touchton et al. (1980) found relatively low concentrations of boron in the topsoil. Pinyerd et al (1984) observed that the quantity of hot-water soluble boron leached from the A to the B horizon of an Ultisol was linearly correlated with the amount of water added during the studied period.

In relation to the highest value observed (0.61 mg dm^-3 for theLE₁ surface horizon), it is important to point the fact that this soil has the highest clay and carbon contents (TABLE 2) among the samples (13 g kg^-1). This HCB content is close to that obtained for a clayey dark red Oxisol in the Paraná State, Brazil by Correa et al. (1985).

The pH in CaCl₂ (0.01 mol L^-1) varied from 3.2 to 5.3 (TABLE 2) and, in this acid range, the pH has little influence over the soluble boron content (Casagrande, 1978; Elrashidi & O'Connor, 1982). The correlations between HCB and pH were not significant since the results presented high dispersion. For example, in the LE₁, pH values were almost the same at both depths (3.7 and 3.5, respectively), whereas the soluble boron decreased from 0.61 mg dm^-3 in the A horizon to 0.25 mg dm^-3 in the B horizon. Similarly, the same value of soluble boron (0.17 mg dm^-3) was found in the LV B horizon (pH = 3.6) and in the LE₂ A horizon (pH = 4.4). Over a wider pH range, for the same order of soils, Brasil Sobrinho (1965) also did not obtain significant correlations between these properties. The low correlation coefficients found in this study (r = -0.20 to -0.30) confirmed some of the results found by Gupta (1968), where r was equal to 0.10 for slightly alkaline soils. In twenty-six Egyptian alkaline soils (pH varying from 7.5 to 8.1), no significant correlation was observed between boron soluble in hot water and pH (Elseewi & Elmalky, 1979).

The correlation between soluble boron and carbon content was performed for the five soils and, then, separately for the Oxisols, due to the high textural gradient between the A and B horizons of the PV. The correlation for the Oxisols was statistically significant (r = 0.75^*). When the PV samples were considered, the correlation was not significant. As shown in TABLE 2, the organic carbon content was practically the same in both layers, whereas HCB content increased, possibly due to the fact that the clay content increased six-fold.

Among the other chemical attributes, only total CEC presented a positive correlation with soluble boron (r = 0.63^*). As the CEC includes hydrogen linked in a covalent form to the colloidal surfaces, whose charges are largely determined by the clay content, the obtained significance was likely a surface phenomenon instead of the composition of the cations of exchange complex. This hypothesis seems feasible since the correlation between soluble boron and the effective CEC was not significant, and the values corresponding to the difference between total CEC and effective CEC were positively correlated HCB (r = 0.64^*).

The literature is quite scarce in relating soluble boron contents to oxide contents. The correlation obtained in this work for boron and soil sesquioxides (TABLE 3) was only significant for the Al₂O₃ (r = 0.70^*). A higher affinity occurred between boron and aluminum hydroxide as compared to iron hydroxide, (Sims & Bingham, 1968; Elrashidi & O'Connor, 1982), since these authors obtained a higher correlation coefficient between soluble boron and Al₂O₃ than that between soluble boron and Fe₂O₃.

For the PV soil, the quantity of HCB in the B horizon was higher than in the A horizon. This may be attributed to the higher clay content in the subsurface horizon, since the organic carbon content was about the same for both horizons. The correlation between soluble boron and clay content was positive (r = 0.69^*). This result was expected, because it is common to find higher soluble boron contents in clayey soils than in sandy textured soils (Brasil Sobrinho, 1965; Gupta, 1968; Pinyerd et al., 1984). For similar soils, Casagrande (1978) found a positive correlation (r = 0.51^*) between the clay content and boron soluble in hot water. In a study with CaCl₂ (0.01 mol L^-1), high correlation coefficients were also obtained between soluble boron and clay content (Gupta, 1978; Adams et al., 1991). It can be concluded that soil clay can minimize boron leaching from the soil profile and will maintain a higher boron level in the soil. However, Jin et al. (1988) pointed to the possibility that boron availability may decrease due to boron adsorption on clay surfaces.

The correlation coefficient between the product (carbon x clay) and soluble boron contents was highly significant (r = 0.78^**). In the LE₁ A horizon, carbon x clay value was 9360 g² dm^-3 kg^-1, and this sample had the highest value of soluble boron (0.61 mg dm^-3). In the A horizon of PV, the lowest soluble boron (0.07 mg dm^-3) and carbon x clay (300 g² dm^-3 kg^-1) values were observed. Another attribute positively correlated (r = 0.68^**) with soluble boron was the soil specific surface area (SSA). This result was foreseen, because SSA is a surface parameter and is largely a consequence of organic matter and clay contents.

Since the simple correlations only measure associations between individual variables, a multiple regression analysis was performed among soluble boron, pH, organic carbon, exchangeable cations, base saturation, aluminum saturation, iron and aluminum oxides (in free, amorphous and total forms), clay content and specific surface area, obtaining the following equation (1):

HCB = -0.09 + 6.63.10^-3 Ca_exch + 7.45.10^-3 Al_exch + 1.69.10^-3 Al_T

where HCB = soluble boron; Ca_exch= exchangeable Ca; Al_exch = exchangeable Al and Al_T = total Al. Total aluminum oxide content accounted for 49 % of the variation of the soluble boron values. With the addition of the exchangeable calcium and aluminum values, the above model represents 85 % of the variation of the soluble boron content. For the exchangeable cations, the simple correlation coefficients were very low and not significant, whereas in group analysis, calcium and aluminum represented 14 and 22 % of the variation of the soluble boron contents, respectively.

CONCLUSIONS

• Boron soluble in a hot CaCl₂ 0.01 mol L^-1 solution (HCB) exhibited statistically significant simple correlations with specific surface area (0.68^*), clay content (0.69^*), CEC (0.63^*), total aluminum (0.70^*) of the five soils, and with the organic matter of the Oxisols (0.75^*).

• In the group analysis, total aluminum, exchangeable calcium and aluminum were the variables that correlated significantly with HCB, explaining 85% of its variation.

• The product (carbon x clay) takes into account the effect of the textural gradient in soil profiles and shows high positive correlation with hot-CaCl₂ (0.01 mol L^-1) extractable boron.

ACKNOWLEDGMENT

The authors express their gratitude to Dr. Sabine Goldberg, from the U.S. Salinity Laboratory, and Dr. Raymond Pacovsky, for reading and suggestions in the manuscript.

The first and the second authors wish to thank the Brazilian National Council for Scientific and Technological Development (CNPq) for a research fellowship.

Recebido para publicação em 09.06.98

Aceito para publicação em 08.09.98

• ADAMS, J.A.; HAMZAH, Z.; SWIFT, R.S. Availability and uptake of boron in a group of pedogenetically-related Canterbury, New Zealand soils. Australian Journal of Soil Research, v.29, p.415-423, 1991.
• BAKER, A.S.; MORTENSEN, W.P. Residual effect of single borate application on Western Washington soils. Soil Science, v.102, p.173-179, 1966.
• BATAGLIA, O.C.; RAIJ, B. van. Eficięncia de extratores na determinaçăo de B em solos. Revista Brasileira de Cięncia do Solo, v.14, p.25-31, 1990.
• BERGER, K.C.; PRATT, P.F. Advances in secondary and micronutrient fertilization. In: MALCOM, H.M.; BRIDGER, G.L.; NELSON, L.B. (Ed.) Fertilizers technology and usage Madison: Soil Science Society of America, 1963. p.287-340.
• BINGHAM, F.T. Boron. In: PAGE, A.L.; MILLER, R.H.; KEENEY, D.R. (Ed.) Methods of soil analysis: chemical and microbiological properties. Madison: American Society of Agronomy, 1982. pt.2, cap.25, p.431-447.
• BRASIL SOBRINHO, M.O.C. Levantamento do teor de boro em alguns solos do Estado de Săo Paulo. Piracicaba, 1965. 135p. Tese (Livre-Docęncia) - Escola Superior de Agricultura "Luiz de Queiroz", Universidade de Săo Paulo.
• CAMARGO, O.A.; MONIZ, A.C.; JORGE, J.A.; VALADARES, J.M.A.S. Métodos de análise química, mineralógica e física de solos do Instituto Agronômico de Campinas Campinas: IAC, 1986. 94p. (Boletim Técnico, 106).
• CASAGRANDE, J.C. O boro em solos do município de Piracicaba. Piracicaba, 1978. 122p. Dissertaçăo (Mestrado) - Escola Superior de Agricultura "Luiz de Queiroz", Universidade de Săo Paulo.
• CIHACEK, J.L.; BREMNER, J.M. A simplified ethylene glycol monoethyl procedure for assesment of soil surface area. Soil Science Society of America Journal, v.43, p.821-822, 1979.
• CORREA, A.E.; PAVAN, M.A.; MIYAZAWA, M. Aplicaçăo de boro no solo e respostas do cafeeiro. Pesquisa Agropecuária Brasileira, v.20, n.2, p.177-181, 1985.
• ELSEEWI, A.A.; ELMALKY, A.E. Boron distribution in soils and waters of Egypt. Soil Science Society of America Journal, v.43, p.297-300, 1979.
• ELRASHIDI, M.A.; O'CONNOR, G.A. Boron sorption and desorption in soils. Soil Science Society of America Journal, v.46, p.27-31, 1982.
• EVANS, C.M; SPARKS, D.L. On the chemistry and mineralogy of boron in pure and mixed systems: a review. Communications in Soil Science and Plant Analysis, v.14, n.19, p.827-846, 1983.
• GUPTA, U.C. Relationship of total and hot-water soluble boron, and fixation of added boron, to properties of podzol soils. Soil Science Society of America Proceedings, v.32, p.45-48, 1968.
• GUPTA, S.K. Effect of soil-properties on the extractable boron contents Schweizerische Landwirtschaftliche Forschung, v.17, n.1/2, p.45-50, 1978.
• JIN, J.; MARTENS, D.C.; ZELASNY, L.W. Plant availability of applied and native boron in soils with diverse properties. Plant and Soil, v.105, p.127-132, 1988.
• MILJKOVIC, N.S.; MATTHEWS, B.C.; MILLER, M.H. The available boron content of the genetic horizons of some Ontario soils: I. The relationship between water-soluble boron and other soil properties. Canadian Journal of Soil Science, v.46, p.133-138, 1966.
• PARKER, D.R.; GARDNER, E.H. Factors affecting the mobility and plant availability of boron in some Western Oregon soils. Soil Science Society of America Journal, v.46, p.573-578, 1982.
• PINYERD, C.A.; ODOM, J.W.; LONG, F.L.; DANE, J.H. Boron movement in a Norfolk loamy sand. Soil Science, v.137, p.428-433, 1984.
• RAIJ, B. van. Conceitos fundamentais na interpretaçăo da análise do solo. In: PEREIRA, J.R.; FARIA, C.M.B. (Ed.) Fertilizantes: insumo básico para agricultura e combate ŕ fome; anais Petrolina: EMBRAPA-CPATSA / SBCS, 1995. p.34-50.
• RATNER-ZOMAR, Y.; BANIN, A.; CHEN, Y. Oven drying as a pretreatment for surface area determinations of soils and clays. Soil Science Society of America Journal, v.47, p.1056-1058, 1983.
• RUIZ-NETO, A.; BARAHONA, E.; JAIME, S.; AGUILAR, A. Relations between B (total and available) and several constituents of cultivated soils. Anales de Edafologia y Agrobiologia, v.47, p.1281-1288, 1988.
• SIMS, J.R.; BINGHAM, F.T. Retention of boron by layer silicates, sesquioxides and soil materials: III. Iron and aluminum-coated layer silicates and soil materials. Soil Science Society of America Proceedings, v.32, p.369-373, 1968.
• TOUCHTON, J.T.; BOSWELL, F.C.; MARCHANT, W.H. Boron for soybeans grown in Georgia. Communications in Soil Science and Plant Analysis, v.11, p.369-378, 1980.

Publication Dates

• Publication in this collection
  08 July 1999
• Date of issue
  1999

History

• Received
  09 June 1998
• Accepted
  08 Sept 1998