RATIONALIZATION OF IONIC STRENGTH AND CATION EFFECTS ON PHOSPHATE SORPTION BY SOILS

The amounts of inorganic phosphate (P) sorbed by four contrasting unfertilized soils during 40 h were influenced by the ionic strength and cation species of the contacting solution (support medium) used, as indicated by isotherms over the final P concentration range of 0 to 1 μg P/ml and 0 to 10 μg P/ml. An increase in ionic strength enhanced P sorption during 40 h but the species of cation also influenced the amount of P sorbed, as shown by the isotherms obtained in 10^?2M Ca and 3 × 10^?2M Na systems. Although pH affected the amounts of P sorbed, pH effects alone could not adequately explain the differences in P sorption. Kinetic studies indicated that within the range of P addition used for each soil, the equilibrium P concentration, at infinite time, was independent of ionic strength and cation species. Consequently, the composition of the solution affected only the rate at which equilibrium was attained. The results are attributed to the effects of ionic strength on the surface charge of retaining components and the thickness of the diffuse double layer, and the effects of specilic sorption of a divalent cation on surface charge, as they relate to the rate of P sorption.     

Development and evaluation of a kinetic model to describe phosphate sorption by hydrous ferric oxide gel

Isotherms for the sorption of inorganic phosphate (P) by hydrous ferric oxide gel (Fe gel) were described by a three-equation Langmuir sorption model. Each equation described sorption within a distinct concentration range or region (I, II, and III) of the overall isotherm. Regions I and II involved chemisorption, whereas region III involved a more physical sorption type. With increasing sorption time between 0.7 and 28.7 days, the extent of sorption in region I increased by more than 30%. In contrast, the extent of sorption in regions II and III remained essentially constant. An equation was developed, based on the change in the sorption maximum of region I (b_I) with increasing sorption time, which described the change in solution P concentration with time. The increase in b_I with time, evaluated by the closeness of fit of this relationship to experimental data, was found to depend on two factors: first, the extent to which P was chemisorbed, and this was affected by pH and ionic strength; second, the batch of Fe gel used. For two different levels of P addition, the proportion of sorbed P which remained extractable in 0.1M NaOH, decreased with increasing sorption time. After 30 days only 88% of the sorbed P remained NaOH-extractable. The data obtained indicated that the increasing chemisorption of P with increasing sorption time involves the diffusion of sorbed P into the bulk of the Fe gel particles. This concept is discussed in relation to mechanisms proposed by previous workers to explain the time-dependence of P sorption.     

EFFECTS OF SOLUTION: SOIL RATIO ON PHOSPHATE SORPTION BY SOILS

Isotherms for the sorption of inorganic phosphate (P) by three contrasting soils during 40 h showed a dependence on solution: soil ratio. Above a final solution P concentration of 0.5 μg ml^?1, more P was sorbed at a solution: soil ratio of 5:1 than at 40:1 for a given level of P in solution. With time up to 146 h, the effect of solution: soil ratio on P sorption was kinetically controlled. Equilibrium solution P concentrations, estimated by extrapolation of the linear relationships between solution P concentration and the reciprocal of time to I/t= o (i.e. t=∞)appeared to be coincident for each solution:soil ratio at high and the low levels of added P. Consequently, sorption isotherms at equilibrium would be coincident, irrespective of the solution soil ratio used. The kinetic control of solution: soil ratio on P sorption is interpreted in terms of the number of P sorbing sites and initial solution P concentration on the rate of P sorption by soils.     

Predicting phosphorus sorption isotherm parameters in soil using data of routine laboratory tests

Knowledge of phosphorus(P) sorption dynamics across different soil types could direct agronomic and environmental management of P. The objective of this study was to predict P isotherm parameters for a national soil population using data of routine laboratory tests. Langmuir and Freundlich sorption parameters were calculated from two different ranges(0–25 and 0–50 mg P L~(-1)) using an archive of representative agricultural soil types from Ireland.Multiple linear regression(MLR) identified labile forms of aluminium(Al) and iron(Fe), organic matter(OM), cation exchange capacity(CEC), and clay as significant drivers. Langmuir and Freundlich sorption capacities, Freundlich affinity constant, and Langmuir buffer capacity were predicted reliably, with R~2 of independent validation 0.9. Sorption isotherm parameters were predicted from P sorbed at a single concentration of 50 mg P L~(-1)(S_(50)). An MLR prediction of P sorption maximum in the 0–50 mg P L~(-1) range was achieved, to an accurate standard, using S_(50), OM, and Mehlich-3 Fe(R~2 of independent calibration and validation being 0.91 and 0.95, respectively). Using Giles' four shapes of isotherms(C, L, H, and S), L non-strict-and C-shaped isotherm curves accounted for 64% and 27% of the soils, respectively. Hierarchical clustering identified a separation of isotherm curves influenced by two ranges of Mehlich-3 Al. Soils with a low range of Mehlich-3 Al(2.5–698 mg kg~(-1)) had no incidence of rapid sorption(C shape). Single point indices, Al, or available soil data make the regression approach a feasible way of predicting Langmuir parameters that could be included with standard agronomic soil P testing.     

Specific sorption of trace amounts of cadmium by soils

Abstract

Sorption of trace quantities of Cd in four soils of different chemical and mineralogical properties, was studied. Initial Cd concentrations were between 15 to 150 μg. 1^?1. The sorption isotherms were linear and had a positive intercept in three of the soils, indicating a constant partition‐high affinity sorption isotherm (Giles et. al⁶). The data also followed the Freundlich sorption isotherm, and the Freundlich K parameter was taken as a measure of the relative affinity of the different soils for the Cd metal sorbed. Cadmium sorbed was extracted by IN‐NH₄C1 followed by 0.1N HC1, and the fraction remaining in the soils was considered specifically sorbed Cd. This fraction also followed a linear sorption isotherm, and was around 30% for the four soils studied. The sorption order for the amount of specifically sorbed Cd showed that the Boomer soil (kaolinite‐iron oxides) had the lowest affinity for specific sorption of this metal. This was taken as evidence that kaolinite and iron oxides have a lower capacity for retaining cadmium through specific sorption mechanism(s) than the materials present on the other soils (2:1 layer silicates and humic substances). The existence of specific mecha‐nism(s) responsible by the sorption of trace quantities of Cd in soil solutions has important implications on soil‐plant relationships, Cd mobility in soil profiles and control of Cd activity in soil solutions.     

印度Chattisgarh州两种新开发土和一种淋溶土中磷酸盐吸附-解析特征研究

Two Inceptisols and an Alfisol of the Indravati Catchment area in Chattisgarh, India, comprising several gradients in physical and chemical properties were studied to relate phosphate sorption and desorption to soil properties. From the P isotherm curve, the standard P requirement （SPR） of the soils was determined. Phosphate sorption data were also fitted both to the Langmuir and Freundlich Equations. The mean sorption maximum values for three different soil series were： Bastar 〉 Geedam 〉 Mosodi. The fraction of added phosphate sorbed for the 3 series followed this same trend as did SPR; the phosphate sorption maximum and the maximum phosphate buffering capacity, which were estimated by the Langmuir isotherm; and the Freundlich constant 1/n. However, phosphate desorption, as well as the maximum recovery percent did not follow this order. The phosphate affinity constant （K） was also different following the same progression for the 3 soil series as the Freundlich constant K＇, which measured sorption strength. Meanwhile, an inverse order existed for K and K＇ versus the percent desorbed relative to the sorbed as well as the maximum recovery percent. In addition, significant correlation coefficients among sorption parameters of P and soil factors were found.     

DESORPTION AND ISOTOPIC EXCHANGE RELATIONSHIPS OF PHOSPHATE SORBED BY SOILS AND HYDROUS FERRIC OXIDE GEL

Sorption of added inorganic phosphate (P) was irreversible in four contrasting soils and hydrous ferric oxide gel during 16 h desorption after 40 h sorption at the same (iso)pH. Irreversibility increased with increasing time of desorption above 30 h. When the amount of P which was chemisorbed during the sorption step was subtracted from the amount desorbed, the latter fell on the isotherm describing the more-physical, potential-determining sorption. No pH change occurred during desorption and net negative charge decreased by 1 equivalent per mole P desorbed. These results suggested that only more-physically sorbed P was desorbed at the iso-pH. The more-physically sorbed P was also reversible with respect to changes in the ionic strength and cation species of the desorbing solution. The isotopic exchangeability of the more-physically sorbed P was at least ten times greater than that of chemisorbed P. With increasing sorption time, both the ease of desorption and exchangeability of sorbed P decreased. Subsequent to desorption, the exchangeability of the remaining sorbed P also decreased. These observations are interpreted in terms of the concurrent changes in the amounts of chemisorbed and more-physically sorbed P.     

REACTIONS OF SOLUBLE PHOSPHATE WITH ACID SOILS: THE INTERPRETATION OF ADSORPTION-DESORPTION ISOTHERMS

Adsorption-desorption and isotopic-exchange isotherms for acids soils of known clay mineralogy indicated that when all the surfa? P accessible to the solution was considered, the bonding energy for phosphate decreased exponentially with site coverage. The soils differed markedly in the quantities of surface P held at sites of very high bonding energy, defined as those in equilibrium with a solution concentration < 1 μM; the values ranging from 9 μmol g^?1 in a soil with 40 per cent gibbsite in the clay fraction, to 1 μmol g^?1 in a soil with only 1 per cent gibbsite and no interlayered material. The reversibility of current P adsorption on soil surfaces depended on the immediate history of the soil sample. Complete reversibility occurred when soil, previously enriched with soluble P and stored for 4 to 5 months at constant temperature, was suspended in solutions of constant pH and ionic strength. It is suggested that irreversibility occurs due to incomplete attainment of equilibrium during the adsorption phase, a condition predisposed by high initial P concentrations (>1mM), and the use of soils that are naturally in dis-equilibrium because of recent fertilizer additions or severe depletion of P by plant uptake.     

SORPTION OF INORGANIC PHOSPHATE BY IRON- AND ALUMINIUM- CONTAINING COMPONENTS

The amounts of inorganic P sorbed by a range of Fe- and Al- containing components varied appreciably and decreased in the order allophane > fresh Al gel > Fe gel pseudoboehmite > aged Al gel > dried Fe gel > Fe-coated kaolinite > haematite > goethite > akaganeite > gibbsite = ground kaolinite > dispersed kaolinite. Al gel sorbed 30 to 70 times more P than gibbsite, and Fe gel sorbed approximately 10 times more P than its crystalline analogues (haematite, goethite. and akaganeite). Despite large differences in the extent of P sorption, the form of the isotherm was essentially the same for each sorbent. The ability of freshly-prepared Al gel suspensions to sorb P decreased with ageing, a property not shown by Fe gel. Drying of Fe gel at 80°C, however, caused an approximately 4-fold decrease in P sorption. Precipitation of Fe gel (2% Fe) on the surface of kaolinite increased P sorption by a factor of 10. The occurrence of Fe gel as a coating apparently presents more sorption sites to solution per unit weight of Fe gel than Fe gel alone. A linear relationship (r= 0.98) was obtained between the amount OH^? sorbed per unit increase in pH value (‘hydroxyl buffering’) and the overall P sorption maximum for each sorbent. Hydroxyl buffering provided a better index of P sorption potential than specific surface area. Except for the crystalline Fe sorbents, isotherms obtained by plotting fractional sorption saturation against final solution P concentration for the sorbents were essentially coincident with those for several contrasting soils. For crystalline Fe components a lower relative amount of weaker sorption, as opposed to chemisorption, of the overall sorption maximum was obtained. Differences in the extent of P sorption. however, appear to be primarily related to the number of functional M-OH groups presented at the solid-solution interface.     

Correlation of solution and extractable phosphorus with vegetative growth of teff

Abstract

Three soils, Hiwassee loam (clayey, kaolinitic, thermic, Typic Rhodudults), Vaiden clay (very fine, montmorillonitic, thermic, Vertic Hapludalfs) and Marvyn sand (fine, loamy, siliceous, thermic, Typic Hapludults), were used in this study. Phosphorus sorption and desorption isotherms of the soils were determined in the laboratory. Average P sorption maximum (b) was calculated using the Langmuir isotherm equation. Based on the P sorption capacity, each soil was fertilized with different P rates and teff [Eragrostis tef (Zucc)] was grown in the greenhouse. The amount of P sorbed and desorbed increased as the concentration of equilibrating solution P increased. Phosphorus sorption maximum from sorption and desorption isotherms was 278, 251, and 37 mg P/kg for Hiwassee, Vaiden, and Marvyn soils, respectively, and for maximum dry matter yield of teff the soils needed a minimum of 0.029, 0.048, and 0.065 mg/L soil solution P, respectively, which were all below the soil solution P at P adsorption maximum.

Soil P was extracted by the Mehlich I, Mehlich III, modified Mississippi, Bray P‐l and Olsen methods. Plant‐available P extracted by each of the five methods was significantly correlated with teff dry matter yield, and statistically any of the five methods can be used as the basis for predicting teff yield responses.     

Differential Sorption Behavior of Cadmium,Lead, Zinc,and Copper in Some Tropical Soils and Their Environmental Implications

This study evaluated cadmium (Cd), lead (Pb), zinc (Zn), and copper (Cu) sorption characteristics of three tropical soils. Data obtained conformed to Freundlich sorption model and the S-shaped isotherm curve. Sorption efficiency of Zn and Pb were highest in alkaline soil while slightly acid soil had the highest Cd and Cu sorption efficiency for monometal sorption. In competitive sorption, metals were more sorbed in slightly acid soil while the least efficiency was recorded in acid soil. Distribution coefficient; K_d (average across soil types) in monometal sorption followed the order: Pb > Zn > Cd > Cu. For competitive sorption, the order was Zn > Pb > Cu > Cd. When in competition, Cd was preferentially sorbed in slightly acid and alkaline soils and Zn for acid soil. Conclusively, lead is more in equilibrium solution when in competition with Cd, Zn and Cu making it potential agent of soil and groundwater pollution.     

THE CONTENTS AND SORPTION OF CADMIUM IN SOME AGRICULTURAL SOILS OF ENGLAND AND WALES

Cadmium sorption was measured in 10 agricultural soils with pH ranging from 4.5 to 7.9, and total Cd content from 0.27 to 1.04 μg g^?1 dry soil. With initial Cd concentrations of 0.5 to 100.0 μM, sorption from 0.002 M CaCl₂ was described by the Freundlich adsorption equation but the gradients of the isotherms increased when the initial concentrations were below 0.5 μm. This indicates that there are specific sites of differing sorption energy; differences between soils in the gradients of the isotherms at low initial concentration could largely be accounted for by their contents of ‘free’ Fe₂O₃. When initial concentrations were below 0.5 μm there was a linear relationship between the quantity of Cd sorbed and the final concentration in solution. This relationship held with all soils except that of lowest pH from which there was a net loss of Cd to the solutions. Desorption was measured from three soils with contrasting pH. With the soil of lowest pH, over 80 per cent of sorbed Cd was desorbed to 0.002 m CaCl₂ and up to 30 per cent to 100 or 500 μm solutions of heavy metal chlorides. In contrast, only very small proportions (<1.25 per cent) were desorbed from the other soils with pH 6.7 and 7.8. The results indicate that Cd is strongly sorbed by soils of pH of above 6.0 when added in amounts comparable to additions in sewage sludges or phosphatic fertilizers, and illustrate the importance of liming as a means of reducing the mobility of this metal in soils.     

石灰性土壤与磷酸盐的反应及吸持态磷的同位素交换性

本文研究了种石灰性土壤（Ｌｏｕ土）与磷酸盐的反应动态过程，短期反应的等温吸附研究表明，在低磷浓度下，以吸附反应机制为主，吸持态磷的同位素交换性随着吸持量的增加而增加；在高磷浓度下，以形成磷酸盐的沉淀反应机制为主，吸持态磷的同位素交换性随着反应时间的延长和吸持量磷数量的增加而了降低。认为在低施磷水平下，土壤中铁，铝氧化物对磷的吸持起重要作用。本文还探讨了在长期（２６０天）恒温恒湿培养过程中，土壤可溶     

Relationships between phosphate sorption and iron oxides in Alfisols from a river terrace sequence of Mediterranean Spain

Phosphate sorption isotherms for samples of the A- and upper B-horizons of Alfisols situated on a sequence of terraces of the River Guadalquivir, southern Spain, showed that phosphate sorbed at an equilibrium concentration of 0.3 μg P ml^?1 was correlated with several soil properties. Crystalline Fe-oxides (goethite and hematite) appeared to be the most important P-sorbing components of the soil samples. Sorption was highly correlated with percent clay and with dithionite soluble Fe. This last property was as good a property to predict sorption as the specific surface of crystalline Fe-oxides (which was estimated by line profile analysis of the X-ray diffractograms), probably because the range of surface areas for goethite and hematite was relatively narrow in the samples and because of the limited accuracy of the method of line profile analysis.Well drained and imperfectly drained soils differed little in sorption properties. In imperfectly drained, hematite-free soils, chroma was correlated with sorption. This is promising for it would allow rapid field estimation of sorption.The sorption maximum, calculated from the fitted Langmuir isotherm, was about 1.7 μmol P m^?2 of Fe-oxides, a value similar to those reported in the literature for natural and synthetic oxides.     

Sorption Behavior of Ofloxacin to Kaolinite: Effects of pH,Ionic Strength,and Cu(II)

Sorption of antibiotics to clay minerals is a key process controlling their transport and fate in environment. In this study, the effects of pH, ionic strength, and Cu(II) on ofloxacin (OFL) sorption to kaolinite were investigated by batch sorption experiments. The results of sorption edge experiments suggested that OFL sorption to kaolinite was pH and ionic strength dependent. Cation exchange was a major contributor to the sorption of OFL⁺ to kaolinite. The decreased OFL sorption with increasing ionic strength indicated the formation of outer-sphere complexation. When solution pH was lower than 7.0, Cu-OFL complexes facilitated OFL sorption through electrostatic attraction or formation of kaolinite-Cu-OFL and kaolinite-OFL-Cu ternary surface complexes. However, existence of free Cu(II) cation in solution competed for sorption sites, and thus suppressed OFL sorption. When solution pH was higher than 7.0, Cu(II) existed as Cu(OH)₂, and the Cu-OFL complexes in aqueous phase and solid phase (precipitation) enhanced OFL removal efficiency from solution. The results imply that Cu(II) effects should be taken into account in the evaluation of OFL mobility in environment.     

A PHOSPHATE SORPTION INDEX FOR SOILS

Phosphate sorption isotherms covering a wide concentration range (10^?6 to 5 × 10^?3m phosphate) were determined for 42 soil samples at 20 °C by a standardized technique. The slope of a plot of the sorption, x, against the logarithm of the equilibrium solution phosphate concentration, log c, measured at c= 10^?4M, proved a suitable reference index to characterize the phosphate sorbing properties of the soils. Several single-point methods were tested by statistical correlation against this reference index. Of these, the sorption, x, from one addition of 150 mg P/100 g soil gave r= 0.951, but r= 0.974 when the equilibrium concentration was also taken into account in the quotient x/log c. This quotient is therefore suggested as a simple yet adequate way of indicating a fundamental soil property, its phosphate sorption isotherm.     

Phosphorus retention and release of anions and organic carbon by two Andisols

Andisols can absorb large amounts of phosphorus rapidly, and then release it slowly, yet the mechanisms by which they retain P and release it for plant growth are poorly understood. Ligand exchange of organic compounds from Al–humic complexes by P and/or Si release – due to breakdown of allophanic microstructure to provide sorption sites – might account for the retention of P, but its extent is not known. We applied a soil column flow-through technique to quantify the release of anions and organic carbon (C) associated with P sorption by two andic soils, and we related the anion release to possible mechanisms for the retention of P. Phosphate (H₂PO₄^–, HPO₄^2–) sorption and concurrent anion desorption were obtained by passing a 1-g P 1^–1 (32 mmol KH₂PO₄ in 1 mm CaCl₂) solution through the soil columns (25 cm³). Total dissolved P, Fe, Al, S, Ca, Mg, K, Mn, organic C and pH were determined in the eluent. Changes in eluent pH and the patterns of the retention of P and corresponding concentrations of Al, Si and organic C in the eluent were similar for the two Andisols. The general pattern and changes in pH of the eluent coincided with changes in the patterns of release of organic C and Si and the rate of P retention. Release of silica accounted for < 6% of the P sorbed and had only a minor role in P retention in these two Andisols. Release of organic C, however, accounted on a molar basis for 40% and 83%, respectively, of the P sorbed. Direct measurements of the pH of the eluent and release of anions and organic C concurrent to P retention contribute to rapid assessment of the controlling mechanisms of P retention. The results indirectly confirm the hypothesis of ligand exchange of solution P with organic complexes held on allophanic surfaces. The organic C release, however, is not specifically related to either the fast or the slow P retention phase. The shift in the controlling P retention reaction associated with a change from the fast to the slow P retention phase is clearly indicated by an abrupt change of the pH of the eluent. This shift, in previous studies identified graphically by a change in slope of the P sorption isotherm, can be identified directly by measuring the pH of the matrix.     

Simple modelling of the sorption of iron‐cyanide complexes on ferrihydrite

The sorption of the iron‐cyanide complexes ferricyanide, [Fe(CN)₆]^3—, and ferrocyanide, [Fe(CN)₆]^4—, on ferrihydrite was investigated in batch experiments including the effects of pH (pH 3.5 to 8) and ionic strength (0.001 to 0.1 M). The pH‐dependent sorption data were evaluated with a model approach by Barrow (1999): c = a exp(bS)S/(S_max‐S), where c is the solution concentration; S is the sorbed amount; S_max is maximum sorption; b is a parameter; and a is a parameter at constant pH. Ferricyanide sorption was negatively affected by increasing ionic strength, ferrocyanide sorption not at all. More ferricyanide than ferrocyanide was sorbed in the acidic range. In the neutral range the opposite was true. Fitting the pH‐dependent sorption to the model resulted in a strong correlation for both iron‐cyanide complexes with a common sorption maximum of 1.6 μmol m^—2. Only little negative charge was conveyed to the ferrihydrite surface by sorption of iron‐cyanide complexes. The sorption of iron‐cyanide complexes on ferrihydrite is weaker than that on goethite, as a comparison of the model calculations shows. This may be caused by the lower relative amount of high‐affinity sites present on the ferrihydrite surface.     

Effects of equilibrating salt solutions on phosphate sorption by soils

Abstract

Several equilibrating salt solutions have been used in the studies of P sorption by soils and sediments. This study was conducted to evaluate the effects of 10 salt solutions on estimation of P sorption by soils. Results obtained showed that, when the equilibrating solution was made to contain 0.01M with respect to CaCl₂, Ca(NO₃)₂, CaSO₄, MgCl₂, KCl, LiCl, Nacl, or KHCO₃, the amount of P sorbed by soil always exceeded the amount sorbed from the soil‐water system. In comparison with the amount of P sorbed from water, 0.01M NaHCO₃ reduced P sorption by soils. Use of THAM buffer (0.05M pH 7.0) to control the pH increased P sorption by some soils and decreased P sorption by others, relative to that sorbed from the soil‐water system. The results indicated that inclusion of salts in the equilibrating solution for P‐sorption studies should be avoided, especially in studies related to water quality.     

Phosphate sorption by calcareous Vertisols and Inceptisols as evaluated from extended P-sorption curves

The plant availability of phosphate applied to calcareous soils is affected by precipitation and adsorption reactions, the relative significance of which is not well known. We used extended P-sorption curves obtained at phosphate addition rates up to 340 mmol P kg^?1 soil to examine the relative contribution of precipitation and adsorption by 24 calcareous Spanish Vertisols and Inceptisols. Adsorption was dominant at 1 day and at small rates of addition (10–35 mmol P kg^?1). With increasing clay and Fe and Al oxides contents of the soil, more phosphate was sorbed before the sorption curve bent upwards, as a result of Ca phosphate precipitation. Sorption curves showed a nearly vertical intermediate region, the length of which increased with time, suggesting that a Ca phosphate buffered the concentration of P in solution. The buffering concentration decreased with time, suggesting a progressive transformation of more to less soluble forms of Ca phosphate. A phase less soluble than octacalcium phosphate seemed to control the concentration of P in solution at 180 days in most soils. The apparent solubility of this phase decreased with increasing carbonate content in the soil. Precipitation of poorly soluble Ca phosphates apparently predominated up to a P addition dose ranging from about 30 mmol P kg^?1 in some soils to more than 340 mmol P kg^?1 in others. At larger doses, the way additional P was bound to the solid phase was different; phosphate was probably adsorbed, at least in part, to low-affinity sites on silicate clays and oxides. The proportion of sorbed phosphate that was isotopically exchangeable decreased with time, soil carbonate content and P addition dose for doses <100 mmol P kg^?1. This is consistent with the idea that P in Ca phosphates is less isotopically exchangeable than P adsorbed on mineral surfaces. At larger additions of P, isotopic exchangeability was unrelated to the soil properties measured, probably because there was a variety of sorbed P forms influenced in turn by different soil components.