Predicting Boron Adsorption Isotherms by Midwestern Soils using the Constant Capacitance Model

concentrations in saline soils are often caused by lack of drainage and are associated with elevated salinity. Prediction of B adsorption and transport has required detailed The narrow range between deficiency and toxicity studies of B adsorption and subsequent determination of model panecessitates accurate quantification of soil solution B rameters. In this study we tested a general regression model previously concentrations. Soil solution B concentrations are condeveloped for predicting soil B surface complexation constants from easily measured soil chemical characteristics. The constant capacitance trolled by B adsorption–desorption reactions on soil minmodel, a chemical surface complexation model, was applied to B erals. Soil mineral and organic surfaces constitute the adsorption isotherms on 22 soils from the A and B horizons of 16 sinks that adsorb B and the sources that release B to soil series from Oklahoma and Iowa. The measured chemical propersoil solution for plant uptake. Because plants respond ties were surface area, organic C (OC) content, inorganic C (IOC) only to solution B concentrations (Keren et al., 1985), content, and Al oxide content. The prediction equations of Goldberg soil minerals can act to attenuate potentially phytotoxic et al. (2000) were used to obtain constant capacitance model values soil solution B concentrations. Boron occurs as boric for B surface complexation constants thereby providing a completely acid over most of the soil pH range. Boric acid acts as independent evaluation of the ability of the constant capacitance a Lewis acid by accepting a hydroxyl ion to form the model to describe B adsorption. The model was well able to predict tetrahedral borate anion, pKa 9.2. B adsorption isotherms on the majority of the soils. The regression Various modeling approaches have been used to demodel was used to obtain the parameters for the constant capacitance model. Then the constant capacitance model was used to predict the scribe B adsorption on soil materials. These include soil specific B adsorption. This is in contrast to regression models applications of chemical models called surface complexthat fit adsorption of a series of soils. The distinction is that using ation models (Goldberg and Glaubig, 1986; Goldberg, the combined regression equations and the constant capacitance 1999; Goldberg et al., 2000). The advantages of surface model only soil properties and not adsorption are needed to predict complexation models over empirical adsorption models, soil specific B adsorption data. The prediction equations developed such as distribution coefficients, Kd, and adsorption isofrom a set of soils primarily from California, were able to predict B therm equations, such as Langmuir and Freundlich apadsorption on a set of soils from different parts of the country. This proaches, are that they define specific surface species, result suggests wide applicability of the model prediction equations chemical reactions, mass balances, and charge balances developed previously, for describing B adsorption both as a function and they contain molecular features that can be given of solution B concentration and solution pH. thermodynamic significance (Sposito, 1983). In a prior study (Goldberg et al., 2000), we developed a general regression model to obtain soil B surface complexation B is both an essential micronutrient element constants for use in the constant capacitance model to required for plant growth and a toxicant at elevated predict B adsorption. The constant capacitance model concentration. The range between B deficiency and toxparameters are obtained from easily measured soil icity is narrow, typically 0.028 to 0.093 mmol L 1 for chemical properties: surface area, OC content, IOC consensitive crops and 0.37 to 1.39 mmol L 1 for tolerant tent, and free Al oxide content. These are also soil crops (Keren and Bingham, 1985). Yield losses can ocproperties that correlate with soil B adsorption capacity. cur both under conditions of B deficiency and B toxicity. The prediction equations were well able to predict B In regions of plentiful rainfall, B deficiency occurs priadsorption behavior on 15 additional soils primarily marily on coarse-textured soils. Deficiency symptoms from California, providing a completely independent can also be triggered by liming of acid soils because of evaluation of the ability of the constant capacitance increased B adsorption at higher soil pH (Reisenauer model to describe B adsorption. The objective of this et al., 1973). In arid regions, B toxicity occurs because approach is to avoid the necessity of performing timeof high levels of B in the soil solution and from additions consuming detailed adsorption studies for each specific of B via the irrigation water. Excessive soil solution B soil. The B adsorption data used consisted of adsorption envelopes defined as: amount of B adsorbed as a function of solution pH per fixed total B concentration. The S. Goldberg, D.L. Suarez, and S.M. Lesch, USDA-ARS, George E. prediction equations have not yet been tested for their Brown, Jr. Salinity Laboratory, 450 W. Big Springs Road, Riverside, CA 92507; N.T. Basta, Dep. of Plant & Soil Sciences, Oklahoma State ability to predict B adsorption isotherms defined as: University, Stillwater, OK 74078. Contribution from the George E. amount of B adsorbed as a function of equilibrium soluBrown, Jr. Salinity Laboratory. Trade names and company names are tion B concentration. In contrast to Kd, Langmuir, and included for the benefit of the reader and do not imply any endorsement Freundlich modeling, our constant capacitance model or preferential treatment of the product listed by the USDA. Received 19 May 2003. *Corresponding author ([email protected]). Abbreviations: AMSE, average mean squared error; EGME, ethylene Published in Soil Sci. Soc. Am. J. 68:795–801 (2004).  Soil Science Society of America glycol monoethyl ether; ICP, inductively coupled plasma; IOC, inorganic C; OC, organic C. 677 S. Segoe Rd., Madison, WI 53711 USA