Factors affecting fertility of selected brown sand soils of Guyana

Material Information

Factors affecting fertility of selected brown sand soils of Guyana
Downer, Alfred Victor, 1937-
Copyright Date:
Physical Description:
xiv, 235 leaves : ill. ; 28 cm.


Subjects / Keywords:
Aluminum ( jstor )
Nutrients ( jstor )
Oxides ( jstor )
pH ( jstor )
Phosphates ( jstor )
Sand ( jstor )
Soil horizons ( jstor )
Soil organic matter ( jstor )
Soil science ( jstor )
Soils ( jstor )
Dissertations, Academic -- Soil Science -- UF
Sandy soils -- Guyana ( lcsh )
Soil Science thesis Ph. D
Soil fertility -- Guyana ( lcsh )
Soils -- Guyana ( lcsh )
bibliography ( marcgt )
non-fiction ( marcgt )


Thesis (Ph. D.)--University of Florida, 1972.
Bibliography: leaves 207-234.
Additional Physical Form:
Also available on World Wide Web
General Note:
General Note:
Statement of Responsibility:
by Alfred Victor Downer.

Record Information

Source Institution:
University of Florida
Holding Location:
University of Florida
Rights Management:
Copyright [name of dissertation author]. Permission granted to the University of Florida to digitize, archive and distribute this item for non-profit research and educational purposes. Any reuse of this item in excess of fair use or other copyright exemptions requires permission of the copyright holder.
Resource Identifier:
028549942 ( AlephBibNum )
09160244 ( OCLC )
ABU5604 ( NOTIS )


This item has the following downloads:

Full Text


FACTORS AFFECTING FERTILITY OF SELECTED BROWN SAND SOILS OF GUYANA By ALFRED VICTOR DOWNER A Dissertation Presented to the Graduate Council of the University of Florida 5n Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy UNIVERSITY 0? FLORIDA 1972




... to those for whose interest and guidance my expressions of gratitude can never be truly adequate ... Mrs. Irene A, Wilson Mr. R. Rueben 3aird Mr. Aubrey P. Alleyne (dec'd.)


ACKNOWLEDGEMENTS The author wishes to express his sincere appreciation to Dr. W. G. Blue for his interest and guidance during the course of this study, and for his assistance which was always readily available. Appreciation is extended to the members of the Supervisory Committee Drs, T. L. Yuan, V. W. Carlisle, J. E. Moore, and G. 0. Mott for their willingness to help, their constructive comments, and their critical reviews of this manuscript. The supply of planting material provided by Dr. G. B. Killinger is also gratefully acknowledged. The author wishes to thank the Center for Tropical Agriculture and the Soil Science Department of the University of Florida for the award of the assistantship which made this investigation possible. In conclusion, the author wishes to acknowledge the importance of his family in this endeavour. The patience and understanding of his wife were particularly gratifying. iii


Abstract of Dissertation Presented to the Graduate Council of the University of Florida, in Partial Fulfillment of the Requirements for the Degree of Doctor of Philosophy FACTORS AFFECTING FERTILITY OF SELECTED BROWN SAND SOILS OF GUYANA By Alfred Victor Dovmer March, 1972 Chairman i Dr. W. G. Blue Major Department J Soil Science Samples of four soils of the "white sand" plateau of Guyana were used in an effort to delineate fertility characteristics and describe means of improving crop production levels. Laboratory studies indicated that kaolinite was the clay mineral in all the soils but that its contribution to cation exchange capacity was negligible; more than 90 per cant of the cation exchange sites were supplied by the organic fraction. Exchangeable bases were low in all soils, and K and Mg were increasingly leached from the soils as pH increased. The three brown sand soils Tabela sand, Kasarama loamy sand, and Ebini sandy loam differed as a group from the Tiwiwid sand. The former soils were relatively rich in amorphous material with fulvic acids dominant in the organic fraction while the latter soil was poorly supplied with amorphous material and its organic fraction was dominated by humic acids. The Tiwiwid sand had a lower pH but a larger supply of plant nutrients than the brown sand soils. iv


Substantial amounts of P were extractable from the Tiwiwid sand, and virtually all added P was recoverable by the Bray #1 extractant. The brown sand soils had very small amounts of P and less than 70 per cent of added P was recoverable. Plant growth under controlled moisture was better on the Tiwiwid sand than on the brown sand soils, with N appearing to be the nutrient which evoked the largest response. On the latter soils, P appeared to limit plant growth but positive responses to added micronutrients were also obtained. Pigeon pea responded positively to lime on the Tiwiwid sand, but pangolagrass yields were depressed by lime on the brown sand soils. This negative response was apparently due to fixation of P at higher pH levels, as fulvate ions of fulvate-Al complexes were replaced by P. Practices aimed at the improvement of the level of fertility in these soils must minimize the loss of nutrients by leaching. The use of lime on the brown sand soils will need to be carefully considered in relation to crop requirements and soil conservation.


TABLE OF CONTENTS Page ACKNOWLEDGEMENTS iii ABSTRACT iv LIST OF TABLES, ix LIST OF FIGURES xiv INTRODUCTION , 1 REVIEW OF LITERATURE k Guyana* •••••••••••••••• 4 Ebini Area 4 Location . c . . . . . 4 Climate . . • 4 Vegetation. .•••••••••••• 5 Physiography Geology 7 Soils 7 Soil Utilization 8 Brown Sands Fertility Concepts Concerning Red-yellow Soils of the Tropics 13 Adsorptive Complex. ........ 16 Inorganic Soil Colloids 17 Kaolinite. . . s 17 Hydrous oxides ........ 18 Organic Soil Colloids 21 Inorganic Interaction 27 Organic-Inorganic Interaction ... 28 Nutrient Retention 31 Cation-Sorption 33 Cation Retention Patterns 35 Anion Sorption. ................ 38 Phosphate Retention in Soils 41 Nutrient Availability and Plant Uptake 48 Amelioration of Fertility in Red-yellow Soils 54


Page Soil Acidity and Plant Growth. •••••'•••• 55 Correction of Acidity on Red-yellow Soils 61 Indicator Crops 66 Pangolagrass . „ . 66 Pigeon pea , 67 MATERIALS AND METHODS 70 Soils , 70 Soil Analysis 71 Plant Analysis 74 Laboratory Experiments . ..... 74 Exp. 1. Incubation Studies 74 i. Leaching studies 75 ii. Exchangeable cations 76 iii. Extractable phosphorus 76 Exp. 2. Studies on the Organic Fraction of the Soils 76 Greenhouse Experiments ...» ...... 78 Exp. 1. Limiting Nutrients Study 73 Exp. 2. Optimum Levels of Ca, P, and K . . . . 79 RESULTS AND DISCUSSION 82 Profile Characterization 82 Particle-size Analysis ..... 82 Clay Fraction 84 Organic Fraction . . ......... 92 Chemical Characteristics of the Soils. ..... 106 Surface Soils 114 Buffering Capacity 114 Cation Exchange Capacity ............ 114 Incubation Studies 121 Cation retention. ....... 121 Phosphate retention ... 125 Soil Properties in Relation to Potential Soil Fertility 132 Plant Growth 138 Limiting Nutrients . 138 Pangolagrass 138 Pigeon pea ....... 141 Central Composite Studies 150 Pangolagrass. 150 Pigeon pea. 174 Soil Fertility and Amelioration 184 vii


Page ;d conclusions. ....... imrnmt ~ 200 APPENDIX ...... LITERATURE CITED ......•••••••• 20? BIOGRAPHICAL SKETCH. ,..,...... viil


LIST 0? TABLES Table i'age 1. Particle size distribution in the soils, .... 83 2. Aluminum and Fe contents of clay separates e . . 91 3. The composition and distribution of the amorphous material in the soils ....... 93 4. Data on the organic fraction of the soils. ... 94 5. Percentages of soil organic C extracted by 0o5N NaOH with humic-C/fulvic-C and C/N ratios. 98 6. Patterns of precipitation of C in NaOK extracts with variation in pH. 99 7. The concentration of Al in the supernatant of the 0.5N NaOH extract at various pK levels. . 101 8. Concentration of Fe in the supernatant of the 0.5N NaOH extract of surface horizons at various pH levels 103 9. Concentration of various cations in 0.5N NaOH extracts of soils 105 10. Total contents of various elements in the soils 107 11. Cation exchange capacity and exchangeable cations in the soils 108 12. pH values in the different horizons of the soils and the relationship between values by different methods of measurement 110 13. Amounts of Al extracted from soils by different reagents at different pH levels . . Ill ix


Table Page Ifc. Contributions to soil cation exchange capacity by the clay and organic matter fractions ... no -I -1 Q 15. Cation exchange capacity measurements 16 Ratio Ca leached: exchangeable Ca at different pH levels and rates of P application J-" 17 Pattern of retention of exchangeable cations against leaching as affected by pH and rate of P application on Tiwiwid sand ^ Pattern of retention of exchangeable cations against leaching as affected by pH and rate of P application on Tabela sand ^-° Pattern of retention of exchangeable cations against leaching as affected by pH and rate of P application on Kasarama loamy sand • • 20 Pattern of retention of exchangeable cations against leaching as affected by pH and rate of P application on Ebini sandy loam. . . 1^0 18. 19. 21. 27. 26. 129 131 Water soluble phosphorus in surface soils incubated at different pH levels and rates of applied P 22 Bray #1 extractable P in soils incubated at different pH levels and with different rates of applied P 23. Some characteristics of' the four surface soils, summarized • • • Zk. Average shoot weights (oven-dried) from two harvests of pangolagrass W 25. Shoot growth (oven dried) of pangolagrass in response to fertilizer applications .... ^ 26. Root weights (oven-dried) of pangolagrass ^ after two harvests Shoot weights (oven-dried) of pigeon pea 1^3 Shoot growth (oven-dried) of pigeon ?ea in response to fertilizer applications


Table Page 29. Root weights (oven-dried) of pigeon pea after one harvest . . « «... 145 30 1, Root growth (oven-dried) of pigeon pea in response to fertilizer applications ..... 1^7 31. Dry matter yield of pangolagrass on the brown sand soils 151 32. Comparison of the regression coefficients for dry matter yield (g/pot) of pangolagrass on the brown sand soils 152 33* Concentration of phosphorus in shoots of pangolagrass grown on the brown sand soils 157 34. Total uptake of phosphorus by pangolagrass from the brown sand soils .......... 158 35. Comparison of the regression coefficients for total uptake of phosphorus (g/pot) by pangolagrass from the brown sand soils. . . . 159 36. Concentration of calcium in shoots of pangolagrass grown on the brown sand soils 16*1 3?. Comparison of the regression coefficients for total uptake of calcium (g/pot) from the brown sand soils. 165 38. Concentration of magnesium in shoots of pangolagrass grown on the brown sand soils ..... 166 39. Comparison of the regression coefficients for total uptake of magnesium (g/pot) by pangolagrass from the brown sand soils. . . . I67 kQ. Concentration of potassium in shoots of pangolagrass grown on the brown sand soils. . 169 hi. Comparison of the regression coefficients for total uptake of potassium (g/pot) from the brown sand soils 170 42. Concentration of nitrogen in shoots of pangolagrass grown on the brown sand soils 172 xi


Table Page 43. Comparison of the regression coefficients for total uptake of nitrogen (g/pot) from the brown sand soils 173 44. Concentration of aluminum in roots of pangolagrass grown on the brown sand soils « . « 175 45. Concentration of phosphorus in roots of pangolagrass as affected by the addition of nutrients to the soil 176 46. Ratios of AlsP in pangolagrass roots as affected by added nutrients 177 47. Dry matter yield of pigeon pea in relation to nutrient supply 179 43, Concentration of phosphorus in pigeon pea as affected by the addition of nutrients to the soil 180 49. Nutrient concentrations in pigeon pea in relation to treatment applied to the Tiwiwid sand 182 50. Nutrient concentrations in pigeon pea in relation to treatment applied to the Kasarama loamy sand 183 51. Mean levels of dry matter yield, nutrient concentrations, and uptake by pangolagrass on the brown sand soils. 185 52. Solutions of partial derivatives of fitted regression equations for dry matter yield of pangolagrass I87 53. Shoot weights (oven-dried) from first harvest of pangolagrass 201 54. Shoot weights (oven-dried) from second harvest of pangolagrass. .......... 202 55 • Analysis of variance for data shown in Table 24 203 xii


Table Page 5&» Analysis of variance for data shown in Table 2?. ........ , 205 57. Analysis of variance for data shovm in Table 29 206 xiii


Figure LIST 0? FIGURES Page 1. X-ray diffraction patterns of K-saturated clay from the Ebini sandy loam 85 2. X-ray diffraction patterns of heated K-saturated clay from the ETbini sandy loam 86 3. Electron micrographs of clay separates from the surface horizons ,....«« 87 h. Differential thermal analysis traces of Mg-saturated clay of the Ebini sandy loam 89 5. Absorption spectra of 0.5N NaOH extracts 97 6. Titration curves of surface horizons in N KG1 modified with HC1 or KOH 115 7. Cation exchange capacity in relation to humiccarbon . . 120 Effects of nutrient elements on the development of nodules on pigeon pea roots in Tabela sand 148 Effects of nutrient elements on the size and coloration of nodules developed by pigeon pea roots in Tabela sand !^9 10. Dry matter yield of pangolagrass on Tabela sand • • • -*-" 11. Dry matter yield of pangolagrass on Kasarama loamy sand -^ 12. Dry matter yield of pangolagrass on Ebini sandy loam ^55 13. Phosphorus uptake from Tabela sand 160 14. Phosphorus uptake from Kasarama loamy sand. ..... 161 15. Phosphorus uptake from Ebini sandy loam 162 xiv


INTRODUCTION "The soil cones first. It is the basis, the foundation of farming; without it, nothing; with poor soil, poor living; with good soil, good farming and living. An understanding of good farming begins with an understanding of the soil." At first glance, the above statement made by Ahlgreen (3« p. ^25) might seem to be a trifle platitudinous. Given some thought, however, it amounts to an effective summary of the rationale behind the major portion of man's endeavours on this planet. The statement can be construed as an immediate basis for the broad division of the world's surface into developed areas— the soil is understood to a considerable degree—and underdeveloped areas — the soil is understood only to a limited degree or not at all. If considered in conjunction with the Malthusian or "Dismal Theory" and available evidence on the relative rates of population growth and increase in food production, the statement makes an overwhelming case for the need to understand the soil. In the so-called underdeveloped regions of the world, there are greatly overpopulated areas at one extreme and vast areas of uninhabited and unutilized land at the other extreme. The Amazon area lies at the latter extreme. This area supports vast stretches of apparently lush evergreen forests, but makes little or no contribution to the production of food or fiber.


Guyana falls largely writl phlcal unit comprised essentially of the Amazon basin (283) ugh it has been the source of several agricultural products for some ^00 years its soils are clearly not understood to any appreciable degree. Abcut 8.5% of the land area of Guyana is composed of coastal sediments, and it is this area which has supported the major part of the population and provided the bulk of the agricultura throughout its existence. Soils found on the remainder of the land area of Guyana, whether formed from in-situ weathering of frecambrian parent material or from transported material from si] urces, have been relatively unused. Extensive systems of cattJ have been practised on areas of the non-coastal soils sine v half of the nineteenth century with varying degrees of success. While production in the Rupununi has been maintained at relatively low levels of efficiency, all early efforts in the Ebini area failed completely. Several studies have been undertaken to delineate the reasons for early failures, but complete understanding of the phone: hich determine the level of fertility of soils at Ebini has still to be achieved. This has encouraged neither the modification of methods of production nor the increase in productivity of these soils, and the non-coastal soils of Guyana have remained virtually unused. However, data compiled by the F. A. 0. Soil Survey project and published in 19^5 indicated the potential value of the Ebini area as a launching site for the development of a viable agricultural program based on the utilization of non-coastal soils. Accordingly, four soil types were selected from the Ebini area for study. The immediate objectives being!


(a) delineation of the fertility characteristics of the soils. (b) description of possible methods of optimizing or ameliorating crop production levels. The extremely low level of inherent fertility and the free leaching to which Wagenaar (315) found the soils to be subject suggested that the objectives might best be realized by an examination of characteristics which influence the retention of added nutrients. This examination could then be supplemented by greenhouse studies aimed at establishing the optimal levels of some of the major nutrients which could be beneficially applied to the soils. Pangolagrass ( Diftitaria decumbens Stent) and pigeon pea ( Ga.janus cajan (L) Druce) were used as indicator crops. Selection of a grass and a legume for these purposes was influenced by the idea that the ultimate system of agriculture would most likely revolve around pastures. Cattle would provide a means of recycling nutrients, and grass or grass-legume associations a means of maintaining desirable levels of organic matter in the soil.


REVIEW OF LITERATURE Guyana Guyana covers an area of about 225,000 kci extending from 1 10' N to 8° 35* N, and from 56° 33' • to 6l° 28* W. Its narrow coastline extends along the Atlantic coast of South America from just west of the Orinoco Delta to the Corentyne River. Deposit ion of suspended material borne by the numerous rivers within Guyana, and by the Amazon, has produced a thin strip of "coastal soils," but by far the greater portion of the country is made up of material from the Guiana Shield. The non-coastal soils have not been very widely used for agricultural production, but interest in their utilization is increasing steadily. Ebinj_.Arga Location Ebini is situated at about 5° 40* N and 57° 40' W in the northeastern portion of Guyana. It lies within the area of the Berbice savannas (25l) which because of their position, relative to the coastland and to the Rupununi savannas, are known colloquially as the "Intermediate Savannas." Climate Vagenaar (315) summarized available data on the climate of the Ebini area. He stated that the climate, broadly categorized as


equatorial, could be subclassed as a "tropical rainforest climate, continuously racist." Mean annual temperature (for 19&2 only) was 26. $C with the highest monthly mean— -28.2C— -in October, and the lowest— 25. 5C--in January. The absolute minimum temperature in that year was 17. 8C, recorded in September-October. Day-length was thought to vary by less than 22.5 ni n * from the mean of 12 hours 7.5 min. recorded in Georgetown. The mean annual rainfall is approximately 2,200mm with a recorded minimum of 1,550mm and maximum of 3;^35rcm» The pattern of rainfall distribution was described ass A long rainy season (mid April to mid August) . A long dry season (mid August to late November). A short rainy season (late November to early February). A short dry season (early February to mid April). The long rainy season accounts for about 50/S of the average annual rainfall. He estimated that the annual evaporation was about 127cm, and suggested that because of its unreliability, rainfall was probably the most important climatic factor in agriculture in the Ebini area (315) • Vegetation Sombroek (283) argued that the Guianas fall within the northern boundary of the phytogeographical unit of Amazonia, and this is supported by Vann (303) who reported a marked similarity in vegetation across the Guianas. Richards (251) indicated that the savannas in


Surinam are very much like the Berbice savannas. These areas are dominated by grasses with "islands" of bushes and trees as well as strips of fringing forests located in stream valleys. The most common grass species is Trachypogo n plumosus, with admixtures of Ax on opus aureus , A xonopus stragalus , and Andropogon virga tus. The most common trees are Curatella americana and Byrsoni m a crassi folia (^5). Physiography The Ebini area lies within the northern regions of the physiographic unit designated as the "white sand" plateau. This plateau is of a somewhat uniform monotonous topography with gently undulating relief. It rises from an elevation of about 16m in the north to about l60m in the south. Its surface is dominated by irregularly distributed areas of white quartz sand (white sand) and brown loamy sand to sandy clay (brown sand). These sands are thought to have developed from sediments deposited by rivers flowing from the upper Rio Branco Basin, and Vann (303) concluded that the time deposition was approximately 227,000 years ago during the third interglacial period. Bleackley and Khan (32) stated that the white sand plateau constitutes part of a soil formation which covers a wedge-shaped area of some 45,000 km in Guyana, Surinam, and French Guiana. Within Guyana the white sand plateau extends over about 26,000 km and of this the brown sands occupy 2 approximately 9,000 km . Most of the brown sands support forest; only 2 about 770 km support savannas. They reported that the white sands occupy the topographically highest point in a given locality within the confines of the plateau as well as areas sloping towards waterways.


Richards (25l) observed that the soils of the Berbice savannas are mainly brown sands but these change to bleached sands Kith changes in vegetation. Geology. Hardy and Follett-Smith (119) estimated that the non -marine sediments in Guyana were derived initially from an Archaean gneiss granite basement complex associated with a volcanic series. The Geological Survey Department of Guyana lists several stratigraphic units, and among these is a group of Tertiary to Pleistocene and Holocene continental -deltaic sediments — the Corentyne group. The white sand plateau which seems to be identical to the Berbice formation forms part of the Corentyne group of sediments. Soils The soils of the white sand plateau include regosols, red-yellow latosols, red-yellow podzols, and ground water laterites displaying various phases and degrees of intergrading. Red-yellow latosols and regosols are the dominant soils. Brinkman (45) reported the occurrence of at least 15 soil types in the intermediate savannas. The soils of most extensive occurrence include the Kasarama loamy sand (56,600 ha), Tiwiwid sand (40,480 ha), Tabela sand (27,640 ha), and Ebini sandy loam (26,300 ha). An apparent characteristic of the area is the occurrence of "black water" streams, the coloration of which is thought to be due to the movement of organic matter down the soil profile and into the drainage system.


8 Soil Utilizatio n Richards (25l) cited burning and grazing practices as factors which may have influenced the development of vegetation in the Berbice savannas. This would suggest that shifting cultivation may have been the practice on these soils at some time, but no data are now available to establish or refute this. Inaccessibility of relevant records, at this time, also precludes any meaningful discussion or interpretation of the effects of settlements in the area during the seventeenth and eighteenth centuries. According to Mayers (186), introduction of cattle to the Rupununi savannas in the 1860s, or thereabouts, led to the utilization of the Berbice savannas for the grazing and resting of animals being driven to market. Subsequently efforts at more long-term utilization of these savannas revealed that the prolonged grazing of cattle resulted in infertility and frequently in death of the animals. Solutions to this problem were actively sought and in 19*U the Ebini Livestock Station was established with the primary objective of facilitating the development of extensive systems of beef production. This objective was modified in 1957 to include the solution of problems likely to occur in intensive systems of production. Exploratory studies on pasture grasses and grass/legume associations were initiated, and later field crop studies were commenced (186). More recently, there has been some crop production in the area, but the extent of utilization of the Berbice savannas remains limited.


Brown San The terra "brown sands" has been in use for some time as a broad description of the soils of the Berbice savannas. The broad grouping permits comparisons with those portions of the id plateau wl extend into Surinam—the Zanderij formation—, and French Guiana—Sols Ferraliques typiques — . However, the description of soils on an individual basis is to be preferred since this provides the opportunity for comparison with similarly developed profiles in other parts of the world. Such comparisons are necessary if only because of the very limited amount of analytic data published about the brown sands. From individual profile descriptions, the bj id s oils are seen to be in fact predominantly red-yellow soils. The Tiwiwid sand appears to be an exception, but Bleackley and Khan (31; reported that humic B horizons of up to 30cm in thickness or reddish horizons (32) were found at depths of less than 12m from the surface of this soil. They described lateral transitions from white sand through pale yellow sands to red sandy loams, and noted that the boundaries of the bleached sands tended to coincide with the presence of humic layers. They also noted that the pans of precipitated Fe and humic matter were to be found at about the level of the water table (3l). Comparable findings were reported by Andriesse (l^) for red-yellow soil associations in Sarawak. Richards (251) described the Tiwiwid sand as a "lowland tropical podzol" and compared it to the "Padang" soils of Borneo described by Hardon. He suggested that the bleached sands were the products of degradation of less sandy soils, and attributed the degradation to


10 the effects of organic materials \*hich were not precipitated because of the low base status of the soils. There is, however, some controversy as to the true parent material of the Tiwiwid sand. It would appear that there are three possible sources of parent material. The first possibility is brown sandy soils as proposed by Richards (25l)» and supported by the data adduced by Bleackley and Khan (31, 32). The second possibility is sand dunes or other coastal deposits, and this is supported by the occurrence of a series of bauxite deposits along the southern extremeties of the white sand plateau in Guyana and Surinam. Those deposits coincide roughly with the coastline of South America prior to the first interglacial period, but Vann (303) argued against the possibility of parent material of marine origin. Ke listed several reasons among which were the lack of "ridges" or other shoreline features, the absence of shell remains, and the close similarity to mineral assemblages of inland rocks. The third possibility is based on a conclusion arrived at by Klinge (l57)» He observed a catena-like relationship between white sand regosols and red-yellow soils in the Manaus area of Brazil, and concluded that both of the sources indicated above may have contributed to the parent material from which bleached sands developed. Despite the lack of a characteristic red-yellow coloration, there may be some genetic relationships to the brown soils proper — the Tabela sand, the Kasarama loamy sand, and the Ebini sandy loam. These relationships can be appreciated in light of available data on the role of organic materials in the alteration of soils. Andriesse (lk) distinguished between eiaphic podzols in Sarawak, the formation of which is


11 dependent on the porous nature of the parent material, and the climatic podzols of the temperate zone which have their origin in the surface accumulation of organic residues. Though such a clear distinction is unrealistic in this case, the effects of percolating water cannot be ignored. Kemper (150) attested to the disruptive effect on soil aggregates of moving water and the abrasive action of solid particles suspended in it. Thorp et al . (298) related the porosity of the soil to the movement of bases, sesquioxides, clay particles, and clay particle constituents. The solvent effect of dissolved materials has been reported by various workers. Bloomfield (33) found differences in the degree of dispersion of soil particles caused by polyphenols, and established that the polyphenols could mobilize ferric oxides as ferrous complexes of considerable stability. Davies et al. (?l) characterized the Fe-polyphenol compounds as easily transported ncn= ionic complexes. Malcolm and McCracken (180) categorized the effects of canopy drip from different plant species in terns of pH, polyphenol composition, and capacity for complexation of Al and Fe. They determined that polyphenols, reducing sugars, and organic acids were the principal agents of Fe mobilization, but that lower concentrations of dissolved 2+ material enhanced mobilization. Their suggestion that the Ca status of the soil had no effect on the mobilization of Al and Fe differed 2+ from that of DeLong and Schnitzer (75) who reported that Ca greatly decreased the solubility in water of Fe-organic matter complexes. The importance of hydroxy -acids, e.g., tannic acid, in the mobilization of Fe and the strong coloration of drainage water (298) lends support to the argument advanced by Richards (251). He proposed that the high


12 tannin content of the residues of the legumes, e.g., Eperua s>vp., supported by the Tiwiwid sand was largely responsible for degradation of brown sandy soils, and development of the Tiwiwid profile. Both parent material and climate will have influenced the nature and composition of the soil mass. Bennema (2.6) indicated that under conditions of high temperature and rainfall, red-yellow latosols with transitions to regosols and red-yellow podzols s,re the soils likely to be formed. Such soils characteristically have an inorganic portion comprised of small amounts of silt, 1:1 silicate clays, set qui oxides (usually less than 10?o ( w / w ) ° f the clay fraction), and larger quantities of quartz or other minerals which are highly resistant to weathering (26), The climatic conditions encourage intense oxidation of organic substances, but low pH and moist conditions inhibit condensation of the decomposition products, thereby permitting appreciable quantities of low molecular-weight compounds to exist in the organic fraction (159). The properties of any soil arise from the composition of the soil mass. Structurally, red-yellow soils display fine granules adhering closely to each other to produce a friable mass (26) . They are physically very suitable for plant growth, as indicated by Ignatieff and Lemos (136), but are usually very poorly supplied with plant nutrients. They have low cation exchange capacities, and relatively large capacities for anion sorption and phosphate fixation. Sombroek (283) provided detailed description of the chemical properties and nutrient status of several red-yellow soils in the Amazon basin.


13 rning the Red-yellow The concept ox refers to the capacity of the soil to supply nutrients i . quantities to growing plants as required, This ir.plies that the ter a store of nutrients in the form of weatherablo mi ropriate composition or that it has the capacity to retain adeqi ties of any nutrients added to that soil* The node of f< 3 red-yellow soils of the tropics and the chare,cterist;" bion of the soil mass imply that readily weatherable minerals supply nutrient elements required for plant growth are absen . tually so. The level of fertility encountered in these soi] ore, is dependent on the store of nutrients which is or Lnst leaching in forms available to plants. Such sto: : i bs are generally extremely low. Despite their 1 I status many red-yellow soils support lush evergreen forests. This anomaly has been explained by Nye and Greenland (221) in tei rient cycling. Vegetation plays a vital role in the sir. nutrients which are returned to the soil in organic form as lit to: h, and root excretions (219). Mineralization of these materials provides inorganic ions for plant uptake. The theory was also applied to savannas. Grasses were found to contain, prior to leaf fall, almost the entire nutrient reserve in the system (22l) . This is essentially a closed cycle, and differs only in the mineralization phase from the "direct nutrient cycle" proposed by Went and Starke ( 321). Their theory involved endotrophic mycorrhizae with fungal hyphae providing continuous conduits from litter to root


and leaving little or no inorganic nutrients to be lost by leaching* Both mechanisms provide an explanation for the rapid loss of fertility seen on the cultivation of highly weathered tropical soils. Not only are exchangeable and mineralized nutrients lost, but soluble products of decomposing organic material move down the profile selectively removing ions from the soil mass. Large amounts of Ca, Kg, and F and traces of K and Fe have been mobilized by non-hydroxy organic acids, while hydroxy -acids removed more Fe, less Ca, Kg, and K, and still less P (293). Jones (1^3) alluded to a cycle of fertility within the soil profile, irhich embraced changes in soil structure and nutrient status, consequent upon the destruction of organic matter, as well as the movement of nutrients down the profile during periods of cultivation. The nutrients which penetrate to lower horizons are returned to the surface during fallow periods. This is in effect an open cycle and probably constitutes the philosophy of shifting cultivation. It does not provide, however, any information as to the relative quantities of the nutrients which are or can be retrieved. Simonson (2??) regarded the initially low base status concomitant with the prolonged weathering and leaching of ultisols as the key to sustained crop production under systems of simple technology, but it is apparent that the potential effect of the current leaching regime must be listed as a factor of basic importance. Its significance will be enhanced by the high infiltration rates (l?9) and low water holding capacities (225) of these soils. Effects of the leaching regime are appreciated through the changes induced in the natural vegetation


15 by progressive loss of nutrients. The litter produced changes to materials of lower "base status and wider C/tf ratios (l4), reducing the richness of the closed nutrient cycle (89). Fungi instead of bacteria become the principal agents of decomposition in accordance with the changed microclimate as more highly acid organic residues accumulate (89). Nye and Greenland (22l) observed that the nutrient level of the soil solution is dictated by the anion concentration. Walker (318) stated that though there might be an initial accumulation of P, N, and S in the organic fraction, these elements can be progressively reduced by leaching, and this culminates in reduction of N -fixation by colonizing legumes. Since the parent material is the only source of P in these soils, he concluded that P was the key element in pedogenesis and that its level in the soil at any point in time wou! place a ceiling on crop production. The red-yellow soils of the tropics are generally accepted as being inherently infertile. Their utilization poses the problem of the preservation of fertility, and would appear to justify the definition of soil fertility as a function of 'the level of technological skills in, and available to, the farming population (37) • Leaching and erosion need to be controlled, and cropping programs carefully considered. The initial level of fertility will reflect the balance bet ween the movement of nutrients down the profile, and the root volume of the cultivated species. The rate of movement of nutrients within the soil will be determined by the effective capacity for nutrient retention, that is to say, by the nature of the adsorptive complex within the soil. Wagenaar (315) reviewed available data on the brown sand soils


16 of Guyana and reported from his studies on the Tabela sand and Kasarama loamy sand that fertilizers, applied over a period of years as well as during the period of his investigations, had caused no appreciable change in the nutrient status of the soils. Removal of bases and the consequent destruction of soil clays foster the development of acidity through increases in amounts of active Al in soils. This adds a further dimension to the fertility status of these soils. Problems of toxicity and reduction of exchange capacity by Al could possibly be encountered. Reeve and Sumner (2^8) suggested that the base retention properties of Natal Oxisols were influenced by the factors which determined the equilibrium position of the various forms of Al in the soil. Adsorptive Complex The adsorptive complex in any soil comprises the sum of the adsorptive properties of the charged constituents of the colloidal fraction of that soil. The functional characteristics of this fraction as measured by reaction, buffering capacity, base saturation, etc., are determined by the nature of the individual constituents, and the proportions of each in the total. The individual constituents may be organic or inorganic. Organic colloids are less thoroughly understood than inorganic colloids, but it is known that variations in composition and, therefore, in properties can occur among these colloids. They are, however, always entirely negatively charged or potentially so. Inorganic colloids may consist of various proportions of silicate clays and crystalline or amorphous hydrous oxides of Al and/or Fe. These


17 colloids nay display negative or positive charges, or both. In red-yellow tropical soils 2sl silicate clays and allophanes are absent or nearly so, Bennema (26), and Ignatieff and Lemos (136) averred that the clay fraction of such soils consists essentially of kaolinite and sesquioxides. The contribution to the adsorptive properties of the soils attributable to the clay fraction is dependent on the degree of interaction between crystalline and the amorphous constituents* Superimposed upon this contribution, is that of the organic colloids, modified by interaction with the inorganic colloids. The mechanisms of interaction will thus determine the surr: of the adsorptive properties of these soils and their susceptibility to change with changes in the conditions within the soil mass. Inorganic Soil Colloids Kaolinite The structure and properties of kaolinite have been described by several workers (68, 112, I85) . Schofield and Samson (266) showed that the kaolinite crystal could have both positive and negative charges at low pH levels. Increase in pH caused an increase in negative charges, and a decrease in positive chax-ges up to an inflection point on the alkaline side of neutrality. They attributed the negative charges partially to dissociation of hydroxy 1 groups in the tetrahedral layer, acknowledging that such dissociation would be negligible in neutral and slightly acid conditions. An additional source of negative charges was thought to be isomorphous substitution in the tetrahedral layer. Fripiat (lOl) discussed this latter source of charge in terras of Al


18 in krather than 6-coordination and suggested that the proportion of Al in 4-coordination and, therefore, the degree of isomorphous substitution would be determined by the pH at which clay synthesis occurred* At lower pH levels there is less Al in ^-coordination. Van Olphen (223) indicated that isomorphous substitution and other lattice imperfections vrould be relatively small in kaolinite. Charges on the clay particles would be due mainly to broken bonds on the edge of the crystals. Those charges due to isomorphous substitution would be negative, and exert a permanent effect manifest on the flat surfaces. Charges due to broken bonds would vary in nature with the pH of the medium and in the pH range in which they would be active. Broken bonds in the octahedral sheet would bear positive charges at low pH while those in the tetrahedral sheet would bear negative charges at high pH. The charges contributed by broken bonds thus exert a "pH-dependent" effect. The bulk of the charges on the kaolinite crystal were thought to be due to broken bonds (112), and their pH-dependent effect was diagrammatically represented by Rich (250), The latter author remarked that kaolinite in soils tends to be less crystalline than specimen material and may also display more evidence of isomorphous substitution. Hydrous oxides Correns (68) described the hydrous oxides as aqueous systems — Si0 2 -H 2 0j Al 2 0-3-H 2 0j and Fe 2 0^-H 2 0~and observed that the Fe and Al systems were somewhat hydrophobic, tending to form gels which displayed little or no crystalline structure by X-ray diffraction.


19 The actual form of a given hydrous oxide in the soil is important since I the reactive potential of the oxide as shown by Van Schulenborgh (306) and Sumner (288). Bonn (36) showed by solubility studies of Fe 2 0~-H systems in acid soils that in the pH range k to 8 Fe(0H) 2 + was the predominant form of the oxide. He argued that some Fe(OH). must \>o encountered as such in soils because of the rapid turnover of Fe due to factors such as microbial activity, redox reactions, eomplexation by organic colloids, and the time lag which is required 2+ for the precipitation of Fe(OH) from oxidized Fe compounds relative to the rate of conversion of Fe(OK)~ to less soluble materials, such as Fe ? CL or Fe„0 ;ic . Correns (68) emphasized the consequences of the fact that Fe can exist in two oxidation states. In aerobic conditions Fe is practically insoluble, and can be transported only as Fe(OH)~ sols. In anaerobic conditions FeCCh or Fe-huraates of low redox potential can occur, end then on oxidation produce FeO(OH). Patrick (226) showed the existence of a linear relationship between pK and redox potential, and between the oxidation states of Fe and redox potential in anaerobic soil conditions. Such conditions do not always obtain, but it would seem that Fe supplied by the ultimate source — ferromagnesian minerals— 3+ accumulates in soils by the precipitation of Fe forms transported in the reduced state as complexes with organic colloids (307). The precipitated forms can, with time, crystallize as geothite or lepidocrocite and may simultaneously incorporate co-precipitated Al(OH) into the structure. Hydrous oxides of Fe may be dehydroxylated to .Fe 2 0o or Fe^O. or react with Si0 2 gels to form Fe-silicates (68). The Al 2 03-H system has been widely studied. The initial sources


20 of Al are the K-f els pars and plagioclases (307), and possibly kaolinite subjected to intense leaching or the action of humic acids (223, 1^9). The forms of hydrous oxides of Al in soils are determined by the degree of hydroxy lation and, therefore, by the pH of the system. Van Schuylenborgh (307) considered that at pH 6 and above Al(OH)^ was likely to be the dominant form. Jackson (138), however, indicated that hydrous forms of Al began to polymerise at about pH 5, and that the isoelectric point of amorphous Al gels was at pH 8.3 in contrast to pH ^.8 for crystalline Al(OH) . Somewhat at variance with this is the report that Al(0H)~ precipitates only near neutrality (68). The most prevalent form of hydrous Al-oxide in soils is perhaps gibbsite, but Cate (52) concluded that its occurrence was determined by the leaching regime during the formation of the soil, with a period of dessicatlon required for crystallization. In the presence of Si0 2 reactions may occur with Al hydrous oxides. Correns (68) stated that at ratios (Al o~:SiO ) of 1»2, kaolinite minerals tend to form, and usually the pH in such conditions is 4. With ratios of 1*3 or 1«4, 2»1 clays may develop. There seems little doubt that there is some association between amorphous hydrous oxides in soil, and that the nature of the association varies. Yuan (3^) calculated that to 16% of the total extractable Si and 1? to 6k% of the Al in some Florida soils could be removed by citrate -bi carbonate -dithionite extraction. Weaver et al . (319) reported the extraction of greater amounts of Si, Al, Fe, and Mg by the same extractant from acid than from neutral or alkaline soils. The differential solubility of the amorphous materials vas also reported by Tweneboah et al. (301), who extracted appreciable


21 quantities of Al but only small amounts of Fe and Si with 0.5M CaCl 2 at pH 1.5, Acquaye and Tinsley (2) mentioned the possibility of occlusion of Si0 2 by hydrous oxides of Al and Fe. Organic Soil Col loids Knowledge of the structure of the organic soil colloids — humus — is relatively less detailed than that of the inorganic colloids. This is due in part to the complexity of the material a,nd in part to the indirect and destructive methods of study. The organic fraction of the soil is usually measured by oxidative processes, but none of the methods available can be accepted as capable of providing accurate data. Weight loss detected on ashing oven-dry (105G) soil at temperatures of 100-200C, 375C or 85OG was due principally to the oxidation of the organic fraction (18, 2^3). The loss of structural water may, however, contribute in varying degrees to the loss in weight of soil. Wet oxidation techniques (ll) can provide precise measurements of the organic C content of the soil, but this has then to be converted to organic matter. Variations in the elemental composition of the organic fractions of different soils and different horizons within the soil are known to exist (159), and Broadbent (^*6) recommended that the standard correction factor of 1.724 proposed by Van Bemmelen be replaced by a factor of 1.9 for surface soils and 2.5 for subsoils in California. Ranney (243) concluded that 2.0 was a more appropriate correction factor for surface soils in Pennsylvania. Fractionation of the organic material in soils has been based on alkaline extraction, usually with O.5N NaOH. Martin and Reeve (l8l)


22 suggested that the alkali accomplished the prerequisite depolymerization of the large organic molecules and increased the dispersion of smaller polymers through the rupture of H-bonds e They regarded the conversion of Al-* + to A10. "" as the main step in the process since this prevented the flocculating action of Al. Levesque and Schnitzer (167) observed variations in the ash and C content of extracts with variations in the alkali concentration of the extractant. At alkali concentrations of 0.1N 7075 of the soil C was extracted from the Eh of a podzolf but at 0.52. only h$% was extracted. Posner (236) concluded that there was an upper limit to the proportion of soil C which could be extracted, and that 0.5^ NaOH extracts had the highest ash and C content as compared to pyrophosphate and bicarbonate extracts. About 1Q# of the soil Al was extracted by 0.2N NaOH (167), and Yuan and Fiskell (3^5) considered it likely that some of the Al would have come from the hydrous Al oxides in the soil. Though determinations of the organic fraction are, at best, only approximations, the fractionation of the extractable organic material meaningfully reflects variations in composition. The components have been grouped according to variations in color, solubility, and molecular size. Felbeck (90) described them ast(1) fulvic acids lowest in molecular weight, lightest in color, and soluble in both acid and alkali. (2) humic acids medium molecular weight and color, soluble in alkali and insoluble in acid. (3) humins highest molecular weight, darkest in color, and insoluble in both acid and alkali.


23 Humins have received much less attention than the humic and fulvic acids. According to Kononova (159) they appear to "be complexes of humic and fulvic acids not readily extracted by dilxjte alkali. This insolubility has been attributed to linkages with the mineral soil rather than to the occurrence of chemical alteration. Humins removed from soil have yielded humic acids with lower C, and higher H and contents, and possibly of a simpler chemical nature than those readily alkali soluble humic acids (l59)» Fulvic acids have been shown to be less aromatic (300), more acidic (336), and have less C (28, 159, 336) than humic acids. Schnitzer (26?) postulated that 6l# of the weight of the fulvic acid molecule was made up of functional groups with the ratios -C00H/-0H and -COOK/phenolic OK having values of approximately 1 and 3» respectively. He stated that fulvic acids were extremely resistant to biological degradation and usually contained appreciable amounts of carbohydrates. Tan and Clark (29l) found a close association of fulvic acids and polysaccharides but virtually none with humic acids. They also noted that fulvic acid formed under grass was less lignoid and associated with more polysaccharides than that formed under pine. Schnitzer and Desjardins (268) estimated that the molecular weight of organic material extracted from the A horizon of a podzol was 1,684 as compared to 669 for that from the Bh horizon. Sach molecule from the A Q horizon was considered to have 6 carboxyls, 2 phenolic and 3 alcoholic hydroxyls while molecules from the Bh horizon had 3 carboxyls, 7 phenolic and 5 alcoholic hydroxyls. Wright and Schnitzer


24 (33^) reported that the total acidity of 0.5N NaOH extracted material was 890 meq/lOOg and 1,180 meq/lOOg for the A Q and 3h horizons, respectively. The former had 280 and the latter 210 meq/lOOg as phenolic hydroxyls. McKeague et al . (188) estimated 7^0 meq/lOOg as carboxyls and 360 meq/lOOg as phenolic hydroxyls in extracts from a spodic horizon. Wagner and Stevenson (316) used extracts from a brunizem soil and found that y& of 390 meq/lOOg as carboxyls in the humic acids were close enough to form cyclic anhydrides, while > 60/S of the 290 meq/lOOg as hydroxyls were phenolic. Schnitser and Skinner (2?l) established that alcoholic hydroxyls did not participate in reactions with metals and that the reactivity of the carboxyls varied. They indicated that tvro types of reactions which might occur with metals may include a major reaction with "salicylate structures" and a minor reaction with the less acidic carboxy3.s. Martin and Reeve (183) described the flocculating effects of Al on organic matter, and suggested that some reactions might involve electrostatic bonding while others utilized coulombic forces (182). The reactions in which humic acids engage are influenced by the pH and the cation. More Fe and Cu and less Al reacted with humic material at pH 5.0 than at pH 3.5 (271). Posner (235) observed that while Ca 2+ flocculated humic acids at ionic strengths > 0.0023M, K + formed no specific complexes. Ling Ong and Bisque (175) reported that the flocculating effects of cations increased as the radius of the hydrated ion decreased. They established that ionic concentrations required for flocculation varied with valence in the order: monovalent : divalent : trivalent = 1.0 : 0.014 « 0.0014. The effect


25 of anions associated with K decreased in the order SO^ > NO3 > CI. The variation in acid strength of the car boxy Is in humic acid (269) was found by Posner (235) to follow a Gaussian distribution about a mean pK value of 5.5 ± !•?• He concluded that this phenomenon was responsible for the atypical electrolytic behavior of humic acids, Gilmour and Coleman (l07) regarded the smaller charge density and the micromolecular structure of humic acids in the presence of monovalent cations as the reason for the atypical behavior. Martin and Reeve (182) reported that the apparent pK values of humic acids varied with the C/Al ratio of the material. At a ratio of 4.9 the pK value was 6.2, but dropped to 4.1 at a ratio of 152. This implies that the more acidic functional groups are the first to be neutralized, but Posner (236) was of the opinion that lower pK values were obtained when humic acids were bound to Al, Fe, or Ca. Various functional groups in the organic fraction have been estimated to contribute different amounts to the total charge. Broadbent and Bradford (4?), and Schnitzer and Skinner (269) attributed 55/5 of the charge to carboxyls, while 35% have been attributed to phenolic hydroxyls, and 10/o to imide-N (106). The contribution made to soil reactions by the different categories will vary with conditions. Lewis and Broadbent (169) suggested that some carboxyl groups may have 2+ reacted with Cu(OH) + while phenolic groups coraplexed Cu . Schnitzer and Skinner (2?l) pointed out the simultaneous action of some carboxyls and phenolic hydroxyls. This and the distribution of pK values emphasize the importance of soil reaction in this context. According to Broadbent (46) the composition of the organic fraction


26 varies with the environment and the material from which the humus forms. Absence of periods of dessication and the predominance of hydrolytic processes inhibit condensation to humic acids and favor the formation of fulvic acids (l59)» Coulson et al. (69) reported that acidity inhibited the oxidation and polymerization of polyphenolic products of decomposition, but Wright and Schnitzer (336) concluded that humic acids could be converted to fulvic acids by an increase in the number of O-containing functional groups, eog9,~C00H. Alexandrova (9) found that the process of huinification involved an increase in carboxyl content and a simultaneous decrease in phenolic hydroxy Is. These changes proceeded simultaneously but at different rates and to different degrees, depending on the type of material involved. The elemental composition of the final products of humification was determined primarily by the type of material and its chemical composition (lO). Differences in the relative amounts of humic and fulvic acids have been shown to exist in the soil and to be related to the horizons (292, 33^» 3^3) • Several workers have also attempted to relate the composition of the organic colloids to soil type. Tan and Van Schuylenborgh (292), and Tan (290) found that the latosols of Indonesia had much higher proportions of fulvic acids than of humic acids. Niu Ching We*n (21^-) reported similar findings for lateritic and red earth soils of tropical and subtropical China, as did Tokudome and Kanno (300) for red-yellow soils of Japan. Kononova (159) described similar properties in acid, leached soils on the Black Sea coast. Ponoraareva


27 (233) suggested that brown, red or yellow soils tended to have higher proportions of fulvic acids than did grey or black soils. Zonn (3^) extended this idea and classified the forest soils in the U.S.S.R. in terms of their proportions of fulvic and humic acids. Inorganic Interaction Sumner (288) demonstrated that ferric oxides and kaolinite form complexes which differ in charge from the original kaolinite component. The charge en the mineral is normally constant from pH Z,k to 5» but addition of ferric oxide caused a decrease in the negative charge at pH < 6, and an increase at higher pH levels. Follett (92) reported that the Fe~cxide was sorbed on the tetrahedral surfaces only, by an apparently crystal lographically specific reaction which was unaffected by excesses of NI-L , Ca or Al . Greenland and Oades (110) found that the form of the oxide determined the mechanism and degree of sorption that took place. The purity of the oxide was also important since sorption of SiO^ or organic matter could produce a net negative charge on the Fe -oxide and inhibit reaction with clay crystals. Host of the free Fe-cxides occurred in the fine clay soil fraction (282). Jackson (138) indicated that the polymers of hydrous Al oxides could be adsorbed on clays through oxygen atoms in functional groups. Precipitation of hydrolyzed exchangeable Al on exchange sites was considered likely (6k), as was the deposition of hydroxy -Al units on the surfaces of clays or other crystals (2^0). These mechanisms and the possibility of hydrous Al oxides occurring as impurities in hydrous Fe oxides represent the major means of Al involvement in interaction of


28 inorganic colloids in soils. Organic Inorganic It 1 Interaction between kaolinite and organic colloids was studied by Evans and Russell (88) « They reported that kaolinite sorbed more fulvic than humic acids and that the sorption of fulvic acids was enhanced by ions in the order N > Ca > K + , while humic acid sorption 2+ + + was enhanced by ions in the order Ca > H > K . They suggested* as did Martin and Reeve (183) presence of exchangeable Al facilitated the adsorption of organic colloids on the clay. The latter authors noted in Al-bei • that the sorption of organic material was complete at pH 2 to 3* but decreased to about 50^ at pH 6. The likely natures of bonds between organic colloids and clays were described on the basis of data from humic acid-Ascanit models by Alexandrova (8) • She characterised the interaction as aggregation by coalescence with the involvement of some intermolecula.r and H-bonds. 2+ Coalescence was enhanced by Ca and reduced by adsorbed water. The involvement of intermolecular and H-bonds was also suggested by Fripiat (lOl) . Mortensen and Himes (204) stated that hydroxyl groups, mainly on the basal surface, were possibly involved through the formation of polar and non-ionic compounds, Fripiat (lOl) indicated that bonds likely to be formed would have degrees of stability varying in the order: Si-O-H > Si-0-N > Si-0-C, and that monomolecular films of water would form bridges between the clay and the organic colloid in the complexes. The possibility of bonding between carbohydrates and exposed lattice


29 Al, or of coordination with adsorbed or lattice Al, was mentioned by Mortenfen and Himes (204). Purely ionic bonds between organic colloids and clays seem to have been ruled out. With hydrous oxides and organic colloids, interaction is regarded as involving the displacement of H from carboxylic groups by R(0H) , where R represents either Al or Fe (10, 270). Greenland (109) classified the mechanisms of interaction as coulombic attraction, coulombic forces, and ligand exchange reactions. Infra-red spectra indicated that electrovalent bonds were formed (270), but Alexandrova et al . (10) regarded the bonds as being coordinate in nature. Schnitzer and Skinner (270) showed that the molar ratio — metal/organic colloid— varied with pH. At pH 5 the molar ratio was about 3» but increased with pH as the organic colloid was replaced by OH ions. With a molar ratio of 3# R(0K) was the postulated ionic form of the hydrous oxide. Examination of extracts from a podzol B horizon gave molar ratios of 2.8 (336), and hydrous oxide of form R(0H) + (188). Organic colloids of the podzol B horizon were found to contain 85?5 fulvic acids and to be immobilized only when all the functional groups were engaged (33&). Ponomareva (233) stated that the organic matter-hydrous oxide complexes had pH values in the range 4 to 6, essentially agreeing with Waksraan*s (317) concept of 3-humus which gave the soil its buffering properties, was precipitated at pH 4.8, and consisted largely of ash. Organic matter has been shown to be flocculated by Al (183), and by Fe (184) but Tan (290) found greater saturation of organic colloids by Fe in the surface horizons, and by Al in the lower horizons of some Indonesian latosols. This confirmed the earlier conclusion (292) that


30 Fe was preferentially complexes "by humic acids and Al by fulvic acids. Titova (299) established that humic-Fe complexes were immobile in an electric field at pH 8.5 vrhile fulvic-Fe complexes in similar conditions yielded a precipitate of Fe(C. T l) and a mobile Fe-organic complex. D'Yakonova (83) indicated that complexation of Fe by organic colloids increased from about pH 2 and decreased above pH 4, but above pH 4,0 the proportion of mobile fulvic-Fe complex decreased and that of immobile Fe -humic complex increased, Alexandrova et al . (10) described the OM-metal complexes as "complex heteropolar salts" and indicated that complexed Fe lost its cationic properties completely and complexed-Al only partially. Part of the Al remained exchangeable. They showed that the OM-metal complexes displayed a "pH effect" during potentic.^ctric titration. The pH range of buffering differed with the .metal and the organic colloid. Humic-Fe complexes were buffered in the pH range 3 to 6, and humic-Al complexes in the pH range 6 to 10. Dialysis of the complexes raised the range of buffering to pH 6 to 10 for both complexes. Fulvic-Fe complexes were buffered in the range pH 2 to 6 and fulvic-Al complexes in the range pH 2 to 5< Dialysis of these complexes shifted the range of buffering to pH 6 to 10. Yuan (3^l) reported that on acidification of 0.5M NaOH extracts from the organic pans of three Florida soils, precipitation began at pH 7, and increased with increasing acidity to a maximum at pH 4.0 to 5.5. He considered this precipitation to be due to Al. The effect of complexation of the hydrous oxides by the organic colloids was also shown by Alexandrova et al . (10) to result in a marked


31 depression of the negative charge on the organic colloid,, This effect was much stronger where Al was complexed than was the case with Fe« The stability of these complexes was shown to vary not only with pH but also with the redox potential of the system (8), and with the cation (48). Complex formation was found to be a rapid process, limited by the rate of diffusion of cations into the solid phase (48). Khanna and Stevenson (l5l) concluded that th tals masked the chelation sites of the organic mater' "' .Nutrient Reten ' The net results of the interaction of soil colloids are reflected in the physical properties of soils throi and persistence of their cementing agents, and, therefore, '. their structural conditions. The effects are, however, more re seen in chemical phenomena, principally pH, the relative importance of the various forms of Al, and the capacity of the soil mass to retain nutrients. Retention of nutrients is effected through phenomena by which fixed charges on the surface are countered by ions of opposite charge in the adjacent layers of solution, at the interface between the solid particle and the liquid. This is the electrokinetic effect— the double layer (43, 223). Retention is also effected by moans of Donnan effects involving macromolecules which tend to remove electrolytes from the solution in order to maintain electroneutrality in the medium. Charges are fixed within the macromolecule as well as at the surface. The internal charges are neutralized by the diffusion of counter-ions into the gel, and the nature of these charges determines whether the external


32 solution develops an alkaline or an acid reaction (^3) • The relative amounts of the different soil colloids determine the contribution to nutrient retention mechanisms attributable to electrokinetic or Donnan effects. The clay minerals invoke electrokinetic effects while the organic colloids (106) and the hydrous oxides of Al and Fe are gelatinous and likely to invoke Donnan effects (43, 68) e Thomas (296) described some red-yellow soils of North Carolina as being both salt sorbers and cation exchangers. He reported the imbibition of electrolyte by these soils with the consequent reduction in salt movement and salt concentration in the soil solution, as well as the influence of anions on the retention of sorbed cations. His (. : suggested that a prominent role is played by organic matter and the products of its interaction with hydrous oxides. Oliveira etal. (222) concluded that the retention of K by some Brazilian Ultisois vras possibly associated with amorphous alumino-silicates. Van Reevwijk and DeViiliers (30*0 showed that such an association was possible, and apparently involved norma,! exchange sites rather than precipitation or other specific chemical reactions. They reported that the retention of K by this mechanism was pH-dependent and decreased in extent with increase in Al content of the Al-Si gels. Muljadi et al . (208) reported the penetration of K-phosphate into less crystalline portions of kaolinite. Slectrokinetic effects are generally considered to provide the most important mechanisms of retention of both anions and cations. Such mechanisms do not normally result in alteration of the chemical form of the charged surface, and ions are reversibly adsorbed.


33 Cation S o rption The capacity of a soil for retaining cations in exchangeable form is usually described as its cation exchange capacity — CSC. This capacity was shown to be the sum of sites of permanent negative charge (265) » and of those sites which develop a negative charge at a given pH. The cations which usually counter the negative charges in soils include (Al) + , H + , Ca 2+ , Kg 2 " 1 ", and K + . Of these, the various forms of Al indicated above as (Al) and H are considered sources of acidity and, therefore, undesirable. The remaining cations vary in quantity depending on soil type and soil reactions, but collectively constitute the degree of base saturation of the soil. The generally undesirable consequences of acidity underline the importance of the degree of base saturation and the controversy which surrounds the pH level most appropriate for CSC measurements. Bradfield and Allison (4l) found that soils in equilibrium with the atmosphere and having a pH value of about 8.2 were completely base saturated. Hehlich (192) and Clark and Hill (58) observed that base saturation -pH relationships varied .with soil type, but the buffered BaCl 2 TEA method of CEC determination at pH 8.2 was shown to give maximum values and to include a component due to the anion exchange properties in soils rich in hydrous oxides. Coleman et al . (63) measured the "effective" CEC at the pH of the soil with a neutral unbuffered salt and found that it included some pH -dependent sites. This was confirmed by Bhumbla and McLean (29) who compared the H KC1, NH^OAc (pH 7.0) and BaCl 2 -TEA (pH 8.2) methods of CEC determination. The N KC1-CEC accounted for 30 to 1§% of the BaCl 2 -TEA CEC at pK 8.2


3^ in red-yellow soils, and the actual proportion was believed to be determined by the extent to which the soil particles were coated with Fe-oxides (66, 190) or Al oxides (65). The NH^OAc-CSC method at pH 7.0 gave values intermediate between the N KC1 and BaCI -TEA (pH 8,2) methods (29) probably because it included more pH-dependent sites than the N KG1 method, but fewer than the BaCI -TEA (pH 8.2) method. In addition to the pH of measurement, the saturating ion also affects the measured CSC values. Variations in values with the valence of the saturating ion were reported by deEndredy and Quagraine (?6). With monovalent ions, C3G was independent of pH in the range 7 to 8, but varied with pH in this range when divalent ions were used. They concluded that the ratio monovalent C2C:divalent CSC was dependent on the clay type, and to a lesser extent on the organic colloids. However, Bhumbla and McLean (29) reported that the difference between + 2+ values obtained using NK. and Ba as saturating ions was highly correlated with the amount of organic matter in the soil. This correlation can probably be explained by the idea (76) that monovalent saturating ions may have caused the dissolution of some of the potential exchange sites in the soil since the alkali salts of organic colloids are soluble, whereas similar alkaline-earth salts are relatively insoluble. Alternatively, at the higher pH levels incompletely hydrolyzed hydroxides of divalent alkaline earth metals may be adsorbed (53) » possibly by phenolic hydroxyls of the organic colloids (169, 213), -thereby over estimating the CEC. The likelihood of this is strengthened


35 "by Posner's (235) findings that huraic acids had a higher C3C at pK 3 to 7 with divalent than with monovalent cations. Reeve and Sumner (248) introduced the concept of "net" CEC based on an Al-equillbrium rather than on pH, per se. The criterion here is exchangeable Al, but in effect the concept seems to be closely parallel to "effective" CSC. Their allusion to free hydrous Al-oxides, exchangeable Al, and Al-organic matter complexes in the context of CSC is apparently justified by the findings by Frink (102) that the change in CSC on removal of sequioxides from the soil by citrate extraction was highly correlated with the Al removed, but not with Fe removed. Pratt (237) stated that neutral salts did not replace pH-dependent acidity, and that CSC was only increased after the acidity was released by an increase in pH. Bhumbla and McLean (29) noted a twofold increase in N KC1-CSC after liming, and the change was attributed to the release of Al from previously occupied exchange sites. The efficiency of K as the saturating cation is perhaps related to the large affinity for K displayed by acid clays either with a low degree of K saturation or largely Al-saturated (218). Cation Retention Patterns The total amount of adsorbed cations is determined by the CEC of the soil, but the proportions of the individual cations vary with the bonding preferences obtaining in the soil. The strength with which a given cation is held will vary with the nature of the colloid, the degree of saturation of the colloid, and the nature of the complementary ions (84, 325). The type of colloid influences the strength of bonding through the location, and density of charge (325). Wiklander (325)


36 showed that the acid strength of the site varied with the location of the charge and that at pH 6.1 hydrolysis occurred more readily than at pH 4.9 to 6.1, so that cations held by sites which are functional only at pH values higher than 6.1 are likely to be more easily lost (324). In the case of crystalline clays all negative charges act at the surface (223, 250), through coulombic attraction, and the bonding strength for a given ion at a given site varies with valence and hydration radius of the ion. The order of strength of bonding can be expected to be Al 3+ > H + > Ca 2+ > Mg 2+ > K + (325). Wiklander and Ghosh (326) showed that kaolinite bonded Ca relatively weakly. 2ck et al . (84) concluded that at saturations greater than 45 to 60%, kaolinite 2+ readily released Ca to plants. With organic colloids, the functional groups are not necessarily superficial and the solubility or stability patterns of organic mattermetal complexes may become more important than the ionic double layer. Alexandrova et al . (10) indicated that Ca , Kg , and K + can be bound by organic colloids in soluble or insoluble heteropolar salts, and Kononova (159) pointed out that the fulvates and humates of alkali metals are soluble at all pH levels, while alkaline -earth humates are less soluble, Ca-humate more so than Kg-humate. Schnitzer and Skinner (272) reported that fulvic acids complexed divalent ions with the 2+ 2+ 2+ 2+ order of stability Cu > Fe > Ca > Mg regardless of pH, but that the molar ratio, metal/fulvate, increased with pH. Broadbent and Ott (48) obtained a similar order of stability, Cu 2+ > Ba 2+ > Ca 2+ > Mg 2+ , with organic matter extracts. It is noteworthy that only the carboxyl


37 groups of the organic colloids are dissociated at pH less than 7.0 (159). Taking the soil as a whole, Spencer (28^) established that in soils in which more than 75% of the CSC was contributed by the organic fraction, cation adsorption patterns were different from those in soils with larger proportions of inorganic CSC, Yuan et al. (3''J-6) related the size of the contribution to the soil CSC made by the organic fraction to s:>il type. They obtained values of $6% for Spodosols as compared to 67% for Mollisols in Florida. The pH-de pendent nature of the contribution to the soil CSC by organic matter was demonstrated by Helling et al . (120). In soils of high organic CSC, Ca 2+ and K + compete strongly for exchange sites excluding K completely, and Mg to a large degree (289). McGeorge (187) demonstrated that the affinity of organic matter for K was very low compared to that for Ca and'Mg ions. Mehlich (193) considered that K + was barely held by exchange sites in organic matter and Gammon (104) concluded that most of the K in cultivated sandy soils in Florida was in the form of soluble salts rather than adsorbed on the exchange complex. Organic matter held Ca more strongly than kaolinite (200), but substantial quantities of water soluble organic matter-Ca complexes have been found in soils (213). Hydrous oxides also modify cation retention patterns. Mehlich (195) 2+ noted that hydrous Al-oxides increased the amount of Ca in solution and suggested that the oxide affected the nature of the bond between Car and the exchange site. There seemed, however, to be no effect 2+ on the ease of release of Ca , by the oxide. Hunsaker and Pratt (134) observed that there was a marked preference for Ca 2+ over Mg 2+ adsorption


38 in red latosols, but no indications were given as to the relative effects of the organic natter and the hydrous oxides present. Hydrous oxides were considered to be involved in the equilibrium, exch-K k=t— non-exch-K, thought to exist in Brazilian Ultisols (222). The low supply of exchangeable K in highly weathered soils may possibly be related to the existence of such an equilibrium or to the capacity of these soils for electrolyte sorption (296), but Graham and Fox (108) found high correlation between labile K and rainfall in similar soils. This conceivably derives from the soluble nature of salts of K and the organic colloids, and to the greater freedom of movement of K in Al saturated sandy soils (297). In sandy soils the movement of K was increased by increase in soil pH (313» 32*0, with K moving primarily as EIO^ and K SO^ (313). Anion Sorption The sorption of anions by soils has been described by Wiklander (325) as comprising the processes of negative and positive adsorption. He indicated that negative adsorption of anions is due to mutual repulsion of similarly charged particles and that it is influenced by the nature of the charged surface, the hydration of the ions, the salt concentration, and the pH level of the soil. Negative adsorption results in a deficit of anions in the immediate vicinity of the charged surface, and thereby facilitates the loss of anions by leaching (325). Positive adsorption is due to electrostatic attraction and chemical bonding, with bonding strengths increasing in the order NO, = CI < SO^ < PO^ (27). In addition to the role of


39 the ion itself, positive anion adsorption is influenced by the nature of the charged surface and the pH of the medium (325) • Berg and Thomas (2?) found that all sites which held anions were pH-dependent in effect, and considered the process of adsorption to be the result of displacement of hydroxyl groups. Samson (264) demonstrated that the process of positive anion adsorption involved the exchange of hydroxyl groups of the octahedral layer of clays by the anion, and that the process was limited by the availability of hydroxyl groups. His computations showed kaolinite to have a higher concentration of hydroxyl groups than montmorillonite. Schofield and Samson (266) attributed the increase in anion adsorption at low pH levels to proton sorption, and, on the basis of this, Muljadi et al . (206, 207) proposed the reaction: . K, ^^(H 2 0) Step 1. JTA1 OH + H + + OH" O ai + .*, '"••OH" \ ^(H 2 0) '_ K 2 .^-(HpO) Step 2. _J Alt + H ? P0, "> Al*.r7 + ° H '•••OH" C ' •-..(H 2 P0^-) in which — • indicates coordinate, and , electrovalent linkages. Hingston et al . (123) constructed adsorption 'envelopes* for different anions on goethite and deduced that the mechanism of adsorption varied with the anion. They found that non-protonated ions did not influence the charge on the surface, and were sorbed in the diffuse double layer so that the proportion of the anion in solution was the same as that sorbed. When the anion was partially protonated, it had the capacity to either accept or donate a proton to the surface, and could, therefore, modify the charge on the surface. When the partially protonated anion increased the negative charge on the surface, it was


40 bound to the surface to a greater extent than would be expected from its proportion in the solution. The anion formed a coordinate linkage with the metal rather than with the hydroxyl, and the extent of its sorption was maximal at the pK value of the corresponding acid species (123). Jones and Handreck (144) also implicated protons in the sorption of Si(OH) , but considered this different to the sorption of H 2 P0. ~ which was bonded through oxygen atoms. The effects of anion sorption in soils have been reported by various workers. Schuffelen and Fdddleburg (273) characterized the release of OH" by anion sorption as "exchange alkalinity." Mehlich (194) observed an increase in negative charge and formulated the concept of the CSC: AEG ratio. He regarded CSC measured after phosphating the soil as the maximum CSC (196), and later concluded that anions were differently bound. His conclusion was that CI" remained in the ionization sphere countering positive charges while SO^ ""* and H 2 P0^" entered the coordination sphere (l97)» hut H 2 P0 2+ " did not add very much to the anion exchange acidity of the soil (198). Bar-Yosef et al . (24) theorized that either the anion did not replace OH" or the displaced OH" remained bound to the surface. Kin jo and Pratt (152) found that NO " sorption on a Brazilian Oxisol decreased to zero with increase in pH to 5«5» hut that the neutralization of replaced NO ~ by H was not stoichiometric. They established that amorphous materials were responsible for N0~~ sorption, and considered the non-stoichiometric relationship to be due either to an inherent difference in anion sorption by soils dominated by Al and Si rather than by hydrous oxides or to the CEC:AEC ratio of the


41 2soil. They found also that though SO^ replaced NO in a stoichiometric manner, there was a 0.5 to 1.0 unit increase in pH. A 111 replacement of N0~~ by H 2 H\" was obtained initially, but this changed to values of 1:3 or 1:4 indicating that P reacted with two or more sites not available to NO ~ (153). They noted that P enhanced the negative adsorption of NO " at low pH levels, and concluded that some charged sites had different affinities for different anions. Muljadi et ai . (209) reported differences in affinity for P by different sites, and concluded that the process of adsorption resulted in an entropy change. Phosphate Retention in Soils Phosphate retention is frequently used interchangeably with phosphate fixation, but it seems preferable to reserve the latter term for those processes which reduce the availability to plants of retained phosphates. Bradfield et al . (42) described P retention as being due to a wide range of reactions which involved distinct but overlapping mechanisms. Kardos (148) summarized the mechanisms as adsorption, double decomposition, and isomorphous replacement. Low and Black (l?6, 17?) reported that P replaced Si in kaolinite in acetate systems at 60C buffered at pH 4.5, and at 45C buffered at pH levels in the range 4.5 to 5.1 by dissolution and precipitation according to solubility product principles. Reifenberg and Buckwold (249) obtained replacement of Si by P at room temperature in the pH range 6.8 to 8.5. They noted that the degree of replacement was higher in finer than in coarser


42 textured soils, and that at pH 7.5, the effect of pH wag minimal. The replacement of Si increased, with increase in P concentration, and with time. Kafkafi et al . (146) found P retained by kaolinite to be isotopically exchangeable initially, but to become irreversibly bound on washing. With larger amounts of P, some Si was replaced and more of the added P was irreversibly held. Hsu (129) interpreted his data to mean that Fe-P formed on decomposition of Fe~silicates, but considered the process a surface reaction at pH 6.4 to 7«0. Klttrick and Jackson (156) examined the reaction of P and hydrous oxides and concluded that P retention was generally due to chemical precipitation. Hsu and Rennie (l3l) reported that P retention comprised two reactions which differed largely in rate; the first and main reaction was rapid, and was followed by a slow double decomposition process. Coleman (6l) also detected two reactions, but found the slow reaction to occur at pH < 5, and the rapid reaction at pH > 5, He did not observe any replacement of Si from the clay minerals, and attributed all P retention to the hydrated oxides coating the clay minerals. Hsu (130) reasoned that P retention by hydrous Al oxides did not involve precipitation, but rather surface adsorption, since there was no obvious destruction of the initial surface. Eriksson (86) concluded that P retention was probably due to a physico-chemical process by which OH ions, possibly from sesquioxides, were exchanged for P ions, or to a purely physical process of polar adsorption to soil particles. Bache (l?) distinguished three stages in P sorption: a high energy chemisorption process, precipitation of



44 despite increases in the amount of sorbed P, or the number of sites was large compared to the amount of P sorbed. The pH of the medium did not appear to influence the proportion of sites in regions I and II. Though the mechanism of P retention is apparently the same for different surfaces, the amount of P retained varies with the size of particles and their degree of crystallinity. Colwell (67) attributed P retention to the extent of 7»34# and 0.05^ ( w / w ) ty amorphous hydrous Al oxides, and gibbsite, respectively, to differences in particle size. Muljadi et al . (207) found that at pH 3.0 kaolinite sorbed 2.95mM P/lOOg while gibbsite sorbed 45mM P/lOOg. Fox et al . (93) reported that the intensity of P retention decreased in the order: amorphous hydrated oxides > cryptocrystalline goethite > gibbsite > kaolinite > 2:1 clays j with pH 5 to 6 permitting optimum P solubility. Ross and Turner (258) found the rate of ciy stall! sat ion of-Al (OH) to be affected by the presence of anions. The effects varied with the size and structural complexity of the anion. Juo and Ellis (145) indicated that colloidal precipitates of Al-P and Fe-P formed rapidly. The Al-P colloids crystallized more slowly and had greater surface area than Fe-P. Sorption of P by organic matter-Fe complexes was reported by Weir and Soper (320), and by Levesque and Schnitzer (168). The latter authors indicated that P was bonded to the metal — Al or Fe — in fulvate complexes and could replace fulvate groups. The amount of P thus retained varied with the G content of the complex, lower C contents permitting higher P sorption.


^5 The degree of overlapping of the different mechanisms is, no doubt, dependent on the composition and reaction of the soil. Soil reaction governs the ionic form of the hydrous oxides present and the availability of divalent cations. Its importance is reflected in the postulate by Bradfield et al. (42), that at pH 2 to 5» P was retained in soils as precipitated Al and Fe phosphates; at pH 5 to 7, P was adsorbed on clay surfaces; and at pH 6 to 10, P reacted with divalent cations. Wild's (327, 328) conclusion differed somewhat from this, but reiterated the influence of pH on the mechanisms of P retention. He concluded that at pH levels < 3 to 3«5 P "as retained by Fe, and that Al-P formed at pH 6 to 7. Chang and Jackson (55) observed that on addition of phosphate to a podzol at pH 5 Al-P and Fe-P increased. They also observed the formation of Ca-P at higher pH levels as suggested by Bromfield (^9). Lucas (178) reported increases in Al-P and Ca-P with addition of P to an amorphous soil. Increases were more marked for Al-P than for Ca-P, while Fe-P decreased. Yuan et al . (3^7) noted initially that the Al-P and water soluble P increased while P additions left Ca-P unchanged, but decreased the percentage of P as Fe-P and Ca-P. Shelton and Coleman (275) found that added P was retained in red clay soils as Al-P and Fe-P for more than 6 months with a greater fraction initially as Al-P. Chang and Chu (56) considered the moisture content of the soil to influence P retention. On relatively dry soil, P was retained mainly as compounds, with relative amounts decreasing in the order Al-P > Fe-P > Ca-P, regardless of pH. Prolonged contact caused Ca-P and Al-P to change to Fe-P, and the rate of change increased with


46 increased moisture. Eriksson (86) distinguished between octo-P and ortho~P of Ca, pointing out that the latter, though more soluble, formed only in the presence of relatively large excess of Ca. The possibility of formation of Ca-Al-P compounds was indicated by Taylor et al . (295) and by Huffman et al . (133). The role of hydrous oxides in P retention by soils is influenced by several additional factors, such as, the form of the oxide, its relative quantity and solubility, and the presence of competing ions. Hsu and Rennie (132) observed that exchangeable Al, held by a resin, was precipitated by P, and Coleman et al . (62) obtained high positive correlation between exchangeable Al and P sorbed by soils. Tandon (293) estimated that 5 or 6 Al atoms were involved in the retention of one P atom* and suggested that the Al was in polymeric form. He obtained better correlation between NH. F extractable P than total P, and NH^F extractable Al, indicating that all P was not retained by Al. Williams (330) reported positive correlations between P retention and hydrous Aloxide content of soils, but noted that the role of the oxide was modified by pH and organic matter content. The A^O^iFeJ)^ ratio also affected the role of hydrous Aloxides in P retention (295* 330) . Having observed some correlation between P retention and organic matter content, Williams (330) suggested that Fe-organic matter complexes may have been involved. LeMare (l66) also obtained correlations with pH, organic matter content, and P potential. Halstead (118) found that organic P was associated with Fe-P, but that Al-P was the major form of P in the silt and fine clay fraction, except in podzols where Fe-P predominated. He found by the


47 fractionation procedure of Chang and Jackson (54) that saloid-P, Al-P, Fe-P, and Ca-P accounted for 62 to 92;3 of the inorganic P with Ca-P being the major form of P in all soils except podzols. Ramulu et al . (244) found a closer relationship "between oxalate extractable Fe and P "fixation" than with dithionite extractable Fe. Boiling 0.1N NaOH extractable Al was highly correlated, 0.1N HG1 extractable Al very poorly correlated, and N KC1 (pH 2.0) very well correlated with P "fixation," However, N KC1 extracted < 1.07a of the soil Al (244). Yuan (342) found 0*1N HC1 extractable Al to be highly correlated with the P sorption capacity of some Florida soils, and Lucas (178) obtained good correlation with P sorption and both N NH. OAc (pH 4.8) and 0.1N HC1 extractable Al. Correlation with the latter was somewhat better. Puri and Swarnakar (238) reported similar trends in extraction of Al, Fe, and P by N KC1 at different pH levels. Minimum values were obtained in each case at pH 6 to 7. Halstead (118) observed that oxalate soluble Al and Fe were related to P retention, but closer relationships were more evident with Al in neutral soils and with Fe in acid soils,, The influence of competing ions on P sorption was shown to vary with the pH, the anion, and the soil type by Deb and Datta (?3» 74). The pH of maximum effect of any anion was the pK of maximum stability of the complex it formed with the adsorbing surface (73) • At pH 6 Si0 2 was most effective in replacing P, but its effect was greater in alluvial than in red soils. Citrate was most effective at pH 7 (74). Acquaye and Tins ley (2) suggested that Si02 might influence the P status of a soil, but that this influence was reduced by humus at pH 6 to 9»


since humus enhanced the solubility of SiCv, in that pH range. Beckwith and Reeve (25) reported the increased sorption of Si with increase in pH, but Jones and Handreck (l44) doubted the efficiency of Si(OH). in reducing the sorption of P. Hingston et al. (123) demonstrated that the competition between P and other anions was more akin to desorption than to exchange. The desorptive capacity of an anion was dependent on the pK value of the corresponding acid. Silicic acid had a pK of 9.2 while ^PO^" had a pK of 6.5 (123). Fulvic acid was shown by Leaver and Russell (163) to compete effectively with P at pH 5.5, but Levesque and Schnitzer found that the effectiveness of fulvates decreased as pH increased (168). This agreed with the concept of maximum adsorption near to the pK value (123) . Hesse (122) concluded from studies in mangrove swamps that P was retained by organic matter, but the necessary role of a metal was stressed by Wild (328) and Levesque and Schnitzer (l68). Fox and Kamprath (96) failed to obtain any appreciable P retention by organic matter in the absence of complexed cations. Nutrient Availability and Plant Uptake The availability of nutrients to plants is governed to a large degree by the amount and chemical form of nutrients retained by the soil. This in turn influences the amounts taken up by the plant. The stage of growth (302) and the moisture regime which obtains during the development of the plant (339) also modify nutrient uptake. Generally, the plant exerts some influence on the chemical form and concentration of nutrients in the immediate vicinity of the roots (20, 253, 259,


49 278), and may thus effectively increase the availability and uptake of nutrients. Rovira and McDougall (259) stated that a layer of mucilaginous material on the epidermis extended the ion-exchange zone of the root by virtue of its COOH groups, and that root excretions could help to increase the solubility of several nutrients in the soil. Root excretions are largely organic acids (259, 278) which Stevenson (286) indicated in his discussion, were capable of lowering the pH and forming complexes with Ca, Mg, Al, and Fe. Their effects may be important only in localized zones near decomposing material or actively growing roots (259), but do not significantly alter conditions in the soil as a whole. Riley and Barber (253) reported salt concentrations in the rhizosphere of soybean plants 5 to 15 times greater than those in the non-rhizosphere soil and suggested that the concentration of cations in the roots was regulated by the physiology of the plant rather than by the supply to the roots. Barber and Ozanne (20) found 2+ an increase of 1.0 pH unit and a 7-fold increase in Ca in the rhizosphere of soybeans and concluded that HCO " had been excreted by the roots. Rains et al . (2^l) mentioned the possibility of competition between HC£u"\ OH", and other anions for plant uptake at higher pH levels . The high number of variables in P availability and the frequency with which this element has been found to limit plant growth have resulted in the expenditure of much time and effort in the study of P availability. Jackson et al . (139) suggested that the rate of plant uptake of P was limited by the rate of oxidative phosphorylation within the roots and that the supply of P was quantitatively restricted by the


50 properties of the external solution. Williams (33l) defined parameters of the P status of the soil as: (i) capacity factors, vrhich determined the potential supply of P. (ii) intensity factors, which determined the strength with which P is held "by the soil. (iii) rate factors, which are determined by the degree of saturation of the soil with P, the rate of release of P by the soil, the rate, of diffusion of P from within reserves, and the rate of plant uptake. Barber (19) stated that diffusion was the dominant factor controlling P availability, and Hagen and Hopkins (ll6) attributed decreases in P uptake by excised barley roots with increasing pH to increases in the concentration of divalent P species. Place et al. (23l) found maximum availability of P at pH 5.1, and noted that pH affected diffusion particularly in kaolinite systems. Water soluble P was highly correlated with the diffusion coefficient, but pH > 5.6 and the presence of Fe-P decreased the diffusion coefficient. Phillips et al . (229) reported that the effect of Fe-P was significant only in kaolinite systems, and Peaslee and Phillips (22?) found that associating ions influenced P diffusion. Ammonium ions depressed P diffusion more than Ca while SO. facilitated diffusion more than Cl~ or NO ", Birch (30) considered the degree of base saturation of the soil important in P availability because at higher base saturation, more P was likely to be in saloid than in colloidal form. Lewis and Quirk (170) observed variations in P diffusion with levels


of P applied. Higher levels of application resulted in greater diffusion. Murraman and Peech (210) concluded that the amount of H 2 0soluble P in the soil was dependent on pH and on the size of the labile pool, rather than on the dissolution of crystalline forms of P. Gtuiary and Sutton (113) suggested that the size of the pool of labile P and the rate of root extension were the important factors in plant uptake of P. Lewis and Quirk (172) demonstrated that sorbsd P could be utilised by plants, and that plant uptake varied with the rate of P application (l?l). Gerrotsen (105) concluded that isoluble P could be solubilized in the rhizosphere and Dalton et al . (70) found that added organic materials enhanced P availability probably because decomposition products complexed Al and Fe. Riley and Barber (254) reported decreased P uptake by soybeans at pH > 4.0, and higher P uptake in presence of NH4 4 " than with N0o~. The relative effect of NH^ varied with the rhizosphere pH which they suggested depended on the rate of P adsorption. The rate of release of P by the soil materials, is thus dependent on the form of P in the soil, the degree to which the sorption reaction is reversible, the ionic contents and reaction of the soil solution. Since P retention tends to result in the formation of relatively insoluble compounds, the solubility of these compounds was thought to be a suitable measure of plant availability of P. Pratt (237) concluded that soil P was most soluble at pH 4.2. Several studies have been reported on the effects of individual cations, salt concentrations, organic matter, and clays on the solubility of soil P. Bradfield et al . (42) suggested that the dissociation of H^PO^


52 was facilitated by clays and Ca(0H) 2 so that at pH 7.0, P0^~ was likely to be the dominant ionic form. Wild (328) noted that Ca 2+ depressed the solubility of Al-P. Lebr and van Wesemael (l64) indicated that the depressing effect of cations on P solubility in soils of pH < 5 "as in the order K > Mg > Ca , and that the depression due to any of these cations was independent of pH. In the 2+ presence of exchangeable Ca the formation of Al-P and Fe-P was not influenced by salt concentration, but Ca-P formation could be superimposed (308). Nye and Bertheux (220) stated that organic matter prevented the precipitation of P by Al and Fe, and Weir and Soper (320) found that about 68% of the P retained by Fe-organic matter complexes was available to plants. Swain (289) cited reports by Apuski on the influence of organic matter on the solubility of soil P. Alkali humates were reported to impede the precipitation of Ca-P at pH h,Z to 8, while raw humus dissolved P from Ca-P. Humic acids leached P from Al-P and Fe-P (289). Huffman et al . (133) described several complex forms of Al-P and Fe-P that could be formed in soils and stated that they all dissolved in water, ultimately yielding Al-P or Fe-P. Minimum solubility of Al-P was considered to be at pH 6.0 (210). Lindsay and Taylor {VJk) concluded that plant availability of P was correlated with neither the solubility nor the insolubility of P compounds, but was instead related to the rate of formation of the compounds. Materials which formed slowly, e.g., basic Al-P, were more stable and P in them less available; materials which formed rapidly,


53 e.g., Ca-P provided the most readily plant available P. Juo and Ellis (145) indicated that colloidal precipitates of Al-P and Fe-P formed rapidly and had more surface area than the crystalline materials. They considered the P in the colloidal precipitates more readily available to plants, and the P supply from such compounds to be controlled by the degree of crystallinity in the order: strengite < varisite <« colloidal Fe-P = colloidal Al-P. Taylor et_al. (294) found colloidal Al-P and Fe-P to be fairly good sources of P, but that plant uptake from these sources was decreased by incubation. Williams (331) stated that most of the plant available P in soils was adsorbed in non-stochiometric complexes on Al and Fe of hydrous oxides, humates, and clays, and in easily acid-soluble Ca-P compounds. Al-Abbas and Barber (?) found Fe-P most highly correlated with plant available P, but Halstead (118) and Lucas (178) concluded that Al-P was possibly the most important source of plant available P. Shelton and Coleman (275) suggested that the maintenance of high levels of plant available P for long periods of time depended on the relative proportions of Al-P and Fe-P, and on the rate of conversion of Al-P to Fe-P. Stelly and Ricaud (285) compared plant availability of P to extractability and obtained best correlations with the Bray #1 extractant. Griffin (ill) found this extractant to be most uniformly consistent on several Connecticut soils, but observed that it gave 23?S lower P extraction at pH 7.7 than at pH 5.3. Levesque and Schnitzer (l68) reported that the Bray #1 extractant removed less P from complexes with Al and Fe-fulvates at higher metal :P ratios.


54 The most meaningful measure of plant availability of P is plant uptake, Lehr anc" van Wesemael (165) noted that P uptake by plants was dependent on the pH and on the dominant cation in the soil. Calcium ions reduced P uptake relative to Na + . Fox et al . (°A) showed that maximum plant P uptake varied with soil pH, soil composition, and plant species. Hingston et al . (123) suggested that P uptake was influenced by the capacity of excretions from plant roots to desorb P from the soil mass. Wright (333) reported the inhibition of P uptake by Al in the soil. Graham and Fox (108) established that labile K in highly weathered soils was highly correlated with rainfall. The pool of labile K was ?-r small presumably because of the high relative activity of Ca and K concentration in plants was about 0*5fo, Sartlett and Mcintosh (23) considered K availability to be influenced by the degree of K saturation of the soil, but Koch et al . (158) reported that the uptake of K and the concentration of K in corn grown on Natal oxisols were not influenced by the K status of the soil. Oliveira et al . (222) found that plant availability and uptake of K markedly exceeded exchangeable K contents in some Brazilian soils. Plant uptake amounted to < 22?S of the total soil K, and even though the concentration of K in plants was about O.kfo, the ratio plant K uptake: soil exchangeable K, varied from 3.3 to 7.2 in Ultisols, and from 3.8 to 9.5 in Oxisols. Amelioration of Fertility in Red-yellow Soils The low level of inherent fertility encountered in red-yellow


55 soils of the tropics is generally attributed to the low base status and low CEC, high P fixation, high permeability, and low water holding capacity (26, 136, 225, 203). In discussing some of the red-yellow "upland soils" of the Amazon basin, Sornbrosk (283) observed that whether under forest or savanna, practically all of the CSC was due to the organic fraction. The base saturation relative to CSC at pH 7.0 (potential CEC) vras always less than kO/o with an average value of 15/S in the A horizon. Low base saturation is associated with high acidity, and methods devised for the amelioration of the inherent fertility status must, of necessity, achieve the correction of acidity and improve the degree of base saturation. Applications of line have been regarded as the best method of correcting soil acidity and the addition of the appropriate fertilizers has served to increase the base saturation. On red-yellow soils, liming has been found to have adverse effects if pH levels above about 5*5 were produced (51» 178, 2^6). Soil acidity and plant growth Soil acidity has been described by Schwertmann and Jackson (27^) in terms of pH ranges of buffering. They indicated that at pH levels lower than 4.0 HoO vras the acid entity encountered, but in the range pH 4.0 to 5»6 ionic Al was the source of acidity. Yuan (3^0 ) reported a narrower range, pH 4.3 to 5.4. Beyond pH 5.6 hydroxy-Al compounds were considered to be the source of acidity (274) . The effects of soil acidity, per se, on plant growth have been studied in some detail. Several attempts have been made to isolate


56 the effects of H^O from those of Al in its various forms. Hackett (ll4) reported that the germination of several grass species was not affected by H_0 , but that shoot weight decreased as pH fell from 6.$ to 5.5. Rains et al . (24l) suggested that HO* might have impaired overall nutrient uptake through competition for carrier sites or derangement and/or damage to the absorptive mechanisms. The latter may have been due to suppressed dissociation of weak acids at low pH or to progressive and ultimately irreversible alteration in cell structure and function. The addition of Al to acid media decreased and delayed germination of several grasses (ll4); at low Al concentrations (< 5 ppm) root weights were increased in some cases, but generally there were decreases in root and shoot weight with increasing Al concentration. Hackett (ll^) concluded that few grass species were tolerant to both low pH and Al. Vlamis (312) pointed out that when plants were susceptible to low pH (Ho0+) the effects of Al were difficult to detect. Since, however, the pH of most soils is > U-, it would seem that the effects of acidity which are of practical significance are likely to be due to Al. The adverse effects of Al have not been completely elucidated, but several possible mechanisms have been suggested. It is apparent that the adverse effects of Al include Al uptake and its consequences on the uptake and translocation of nutrients, induced metabolic disorders, and the adsorption and/or precipitation of Al on root surfaces. The effects of Al are normally expected at pH < 5.6, but Soileau et al . (28l) observed Al uptake by cotton plants at pH 6.0 to 6.5, and attributed this to the resolubilization of precipitated Al(OH).


57 in the microenvironment of the root. Rees and Sidrak (2^5) encountered Al toxicity at pH 7.0 to 8 .5 on fly-ash soils. These data suggested that Al might he taken up in several forms. Johnson and Jackson (l^l) reported the uptake of ionic Al by wheat seedlings and DeKock and Mitchell (?8) suggested that chelated Al might be taken up by plants. They countended that the charge on the chelate influenced its uptake, lower charges being associated with easier uptake. After uptake the metal is split off by metabolic processes (7?). The uptake of Al has been shown to increase water absorption, reduce uptake and translocation of Ca (l4l), and disturb the K/Ca balance within the plant (2^5) . Foy et al . (99) established the occurrence of Al-induced Ca deficiency in some soybean plants by autoradiography, and suggested that this differed from absolute Ca-deficiency in terms of the distribution and chemical form of Ca within the plant. Johnson and Jackson (l^l) found, however, that precipitated or chelated Al exerted no appreciable influence on the uptake of Ca by wheat roots. The capacity for uptake and translocation of Al varies with plant type. Ahmad (4) reported greater immobility of Al in roots of maize than in those of the cowpea. Ouellette and Dessureaux (224) observed differences in the amounts of Al translocated in different clones of clover. Jones (l^2) concluded that the amount of Al within a plant was determined by the nature of the buffer system of that plant. Immobilization of Al by the roots does not lessen the adverse effects of Al on plant growth. Wright (332) established that the


58 ratio Al(roots)/Al(tops) was highly correlated with yield of barley even though Al (tops) was virtually constant. Generally, the visual symptoms of Al toxicity are localized in the roots. Inhibition of branching and the development of a brownish discoloration (332) of coralloid or stunted roots (k) are now regarded as distinctive symptoms. Immobilization of Al in the roots has been described as an accumulation of Al in the protoplasmic nuclei (l89)» or in the vacuoles (256). In the latter case toxicity is thought to be controlled. The mechanisms of immobilization seem to involve adsorption on exchange sites and/or precipitation on root surfaces. Rorison et al . (257) considered Al toxicity to be due to a surface reaction involving saturation of the exchange sites rather than a process of active absorption occurring in the presence of ionic Al. Greatest effects •were believed to occur early in plant growth, when the source of nutrients was being changed from the seed to the soil. Clarkson (59) concluded that Al accumulated in the Donnan free space of the roots and prevented exchange reactions with other cations. The adsorbed Al became firmly bound to the cell wall either by precipitation, outside the plasmalemma, of Al(OH)^ formed on the hydrolysis of Al(OH) + at pH < 6.7 (59), or by reaction of Al with -COOH groups causing the crosslinking of pectic materials (60). Lance and Pearson (162) suggested that the plasmalemma was the site of injury, Al causing changes in the structural configuration of the membrane. The immediate consequence of the immobilization of Al in the roots,


59 is the restriction of movement of P. Wright (332* 33 ; reported the inactivation of F by Al on roots in contact with ionic or precipitated Al, and it was later demonstrated that this P was held in inorganic form (335). The inactivation of P creates conditions of P-deficiency in the shoots (242), and merlstemmatic regions of the root (335). and restricts cell division by impairing the synthesis of DMA (263), and the hexose phosphorylation processes (59). Clarkson (60) reported that the molar ratio of HgPO^/Al in inactivated form tended to a maximum value of 0.3. The susceptibility of plants to Al toxicity is known to vary with plant type, species, and variety. Rains e_t_al ._ (24l) found differences in tolerance to Al among grasses. Foy et al. (98) reported that wheat varieties varied in tolerance to Al. Ahmad (4) indicated that maize was more susceptible than cowpea to Al toxicity. The evidence suggests that susceptibility is a genetic trait, and various mechanisms have been proposed. Foy and Brown (97) concluded that Al tolerance was related to the capacity for absorption and utilization of P in the presence of excesses of Al. Jones (142) suggested that the buffer systems of Al tolerant plants contained an excess of organic acids which either caused the precipitation of Al in the root zone or maintained the mobility of Al within the plant. The implied principle of exudation of organic acids was also suggested by Foy et al. (93). They regarded wheat plants which caused greater acidification of the root zone as being more sensitive to Al toxicity. Vose and Randall (314) found better relation of Al tolerance to root CSC than to soil type or geographic origin. They concluded that low root CSC was


60 associated with tolerance, but this seems to conflict somewhat with the concept that saturation of the exchange sites resulted in marked Al toxicity symptoms (25?). Different forms of Al have been shown to be conducive to Al toxicity and to impaired nutrient utilization. Various attempts have been made at characterization of Al in relation to extractability and relative "activity." Ramulu et al. (244) used 0.1N NaOH (100C) and reported the extraction of free Al and some Al from Al -silicates. Levesque and Schnitzer (167) used NaOH varying in concentration from 0.1 to 1.0N and considered Al extracted to be that complexed by organic matter. Tandon (293) extracted Al with N NH. F at pH 8.2. Several workers (174, 190, 218) used N KC1 at pH 5,7 to remove exchangeable Al and others adjusted the pH to 2.0 (238, 244). Yuan (342, 345) used N NH^OAc at pH 4.0 and 4.8, and 0.1N HC1, while Tweneboah et al. (301) used 0.1M CaCl 2 (pH 1.5) to estimate "active" Al. Lin and Coleman (l73)» and Coleman and Thomas (65) indicated that red-yellow soils contained large amounts of potential acidity which could not be displaced by N KC1. Pioncke and Corey (230) postulated that the amount of acidic Al in a soil was constant and comprised exchangeable and non-exchangeable forms in equilibrium with each other The amount of extractable Al was reported to vary with vegetation (337) and with soil type (342). Richburg and Adams (252) concluded that the solubility of Al(0H)~ varied with soil type and with the nature of the Al-polymers present. The polymers may have condensed on solid surfaces (240) or formed part of complexes with organic matter (159).


61 Nye et al . (218) suggested that the concentration of Al in the soil solution would be low if the soil solution had a low electrolyte concentration and the degree of Al saturation was < 60;"'o. At a given pH increasing fertilizer applications increased the Al concentration in the soil solution (51), particularly at pH < 5.8 (191), while increasing organic matter decreased the Al concentration (87)* Ragland and Coleman (239) found that exchangeable Al varied with drainage conditions. The relationship between exchangeable Al and pH was described as hyperbolic by Popenoe (23*f) and by Burgos (51), but Hutchinson and Hunter (135) obtained a linear relationship in surface soils and a curvilinear relationship in subsoils. Correction of Acidity on Red-yel low Soils Observing the destructive effects of the continual loss of bases and the concommitant development of acidity in soils comprised largely of 2:1 clays, Joffe (140) suggested that the application of lime would provide a measure of prevention of these effects. Liming such soils satisfied the high Ca requirement of the clay mineral as shown by Snyder et al . (280), and stabilized the high humic acid content of the soils by the precipitation of Ca-humates (159). In addition, liming 2+ provided a continuous supply of Ca to the soil solution which in such soils tends to be at a lower pH than the solid phase, according to Wiklander and Ghosh (326). When a similar rationale is applied to red-yellow soils, expected results are not realized. Sherman and Fujimoto (276) warned that applications of lime to some Hawaiian soils needed to be carefully


62 limited. Shuffelen and Middleburg (273) observed that whereas small quantities of lime decreased the permeability of lateritic soils, larger quantities increased it. They interpreted this in terms of competition between the peptizing effect of OH" and the coagulating effect of Ca 2+ . Ignatieff and Leraos (136) in reviewing the effects of lime applications to latosols attributed adverse effects partially to the displacement by Ca and leaching out of the root zone of K + , Mg , and micronutrients, and to reduced availability of Fe and Zn at high pH. Sombroek (283) suggested that liming results in decreased availability of plant nutrients, but Reeve and Sumner (2^6) concluded that induced P fixation was responsible for reduced yields. Fox et al . (94) noted that the effects of lime varied with the soil, but reported appreciable reductions in yield at higher liming rates, as did Burgos (51), Monteith and Sherman (202), Lucas (l?8), and Reeve and Sumner (246). Studies aimed at providing a substitute for lime have not been very helpful except for sugar cane (95) • Monteith and Sherman (202) used calcium silicate and found that at comparable rates of application, yield depressions were smaller than with lime. Reeve and Sumner (246) attributed a similar difference to the fact that calcium silicate was less soluble and had a lime equivalent value of about 0.30. Since the supply of Ca in the soil solution of kaolinitic soils is much greater than in montmorillonitic soils (84, 199) and the soil solution in the former soils tends to be higher in pH than the solid phase (326), it would appear that the adverse effects of liming point to the need for a reassessment of the criteria employed in


63 correcting acidity in rsd-yellow soils. Indeed, several criteria have been suggested. Coleman et al . (63) concluded that N KC1 exchangeable Al would constitute a suitable criterion. This was supported by Reeve and Sumner (2*4-7), but several modifications have been offered. Sombroek (283) stressed the need for the neutralization of exchangeable Al and the correction of Mn toxicity. Kamprath (1*4-7) found that a correction factor of 1.5 or 2 was desirable if exchangeabli Al was to be used in this context and pointed out that Ca and Mg should be supplied in addition to the neutralization of exchangeable Al. Vincente-Chandler (31l) reported that reduction of exchangeable Al to levels less than 2 meoyiOOg was optimal. Hoyt and Nyborg (128) suggested that extractable Al would be a valuable supplement to pH in assessing the need for liming. The reduction of exchangeable Al as a criterion for liming is supported by the findings of McLean et al . (190) that lime reacted rapidly with exchangeable Al and only slowly with non-exchangeable Al. The possibilities that these soils are Al-buffer systems (138); that plant growth is limited by exchangeable Al rather than Ca deficiency (87, 100, 1*4-7); and that the P status of the soil is influenced by sorption on sesquioxides (276) are all likely to support the adoption of this criterion. The use of the degree of base saturation as a criterion has also been advocated. Abruna-Rodriguez et al . (l) found 60/t base saturation or < 10$S Al saturation optimal. Kamprath (1*4-7) suggested that < 15% Al saturation was desirable, and Hutchinson and Hunter (135) indicated that lime applications would be beneficial if Al saturation was > 25%.


64 The two criteria have exchangeable Al as the common factor and differ markedly in philosophy from the fractional neutralization of titratable acidity. Optimum response to liming was obtained on some red-yellow soils on the Black Sea Coast, when 25?S of the titratable acidity was neutralized (82). Ayres et al. (l6) showed that exchangeable Al vras reduced to zero at pH 5„0 to 8.4, and Yuan (34l) reported that Al was neutralized at pH < 5.4. Burgos (5l), however, found that although liming reduced exchangeable Al, uptake of Al by millet was increased. Moschler et a l, (205) also obtained reduction of exchangeable Al on liming and in addition reported the reduction of exchangeable K and improved metabolism of N by alfalfa. Koch et al . (158) noted a reduction in the availability of K to corn on Natal Oxisols. The expected increase in exchangeable Ga and CSC of soils were not realized by liming in some cases, Rixon and Sherman (255) could find no significant change in CSC attributable to liming, while Mahilum et al . (l?9) reported that beyond the first increment of lime, Ca was easily leached. An alternative approach to the amelioration of fertility in redyellow soils appears to be the use of P, with Ca supplied only to satisfy plant requirements. Increases in yield on P application in the absence of lime have been reported by Lucas (l?8), and Plucknett and Sherman (232). Younge and Plucknett (338) advocated the use of P in adequate quantities for the quenching of the fixation capacity of the soil. Several desirable effects on the soil have been attributed to the use of P. Ayres and Kagihara (15) reported the


65 increased retention of K. Rixon and Sherman (255) pointed out that the effects of P varied with the soil, but obtained increased CEC and exchangeable Ca on some Hawaiian latosols. Plucknett and Sherman (232) also obtained increased CSC and pH on P application to some bauxitic soils of Hawaii. Russell (260) attributed increases in organic matter and N to the use of superphosphate, and noted that for each 1.0^ increase in soil N there was a 3»5 meq/lOOg increase in CSC (262) . Increases in N were regarded as the consequence of enhanced N fixation in the presence of added P. Donald and Williams (8l) concluded that increases in soil N were directly related to the amount of superphosphate applied. They observed a drop in soil pH, and regarded this as being due to increased CEC concomittant with the build-up of organic matter. Williams and Donald (329) reported that this build up of organic matter permitted the maintenance of constant ratios of C»N:S:P at about 155:10:1.4:1.68, and that S was the limiting factor in the build up of organic matter. Barrow (21) doubted that the accumulation of organic P as noted by Russell (26l) would provide a large enough reservoir for plant growth. The use of P as a means of soil amelioration permits the accumulation of soil organic matter, largely through its positive effect on plant growth, but the rate of accumulation would depend on the ecological conditions prevailing (219), and on the composition of the soil mass.


66 Indicator Crops Pangola pxass The botanical characteristics of pangolagrs.5S ( Dig;itaria decu mbens Stent) were described by Kodges et al . (126). The plant i red the maintenance of fertility in the soil for proper growth (124), Gammon (103) found that pangolagrass had a very high K requirement, but that more than 60% of this could be substituted by Ha without appreciable reduction in growth. This grass is very sensitive to Ca-defici'ency (124), but responds markedly to P (125, 35* 1?8). Ahmad et_al. (6) did not, hovrever, obtain a significant response in P content or yield on Trinidad soils e Phosphorus availability was seen to influence the response of pangolagrass to Ca (127), but Lucas (l?3) detected no interaction between Ca and P. Hodges et al . (126) stated that pangolagrass could vigorous growth at pH 4.2 to 4.5 if all the required nutrients were in adequate supply, but that additions of lime improved growth through increased Mg supply and increased efficiency of nutrient utilization. They considered pH $*$ to be desirable on the acid flatwood soils of Florida (125), but Hortenstine and Blue (127) found pH 6.3 to be optimal on Puletan loamy sand. Liming to higher pH levels depressed yield. Lucas (178) also obtained depressions in yield of pangolagrass on liming an amorphous Costa Rican Entisol from Los Diamantes. Blue et al . (34) observed slow establishment of pangolagrass on newly cleared land in Costa Rica, but subsequently noted appreciable response to N fertilizers in terms of yield and N concentration in


67 the tissues. Ahmad et al. (5) also obtained significant responses in yield and N concentration on Trinidad soils. The selection of pangolagrass as an indicator crop appears justified by the fact that though this species has been "the improved pasture grass" at Ebini for most of the 1960s, its growth and production were, according to Mayers (l86), far from satisfactory. Pigeon pea Whyte et_al. (323) described the botanical characteristics of the pigeon pea' (Ga.jamis cajan (L) Druce). The variety "Norman" which was used in this study was developed from an introduction from Pakistan. It has shown some adaptation to mechanical harvesting and is now being tested in Florida by Killinger (l5^)<> The nutrient requirements of this crop have not been widely studied so far. Krauss (l60) concluded that pH 5 to ? was most favorable, but Lucas (178) obtained no significant response to lime up to pH 6.8. Phosphorus applications gave significant increases in yield. Nichols (21l) indicated desirable levels of concentration in the tissues for the major plant nutrients as well as effects of likely interactions and deficiencies (212). It is not now known whether any attempts at growing the pigeon pea have been made at Ebini, but Kayers (186) reports that the legumes found in that area, e.g., Desmodrum ydxan us, Zorna diphyla, did not thrive when incorporated into "improved pastures." The importance of the legume in soil fertility was discussed by Ellison (85), and its more specific role in pastures was examined by Bryan (50)» Since,


68 however, both the capacity of the legume to supply N to the sward and the content of N in its tissues are determined by the amount of l\ fixation, the focal point in this context is possibly the effectiveness of the symbiotic relationship in which the legume participates, rather than the legume itself. The pigeon pea, therefore, offers a means of assessing the conditions which may have contributed to the failure of those legumes already used. Vincent (310) reviewed the literature on the environmental factors which are likely to influence N fixation by the legume/Rhi zobium symbiosis. He stated that though acidity is generally accepted as the factor of major impact, the broad spectrum Rhizobium .japonicum is capable of withstanding pH < 3*5. Indeed, while the effect of acidity, per se, is more marked in relation to the bacteria than to the host plant (12) nutritional factors normally associated with acidity may affect both symbionts. Inadequate supplies of Ca and Mo, and excessive amounts of Mn and Al are normally associated with acidity. Hallsworth (117) found that liming benefited nodulation when Ca supply in the soil was low, but Andrew and N orris (13) showed that tropical legumes were capable of nodulating in conditions of Ca -deficiency. The tropical legumes vary in their response to added Ca, because of inherent differences in their capacities for extracting Ca from soil (13). It has been shown that the Rhizobium has lower requirements for Ca than for Hg (215, 309), but Vincent (310) cautioned that the need may be for divalent cations rather than for Ca or Mg specifically. Norris (216) classified Rhizobium strains normally associated with legumes adapted to acid soils, as alkali producers. This may be of


considerable importance to the nutrition of the host plant. Dobereiner (80) attributed responses to Ca by beans, to control of the Ca/Mn ratio in the plant, while Foy et al. (99) distinguished between absolute Ca deficiency and Al-induced deficiency observed in different varieties of soybean. Ouellette and Dessureaux (22'J-) considered that the rate of Ca uptake was the important factor in tolerance to Al and Mn by different legumes. The more tolerant types had the highest Ca contents. Henzell (l2l) indicated that differences in the capacity of legumes to extract Ca from the soil were genetic in origin, and that similar differences existed in capacities for extraction of Ca and P, and in tolerances to excesses of Al and Mn. The importance of P to the legume was stressed by Van Schreven (305), and Vincent (310) observed that deficient P supplies to the host plant had indirect adverse effects on both the formation and functioning of nodules. Dobereiner (80) noted that P stimulated N fixation but did not counteract excesses of Mn. It would seem that though the primary purpose of cultivation of the legume is N fixation, this purpose is likely to be defeated in a soil which contains appreciable quantities of combined N. Stewart (287) stated that such conditions resulted in the decrease of the number of root infections by Rhizobium, and in the number and importance to the host plant of those nodules which developed. V/hether those nodules which develop are effective or not, is dependent on the development of hemoglobin in the nodule (217) •


MATERIALS AND :-3TH0D3 Soils Samples of four soils which occur on the White sand plateau were taken by members of the staff of the Ministry of Agriculture. Guyana, and shipped in an air-dry state to Miami, Florida. The samples were quarantined and fumigated at the port of Miami, then transported to Gainesville, where they were screened (5 rr;m), airdried, and stored in polyethylene bags. The soils were described by Brinkman (45) as follows sI. Ebini sandy loam a well -drained soil intergrading from a red-yellow latosol to a red-yellow podaol (Tv_gic normochrult) developed in fine-textured sediments of the Berbice formation under forest and savanna vegetation. This soil occurs often on the lower parts of slopes where residual material from crystalline rocks and unconsolidated colluvium are present. It correlates well with similar soils of the Brazilian states of Sao Paulo and Rio de Janiero, and is usually associated in the Sbini area with the coarser-textured Kasarama loamy sand and Takama sand (45) II. Kasarama loamy sand a well-drained red-yellow latosol ( Ochric ustox) , developed in medium-textured sediments under forest and savanna vegetation. It occurs on gentle slopes, and like the Sbini sandy loam may have numerous anthills 70


71 when under savanna vegetation. This soil is intermediate in texture between the Ebini sandy loam and the Tabela sand with which it may be associated (^5)» III. Tabela sand an excessively drained regosol (U ltic quartzipsamment ) developed in sandy sediments under forest and savanna vegetation. It occurs on undulating or gently sloping relief in association with Kasarama loamy sand, Takama sand, and Tiwiwid sand. It is coarser-textured than the Kasarama and Takama soils and browner in color throughout than the Tiwiwid sand (hs) . IV. Tiwiwid sand an excessively drained regosol (Typic Quartzipsamr.ent) , developed under forest and scrub-tree vegetation in coarser-textured sediments. It occurs in relatively large areas in higher portions of the landscape and in colluvial positions on creek slops in association with Ituni sand, Tabela sand, and Kasarama loamy sand (45) . Samples of the Tiwiwid sand were taken from the surface horizon only. The remaining samples embraced the four uppermost horizons. Soil Analysis Particle-size analysis was effected by the method described by Day (72) using 20-g oven-dried (105G) samples and omitting the filtration process. Samples for mineralogical analysis were pretreated according to the techniques detailed by Kunze (l6l). Sub-samples were prepared by


72 the methods of Whittig (322), and examined for X-ray diffraction patterns as oriented aggregates on a glass slide. X-rays emanated from a Cu-source and were passed through a Ni filter and a 3° beam slit. A proportional counter with a voltage of 1,500 V served as the detector with a detector slit of 0.05°. Samples were scanned over the range of 20 values from 2° to 40° at a speed of 2°/min. Sub-samples of the pre-treated soils saturated with Mg were used for differential thermal analysis over the temperature range 50 to 1,100C at a heating rate of lOC/min. using a soil:burnt asbestos ratio of 1:3 (22). Further sub-samples were saturated, mounted and shadowed as directed by Kittrick (l55)» examined under a Phillips E M 100 B electron microscope, and photographed. Amorphous materials were extracted from the soils by a citratebicarbonate -dithionite solution buffered at pH 8.5 (l6l). Each 10-g soil sample was extracted four times, and the supernatant solutions were made to volume and analysed for Al, Fe, Si, and Mg using a model 303 Perkin-Elmer atomic absorption spectrophotometer. Organic matter contents were determined by weight loss on heating 20-g oven-dried (105G) samples at 3750 for 16 hours as described by Ball (18). Organic G was studied by the Walkley -Black method of wet oxidation outlined by Allison (ll). Organic matter was extracted with 0.5N NaOH (203), and precipitated according to Yuan's method (34l). The total contents of various elements in the soils were measured by the HF-HC10. method of Jackson (137). Soil digests were analyzed for Al, Fe, Ca, and Mg by atomic absorption spectrophotometry and for


73 K by flame emission. Total N was determined by micro-K jeldahl with the salicylic acid modification of Volk and Fontein. 1 Total P was measured by the chlorostannous-reduced molybdophosphoric blue color method in a HG1 system as directed by Jackson (137). Cation exchange capacity determinations were made using N HH^OAc (pH 7.0) and N KC1 (pH 5.7) as saturating solutions. The NH^ + was displaced by N Nad (pH 2.5) as suggested by Chapman (57) and measured by distillation while K was displaced by N CaCL? and measured by flame emission after the method of Shumbla and McLean (29) • Exchangeable bases Ca, Mg, and K were determined in the N NK OAc leachate, while Al, Mg, and Ca were measured in the N KC1 leachate. Exchange acidity was found by titration of the BaClp-TSA extract with dilute HC1, using a mixed bromocresol green -methyl red indicator as recommended by Peech (228) . Titration curves were constructed from pH values after addition of known amounts of HC1 or KOH to 5-g soil samples in 50 ml N KC1 as described by Puri and Swarnakar (238). A Corning Model 12 glass -electrode pH meter was used to determine the soil pH in a 1:1 ( w /v) soil-water suspension and 1:2.5 ( W /v) soil N KC1 suspension. Moisture equivalents were estimated by the method of Briggs and McLane (Zj4). 1 Unpublished mimeographed sheet. Department of Soils, Univ. of Florida, Gainesville.


74 Plant An a lysis One-gram samples of oven -dried (70C) plant tissues were ashed in a muffle-furnace at ^l25G for 12 hours. On cooling, 5 ml of N ffifOwere added to the ash and evaporated to dryness on a hot plate at 90G. The dried samples were then heated to HZ^C for 2 hours and taken up in 2 ml 511 HG1 » This is essentially Peech's technique as described by Jackson (137). The solutions were filtered into 100 ml volumetric flasks, and made up to volume with deionized water. Fivemilliliter aliquots were used for the determination of P by the 1,2,4-aminonaphthol-sulfonic acid-reduced molybdophosphoric blue method of Fiske and Subbarow (9l). Calcium, Kg, Fe, Zn, Kn, and Cu in the solutions were determined by atomic absorption spectrophotometry and K by flame emission. Nitrogen was determined in 0.20-g samples of oven-dried (70G) material by the micro-Kjeldahl method with the salicylic acid modification of Volk and Fontein. Results of the analyses were computed on the oven -dry (70G) basis. Laboratory Experiments Experiment 1. Incubation Studies Five hundred gram samples of the surface horizons of the soils were incubated with various levels of P at different pH levels. Initially 2 kg samples of each soil were mixed in the air-dry state with the levels of CaCO^ required to produce pH (N KCl) levels of 5*5* 6.0, and 7.0 respectively. These samples were divided into 500-g


75 portions which were then moistened with distilled water to field capacity and incubated under laboratory conditions for 2 weeks. Distilled water was added as required to maintain each sample at constant weight over the period of incubation. The pH (N KCl) level of each sample was recorded after 2 weeks of incubation and P added at levels of 0, 50, 100, and 150 ppm, to randomly selected samples to give the effect of a 4 x 4 factorial design. The CaCO treatments were, however, not orthogonal. The source of P was KF^PO^, and this was added with other nutrients to provide final levels of N (100 ppm, from NH^NOj , K (100 ppm, from K^PO and KCl), Kg (40 ppm, from Mg(OAc) , and MgSO^), S (20 ppm, from KgSO. ) , and micronutrients from FTE 504 2 at 30 kg/ha. The samples were then adjusted to field capacity with distilled water, covered and left thus for 12 weeks. Thereafter, they were allowed to attain an air-dry state and stored for subsequent analysis. Subsamples were used for:i. Leaching studies. — Ten -gram samples from each treatment were weighed in duplicate into "leaching tubes" and leached with 200 ml distilled water. No attempts were made to minimize the disturbance of the soil, but the head of water in the tubes was kept at a maximum of 20 cm. The leachates were collected and analysed for Ca and Kg by atomic absorption spectrophotometry, for K by flame emission, and for 2 FTE 504 contains k.OOfo B, 7»00?S Cu, lk.00/o Fe, ?.00?S Mn, 7.00?$ Zn, and 0.0?/o Mo.


76 P by the chlorostannous-reduced molybdophosphoric blue method in an HC1 system (136). ii. Exchangeable ca tions,, — Ten-gram samples from each treatment were weighed in duplicate into 100 ml polyethylene centrifuge tubes and shaken with 50 ml of N NH.OAc (pH 7.0) for 1 hour on a reciprocal mechanical shaker. The tubes were left standing overnight then centrifuged, and the supernatant decanted. The N NH OAc washing was repeated with 10 min. shaking to give a total of four washings per sample. The supernatant solutions were combined and made up to 200 ml with N NH.OAc (pH 7.0), then analyzed for da, Kg, and K by the techniques used in (i) above. Results were corrected for the amounts removed by leaching and categorized as exchangeable nutrients retained against leaching. iii. Extractable phosphorus . — Two-gram samples from each treatment were weighed in duplicate into 100-ml polyethylene centrifug tubes and shaken for one min. with 20 ml of a solution of 0.03 N NH^F in 0.025 E HG1 » according to Jackson's description of the method of Bray and Kurtz (137). Appropriate' aliquots were analysed for P by the technique used in (i) above. Experiment 2. Studies on the Organic Fraction of the Soils Twenty-five gram samples of soil from each horizon were weighed out in duplicate into 100 ml polyethylene centrifuge tubes and shaken for 1 hour with 50 ml 0.5N NaOH on a reciprocating mechanical shaker. The tubes were left standing overnight then centrifuged and the


77 supernatant solution decanted. This 0.5N NaOH extraction was repeated and the supernatant solutions from each extraction of each horizon combined and adjusted to a final volume of 200 ml with 0.5N NaOH. Ten-milliter aliquots from each extract of organic matter were put in 100 ml "beakers and adjusted with 2N H JSQh to pH values of 1, 2, 3, ...... 10. They were then transferred to 100 ml centrifuge tubes with all washings and centrifuged. The supernatant solutions were decanted into 100 ml beakers and evaporated to dryness on a. hot plate at 80C, then heated to 450C for 2 hours to oxidise all the carbon. The residues were taken up in 2 ml 5N HC1 and filtered into 100 ml volumetric flasks, made up to volume with distilled water, and analysed by atomic absorption for Al and Fe. Ten milliter aliquots of the 0.5N NaOH extracts at pH 12 c 2 were also evaporated to dryness, ashed at ^50C, and analysed as above for Al, Fe, Ca, Mg, and Hn, and by flame emission for K. The precipitates formed on acidification of the NaOH extracts and 10 ml aliquots of the original extracts were analysed for C content by the Walkley-Black method (ll). Appropriate aliquots of the NaOH extracts were used for determination of N and P by the respective methods indicated above. Absorption spectra were constructed by the method described by Yuan (3^1, 343) using aliquots of the original NaOH extracts which had been adjusted in volume to provide equal concentrations of C.


78 Greenhouse Ex periments Experiment 1. Limiting: Nu trients i Study, Two-kilogram samples of soil from each of the surface horizons were adjusted to pH 6.0 by the addition of CaCO,, on the basis of previously constructed titration curves. The samples were then kept in a moist condition for 2 weeks in plastic pots. On the basis of a randomly assigned 7p factorial treatment plan, P (100 ppm) as KH P0. , K (100 ppm) as KC1, S (20 ppm) as MgSO^ or N (100 ppm) as NH^NO-j, and micronutrients as FTS 50^ at a rate equivalent to 30 kg/ha, were added to appropriate pots. Magnesium was added as Kg (0Ac) ? to give a final level of kO ppm in each pot. Complete 2? factorial studies were run on each surface soil using pangolagrass, and on each soil except the Ebini sandy loam using pigeon pea. Rooted pangolagrass cuttings were planted in appropriate pots on November 3» 1969, and harvested 10 weeks later on January 1?, 19?0. A second harvest was taken on Karch 15, 1970, after which the soils were allowed to dry. The roots were then removed, washed thoroughly, and dried at 70C. Pigeon pea seeds inoculated with a broad spectrum Rhizobium strain were planted in appropriate pots on December, 3» 1969» and harvested 12 weeks later on February 22, 1970. The roots were removed immediately after harvesting of the tops and examined for nodule development, washed thoroughly, and dried. All harvested material was appropriately labelled and weighed after drying at 70C. Neither study was replicated.


79 Experiment 2. Optimum l evels of Ca. P. and K On the basis of data gleaned from Experiment 1, a central orthogonal composite design described by Mendenhall (20l) was used to determine the levels of Ca, P, and K which were required for maximum plant growth. Each category of soil from the previous greenhouse study was combined with unused samples from the same horizon; FTE 50^was added at the rate of 7. 5 kg/ha, and the whole sample thus obtained thoroughly mixed. The pH (N KCl) value of the coil from each surface horizon were then measured. The total quantities of CaCOnecessary to satisfy the requirements of the design were then thoroughly mixed with suitable quantities of each soil. Limed and unlimed soils from each surface soil were thoroughly mixed in proportions calculated to give the finally desired levels of Ca and pH in each treatment. Two-kilogram samples were then put into plastic pots and kept in a moist state in the greenhouse for 2 weeks, after which appropriate quantities of P (as Kl^PO^and NH^HgPO ), and K (as KCl and KH 2 P0^) were added to the pots. Other nutrients were also added at this time including N (100 ppm) as NHjH P0^ and NH^NOy S (20 ppm) as MgS0^ f Hg (kO ppm) as MgSO^^O and Mg(0Ac) 2 » B (5 ppm) and H^, Mo (5 ppm) as (NH^) 6 Mo ? O-^.^O, Rooted sprigs of pangolagrass were planted (three/pot) on the Tabela sand, the Kasarama loamy sand and the Sbini sandy loam, on July, 1970, and harvested 8 weeks later on September 15, 1970. Urea (100 ppm N) was added to each pot 1 week before harvesting and


80 a second harvest was taken on December 12, 19?0. A similar application of N was made and a third harvest was obtained 8 weeks after the second. The roots were left in the pots until the soil dried then removed, washed thoroughly, dried at ?0C, and weighed. Pigeon pea seeds (five/pot) inoculated with a broad spectrum Rhizobium strain were planted on the Tiwiwid sand and Kasarama loamy sand on July 8, 19?0. Nutrients had been added as indicated above, except for the omission of the N sources, NH^H 2 P0^ and NH^KOo. Tops were harvested on September 26, 19?0, and the pots left to dry before the roots were removed and examined. There were three replications in this experiment, as in the pangolagrass study. The indicator crops were utilized in an attempt to characterize the response surface of each soil for each crop in so far as limitatior in quantity of soil permitted. Dillon (79) observed that response surfaces in crop and livestock production can be very well depicted by studies based on central composite designs. Such designs permit a greater number of treatments and are more conducive to accuracy than conventional factorial studies. The central composite design was developed by Box and Wilson (39) to estimate response functions. The effect of each parameter in the function is reflected by its coefficient. They pointed out that estimates of a coefficient will differ from the true value because of experimental error and because of biases which arise when it is impossible to represent the surface by equations of the type fitted. Nevertheless, it is not recommended that non-significant terms in the fitted equation be dropped, since they provide the best


81 estimate of any given effect from the available data. Hader e.t a l. (115) stated that when the variance of the estimated coefficient was used as a criterion, the central composite design was more efficient than a comparable complete factorial. Box and Youle (40) indicated that the general shape of the response surface and of the area of the maximum were determined by the presence or absence of interaction among the variables. In the presence of interaction a ridge rather than a point corresponded to the maximum, which could be attained by several different combinations of the interacting variables. The existence of interaction amongst plant nutrients is lenoirn (279) as are the additional influences of season (302), and rainfall (339). This design permits a closer examination of more variables, and the values at which they combine to provide a maximum, could be computed from the fitted equation as discussed by Box (33). The use of a central composite design has limitations in that the variance of the response is not uniform over the experimental region, but increases with increasing distance from the center of the region (20l) . This introduces an element of tedium to multiparameter inference making procedures.


RESULTS AND DISCUSSION Profile Characterisation Particle-size Analysis In each profile the proportions of clay and silt inc with depth, while these of sand decreased, as shown in Table 1., increase in silt content appeared, except in the Kas to be largely in the 0.005 to 0.02 mm particle-size ran the decrease in sand content was most narked in the med . 10 to 0.50 mm) fraction. The proportions of very fine sand ( 10 mm) increased continuously down the profile while those of e and very coarse sand increased to a maximum and then dec The increase in relative amounts 01 particles of smaller size uith depth appeared to be a result of the movement of water down the profile and the consequent segregation of smaller particle:.; as structural aggregates were modified (298). While the proportions of inorganic particles of colloidal size increased with depth ; organic colloids decreased appreciably though there was a small increase seen in the B21 over the B-^ horizon of the Kasarama loamy sand, and in the C over the B~ of the Ebini sandy loam. These increases may not be significant, but the cessation of the decreasing trend was perhaps due to a reduction in rate of movement of percolating water as the 62


83 T) rH C «) rJ O o o o H


8^ water table was approached, and possibly to the saturation of the organic colloids by hydrous oxides. Clay F raction The X-ray diffraction patterns of the clay fractions from the Ebini profile shown in Fig. 1 were virtually identical to those obtained for clay fractions from the other soils, regardless of whether the samples were Mg-saturated and glycerol-solvated, or K-saturated and air-dried. The diffraction peaks occurred at ZO values of 10.5° to 11°, 22.7° to 33°, and 35.5° to 33°, corresponding to first-, second-, and third-order diffractions, respectively. These o peaks indicated a diffraction spacing of approximately ?.l to 8.5 A, suggesting the presence of kaolinite, raetahalloysite, and/or chlorite (322). All the diffraction patterns obtained after heating the K-saturated samples to 500G for 2 hours were similar to those shown in Fig. 2. The disappearance of the peaks on heating indicated that kaolinite was the clay mineral in the samples. This was confirmed by electron microscopy. The micrographs a d shown in Fig. 3 were probably not representative of the degrees of crystal imperfection to be expected in a given soil, but reflected the various stages of degradation of the clay crystals that might be encountered. Further evidence of the kaolinitic nature of the soil clays was provided by differential thermal analysis. The DTA traces shown in Fig. h were typical of those obtained for all samples. The endothermic peaks at 580G were probably due to the loss of water in the crystal lattices and the breakdown of crystal structures, while

PAGE 100


PAGE 101

86 '.'. A o-30cm. VuAL B 2 30-75 w ^ *« ^^ ; ^^ ^^^^W 40 30 20 20 10

PAGE 102

(a) Tiwiwid (ij4,000x) (b) Kasararaa (38,000x) Fig# 3. —Electron micrographs of clay separates from the surface horizons.

PAGE 103

(c) Tabela (53,000x) (d) Ebini (44,000x) Fig. 3. —Cont'd.

PAGE 104

Fig. h. —Differential thermal analysis traces of Kg-saturated clay of the Ebini sandy loam.

PAGE 105

90 the exothermic peaks at 1,050C were indicative of the crystallization of corundum, as stated by Mason and Berry (185), The skewed nature of the endothermic peak might have been attributable to the presence of some halloysite (22), but no evidence of halloysite was gleaned by electron microscopy. The poor crystallinity of the kaolinite may, therefore, have been responsible for the character of the peak. The DTA traces demonstrated the absence of any chlorite, gibbsite, goethite, or brucite (22). Gibbsite was also absent from a sandy yellow latosol which had kaolinite as the only clay mineral, and which developed from Quaternary alluvial sediments in the Amazon (52). The total Al and Fe contents of the clay separates are listed in Table 2. Except for the surface horizons of the Tiwiwid and Tabs la sands, and the B horizon of the Kasarama soil, the contents of Al did not vary greatly from the value of 20.9/S indicated by the chemical formula for kaolinite (185). This suggested that the loss of crystallinity involved approximately equal changes in the contents of Al and Si, rather than a process of desilification (52). The small, but virtually constant content of Fe was probably due to isomorphous substitution (223). It is also possible that some Fe may not have been removed by the citrate-bicarbonate-dithionite extractant (283), but several reports attest to the capacity of the extractant (92, 3^) for the complete removal of Fe, and to its greater efficiency in acid than in neutral soils (319). The amounts of amorphous oxides removed by this extractant are

PAGE 106

Table 2. --Aluminum and Fe contents of clay separates, 91 "-==**-tasssasss

PAGE 107

92 shown in Table 3* From the data it is clear that the clay crystals were not saturated with ferric oxides. The 10 to 12 % ( w / u ) requirement (282) was never met. The relatively large quantities of SiO may have been due partly to the progressive degradation of crystalline clays and partly to its accumulation in association with (319) or occluded by (2) hydrous oxides of Al and Fe. The total contribution to the clay fraction of the soils by the amorphous materials tended to decrease with increases in the depth of the horizon, but was in every case less than 10$ ( w /w), as postulated for similar soils by Bennerca (26). Organic Fraction The organic fractions of the different horizons are described quantitatively in Table 4. The decreasing trends with increasing depth of the horizons, revealed by loss of weight on ashing the soils at 375C, paralleled those of organic C content as measured by the Walkley Black method. The B 2 i and C horizons of the Kasarama and Ebini soils, respectively, varied somewhat from the general trend. The loss in weight on ashing was greater in these horizons than in those immediately above them. The ratios of weight loss to per cent organic-C increased with increasing depth of the horizon in each soil. In the surface soils, these ratios were consistently lower than the "Van Bennelen factor" of 1.724 (ll). This was also true for the C, and A-i horizons of the Tabela and Kasarama soils, respectively. The factor of 2.5, proposed by Broadbent (46) for subsoils was exceeded in the two lowest horizons of each profile except for the B-j_ horizon of the

PAGE 108

93 O Pi CnOn r-l -^ O O iH pH On Cn-nO CO C"\ On CN. C"n < I . . HO C-N.-3O-CO 4riO>A on :>cm , h O O r-i rt 00 -* -dvn if-,o)iA4 cc> o -dno -3" Cnir\ \OnO tr,-* O 00 OO vO VA m Cn.. 0> H .j»A 00 CNl CA rH Cn-C-ncOO^ OnOnJ-C^N MJ^OW OO-VAON i> 3" CN} tf> VAUN.1AK) >AtSWO\ i-IVPvCn-Cn) c~\.c} o~nco c)o~\ cn-no no oo cnco C~\ CA Onj^ Cvl -3" c!£n. vr\ cni On 00 "^ CM CA H O CM vO Cn-CQ ON no -3" < cm cnj ca rS cnj run VA Cnr-\ Vf> r\c>co n o <*~> cm o CA CnrHNO I I I O 1AO CA OO ?! A

PAGE 109

'.•'! H " tV ^ CM O^ vr> W\ OS W\CO O r-l CM C^A H CM C">. -3" 0) O VO

PAGE 110

95 Kasararaa loamy sand* The disr 'ween the weight lo. ios, and the conventional conversion factors could be hinge on the true C content of organic fraction, the relative accuracy of the different methods of measurement, and the existence of definite sources of error. Estimates of the C content of the orgai C concentration in organic matter— »(Table k) f shoi nic fraction of the upper two horizons of the soils had mo: ; C, while those of the lower two horizons and the B„ of tl had much less c Residues of vegetation, burnt on th : of the soils may have inflated the estimates for the upper hori: . The extremely low C content of the lower horizons may part to inflation of the weight changes o in the inorganic materials (l8) 6 As much as J»8% of th. material may be lost at 375C (2'K3)» and the substantially higher clay contents of the lower horizons could, therefore, have d the apparent C concentration in the organic fraction. The efficiency of the wet oxidation process in measuring the organic-C contents is determined to some extent by the chemical complexity of the organic material (ll), and by the nature of the bonds by which this material is linked to the inorganic fraction of the soil. The close association of organic material and clay observed in the Tiwiwid sand and the absence of interlayer spaces in the clay point to the likelihood of almost total oxidation of C. The chemical complexity of the organic fractions was compared by the construction of absorption spectra in the range ^50 to ^75 rou

PAGE 111

96 using 0.5N NaOH extracts (pH 12.2) from the different h( adjusted to similar C concentrations. The absorption sj are shown in Fig. 5 (a d). The differences in the slo] I extracts, from the surface horizons, Fig. 5 (3,), indicate material extracted from the Tiwiwid sand had the ha g cee of aromaticity and that from the Tabela sand had the 1 of aliphatic C (159) « The degree of aromaticity in material extracted from the horizons of the profiles decreased with :' n c depth of the horizon, Fig. 5 (b d), reflecting the pre increasing proportions of the less aromatic fulvic acids (30( ) depth in the profile. This trend is reflected by the husic (• ratios, though in the brown sand soils, the Tabela, Kas; Ebini, hunic-C was absent from all but the surface hoi'.' (Table 5). The proportion of organic C not extracted by alkali fro: the soils was greater than 60% in every case, suggesting were substantial amounts of humins present. The marked decline, with depth, in C/N ratios of the extracted material in spite of increasing amounts of fulvic acids, was probably due to the mov< of inorganlc-N down the profile. The wider C/N ratios throughout the Ebini profile perhaps resulted from a poorer supply of inorganic-N in the surface horizon. An examination of the response of alkali -extracted material to variation in pH showed that there was maximum precipitation of C at pH 1 from the Tiwiwid, and at pH 5 from the brown sand soils (Table 6). The material precipitated from the Tiwiwid extract was completely redissolved at approximately pH 2.8, while that precipitated

PAGE 112

97 0,6, (a) Surface soils 50 mg C/liter 0.^ 0.2 0,6 ; (b) Tabela profile 150 mg C/liter 0.4 /A, 0.2 A > mu (a) Nos. 1, 2, 3, and '4are the Tiwiv?id, Ebini, Kasarama and Tabela surface soils, respectively. 675 ////% 575 ^75 X , mu 0.6 (c) Kasarama profile 0.6 (d) Ebini profile 150 mg C/liter & C 575 ^75 A» mu Fig. 5. —Absorption spectra of 0.5 N NaOH extracts,

PAGE 113

98 o co

PAGE 114

99 •

PAGE 115

from the brown sand soils was redissolved at approximately pH 10 in the case of the surface horizons, and at about pH 7.5 to 9»0 in the subsurface horizons. The pH of maximum precipitation of C r different from the pH J+.8 level postulated by tfaksman (31?) a point of precipitation of 3 -humus. This material is known to be of high ash content, and has been described as being largely Ai some organic material (317)» or as consisting of organic entrapped by mineral matter during fractionation (286). The amounts of Al'in the alkali extracts and in the supernal solutions at the different pH levels are shown in Table 7. It is clear that the minimum amounts of Al remained in solution at pH 5« In the brown sand soils, the parallel solubilities of extracted G and Al in relation to pH changes bear some simil to the solubility patterns of C and Al in alkali extracts from the spodic horizons of some Florida soils (3^l)» The insolubility of C and Al in the latter case was considered to have been due to 1 the occurrence of a direct chemical reaction, or to the co; tion of forms of the two elements. Since this pattern was observed in the extracts from spodic horizons (3^l)» and the organic ? such horizons has been shown to be largely fulvic in character (33&)» it would seem that the phenomenon involved fulvic rather than huiaic C Fulvic acids are known to react with Al ( 10, 2?l), and this suggests that the occurrence of a direct chemical reaction is the more likely mechanism. Positively charged monomeric hydroxy -Al ions have a pK value of about 5 (138), and the functional groups of fulvic acids must, because of their similarity to those of humic acids, have pX

PAGE 116

101 si •p 8 * w O -P (H M < O P N 0)

PAGE 117

102 values in the range k.l to 6.2 (182, 235) • Adsorption reactions are likely to be preceded by chelation of Al by "salicylate structures" (2?l), and then by the formation of nuclei from which a dispersed precipitate can develop (43) . The neutralization of charge by adsorption may result in flocculation. The capacity of Al to cause flocculation of organic natter (183) and the effect of the valence of the ionic form on its efficiency in this context (175) support the possibility of the occurrence of direct chemical reaction. Further support is derived from the close agreement of computations based on laboratory preparations and measurements made Oj. on soil samples showing that Al-fulvate complexes involved Al(OH) and had maximum stability when all the functional groups of fulvic acid were neutralized (271) „ Relatively large amounts of organic C (38 to 75?o) were not precipitated at pH 5, even though at this pH the extracts were almost completely decolorized. Possibly the non-precipitated C was in the form of polysaccharides (29l) and/or simple organic acids (286). The behavior of Fe in the alkaline extracts from the surface horizons of the three brown sand soils is similar to that of Al in that there is again a pH of minimum Fe solubility. The data in Table 8 show that minimum solubility of extracted Fe occurred at pH 5 in the Kasarama and Ebini soils, but at pH 6 in the Tabela soil. The solubility pattern was possibly also related to the formation of Fe-organic matter complexes, but the mechanisms involved were much less obvious than those involving Al. The higher pH for

PAGE 118

103 Table 8. —Concentration of Fe in the supernatant of the 0.5N NaOH extract of surface horizons at various pH levels. Soil .v[. Tabela

PAGE 119

1C4 maximum precipitation in the Tabela soil was perhaps due to the approximately 1:1 ratio of FesAl in the alKaline extract as compared to corresponding ratios of 1:2 in the extracts from the Kasarama and Sbini soils ('Table 9) • The more stable Fe-huraates (299) are, therefore, likely to be of greater importance in the Tabela soil, in light of the preferential bonding of Fe by humic acids (290, 292). However, though Fe-fulvates break down at pH > 4.0, this breakdown is only partial (83, 299), thus distinction of the roles or relative importance of the fulvate and humate complexes of Fe is impractical* In the alkali extracts from the profiles, the decreasing amounts of Fe and increasing amounts of Al emphasized the increasing importance of fulvic acids with depth in the profile. Similar trends noted in Indonesian soils were interpreted in terms of preferential bonding of Al by fulvic acids and of Fe by humic acids (290, 292). The progressively higher Al concentrations also suggested the stabilization of fulvic acid down the profile. The amounts of Ca, Mg, K, and Mn removed by the alkali are also indicated in Table 9» The amounts of Ca and Mg were remarkably constant with depth in the profile, while K increased with depth except in the Tabela soil. These three cations could be associated primarily with the organic material or with the amorphous oxides. Amorphous materials in soils have bejjn shown to be involved in the retention of K (222, 304), hut NaOH exerts a solubilizing effect (167), and provides a replacing cation-Na —so that the true source of the cations must be considered obscure.

PAGE 120

105 -do o cj c^ OnCQ O »A vn CO C\J m O CO >-'"\ 0*NO U-\CO riHri riririri H H ri r! IA cn^O C^O^vPiCO O^^^ § S&&3 &&&& §&&& VO W\ OJ CO O CO O *A u~\ CO P^ CO W O NHCMA r\ONN nvO CO CH C^CXVCM C\l-=fr£V0O CJ ^ iJ C^i r-t i-i r-l r-t H H r4 r-i r-i r4 O M -d o OOON Cvl »P»ONH i T S 1 -d

PAGE 121

106 Chemical C ha ract erist ics of the Soils The total amounts of some elements in the soils are shown in Table 10. The amounts of Al and Fe increased with depth in the profile and this can be readily appreciated in relation to the increasing clay content of the lower horizons. The increase in Fe content is also appreciated in terms of the progressively stronger red-yellow coloration of the lower horizons. The Al content of the C"i and C2 horizons of the Tabela soil varied somewhat from the general trend, probably because of the more strongly podzolic nature of this soil 6 The amoimts of Ca were not very different from soil to soil or among horizons of a profile. The amounts seemed to be higher than would be expected from soils which had been subject to such prolonged weathering, and which were virtually free of minerals other than quartz and kaolinite. The amounts of Mg were much less than those of Ca, but much more than those of K. Both Mg and K were present in almost constant amounts, with virtually no variation among soils or horizons. The P contents of the soils were low as indicated, and as expected from the degree of weathering to which the soil materials had been subjected. The amounts of N in the surface horizons of the brown sand soils were strikingly similar, in contrast to the obvious differences in organic matter content. The decrease in N content with increasing depth of the horizon followed the pattern of distribution of organic matter. The pattern of distribution of organic matter is also followed by the CEC, the individual, and the total amounts of exchangeable cations (Table ll) . 3oth groups of cations also decreased with depth

PAGE 122

10? I CO CM CO O rH ISO V\-^ ^JH r-{ O O O d oooo On vO CO OvO r-1 H .-! rH H O OOOO OOOO i « i 7 < o ! ^c VO CO C*"\ C\! O O Q O COOn r-! H r! rH o^ ca m PS OOOO VOHCACO H H rH CO «n C\! VQ O OWO .=> C\2 U"\C\1 0\ cnCM CM H OOOO OOOO OOOO OOOO 0.3mvO riHriH vnco t I rH rHCA,-! CM O vfNO O I I I I O O u~\0
PAGE 123

108 «H C\! vQ ^) v.O ^ r-H OOOO OOOO OOOO 00

PAGE 124

109 down the profile., and the total amounts of exchangeable cati( decreased in the same direction as the organic matter content of the surface horizons. Comparison of the C3G and the total amounts of exchangeable cations in a given horizon indicated that the adsorptive complexes of the different soils were fa: saturated. Because of the obvious unsaturation and th Lable amounts of exchangeable Al, some acidity is to be expected in these soils. The variations in pH within the profiles of the individual soils are shown in Table 12. In the Tabela profile pH H and pH exhibited the same trends and differed from each other by an almost constant value, except for the C~ horizon. In the Kasarama profile, the pHjj of the A horizon was rather high, and this differed wid< from the pH^ value. The remaining horizons of this soil displayed the same parallel relationship between pH.. _ and pH r with differences of the same magnitude as in the Tabela profile. The A and C horizons of the Ehini soil had relatively high pH values which also differed widely from the pH N KQ1 values. The B horizons of this soil were comparable to the upper three horizons of the Tabela soil in their ph\, . and pH relationships. H 2 * n KC1 ^ The difference between pH and pH has been attributed to the effects of the hydrolysis of Al liberated by N KC1 (203) and this would appear to provide a part of the explanation in this case. The amounts of N KG1 exchangeable Al (Table 13) decreased down the profile as pH^ K _ increased, but the virtually constant differences between pH u and pH suggested that the explanation is not complete «2 U £L kci

PAGE 125

110 §1? o © £ c o © JZ u +> 3 3 © ,Q 3 © ( o «. .. flj O Hv.3-3 v/^\a ^t -juv Vf % ir\ u~\ u~\ vr\ vr\ >A>A »A v\ ir\ *A OOOO lAO^iA O >A O O Cv! m O C\! CM -d" m CO C^C^CT^CM OlAOlA C\> -* >A W^T ^pfirf d

PAGE 126

Ill O O K i . O O U"\ O u"\ o O O O O l <^ u"i O 0^00 CO Cn C*\ CO CO C?S OIV CO HHHrl rH CM CM CM CM CM CM CM "~\0 O VA ON 0~1 M CM rH rH rH : t3 o U o CM rH ("S '-A OOCM C\!

PAGE 127

112 The Inadequacy of the hydro] difference In pH values with di reiterated by the low pH levels of the Tiwiwid sand and the small difference of the two values, relative to the amount of N KC1 exchangeable Al present in this soil. In th d soils, 1 pK N KC1 valxies a11 vrere within the ran S e ^7° t0 5«35 S. KC1 which is known to be the range of neutralization o able AI (i6, j The possible importance of Al in these i mined by comparison of the amounts of Al removed by different e solutions. It can be seen from Table 13 cunts were removed by 0.5N KaOH and citrate -b5 .< ionits (CBD). The CBD extracted Al was more than the NaOH ext three and lower two horizons of the Tab3] a and respectively , while alkali extracted Al excee L in all horizons of the Kasarama soil, in the up] izons of the Ebini soil, and in the Tiwiwid sample. Larger of Al were extracted from the latter soil by N KC1 than by I k.8) or 0.1N HOI. In every other case, N KG1 removed the si mount of Al, relative to the three acid extract-ants. In the profiles N KG1 extracted Al decreased as the depth of the horizon increased, while Al removed by N NhVOAc (pH ^.8) showed a generally increasing trend. All other extractants removed more Al from lower horizons. Alkaline extractants tend to remove hydroxyAl (3^5) • The A1 removed by the CBD extractant is associated with amorphous materials (161, 319, 3^2) but should also include Al from organic complexes

PAGE 128

113 since the organic material was destroyed by H^O o prior to Al extraction* The alkali extracts Al as organic complexes (159, l6?) j, and would ba expected to remove less Al than CBD C Possibly, the higher pH of the alkali extractant and its destructive effects on the H~bonds polymer size of the organic substances (l8l) may have faciV .ore comple 1 of Al complexed by organic matter as well as the dissolution of Al from kaolinite in the more advanced stages of crystalline degradation. Two of the acid extractants are also known to remove Al complexed organic matter. Dilute acids extract fulvate complexes of mobile forms of hydrcxy-Al (159) while N NhVOAc (pH ^-.8) removes some organic matter-Al complexes (230, 3^5)0 This latter category of extracted Al, though not very well defined, is regarded as comprising exchangeable and other soluble forms of Al (232), and has been described as acidic-Al (230) c The Al removed by N KCi is exchangeable and bears either a curvilinear (51, 234) or linear (135) inverse relationship to the pH of the soil. This form of Al has been implicated in equilibrium conditions in the soil, with non-exchangeable acidic-Al (230) and with 'net* CSC (248). On the basis of the wide differences between the amounts of exchangeable Al and those extracted by other reagents, it would seem that the brown soils are Al-buffered systems (138), In the Tiwiwid soil, the major portion of the extractable Al was exchangeable, and this suggests that the extent of Al-buffering in this soil is extremely limited.

PAGE 129

114 Surface Soils Buffering Capacity The titration curves shown in Fig. 6, point to the importance of the organic fraction in the buffering of the soils. The greatest degree of buffering occurred in the Tiwiwid, and the least in Tabela soil, while the Kasarama was only slightly less buffered than the Ebini soil. The direction of increase in buffering capacity was the same as that cf increase in humic and of decrease in fulvic-C The trends were most likely due to differences in the acid strength of the organic matter in the different soils (169), since the clays were all of the same type, and only poorly crystalline. The acid strength of fulvic acid is greater than that of humic acid (336), and, therefore, soils higher in fulvic acid content would be expected to have lower buffering capacities. Cation 5xchan,ge_Capacity_ Buffering capacity is directly related to CEC, and, therefore, to the materials which provide the exchange sites. Using the CSC data for the upper two horizons (Table ll) , simultaneous equations of the form Y = aX x + bX 2 , where X 1 and X 2 are the clay and organic matter contents respectively, in the soil, Y is the measured CEC, and a and b are constants, showed that the organic matter supplied more than 90% of the CSC in each of the brown sand soils (Table 14). The CSC of the organic matter varied from 197.5 meq/lOOg soil in the Tabela to 160.7 meq/lOOg in the Kasarama, while the clay fraction had

PAGE 130

115 ::,l

PAGE 131

116 Table Ik. —Contributions the clay and

PAGE 132

11? a maximum of 1*53 meq/lOOg in the Kasarama, and a minimum value of -3*72 meq/lOOg in the Tabela. The negative quantity, no doubt, indicated that th phous hydroxy Al and Fe exceeded the negative charge in the crystalline clay* Multiple regression, analysis of the CSC for the four surface horizons yielded the equation? Y = 0.029 + O.OOtot + 1.650X s r 2 (for > r s end X~) 0.99 where Y represents the CSG, and X and X_ are the clay and organic matter percentages, respectively. / value for the CEG of the organic fraction of the soil was thus 165 J 3q/l00g, while a comparable value for the clay fraction was 0,*+0 meq/lOOg, Measurements of CSC in soils which contain appreciable amounts of sesquioxides have shown that the sesquioxides exert considerable influence on the measured CSG (29, 66). The saturating cation used in the determination also influences the CSG values obtained (76), and as shown in Table 15* the CSG of the soils varied considerably. It is surprising that the CSG values obtained by the N KC1 method were the highest, rather than the lowest. The high values were undoubtedly related to the larger amounts of Al released from the 2+ soil by KC1 saturation and the use of Ca as the replacing cation. The only reason which can be advanced for the removal of different amounts of Al from the same soils by the same reagent is that approximately 2 years elapsed between measurements, but this does not justify increases of -0.1^, 0.38, 0.70, and O.67 meq Al/lOOg soil in the Tiwiwid, Tabela, Kasarama, and Ebini soils, respectively,

PAGE 133

118 , r1 »

PAGE 134

119 In the later measurement of exchangeable Al. Theestablished sequence of measured CEC N KCL < N NH^OAc (pH 7 9 0) < BaGl 2 -TEA (pH 8*?.) (66) incorporates a pH effect, but it has been shown that an additional source of variation still remains (76). It s< differences in the composition of i' .ions of the soils might account fc plained variation. Since the fulvate salts of the saturating ions are all soluble and, therefore, likely to be removed in the initial process of leaching, si; . regression analyses were run on the N KC1 and N NH^OAc estimates of CSC. The curves in Fig. 7 indicate that the m< i a related to the humic acids. When all soils were considex'ed, the humic-C had average values of 215 and 186 meq/lOOg, by the NH. OAc and KC1 methods, respectively. These values are comparable to those computed for the organic fraction as a whole, end suggest that the possible removal of fulvic acids by monovalent saturating cations warrants some consideration. The distribution of the points indicates that the brown sands may, in fact, belong to a different population from the Tiwiwid sand. Indeed the regression equations for all the soils weret Y 1 = 1,90 + 2.15Xj r 0.96 Y 2 « 3.45 + 1.86X; r 0.71 as compared to those for the brown sands only* Y 1 = 1.16 + 5»38Xj r 0.99 Y g . 1.15 + 11.83X; r = O.98

PAGE 135

120 Open points NILOAc 0.4 0.8 1.2 Humic Carbon, g/lOOg soil. Fig. 7. — Cation exchange capacity in relation to humic-carbon.

PAGE 136

121 where Y 1 and Y 2 represent the CSC measurements "by NH^OAc end KC1, respectively, and X represents hum.ic-C. The larger values for C^C of humic-C in the latter equations are comparable to the total CSC of 680 meq/lOOg rep< tevenson (316) for humic acids and 1,100 meq/lOO^. reported b ;. et al . (188) for extracts from a spodic horizon. The higher values in the N KC1 equation probably resulted from the larger amounts of Al removed from the soils. Incubatio n Studies The soils were incubated in order to ensure complete neuti of added lime and to permit the measurement of the resulting changes in pH. It was found that pH changes varied from the predictions based on the titration curves (Fig. 6) by ± 0.3 pH units. Incubation effected at field capacity by the addition of 5.3, 6.0, 8.0, ( w /w) of distilled water to the Tiwiwid, Tabela, Kasarama, and Sbini soils, respectively. Cation re tenti on The effects of pH and added P on the retention of cations added to the limed soils followed the same general patterns. Liming enhanced the leaching of K and Kg from each soil. Substantial amounts of Ca were also in leachable forms (Table 16). The Tiwiwid sand retained < 2?;i of the Ca added in leachable forms while as much as 95, 52, and y0> were leachable in the Tabela, Kasarama, and Ebini soils, respectively. The trend was similar to that of the fulvic acid contents, and suggested that

PAGE 137

122 o o o o o o o o no N N r-H H H !-! oooo oooo oooo CO lAu-^v^ CM C^-On (M rl ri H OOOO oooo oooo oooo rf £3 °^P, P £" JO^ •"-) CM CM CM >AvO (N^t OHHN O)\0 OnN CM -3" U"\. U^ ri W N r» oooo oooo oooo oooo

PAGE 138

123 leachable Ca was present as fulvate salts* 1 ble Ca incr with increase in pH, and v;as probably in the form of Ca-humates. The addition of $0 ppm P to the Tiwiwid sand decreased the amount of Ca. in leachable form except in the case where no lime was added. Further additions of P caused no cha.nge in the relative amounts of leachable Ca. The depressing effect of P on the amounts of leachable Ca occurred in each of the brown sand soils at 150 ppm level of P additions* This effect was apparently not due to the formation of Ca-P compounds (49* 55$ 86) since the insolubility of such compounds would be expected to cause the amounts of exchangeable Ca to remain constant or to decline. The proportions of Ca in leachable form would, therefore, also be expected to remain constant or to incr: The apparent disappearance of the "P-effect', as the 50ppm level was exceeded in the Tiwiwid sand, supports the possibility that Ca-P compounds were not formed. The persistence of the e P-effect e in the brown sand soils suggested that added P made a definite contribution to the negative charges of these soils. The total amounts of exchangeable cations held against leaching followed the same pattern seen for Ca. The data for the Tiwiwid sand are presented in Table 17. However, the individual cations differed in patterns of retention. The amounts of Mg retained increased initially and then declined. The increase appeared to be consequent upon the addition of 50ppm P. Amounts of K retained showed an increasing trend with addition of P, perhaps because KH PO, was the source of P used. Like Mg, the amounts of K retained decreased with

PAGE 139

12/+ OnvO CO C-o o o <-t cvoo/n. O O O r-< NOon 4-004 UN, o O "N, ! ! O O O r-l CM O O C\ ,G fi CAWHN C^<^iovO VO On O <-T\ on ^r O -c}N^OH \Q C\C) o oodr-i woon r»oo4 IAOOIA Rj o n O Xl B s +> S-i 4* 4S uo o a « E-i

PAGE 140

125 increase in pHj thus, the total amount of exchangeable cations retained was mainly Ca, and the proportion of Ca in each total increased with increase in pH, Tables 18, 19, and 20 show comparable data for the Tabela* and Ebini soils, respectively. In each case trends were similar to those seen in the Tiwiwid sand, except for the increase in exchangeable Ca which occurred with each increment of added P provided some lime had also been added. The total amounts of exchangeable cations retained by each of the soils increased with increase in pH, This is readily appreciated as a consequence of increased CEC due to the increased pH-dependent CEC (120) and the freeing of exchange sites previously blocked by Al (29) and/or Fe (;'. . Phosphate .retention The P added to the soils was found to vary in its degree of solubility in water, as shown in Table 21. In the Tiwiwid sand, 33 to G0% of the added P remained water soluble, regardless of the level of P addition. In the absence of added P, and at the 50ppm level, increasing the pH depressed the amount of P in the water soluble form. This depressing trend was not obvious at the 100 and 150ppm P levels. In the Tabela sand < 11$ of the added P remained water soluble, the proportions increasing from 2.3 to 6,6% at the 50ppm level, to 3 to 12% at the 150ppm level. Increased pH seemed to increase the amount of water soluble P in this soil. The proportions of watersoluble P in the Kasarama soil were similar to those in the Tabela soil, while those in the Ebini soil were much lower, and seemed to decrease with increasing pH.

PAGE 141

126 oooo oooo r-{ O O H HriOW moo tA oooo o o o o CO O O f'> o a .O

PAGE 142

127 C^ O O 0^ O O H NOON O 1 ^ O O C> l> C^\ o o c*-\ e"\ o o r> 3* 0) o

PAGE 143

128 «J

PAGE 144

129 4-N\0 CO . OCOONN VO-^O^C^ oooo o o o o 1 1

PAGE 145

130 22 shows the amounts of P removed by the Bray #1 extractant. In the Tiwiwid soil, virtually all the P added was removed at the 50ppm level, while 85 to 100/5 was removed at the • level, and 72 to 90^ at the highest level of P addition. The proportions reyaoved from each of the brown sand soils were ; 32 to $6% at the 50ppm, 3? to 68;$ at the lOOppm, and 37 to t the 150ppm P levels, respectively. The trends of P ibility by the Bray #1 extractant were not very consistent. Extractable P seemed to increase with pH in the Tabela soil except zero F level, but this trend was less obvious at the higher P levels 6 In the Kasarama soil, there seemed to be a decreasing trend at the 50ppm P level, but an increasing trend at the 100 and 150ppm P levels. In the Ebini soil, extractable P appeared to decrease with increase in pH. The Bray #1 extractant removed the more readily soluble portions of each form of available soil P, and may thus have extracted P from Ca-Pfr and Fe~P compounds formed in the soils. The amounts of Ca-P wereunlikely to have been appreciable, as indicated by the increasing amounts of exchangeable Ca found at the higher levels of lime. The amounts of P obtained from Al-P compounds were likely to have been of much greater importance than those from Fe-P compounds, because of the relatively larger quantities of Al than Fe in the soils (330), and because fractionation studies (55, 118, 178, 275, 3^7) have shown that Al-P is frequently the form of soil P which is increased most on the addition of P to the soil. Further,

PAGE 146

131 O CO ON CO OSHHN \Of°\C^CO rHr-!QC^M ri y L» j^j-ar% ;l « irwf) Mp CO C\vO\D N dT\<)\0 1AVAOO OOOO O^AOO O O^O CO OONOJiH u^-d" C\J CO O.-^i/"\vo vPyvovO C*-3" vPivO C^-3" u^\vO vO I III O U"\0 O CO ON\D vr>

PAGE 147

132 the trends of P retention against extraction increased in a direction similar to that of increasing extractable Al content being least in the Tiwiwid sand and highest in the Ebini soil. Several reports attest to the possible reaction of P with extractable Al (2W-, 293), exchangeable Al (62, 132), Al complexed by organic matter, and Al in gibbsite and kaolinite (l?6, 177, 2^9). The absence of gibbsite and the different amounts of Al extracted from the soils by different reagents suggested that Al complexed by organic matter was the most likely source of Al for reaction with P. The possible contribution of Fe-P compounds to the extracted P cannot be ignored, since hydrous Fe oxides (226, 330), and Fe complexed by organic matter (168, 320) are both capable of reacting with P. Soil Properties in Relation to Potential Soil Fertility The absence of samples of subsurface horizons of the Tiwiwid sand precluded valid comparisons of the profiles. However, data adduced by Bleackley and Khan (31, 32) indicated that the Tiwiwid sand has been subjected to a considerable degree of podzolization. The three profiles examined displayed some differences, but these were differences in degree only. It would appear from the trends in the contents of organic matter, amorphous materials, and clay, that the three soils are similar in origin and have passed through the same developmental procesees. The brown sand soils seemed to differ, as a group, from the Tiwiwid sand, and though all the soils were kaolinitic, the apparent differences were reiterated by comparisons

PAGE 148

133 of the surface horizons of the soils. The differences were seen in the composition of the organic fraction, in the contents of amorphous materials, and in the soil contents and patterns of P retention. The similarities in the four soils were seen in the patterns of cation retention and in the potential for loss of cations "by leaching. These similarities originated in the sizeable contributions to the negative charges of the soils, by the organic fraction. The clay fractions were relatively small in each of the soils and contained poorly crystalline kaolinite, probably as mute testimony of the millenia of weathering to which the soils have been subjected (l!9» 303). Indeed, the soils comformed taxonomically to the CEC-source groupings found by Yuan et al. (3^6). They also conformed to the predictions of Spencer (284), since with more than 90% of the CSC due to the organic fraction in each case, Ca was apparently preferentially bonded, while K and Mg were virtually excluded. The high potential for loss of K and Mg by leaching, and to a lesser extent, the relatively large amounts of Ca which were held in leachable form, offer some explanation for the failure of added fertilizers to have any effect on the exchangeable nutrient status of these soils (315). Each soil showed the same type of response in cation retention to pH increase, and though this may have been due in part, particularly at higher pH levels, to the sorption of Ca(OH) ions 2+ rather than Ca (53, 169, 213), the increased contribution of the pH-dependent CEC (29, 120) must have been the major factor. The importance of pH -dependent CEC was reflected by the amounts

PAGE 149

134 of titratable acidity as compared to the amounts of exchangeable Al in the soils* The data presented in Table 23 indicated that though there were considerable amounts of acidity in each soil, less than IJfc of this was attributable to exchangeable Al in any of the soils. Thus, there must be appreciable numbers of exchange sites which became active as pH was increased, either by the release of nonex changeable Al or by the dissociation of functional groups (169). The data presented in Table 23 also indicated the existence of some obvious differences in the soils. The Tiwiwid sand displayed very different degrees of Al and base-saturation as compared to the brown sand soils. The Al-saturation computations based on the more recent measurements in the process of KC1-CSG determinations were approximately 6, 26, 26, and ZZfo in the Tiwiwid, Tabela, Kasarama, and Ebini soils, respectively. The existence of group differences was reiterated by the higher base saturation of the Tiwiwid sand, with its higher CSC, but lower pH. Exchangeable Al was apparently not the major factor in determination of the pH of this soil. The brown sand soils on the other hand contained less exchangeable bases than exchangeable Al but had higher pH levels. The relative amounts of amorphous Al, and of Al extracted by 0.^ NaOH suggested that these soils complied with the Al-equilibriura conditions postulated by Pi on eke and Corey (230). Each soil showed a positive effect of P addition on the amounts of exchangeable Ca retained, but here again there were differences.

PAGE 150

135 • to

PAGE 151

136 The P effect was relatively short-lived in the Tiwiwid sand. Similarly, as indicated by the solubility of added P in Kater, the Tiwiwid sand sorbed much less P than the brown sand soils, at any of the P levels used. This was to be expected from the contents of clay and of hydrous oxides of Al and Fe in the soils. The virtually complete removal of added P by the Bray #1 extractant from the Tiwiwid sand indicated that P sorption in this soil was reversible with pH, probably because the adsorption process was mainly physical (l?), and/or because it was effected largely by kaolinite (207) . The greater amounts of P sorbed by the brown sand soils and the partial irreversibility of adsorption suggested that chemisorption (17) was of greater importance, as were the amorphous hydrous oxides (207) in this context, in these soils. Forms of the hydrous oxides were not definitely established, but they were very probably associated with the organic fraction of the soils. The likelihood of involvement of Al-organic matter complexes in P sorption was bolstered by the removal of comparable amounts of Al from the brown sand soils by the 0.
PAGE 152

13? So also did the parallel patterns of increase in CSC and P sorption on application of P. The adsorption of P has been shown to increase the negative charge on the adsorbing surface (123, 206), and it was apparently this increase which Mehlich (192, 19^, 196) referred to as'^CSC due to AEG." Though the Tiwiwid sand had apparently ample levels of P (Table 2l), each of the soils was very poorly supplied with native P. Their contents of exchangeable nutrient cations were also very low, despite surprisingly high total amounts in the soils. The fertility of a soil is determined by the level and nature of acidity, the buffering capacity, and the capacity to retain added nutrients in plant -available forms. The data presented indicated that the soils should be expected to differ in potential fertility, with the most marked sources of the differences dependent on the nature of the buffering systems and the patterns of sorption of P.

PAGE 153

138 Plant Growth Limiting Nutrients Panftola^crass The variations in average dry matter production of pangolagrass, consequent upon the addition of individual major nutrients, and fritted mi cronutrients— referred to ac Tr— are shown in Table Zh» The data for the individual harvests are presented in Table 53 and 54, The main effects of each treatment are summarized in Table 25, and described in terms of the average rates of change. In the absence of a response, the slope computed for a given treatment should be approximately zero (20l). Negative slopes could indicate either the existence of severe deficiencies, or of toxicities induced byexcessive supplies of the corresponding treatment, as postulated by the generalized Steenberg curve (279). In the Tiwiwid sand, the response to the addition of S was negative while all other responses were positive. The slopes of the response to N and Tr were approximately 0.9 and 0.6, respectively, implying that on this soil, the growth of pangolagrass was limited primarily by the supply of if, and to a lesser extent by the supply of mi cronutrients . The F -tests (Table SS) showed that only the N response was significant. In the Tabela soil, all responses were positive with those to N and K being significant, and that to P highly significant. The slope of the P response was extremely steep, suggesting that the supply of P to pangolagrass was far from optimal. The data for the Kasararaa soil were identical in meaning to those

PAGE 154

139 Table 24. — Average shoot weights (oven-dried) from two harvests of pangolagrass . ========== — ===

PAGE 155

1*K) °i

PAGE 156

UU for the Tabela soil, except for the negative response to micrormtrients. In the Ebini soil, only the response to P was significant, and again, the response to micronutrients was negative. The root weights of pangolagrass (Table 26) were largest in the Tiwiwid sand, and appeared to reflect beneficial interactions of K with the other nutrients supplied. In the brown sand soils, maximum root growth was obtained when N, S, P, and micronutrients were supplied together. Like shoots, root weights in the control treatments increased in the order: Tabela < Ebini < Tiwiwid, and responded in very similar ways to a given treatment In the brown sand soils. Taken as a whole, the data indicated that N was the nutrient of major importance in the Tiwiwid sand while P was the nutrient which limited growth of pangolagrass on the brown sand soils. Additions of N and K to the latter soils also had beneficial effects. Pigeon pea The shoot weight variations of pigeon pea are shown in Table 27, and the responses to the added nutrients are summarized in Table 28. All responses were positive in the Kasarama soil, but those to S and K were negative in both the Tiwiwid and Tabela soils. In the Tiwiwid sand, responses to N and micronutrients were significant (Table 55), but micronutrient additions gave the largest response. In the Tabela soil, the positive responses to N, P, and micronutrients were all highly significant, as was the P response In the Kasarama soil. The N response in the latter soil was positive and significant. The data in Table 29

PAGE 157

142 Table 26. —Root weights (oven -dried) of pangolagrass after two harvests* Trc?1 Soil

PAGE 158

11*3 Table 27. —Shoot weights (oven-dried) of pigeon pea, Treatment

PAGE 159

ikh o

PAGE 160

145 Table 29. — Root weights (oven-dried) of pigeon pea after one harvest. Treatment Soil Tiwiwid Control 4.0 N 3.0 S 4.6 P 7.0 K 2.5 Tr 7.6 NS 1.2 NP 3.7 NK 2.3 NTr 5.6 SP 5.0 SK 1.7 STr 4.7 PK 2.6 PTr 4.5 KTr 3.1 NSP 5.2 NSK 4.7 NSTr 3.1 NPK 3.4 NPTr 4.2 NKTr 6.1 SPK 4.3 SPTr 3.6 SKTr 4.2 PKTr 4.0 NSPK 1.3 NSPTr 4.4 NSKTr 1.1 NPKTr 4.3 SPKTr 2.0 NSPKTr 3.2 'abela

PAGE 161

1H6 revealed that root weights, following the pattern of shoot weights, were highest in the Tiwiwid sand and that those in the Tabela and Kasarama soils were similar. In the latter soils, variations due to treatments were similar. Summaries of the responses in root growth are shown in Table JO, In the Tiwiwid soil, the responses to F and Tr wexe both positive, but only the negative effect of K was significant (Table 57) • The K response was also negative, but not significant in the Tabela, as was the S response in the Kasarama soil. Responses to P and Tr were both positive and significant in the two brown sand soils. The negative, though significant, effect of N in the Tiwiwid sand implied that in this soil nodulation was depressed by the action of added inorganic N on Rhizoblal activity (287). The positive N responses on the Tabela and Kasarama soils suggested that N fixation was sub-optimal in the -N treatments in those soils. The sub-optimal level of N fixation was undoubtedly related to the steep slope of the P response in both of these soils. It was obvious that maximum nodulation occurred only after the addition of P and micronutrients (Fig. 8), and as has been reported^ maximal nodulation was reflected by the size as well as number of the nodules formed (212). The effectiveness of those nodules which formed was dependent on the development of hemoglobin and, therefore, of a reddish coloration within the nodules (217). The absence of P and/or micronutrients, resulted in a reduction in size and in the degree of coloration of the nodules (Fig. 9).

PAGE 162

Ik? CO 0~\ o co o _ o C^v/-\CMC\I -^ ON »A C\2 ONC^iO j-nno ?o ON CM H O O CO PM O Cv. \ Cv-vr\C\JiH CM r-H rH CO W) •••• e e t « CM OCN.O OH CO O VOIA C^H CM On On On O ^J" 0-CN.O UNnO M I ON CO cJ-vO C-J VACOC--0 "OHlTiO Vf\c0cv-0 NiANO W O

PAGE 163

148 (a) Main effects and some first order interactions of added nutrients. (b) Main effects and interactions of added P. Fig. 8. — Effects of nutrient elements on the development of nodules on pigeon pea roots in Tabela sand.

PAGE 164

149 (a) Main effects of added nutrients. (b) Some main effects and interaction of added nutrients . Fig. 9. — Effects of nutrient elements on the size and coloration of nodules developed by pigeon pea roots in Tabela sand.

PAGE 165

150 Central Composite Studies Pang;olagrass The dry matter yields of pangolagrass on the three brown sand soils, resulting fron\ various treatment combinations, are presented in Table 31 • The omission of several treatment combinations permitted by the experimental design, dictated that treatments could best be described by regression coefficients. The coefficients obtained for each of the parameters when the model equation (20l) was fitted to the dry matter yield data are shown in Table 32. It is evident that in each soil there was a significant negative effect on yield induced by the addition of Ca, in contrast to a positive effect by the addition of P. The coefficient for P was appreciable but not significant in the Kasarama soil. Lack of significance was, however, apparently not due to a lesser effect of P on yield, but rather to the absence of any previous applications of P and the wider pH range utilized in this soil. The effect of K addition was negative and non -significant in the Tabela and Kasarama soils, but was positive and significant in the Ebini soil. The o Ca effect was also positive and significant in the last-named soil. The fitted equations for the Tabela, Kasarama, and Ebini soils are diagrammatically represented in Fig, 10, 11, and 12, respectively, with computations based on the K level held constant at lOOppm. The figures show that at intermediate levels of P and K, yield decreased by some 10fo from 9.2 to 8.3 g/pot as pH increased from 6.05 to 7. 30 in the Tabela sand. In the Kasarama loamy sand, the yield dropped by approximately 57% from 15. 7 to 6.7 g/pot as pH increased from

PAGE 166

151 I ;-. o

PAGE 167

152 I

PAGE 168


PAGE 169


PAGE 170


PAGE 171

156 4.90 to 7.15, while a decrease of about 28$ from 12.6 to 9.0 g/pot occurred as the pH of the Ebini sandy loam increased from 5.50 to 7.15. At intermediate levels of K and pH, the application of P resulted in yield increases of 5*4-, 57, and h6% in the Tabela, Kasarama, and Ebini soils, respectively. The yield responses on the Tabela and Ebini soils were sizeable despite previous applications of lOOppm P to the soils in the experiment used to determine the limiting nutrients. The negative effect of increasing pH on yield was more pronounced in terms of P concentration in the tissues of pangolagrass. Data in Table 33» showed that at low levels of P and K the concentration of P in the plant tissues increased somewhat with increase in pH; however, at high levels of P and K, increase in pH resulted in a decrease in P concentration. Similar trends were obtained for total P uptake as shown in Table 34. Coefficients of the fitted equations for P uptake (Table 35) reiterated the highly significant negative effects of pH, and the positive effects of P. The effect of K on P uptake varied, but was not significant. It was negative in the Tabela and Kasarama soils, but positive in the Ebini. The latter soil had a significant effect on P uptake due to the Ca x P interaction. Diagrammatic representations of the fitted equations for P uptake, at the lOOppm K level „ from the Tabela, Kasarama, and Ebini soils are presented in Fig. 13, 14, and 15, respectively. The surfaces for P uptake were strikingly similar to those for

PAGE 172

157 n H I! P ii P It U tr\

PAGE 173

158 n x) n «5 I WM (^ I o C^i • i <• -d-OOOvO vO O O O -d o 1*11 I \r\ i o vO I -d O t • I s* a I • o o o o o O O Q r-\ ONOOOO -d" O O O \T> vOQ^-On m0?4-O N iHQOOCO iH^-vOvO r\^o\o N CA-3CO C\J CM VOVOVONN -d" IA\D C^ C^VMA^ONN I

PAGE 174

159 a i b o

PAGE 175


PAGE 176


PAGE 177


PAGE 178

163 dry matter yields, emphasizing the importance of the supply of P for plant growth on the soils. The negative effect of increasing pH would appear to be due to the decreased solubility of the compounds formed on the addition of P to these soils. This is consistent with the effects of pH > 5 on complexes of the form P-(A1, Fe)~fulvie acids (l68). The concentration of Ca in the pangolagrass tissues increased with increase in pH (Table 36). The coefficients for the Ca uptake equation, shown in Table 37, indicated that pH had no real effect on Ca uptake c The effects of P on Ca uptake were positive but significant only in the Ebini soil; K exerted a negative influence in the Tabela and Kasarama soils. The non-significance of the coefficients for Ca and K in each of the soils, and for P on two soils indicated that the supply of Ca was adequate at all pH levels, and that the release of Ca to the plant roots by the soil did not depend, to any appreciable extent, on the degree of Ca-saturation of the soil. The data in Table 38 indicate that the concentration of Mg in the tissues was depressed by increase of pH and P levels. The uptake of Mg, as shown in Table 39, was subject to a significant negative effect from increased pH. The effects of P addition were positive but significant only in the Ebini sandy loam. Increase in the K level exerted a negative effect which attained significance in the Tabela sand only. The high degree of solubility of Mg at high pH values suggested that Mg uptake would be enhanced by increasing the

PAGE 179

164 o

PAGE 180

165 S3 & C^ ^ ^v
PAGE 181

166 o I I c I I I I O » I « O I O I i i o i : io! > o »r> o CO CM I • ( o CM CM O I O 1 O I O I 1! * O I O I o O ! O I O hi & in CM o I vrs O CM CM O I o NOOOCO -3" O O O va CO O O O CM ONQ O O O CMC^oS iH CM U-\CO CO O O O O O c>o o o c^ u->o o o cJcH CM 5MD MD &SZ. MDMDMDO-C^Jir\M3 C 1 ^ l>lAVOvO NN 3 8

PAGE 182

167 *

PAGE 183

168 soil pH; however, the solubility of Ca also increased with increase in pH, and this permitted greater competition "between the ions for uptake by plant roots. The larger quantities of Ca in the soil solution could thus decrease the activity and uptake of Kg. The positive effect of P may have been a consequence of improved growth consequent on the addition of P to the soils. The greater quantity and solubility of K in the soils probably in< petition between the cations for absorption sites on the plant roots and were thereby responsible for the negative K effect. The concentration of K increased with increases in the levels of K and pH, but decreased at high levels of P, as shown in Table *K). The increase in K concentration with increases in pH was probably the combined result of reduced growth and enhanced solubility of K at higher pH levels. The increase in K concentration with increased supply of K was undoubtedly due to the greater amounts of K in the soil solution which resulted from this treatment. The decrease in K concentration seen at higher P levels was possibly a consequence of the sorption of K with P. The uptake of K suffered a negative pH effect in each soil, but this was only significant in the Kasarama soil (Table kl) , The P effect was positive in each soil but was not significant in the Kasarama soil, while the K effect was positive and highly significant in each case. The interaction of Ca x P exerted a significant effect in the Kasarama and Ebini soils but this was negative in the Ebini. Similarly, the Ca x K interaction was highly significant in the

PAGE 184

169 i c a .1.1 MINI

PAGE 185

170 f)

PAGE 186

171 Kasarama and Sbini soils, "but was negative in the Tabela and Kasarama. The P x K interaction exerted a positive and significant effect in the Tabela and Ebini, and a negative effect in the Kasarama soil. Table k2 shows that the concentration of N in the tissues tended to decrease with increase in pH, K, and P. Uptake of N (Table ^3) was subject to a highly significant negative Ca effect, and to a positive P effect. The P effect was highly significant in the Tabela and Ebini soils. The effect of K on N uptake varied. It was significant in the Tabela and Ebini soils, but negative in the Tabela and Kasarama. The effects of the Ca x K and P x K interactions r> 2 and of the P~ and K terms were significant in the Tabela soil 2 only, vrith the Ca x K interaction and the K effects being negative. The depression of N concentration at higher pH levels was probably due to competition between Ca and N, with Ca uptake being facilitated by the higher amount of Ca in the soil solution. A similar explanation probably applies to the K effect, while the P effect may have been due to dilution concommitant with the greater growth of the plants at constant N and higher P levels. The uniformity of color in the plants was perhaps indicative of better utilization of N at the higher pH levels (205). The effects of acidity on plant growth are usually attributed to high levels of activity of H + and/or to Al toxicity (114, 24l). + In the pK ranges used in this study, H should have been largely neutralized (2?^), leaving Al as the factor most likely to be active.

PAGE 187


PAGE 188

173 in en en en o ( > o

PAGE 189

174 The data in Table 44 indicated that despite the visually normal growth of the roots, substantial amounts of Al were associated with them. The data revealed no trends in the amounts of Al in relation to changes in pHj h rer, increasing levels of K decreased the amounts of Al in the roots . This may have been due to the formation of insoluble Komds (295) which reduced the amounts of mobile Al in the soil. This possibility is supported by similar, but less marked, reduction of root P by increased K levels as shown in Table 45. These data also show that concentration of P in the root increased with increase in pH. The AlsP ratios presented in Table 46 were narrower at higher pH levels in the Sbini and Kasarama soils suggesting that more P might have T ilized by Al in or on the roots (333* 334) at higher pH levels. The tendency for the ratios to have values of about 3 to 4 indicated tl >le formation of Al-P compounds in the roots (60). The inconsistent effects of K on the Al/P ratios probably reflected differences in the amounts of P sorbsd in the soil as potassium phosphate salts (20?) and, therefore, in the degree to which the P sorption capacity of each soil was satisfied. The absence of any really overwhelming trends in the Al relations in the root system, seems to imply that the effects of pH increases on plant growth were related more to changes in soil properties than to impaired plant metabolism. Pigeon pea Pigeon pea plants, grown on both the Tiwiwid and Kasarama soils, developed necrotic areas on the lower leaves soon after abscission of the cotyledons. The necrosis began at the apex of the older leaves,

PAGE 190

175 ii 1/ li a ii « . ii 'd ii II -rf . II 0) In O O C i o I rH I rH I Vf> U>v IT\ N N N H i rH I H O o o O »A H I O i VO t CM CM I CM CM O O O CO 3O O O vr\ O O O rH S H I O I I CM I rH I i a i s> i f CM 8 CM I VA XT\ o tNCM rH ^ I -T> I O H In I ri o I O I I CM I O CM O I 00 rH I CM CM I CA I I I CM I I I v£> I I CM I CO O O O CM OOOOO OOO c-^oooc^ NOVO vriOQO-d" H CVHACOCO rHcM^-vOvO fc: \0\O\O NN ^ vr\\D Cv C^ir-\ mvo tS C^

PAGE 191


PAGE 192

177 1 a IN I iA I I | II P II 0"\ u-\ \o I e I • ^Itfll«) ir\ I CO 1 CM ; I-'-: On ON U*» o C'N I O O I « C\iO O O CO -d-O o O M> •3" I » I ^ COOOOCV! CvJOOOO ONOOOO c^-oooc^ n5 pj o« moooj HWmtOCO HCNi-'d'NONO sOnO vD C^-0^t vnvO DN \r\ir\M3 C^N

PAGE 193

178 and the affected areas expanded along the margins towards the leaf base and inwards to the midrib. The dried margins curled downwards and inwards shortly before the premature abscission of the affected leaves. The symptoms bore some similarities to those described by Nichols (212) for K and Mg deficiencies, but showed no variations in severity, and were alleviated by neither the addition of *K)ppm Mg to each pot, nor the foliar application of each of the micronutrients. The heavy leaf fall during the study rendered the dry weights of the plants little more than inaccurate estimates of dry matter production. In addition to the problem of leaf fall, the soil pH levels after harvest varied considerably from those anticipated. The lowest lime treatment on the Kasarama soil resulted in a final pH which was very close to that produced by the highest level of lime. This similarity in pH was very likely due to incorrect mixing of the limed and unlimed soils in this particular case. The dry weights of pigeon p?a attributable to the different treatments are shown in Table ^7. The plants grown on the Tiwiwid sand showed a very positive response to increased Ca supply, while responses to P and K increased to a maximum and then decreased. Plants grown on the Kasarama soil, ignoring those of the pH 6.10 treatment, seemed to have responded positively to increased P and negatively to increased K. The concentration of P (Table h8) in the plant material from the Tiwiwid soil, increased to a maximum and then declined as the P supply was increased (at the lOOppm K level). There seemed to be no appreciable effect due to K, but P concentration was lowered

PAGE 194

179 to -A

PAGE 195

160 13

PAGE 196

181 "by increased pH. In the plants grown on the Kasarama soil, the concentration of P v:as influenced to a very small degree only. The major effect v:as positive and due to K, while pH and P increases exerted very little effect. Table H-9 shows the concentrations of Ca, Mg, and K in the tissues of plants grown on the Tiwiwid sand. The concentration of Ca increased with increase in P level, but was not affected greatly by changes in pH. The concentration of Mg also increased with increases in P, but decreased as K .and pH levels increased. Increases in P and K levels increased the concentration of K. Plants grown on the Kasarama soil showed a positive effect of K on Ca concentration (Table 50) at low P levels. Increasing P levels depressed the Ca concentration as compared to the increases which resulted from increases in K and pH. The concentration of Kg was Increased by increasing the K level, but was not greatly affected by increases in pH and in P supply. Increasing the K levels increased the K concentration in the plants, while increasing pH and P levels decreased it. Except for the Ca concentration in plants grown on the Tiwiwid sand at low P levels, the concentrations of the cations were very similar in plants from both soils. The levels of 0.44>t Mg and 1.85$ Ca, regarded as adequate (212), were not met by the plants on either soil, though the concentration of 0.25/5, regarded as the point of incipient Mg deficiency was generally exceeded. The K concentrations in all plants exceeded the 1.65/2 level indicated as the threshold of K deficiency, and frequently exceeded the 2.0?^ level

PAGE 197

v ', O I • I I r-l I I I H CM en I I . I I I O I I O I O I t^CM \0 O^S I • I • I N I CJ O I & « * , °. , On CO vr\ Hi O I O I o I O I O I CM O O O CO 40001^ CM O O O CO 4 OOO^ I CM » CM I m 00 IN Iri I 95 ON 1*11 I rH I I CM O O O CO 3" O O O ""\ M)\0\0 NN \0 vflvO NN ^)\0\ONN 182

PAGE 198

183 P P-i II f-i ii 3 a H I rH I iH I I H I I I H I H I CM O O O CO -S OO O m met O cm .* CM cncn 1^ I o VO CM r • i o I O ! O I v£> CO O I O 1 I I O I I CM O O O CO ^h O O O *A in co o cnzjCM CM C^\ I CM I CM I I I CM I I CM O O O CO -do o o *n mCO O CM -* CM en en r-i

PAGE 199

184 regarded as adequate (212) » The P concentrations were higher by about lOOfo in plants grown on the Tiwivrid sand than in those grown on the Kasarama soil. The data did not provide any explanation for the necrosis and fall of leaves, and the fact that the soils were fumigated should have minimized the possibility of root pests such as nematodes* There are, however, no actual data on the latter possibility. Soil Fertility an d Amelioration The patterns of plant growth on the four soils reflected the differences in the properties of the Tiwiwid sand on the one hand and of the three brown sand soils on the other. The strong similarities among the brown sand soils are borne out by the data in Table 51 • The means of the levels of dry matter production of pangolagrass and of the levels of nutrient uptake provided by the fitted regression equations, were used to compute the mean levels of concentration. The mean dry matter production ranged from 9.56 g/pot on the Tabela sand to 11.09 g/pot on the Kasarama loamy sand. Mean uptakes of Ca, K, and N were highest from the Kasarama soil, while P uptake followed the soil sequence Tabela > Kasarama > Ebini. The concentration of P in the plants followed the same sequence as P uptake. Among the elements considered, P displayed the greatest variation with soil in terms of both concentration and uptake by the plants. The concentration of N was virtually constant, while concentrations of Mg and K were largest in the plants grown on the Tabela sand, and lowest in those grown on the Kasarama loamy sand.

PAGE 200

185 o c -p o
PAGE 201

186 These differences appeared to be related to the differences in the pH range studied in the different soils and, therefore, in the amounts of the cations which were present in water-soluble forms. The uptake of P may have also been influenced by the differences in pH ranges, but it increased in the same direction as did the proportion of fulvic acids in the organic fraction of the soils, and decreased with increasing amounts of alkali -extractable Al. Computation of the levels of each of the three nutrients of variable supply in the study, which are required for maximum dry matter production was not conclusive. The data obtained when the partial derivatives of the fitted regression equations were equated to zero and solved, are presented in Table 52. The coded range of nutrient levels was from -1.215 to 1.215, and those extremes were frequently exceeded by the roots of the equations. The computations have no real value, since extrapolation beyond the limits of the experimental area is meaningless (20l). They, however, indicate the directions in which treatment levels should be adjusted for the maximization of yields. When decoded, the roots of the equations indicate that dry matter production on the Tabela sand was nearer to a minimum level and that at that point the levels of K and P were lj-9.5 and *K).5» respectively, and the soil pH somewhat lower than pH 6, Yields on the Sbini sandy loam were also nearer to a minimum than to a maximum. In this soil, excessive levels of K (> l60ppm), and inadequate quantities of P (< 37PPm) combined to produce minimum yields at pH 6.4. The Kasarama loamy sand produced yields which were nearer

PAGE 202

187 .• ' r

PAGE 203

188 to the maximum. The conditions for maximum yields on. this soil were high P levels (> l60ppm) and low K supplies (< 37PPn») at a pH somewhat less than ^.9. The computations underline the importance of the supply of P to plant growth on these soils; they indicate that relatively small quantities of K are needed and that lower pH levels favor higher yields of pangolagrass. Growth of pigeon pea and pangolagrass was appreciably better on the Tiwiwid sand. Cation concentrations in the legume were remarkably similar on the Tiwiwid and Kasarama soils, except for Ca concentration which was higher in the plants grown on the Kasarama soil at the lower P levels even though the latter soil had a higher pH level. The better growth of the pigeon pea on the Tiwiwid sand was paralleled by P concentrations which were approximately 100;2 higher than those obtained on the Kasarama loamy sand. The shortcomings of the pigeon pea study on the Tiwiwid and Kasarama soils did not permit the effective delineation of differences in response by the two crops to given soil conditions. However, it is not unlikely that the grass would affect the soil differently in terms of its residues, exudates, and relative uptake of nutrients. Differences in the capacities of the plants to modify the rhizosphere are likely to affect the availability of P in the later stages of plant growth and no real meaning can be attributed to the differences in the concentrations of P in the legume as compared to the grass. It is clear, nevertheless, that P added to the brown sand soils did not remain available to plants for long periods of time, nor did its availability improve with time. The Kasarama soil used for the

PAGE 204

189 growth of the pigeor pea plants had previously received lOOppm P, hut this treatment did not appear to affect the concentration of P in the plants. Similarly, the Tabela and Ebini soils used for the growth of pangolagrass had been given previous applications of P. The response surface for dry matter yield on the Ebini £-oil suggested that the P fixation capacity of that soil was at least partly satisfied. Levels of P in the plants were, however, not very different from those on the Kasarama soil which received no previous P treatment. Practices aimed at the amelioration of the fertility levels of these soils will need to be based on the properties of the soils. Plant growth on the Tiwiwid sand appeared to be limited by the supply of N, while the limiting factor on the three brown sand soils was P. These limitations in the unamended soils were complicated by the inherently low amounts of exchangeable K, Hg and Ca. Though the Tiwiwid sand had a higher base status than the brown sand soils, there was a common factor among the soils in that the source of sites for cation exchange was primarily in the organic fraction. The addition of N, K, Kg, and S is, therefore, likely to be negated by leaching, if the moisture regime of the soils is not controlled. A similar condition applies to P added to the Tiwiwid sand. Each soil derived some bebefit from the addition of micronutrients and Ca, but the effect of the latter element on the growth of pangolagrass on the brown sand soils dictates the need for caution in its use.

PAGE 205

190 The yield data for pigeon pea or. the Kasarama loamy sand were far from conclusive but indicated a negative effect of increasing pH at low P levels. It is evident that high pH levels depressed dry matter production by pangolagrass , The negative effect of high pH on the brown sand soils was apparently related to decreased P availability to plants. The high correlation between Bray #1 extractable P and plant growth (285) validates the conclusion that only a small fraction of the P added to these coils was available to plants. The effect of increasing pH on the extractability of added P was not clearly demonstrated, but plant uptake of P was obviously lower at higher pH levels. The negative effect of high pK on the availability of P to plants was contrary to the classically accepted principles of soil fertility, but it has been reported to occur in several tropical and subtropical soils. This effect has been observed on Oxisols in Natal (2^6), and on several latosols in Hawaii (9^). A striking demonstration was provided by an amorphous Entisol from Costa Pdca on which a given yield of pangolagrass was attained on the addition of lime, only if amounts of P were applied in proportion to the resulting changes in pH (173). The latosols were generally red-yellow soils and could therefore, be expected to have organic fractions which were dominated by fulvic acids (159, 233, 300). The stability of these acids in the soil depends on their degree of Al saturation (2?0, 336), and the pH levels of the brown sand soils studied suggest that Al-fulvate complexes may have been the principal agents of buffering. The capacity for competitive binding of fulvates and P by Al has been demonstrated

PAGE 206

191 (163, 168), and the preference for the binding of fulvates at lower pH levels— about 5—can be appreciated by consideration of the pK values of the organic matter (182, 235) and of hydroxy -Al (138). At higher pH levels P is increasingly bonded by Al or Fe at the expense of fulvates, and the addition of P to Al-fulvate soil systems could be expected to result in the formation of more fulvate-Al-P compounds at lower pH levels and the formation of more Al-P at higher pH levels (163). The greater ease of extractability of P from fulvate~(Al, Fe)~P complexes than from Al-P or Fe-P (160), explains at least, in part, the decrease in availability of added P as the soil pH increases. Though both Fe-P and Al-P are known to be capable of supplying P to plants (7, 118, 178, 29^, 331), it is also known that Al and Fe compounds are least soluble at pH 6 to 8 (238) . Further, though Al-P tends to become the more important form of P as the soil pH increases (118, 1?8), the solubility of Al-P is depressed by Ca (328) * In addition to their role in P retention, the hydrous oxides influence the relative amounts of exchangeable cations in the soil. The oxides favor the retention of Ca (134), but render K more soluble (I87, 193, 297). The low amounts of exchangeable K in the soils relative to the total soil contents of K and uptake of K by pangolagrass were probably indicative of the operation of a K equilibrium condition, which is controlled by amorphous Al-and Fesllicates (222, 304). The increase in exchangeable Ca in the brown sand soils on addition of P was likely to be due to the sorption of

PAGE 207

192 P by hydroxy-Al and hydroxy -Fe, with an increase in the negative charge on the surface of the sesquioxide (2^, 123). The sources of acidity— H and ionic Al=-are both likely to be neutralized at or below pH $A (l6, 27^, 3^l). Maximum yields of pangolagrass on the brown sand soils are to be expected at pH < 5.*U This, taken in conjunction with the adverse effects of high pH on the retention of K and Mg, and on the availability of P, as well as the low Ca uptake by the grass, suggested that exchangeable Al is likely to be the most meaningful criterion for liming practices on these soils. The replacement of organic matter by OH" in organic matter sesquioxide complexes at higher pH levels (l68) is likely to be conducive to the loss of organic matter by leaching with subsequent structural deterioration, and this makes an additional argument against the use of large amounts of lime. The differences in fertility between the Tiwiwid and the brown sand soils seem to hinge on the proportions of sesquioxides present. These differences are manifest largely in the availability of added P, and the optimum pH levels of the soils. The amelioration of the fertility levels of the soils should have as its immediate objective the increase of the capacities of the soils to retain nutrients against percolating water and the provision of adequate levels of available P. The low levels of pH which have been shown to be desirable in the brown sand soils, suggest that the amelioration of fertility levels can be achieved by the judicious exploitation of the effects of additions of P to these soils. These effects of F addition which serve to increase K retention (15), pH and CSC (232), N fixation

PAGE 208

193 (8l) and promote the accumulation of organic matter (262, 329) suggest that the quenching of the P fixation capacity of the soils should be attempted with close attention being paid to any ecological changes that may occur (219). The source of P likely to be most satisfactory in this context appears to be superphosphate,

PAGE 209

SUMMARY AND CONCLUSIONS The soils of the "white sand" plateau of Guyana are all highly weathered, having developed from materials eluviated from the weathered surfaces of the Guiana Shield, and deposited in their present positions during the third interglacial period. Laboratory studies revealed that the Tiwiwid sand and three brown sand soils were comprised largely of quartz, and that their contents of clay in the surface horizons increased in the order: Tiwiwid < Tabela < Kasarama < Ebini. This sequence is the reverse of the elevation of the soils. The clay contents also increased with depth in the profile in the Tabela, Kasarama, and Ebini soils. The clay fraction of the soils consisted largely of poorly crystalline kaolinite, and had < 10^ of its weight in the form of sesquioxides. Gibbsite was absent from each of the soils. The sesquioxides were amorphous and associated with substantial quantities of SiO . Hydrous Al apparently originated from the degradation of kaolinite crystals, and was retained in the Tabela, Kasarama, and Ebini soils through its interaction with the organic fraction. The sesquioxide content of the Tiwiwid sand was very much lower than that of the three brown sand soils, and most of the hydrous Al was present in an exchangeable form. The soils were very poorly supplied with plant nutrients, and though the total contents of the major nutrient cations were 19^

PAGE 210

195 unexpectedly high, very small proportions of these were in exchangeable forms. The soils can be put into groups on the basis of their level of exchangeable nutrients and sesauioxides. The Tiwiwid sand with its high nutrient supply and lower sesquioxide content constitutes one group while the Tabela, Kasarama, and Ebini soils comprise the other group. The soils of the latter group were also different from the Tiwiwid sand in that their organic fractions were dominated by fulvic acids rather than humic acids, and their pH levels in the natural state suggested that Al-fulvate complexes provided the main source of buffering. The Tiwiwid sand had a substantial portion of its P content in easily extractable forms, while the brown sand soils had < 2ppm P extractable by the Bray #1 solution. A common characteristic among the soils was the source of the soil CSC. More than 9C$ of the GEC in each of the soils was contributed by the organic fraction, almost exclusively by the humic acids. This entailed a similarity in the pattern of retention of cations in the soils, such that Ca was preferentially bonded, while K and Mg tended to remain in solution. Thus, the ease of loss of nutrients by leaching was a property of the soils derived from the essentially organic nature of the exchange complex. Modification of the pH of the soils by the application of lime decreased the amounts of K and Mg retained in exchangeable form, while the addition of P seemed to increase the amounts of exchangeable Ga. Added P remained almost completely extractable from the Tiwiwid sand, but < ?0^ was recoverable from the brown sand soils. In the brown sand soils it seemed that

PAGE 211

196 P was adsorbed on Al-fulvate complexes, possibly displacing fulvates at higher levels of addition. Increasing soil pH enhanced the replacement of fulvate by 0H~ and/or P, and increased the amounts of P fixed by the soils. Plant growth on the soils reflected the broad division of the soils into the groups. Higher yields were produced on the Tiwiwid sand with higher plant P concentration, and the nutrient most likely to limit growth on this soil seemed to be N. On the brown sand soils, plant growth responded markedly to additions of P. Micronutrients were also in limited supply, but it will be necessary to study the effects of the individual micronutrients. Nodulation of pigeon pea was greatly improved by the addition of P and micronutrients to the soils. Both the size of the nodules and the development of hemoglobin within them were improved by the addition of these nutrients. Positive responses to lime are to be expected on the Tiwiwid sand, but on the brown sand soils, higher yields are more likely at lower pH levels. This is particularly true of pangolagrass . On the latter soils, P fixation restricted plant growth even though the plants developed very healthy appearances. The optimum pH level for maximum yields is likely to vary with the type of crop, and with its modification of the conditions in the rhizosphere. Improvement of the fertility levels of the brown sand soils must achieve an increase in the capacity of the soils to retain cationic nutrients against leaching and to supply adequate levels of P. This will probably be best effected by the use of P fertilizers,

PAGE 212

197 e.g., superphosphate, which ultimately will satisfy the P fixation capacities of the soils, and promote the build up of organic matter. The new store of organic matter is likely to be less fulvate in character, and cultural practices must be designed to accomodate such changes. The use of lime will need to be carefully controlled in order to minimizs the possibilities of the soils becoming entirely inorganic in composition. The following conclusions appear to be justified from the results of this study i (a) Kaolinite, the only clay mineral present in these soils exhibited various degrees of degradation and made little or no contribution to the chemical properties of the soils. (b) The absence of gibbsite and the physj cal condition of the clay crystals suggested that much of the amorphous Al in the soils may have been derived from the destruction of the clay crystals. (c) The organic fractions of the soils varied in chemical composition, with those of the three brown sand soils being predominantly fulvic in character, and that of the Tiwiwid, predominantly humic. (d) Complexes of organic matter and Al, principally Al-fulvates, appeared to provide the buffering mechanism of the brown sand soils, while the Tiwiwid soil seemed to be buffered by the organic matter, per se. (e) The nutrient status of each of the soils was very low, and mineral sources of nutrients were virtually absent.

PAGE 213

198 (f) The CEG of the soils was due to the organic fraction, apparently to the huraic acids in particular. The exchangeable cations constituted a very small portion of the total amounts of cations in the soil, probably because of the penetration of the cations into the gel-like organic matter. (g) The organic nature of the cation exchange sites favored the preferential bonding of Ca and the exclusion to a great degree of K and Mg. This effect was more marked at higher pH levels, and increased the possibility of loss of K and Mg by leaching. (h) The addition of P increased the C3C of the soils. Virtually all of the P added to the Tiwiwid soil was extractable, but very little of that added to the brown sand soils. (i) Plant growth was much better on the Tiwiwid sand than on the brown sand soils at comparable nutrient levels when the moisture regime was controlled. (j) The improvement of fertility levels of the Tiwiwid sand can be effected by increasing the pH of the soil, and supplying adequate levels of all nutrients in plant available, but not readily leachable, forms. (k) The fertility levels of the brown sand soils can be improved by the addition of adequate levels of all nutrients as in the Tiwiwid sand, but the pH levels of the soils need to be carefully controlled in relation to the crop to be grown and the capacity of the soils to fix P. (1) The use of P as the principal agent of improvement of the

PAGE 214

199 chemical properties of the brown sand soils seems to hold the most promise of prolonged utilization. (m) Commercial utilization of the brown sand soils will depend almost entirely on the technology available in relation to soil conservation and moisture control.

PAGE 215


PAGE 216

201 Table 53. —Shoot weights (oven -dried) from first harvest of pangolagrass. Treatment

PAGE 217

202 Table 54. — Shoot weights (oven-dried) from second harvest of pangolagrass . Treatment

PAGE 218

203 inH H O (N! ir,or\coo4\o O VO i>--V CO COHO ir\ Q vo o o\-d* 4 or\rico cnvo O O H !>-H CCV! OWO C\! vrv C^ COVOOCA''Ar\ 0\0 0\NO CM H U^vO O H VO II II

PAGE 219

204 UN, r-i OJ C^ m Cvl •^ i-^ C^v ,H H CN00 CV Vn, CN. Cnn£> O O O) M ri in C\ O ON C\! On CM O n} ?-l M O ft W M B H En

PAGE 220


PAGE 221

206 INPMA H CN -d" O O rH OON tr\ O -d" 03 COOW C'"\ 0^ O O O O CD H O O O >A Cv> H CO N \i) 4Cn C~\ £>^NVOON CM U A O C^v C\i rH co 0,0^ co c<^ >o» CM rH oUN. CM c-n wovoonw o o -3O r-l o C-\ CO 0"\ CO CO INi ; -cj o r-l O l>C^ O NO OOt^l O O H O O C\} o o o nilO f-i £ o iz; CO Pi ^! E-i W h HHHrlHVOrl C\! C~\ S$ CO f4 M EH W En

PAGE 222

LITERATURE CITED 1. Abruna-Rodrigues, F,, J. Vincente-Chandler, R. W, Pearson, and S. Silva. 19?0. Crop response to soil acidity factors in UltisolE and Oxisols. 1. Tobacco. Soil Sci. Soc. Amer. Proc. 34:629-635. 2. Acquaye, D. K„, and J, Tinsley. 1965° Soluble silica in soils. InE. C. Hallsworth, and D. V. Crawford (ed.) Experimental pedology. Proc. 11th Easter School in Agr. Sci., Univ. Nottingham. Butterworths, London, p. 126. 3. Ahlgreen, H. L. 1948. Grass. Yearbook of Agriculture. U. S. Dept. Agriculture, p. 425. 4. Ahmad, N i960. Aluminum toxicity of certain soils on the coast of British Guiana and problems of their agricultural utilization. Int. Congr. Soil Sci. Trans. ?th. 11:161-170. L. I. Tulloch-Reid, and C. E. Davis. 1969. Studies on pangola grass (Digitari a decumbens Stent) in Trini lad 1, Description of the experiments and effects of nitrogen. Trop, Agr. Trin. 1*6:173-178. . and . 1969. Fertilizer studies on pangola grass in Trinidad II. Effects of phosphorus, potassium, and magnesium. Trop. Agr., Trin. 46:179-186. Al -Abbas, A. H., and S. A. Barber. 1964. A soil test for phosphorus based upon fractionation of soil phosphorus: 1. Correlation of soil phosphorus fractions with plant -available phosphorus. Soil Sci. Soc. Amer. Proc. 28:218-221. Alexandrova, L. N. i960. On the composition of humus substances and the nature of organo-mineral colloids in soil. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 11:74-81. 1966. Possible ways of fulvic-acid formation in soil. Int. Soc. Soil Sci. Trans. (Aberdeen, Scot.) Soil chemistry and fertility, p. 73-77. 10. , T. Th. Arshavskay, E. M. Dorfman, M. F. Lyuzin, and 0. V. Yurlova. 1968. Humus acids and their organic-mineral derivatives in soil. Int. Congr. Soil Sci. Trans. 9th. (Adelaide, Aust.) 111:143-152. 207

PAGE 223

208 11. Allison, L. E. 1965. Organic Carbon. L (ed.) Methods of soil analysis. Part II. Amer. Soc. A. Madison, Wis. p. 1367. 12. Andrew, C, S. I962. Influence of nutrition on nitrogen fixation and growth of legumes. In A review of nitrogen in the tropics with particular reference to pastures. Bulletin 46* Coram. Agric. Bureaux. Bucks., England, p, 130. 13. _> and D. 8 Norris. 1961. Com; to calcium of five tropical and four temper; species. Aust. J. Agr. Res, 12:40-55. 14. Andriesse, J. P. 1969. A study of the environment cteristics of tropical podzols in Sara. Malaysia). Gec : 2 8 201-227. 15. Ayres, A. S., and H. H. Hagihara. 1953-. : 1 I of the anion on the sorption of potassium by some humic and hydrol .hui i : latosols. Soil Sci. 75 J 1-17. 16. ___, __, and G. Stanford. 1965. Significance of extractable alurdnum in Hawaiian su: ils. Soil Sci. Soc. Amer. Proc. 29037-39217. Bache, B. W. 1954. Aluminum and iron phosphate studies relating to soils. II. Reactions between phosphate and hydrous oxides. J. Soil Sci. 15:110-116. 18. Ball, P. F. 196^. Loss on ignition as an estimate of organic carbon in non-calcareous soils. J. Soil Sci. 15*84-92. 19. Barber, S. A. 1962. A diffusion and mass-flow concept of soil nutrient availability. Soil Sci. 93: 39-49. 20. _, and P. C. Ozanne. 1970. Autoradiographic evidence for the differential effect of four plant species in altering the calcium content of the rhizosphere soil. Soil Sci, Soc. Amer. Proc. 34:635-63?. 21. Barrow, N. J. 1969. The accumulation of organic matter under pasture and its effects on soil properties. Aust. J, Exp. Agr. and Animal Husb. 9:437-444. 22. Barshad, I. 1965. Thermal analysis techniques for mineral identification and mineralogical composition. In C. A. Blade (ed.) Methods of soil analysis. Part II. Amer. Soc. Agron. Madison, Wis. p. 699. 23. Bartlett, R. T., and J. L. Mcintosh. 1969. pH-dependent bonding of potassium by a spodosol. Soil Sci. Soc. Amer. Proc. 33»536-539<

PAGE 224

209 24. Bar-Yosef, B., V. Kafkafi, and N. Lahav. 1969. Relationships among adsorbed phosphate, silica, and hydroxyl during drying and revretting of kaolinite suspension, Soil Sci, Soc. Amer. Proc. 33 J 672-677* 25. Beckwith, R. S., and R. Reeve. 1963. Studies on soluble silica 'oils. 1 8 tion of silicic acid by soils and minerals« Aust. J. Soil Res. 1:157-168. 26c J, 1963» The red and yellow soils of the tropics and sub-tropical uplands. Soil Sci. 95*250-257. jrg, W, A«, and G. W. Thomas. 1959. Anion elution patterns . soils and clays. Soil clays. Soil Sci. Soc. Amer. Proc* 23:3^3-350. 2.8. Bhandari, G. S., M. S. Kaskina, and N. S. Randhawa. 1970. A dy of humic and fulvic fractions of some soils of different -climatic regions of the erstwhile Punjab. Geoderma 3: 239— . D, R., and E. 0, McLean. 1965. Aluminum in soils, VI. ( -'''-: acidity, cation exchange capacity, and le aluminum with additions of lime to acid surface soils, ., Soc. Amer. Proc. 29; 370-374. rch, H. F. 1951. Relationship between base saturation and crop response to phosphate in acid soils. Nature, London ] ' Is 388-389. 31. Bleackley, D., and E. T. A. Khan. 1959. A note on the soils of white sand formation of British Guiana. Proc. 5th Interna Geological Conference. Georgetown, British Guiana. 32. , and . 1963. Observations on the be sand areas of the Berbice formation, British Guiana. J. Soil Sci. 14:44-51. 33* Bloc-field, C. 1957. The possible significance of polyphenols in coil formation. J. Sci. Food Agr. 8:389-392. 34, Blue, Vf. G,, L. Andrade, E. Rey, M. Ramirez, L. L. Larson, and VI. E. Schaefer. 1963. Investigations of the potential for pasture development in the Atlantic zone of Costa Rica. Soil Crop Sci. Soc. Fla. Proc. 23:208-221. 35. Blue, W. G. 1969. Fertilizer response with pangolagrass on Puletan loamy sand, British Honduras, as indicated by pot experiments. Trop Agr. Trin. 46:25-29. 35. Bohn, H. L. 1967. The (Fe)(0H) 3 ion product in suspensions of acid soils. Soil Sci. Soc. Amer. Proc. 31x641-644.

PAGE 225

210 3?. Boserup, E. 1965. The conditions of agricultural growth. Aldine Publishing Company, Chicago. 38. Box, G. E. P. 1956, The determination of optimum conditions. In 0, L. Davis (ed) Design and analysis of industrial experiments. 2nd ed. Hafner Publishing Co. Inc., New York. 39. , and K. 3. Wilson. 1951* On the experimental attainment of optimum conditions. J. Royal Statistical Soc. r. B. 13? 1-45. 40 , and P. V. Youle. 1955* The exploration and exploitation of response surfaces. Biometrics lis 287-322. iH. Bradfield, R., and VJ 6 H e Allison. 1933. Criterion of base saturation in soils. Int. Soc. Soil Sci. Trans Comm. II. A : 63-7942. . _, G« Scarseth, and J. C. Steele. 1935. Factors ting the retention of phosphate by clays. Int. Congr. Soil Sci. Trans. 3rd (Oxford, England) It 74-75. 43. Brey, tf. S., Jr. 1958. Principles of physical chemistry. Applet on-Century -Crofts, Inc., New York. 44. Briggs, L. T», and J. W. McLane. 1907. The moisture equivalent of soils. U. S. Dept. Agr. Bur. Soils Bull. 45. 45. Brinkraan, R. 1964. Soil survey of the Ebini-Ituni-Kwakwani area. Report on the soil survey project, British Guiana. Vol. VI. F.A.O. Rome. 46. Broadbent, F. E. 1953. The soil organic fraction. Adv. Agron. 5:153-183. **?• » and G » R » Bradford. ' 1952. Cation exchange groupings" in the soil organic fraction. Soil Sci. 7^:^7-^57. 48. , and J. B. Ott. 1957. Soil organic mattermetal complexes. 1. Factors affecting retention of various cations. Soil Sci. 83s 419-427. 49. Bromfield, S. M. 1965. Studies on the relative importance of iron and aluminum in the sorption of phosphate by some Australian soils. Aust. J. Soil Res. 3t 31-44. 50. Bryan, W. W. 1962. The role of the legume in legume/grass pastures. In A review of nitrogen in the tropics with particular reference to pastures. Comm. Agr. Bureaux. 3ucks., England.

PAGE 226

211 51 «, Burgos, C. F. 1962. Studies on exchangeable aluminum in three tropical acid soils, M. S. Thesis. University of Florida. 52. Gate, R. B., Jr. i960. Studies on gibbsite accumulations in soils. M. S. Thesis. N. C. State College, Raleigh. 53. Carlson, R. M., and R. Overstreet. 1967. A study of the ion iour of the alkaline earth metals. Soil Sci. 1035 213-213. 54. Chang, S. C, and M. L. Jackson. 1957* Fractionation of soil phosphorus. Soil Sci. 84:113-144. 55 • . and ___. 1958* Soil phosphorus fractions iome representative soils, J. Soil Sci. 9 » 109-119. 56 . _ __„» and W. K. Chu. 1961. The fate of soluble phosphate applied to soils. J. Soil Sci. 12:286-293. 57. Chapman, H. D. 1965. Cation exchange capacity. In C. A» Black (ed.) Methods of soil analysis. Part 2. Amer. Soc. Agron. Madison, His. p. 891. 58. Clark, J e S., and R. G. Hill. 1964. The pH-percent base ration relationships in soils. Soil Sci. Soc. Amer. Proc. 28:490-492. 59. Clarkson, D. T. 1966. Effect of aluminum on the uptake and ; itabolism of phosphorus by barley seedlings. Plant Physiol. 41 sl65-172. 60 « . 196?. Interactions between aluminum and phosphorus on root surfaces and cell wall material. Plant and Soil 27:347-356. 61. Coleman, R. P. 1945. The mechanism of phosphate fixation by montmorillonite and kaolinite clays. Soil Sci. Soc. Amer. Proc. 9:72-78. 62. Coleman, N. T., S. B. Weed, and R. J. McCracken. 1959. Cation exchange capacity and exchangeable cations in Piedmont soils of North Carolina. Soil Sci. Soc. Amer. Proc. 23:146-149. 63. . J. T. Thorup, a.nd W. A. Jackson, i960. Phosphate sorption reactions that involve exchangeable aluminum. Soil Sci. 90:1-7. 6 '+« • G » w » Thomas, F. H. LeRoux, and G. Bredell. 1964. Salt-exchangeable and titratable acidity in bentonite-sesquioxide mixtures. Soil Sci. Soc. Amer. Proc. 28:35-37.

PAGE 227

212 65 . Coleman, N. T., and G. W. The . 1964. Buffer curves of acid clays as affected by the presence of ferric iron and aluminum. Soil Sci. Soc. Amer. Proc. 28:187-190. 66. , and __ . 1967. The basic chemistry of soil"acidity. In R. W. Pearson and F. Adams (ed.) Soil acidity and liming. Amer. Soc. Agron., Madison, Wis. p. 1. 67. Colwell, J. Do 1959. Fhosphon by iron and alumin oxides. Aust, J. Appl. Sci. 10:95-103. 68. Correns, C. W. 1969. Introduction to mineralogy. (Translated by W. D, Johns). Springer-Ve York, Inc. 69. Coulson, G. B., R. I. Davies, and D. A. Lewis, i960. Polyp! in plant, humus and soil. 1, tiols of leaves, lit: and superficial humus from mull and mor sites. J. Soil Sci. 113 20-29, 70. Dalton, J. D., G. C. Russell, and D. H* Sielig. 1952. Effect of organic matter on phosphate availability. Soil Sci. 73:173-181. 71. Davies, R, I., C. B. Coulson, and D. A. Lewis. 1964. Polyphenols in plant, humus and soil. IV. Pac to increase in biosynthesis of polyphenols in leaves and their relationship to mull and mor formation. J. Soil Sci. 15:310-317. 72. Day, P. R. 1965. Particle fractionation and ] -size analysis. In C. A. Black (ed.) Methods of soil analys Part 1. Amer. Soc. Agron., Madison, Wis. p* 545* 73. Deb, D. L., and N. P. Datta. 1967. Effects of associating anions on phosphorus retention in soil. 1. Under variable phosphorus concentration. Plant and Soil 26:303-316. 74. __, and . 1967. Effect of associating anions on phosphorus retention in soils. II. Under variable anion concentration. Plant and Soil 26:432-444. 75. DeLong, W. A., and M. Schnitzer. 1955* Investigations on the mobilisation and transport of iron in forested soils. 1. The capacities of leaf extracts and leachates to react with iron. Soil Sci. Soc. Amer. Proc. 19O60-363. 76. deEndredy, A. S., and K. A. Quagraine. I960. A comprehensive study of cation exchange in tropical soils. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 11:312-320. 77. DeKock, P. C. i960. The effects of natural and synthetic chelating substances on the mineral status of plants. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 111:573-579.

PAGE 228

213 78. DeKock, P. C„ and H. L. Mitchell, 1957. Uptake of chelated metals by plants. Soil Sci. 84:55-62. 79 o Dillon, J. L. 1968r The analysis of response in crop and livestock production. 1st ed. Pergamon Press, New York, 80 e Dobereiner, J. 1966. Manganese toxicity effects on nodulation and nitrogen fixation of beans (Phaseolus vulgaris) on acid soils. Plant and Soil 24:155-166. 81. Donald, C. M., and C. H. Williams. 1954. Fertility and productivity of podzolic soil as influenced by subterranean clover ' tbterranean L) and superphosphate. Aust. J. Agr. Res. 5 1 6 .6-4-687. 82. Dmitrenko, P. A., and N. Ye. Shestidesyatnaya. 1962. Effect of liming on the fertility of the soils of Transcarpathia. Soviet Sci. 1142-1147. 83. D'Yakonova, K. V. 19o2. Iron-humus complexes and their role in plant nutrition. Soviet Soil Sci. 692-698. 84. Eck, P., M. Drake, and J. E. Steckel. 1957. Calcium uptake by tomatoes as influenced by percent Ca-saturation of Ca-H clay systems. Soil Sci. 84:145-154. 85. Ellison, W. 1958, The role of legumes in farm ecology. In E. G. Hallsvrorth (ed.) Nutrition of the legumes. Butterworths, London, p. 308. 86. Eriksson, E. 1952. The physico-chemical behaviour of nutrients in soils. J. Soil Sci. Ji 238-250. 87. Evans, C. E., and E. J. Kamprath. 1970. Lime response as related to percent Al saturation, solution Al, and organic matter content. Soil Sci. Soc. Amer. Proc. 34:893-896. 88. Evans, L. T., and E. W. Russell. 1959. Adsorption of huuic and fulvic acids by clays. J. Soil Sci. 10:119-132. 89. Eyre, S. R. 1968. Vegetation and soils. 2nd ed. Aldine Publishing Co., Chicago. 90. Felbeck, G. T., Jr. 1965. Structural chemistry of soil humic substances. Adv. Agron. 17:327-368. 91. Fiske, C. H., and Y. Subbarow. 1925. The colorimetric determination of phosphorus. J. Biol. Chem. 66:375-400. 92. Follett, E. A. C. 1965. The retention of amorphous colloidal ferric hydroxide by kaolinites. J. Soil Sci. 16:334-341.

PAGE 229

234 93 * Fox, R. L., S. R e Datta, and D. Sherman. 1962. Phosphorus solubility and availability to plants and the aluminum status of Hawaiian soils as influenced by liming. Int. Soc. Soil Sci. Trans. Comm. IV and V (Palmers ton N., New Zealand), P. 571 505. 94. , S. K. Dedatta, and J. M. Wang. 1964. Phosphorus and aluminum uptake by plants from latosols in relation to liming. Int. Congr. Soil Sci. Trans. 8th. (Bucharest, Romania). iv 5595-603. 95. _, J. A. Silva, 0. R. Younge, D L. Plucknett, and G. D. Sherman. 1967. Soil and plant silicon and silicate response by sugarcane. Soil Sci. Soc. Amer. Proc. 31*775-779« 96. __ , and E. J. Kamprath. , 1971. Adsorption and leaching of P in acid organic soils and high organic matter sand. Soil Sci. Soc. Amer. Proc. 35 s 154-156. 97. Foy, C. D., and J. Co Brown. 1964. Toxic factors in acid soils. II. Differential aluminum tolerance of plant species. Soil Sci. Soc. Amer. Proc. 28:27-32. 98* . V. H. Armiger, L. W. Briggle, and D. A. R Aluminum tolerance of wheat and barley varieties. Agrc 1. J. 57:413-417. 99, , A L. Fleming, and W. H. Armiger. 1969. Aluminum tolerance of soybean varieties in relation to calcium nutrition. Agron. J. 61.505-5H. 100. Fried, M., and M. Feech. 194-6. The comparative effects of lime and gypsum upon plants grown on acid soils. J. Amer. Soc. Agron. 38s 61^-623. 101. Fripiat, J. J. 1965. Surface chemistry and soil science. In Hallsworth and Crawford (ed.) Experimental pedology. 11th Easter School, Nottingham. Butterworths, London, p. 3. 102. Frink, C. R. 1965. Characterization of aluminum interlayers in soil clays. Soil Sci. Soc. Amer. Proc. 29:379-382. 103. Gammon, N., Jr. 1953. Sodium and potassium requirements of pangola and other pasture grasses. Soil Sci. 76:81-90. 104. . 1957. The behavior of potassium and potassium fertilizers in Florida soils. Soil and Crop Sci. Soc. Fla. Proc. 17:156-160. 105. Gerretsen, F. C. 1948. The influence of microorganisms on the phosphate intake by the plant. Plant and Soil 1:51-81.

PAGE 230

215 106. Gillam, W. S. 1940. A study of the chemical nature of humic acid. Soil Sci. 49:433' 107. Gilmour, J. T., and tf. T. Coleman, 1971. Buffer curves and acid | nd two resins. Soil Sci. Soc. Amer. Proc. 35.710-714. 108. Graham, E. R., and R. L. Fore. 1971. Tropical soil potassium as related to labile pool and calcium exchange equilibrium. Soil Sci. 111:313-322. 109. Greenland, D, J. 1971. Interactions between humic and fulvic acids and clays. Soil Sci. Ills 3*1-41. 110. Greenland, D. J., and J. H. Oades. 1968. Iron hydroxides and clay surfaces. Int. Congr. Soil Sci. Trans. 9th. ( Aust.) 1:657-668. 111c Griffin, G. F, 1971. Effect of liming on the soil test level of phosphorus as determined by three methods. Soil Sci. Soc. Amer, Proc. 35 « 540-542. 112. Grim, R 6 B. 1963. Clay mineralogy. 2nd ed. McGraw-Hill, New York. 113. Gunary, P., and C. D, Sutton. 196?. Soil factors affecting plant uptake of phosphate, J. Soil Sci. 18:167-173* 114. Hackett, C 1964. Ecological aspects of the nutrition of (pes-' Lexuosa (l) Trin.) 1. The effect of aluminum, pH on germination, J. Ecol, 52:159-167* 115. Hader, R. J., M. E. Harward, D. D, Mason, and P. P. Moore. 1957. An investigation of some of the relationships between copper, iron and molybdenum in the growth and nutrition of lettuce. 1. Experimental design and statistical methods. Soil Sci. Soc. Amer. Proc. 21:59-64, 116. Hagen, C. E., and K. T. Hopkins. 1955. Ionic species in orthophosphate absorption by barley roots. Plant Physiol. 30:193-199. 117. Hallsworth, E. G. 1953. Nutritional factors affecting nodulation. In E, C. Hallsworth (ed.) Nutrition of the legumes. Butterworths, London, p. 183, 118. Halstead, R. L. 1967. Chemical availability of native and applied phosphorus in soils and their textural fractions. Soil Sci. Soc. Amer. Proc. 31*414-419.

PAGE 231

216 119. Hardy, P., and R. R. Follett-Smith. 1931. Studies in tropical soils. II. Some characteristic igneous rock soil profiles in British Guiana, South America. J. Agr. Sci. 21:739-761. 120. Helling, C. S., G. Chesters, and Re B. Corey. 196*1-. Contributions of organic matter and clay to soil cation -ex change capacity as affected by the pH of the saturating solution. Soil Sci. Soc. Amer. Proc. 28s 317-320. 121. Henzell, E. F. 1967. Tropical pasture legumes in Northern Australia. Soil and Crop Sci. Soc. Fla. Proc. 27:322-338. 122. Hesse, P. R. 196l. Phosphorus fixation in mangrove swamp muds. Nature, London. 193*295-296. 123. Kingston, F. J., R. J. Atkinson, A. M. Posner, and J. P. Quirk. 1963. Specific adsorption of anions on goethite. Int. Soil Sci. Soc". Trans. 9th. (Adelaide, Aust. ) 1:669-678. 124. Hodges, E. M., D. W. Jones, and W. G. Kirk. 1951 • Grass pastures in central Florida. Agr. Expt. Sta., University of Florida, Gainesville. Bull. 484. 125. , W. G. Kirk, F. H. Peacock, D. W. Jones, G. K. Davis, and J. R. Neller. 1964. Forage and animal response to different phosphatic fertilizers on pangolagrass pastures. Agr. Expt. Sta., University of Florida, Gainesville. Bull. 686. 126. ___, G. B. Killinger, J. E. HcCaleb, 0. C. Ruelke, R. J. Allen, Jr., S. C. Schank, and A. E. Kretschner, Jr. 1967. Pangolagrass. Agr. Exp. Sta. I. F. A. S., University of Florida, Gainesville. Bull. 718. 127. Hortenstine, C. C, and W. G. Blue. 1968. Growth responses in three plant species to lime and phosphorus applied to Puletan loamy fine sand. Soil and Crop Sci. Soc. Fla. Proc. 28:23-28. 128. Hoyt, P. B., and M. Nyborg. 1971. Toxic metals in acid soil. 1. Estimation of plant -available aluminum. Soil Sci. Soc. Amer. Proc. 35 » 236-240. 129. Hsu, P. H. 1954. Adsorption of phosphate by aluminum and iron in soils. Soil Sci. Soc. Amer. Proc. 28:474-478. 3-30. . 1965. Fixation of phosphate by aluminum and iron in acidic soils. Soil Sci. 99s393-402. 131. , and D. A. Rennie. 1962. Reactions of phosphate in aluminum systems. 1. Adsorption of phosphate by X-ray amorphous "aluminum hydroxide." Can. J. Soil Sci. 42:197-209.

PAGE 232

217 132. Hsu, P. H., and D. A. Ronnie. 1962. Reactions of phosphate in aluminum systems. II. Precipitation of phosphate by exch able aluminum on a cation exch Ln. Can. J. Soil Sci. 42 j 2 10 -221. 133. Huffman, 3. 0., W, E. Cate, M. E. I d K. L. Elmore. i960. Rates and mechanisms of dissolution of some iron and aluminum phosphates. Int. Congr. Soil Sci, Trans. 7th, (Madison, Wis.) II 1 404-412. 134. Hunsaker, V. E,, and P. F. Pratt. 1973 exchange eauilibria in soils. Soil Sci. Soc. Amer. Proc. 35:151-152. 135. Hutchinson, F. E., and A. S. Hunter. 1970, Exchangeable aluminum levels in tvro soils as related to 1 ment and growth of six crop species. Agron. J. 62:702-704. 136. Ignatieff, V., and P. Lemos. 1963. So more important tropical soils. Soil Sci. 95*243137. Jackson, M. L. 1958. Soil chemical analysis. XI H ill, Inc., Englewood Cliffs, N. J. 138. „____» 1963. Aluminum bonding in soils: ing principle in soil science. Soil Sci. Soc. / , tl-10. 139. Jackson, P. C. f S. 3. Hendricks, and B. M. Vasta. 1962. Phosphorylation by barley root mitochondria and phosphate absorption by barley roots. Plant Physiol. 37*8-17. 140. Joffe, J. S. 1935. Behaviour of replaceable cations in soil and their availability. Int. Congr. Soil Sci. Trans. 3rd. (Oxford, England). 1 166-67. 141. Johnson, R. E., and W. A. Jackson. 1964. Calcium uptake and transport by v.-heat seedlings as affected by aluminum. Soil Sci. Soc. Amer. Proc. 28 1 381-386. 142. Jones, L. H, 1961. Aluminum uptake and toxicity in plants. Plant and Soil 13:297~310. 143. Jones, E. 1968. Nutrient cycle and soil fertility on red ferrallitic soils. Int. Congr. Soil Sci. Trans. 9th (Adelaide, Aust.) 111*419-427. 144. Jones, L. H. P., and K. A. Handreck. 1967. Silica in soils, plants, and animals. Adv. Agron. 19:107-149. 145. Juo, A. S. R., and B. G. Ellis. 1968. Chemical and physical properties of iron and aluminum phosphates and their relation to phosphorus availability. Soil Sci. Soc. Amer. Proc. 32:216221.

PAGE 233

218 1U6. Kafkaf5, V., A. M. Posner, and J. P. Quirk. 196?. Desorption of phosphate from kaolinite. Soil Sci. Soc. Amer. Proc. 31:3^8-353. 1*1-7. Karaprath, E. J. 1970. Exchangeable aluminum as a criterion for liming leached mineral soils. Soil Sci. Soc. Amer. Proc, 34:252-25^. 148. Kardos, L. T. 196^. Soil fixation of plant nutrients. In F. E. Bear (ed.) Chemistry of the soil. 2nd ode Reinhold Publishing Corp., New York. p. 369* Hi9. Keller, W. D. 196^. The origin of high aluminum clay minerals. A review. Clays and clay minerals. Proc. 12th Nat*l. Conf . on clays and clay minerals p. 129-151. 150. Kemper, W. D. 1965. Aggregate stability. In C. A. Black (ed.) Methods of soil analysis. Part 1. Amer. Soc. Agron. Madison, Wis. p, 5H« 151. Khanna, S. S., and F. J. Stevenson. 1962. Organo-metallic complexes in soil. 1. Potent iometric titrations soil organic matter isolates in the presence of transition metals. Soil Sci. 93:298-305. 152. Kin jo, T., and P. F. Pratt. 1971. Nitrate adsorption. 1. Some acid soils of Mexico and. South America. Soil Sci. Soc. Amer. Proc. 35*722-725. 153. , and ... . 1971. Nitrate adsorption II. In competition with chloride, sulfate, and phosphate. Soil Sci. Soc. Amer. Proc. 35:725-728. 154. Killinger, G. B. 1968. Pigeonpeas (Cajanus cajan (L) Bruce). A useful crop for Florida. Soil and Crop Sci. Soc. Fla. Proc. 28s 162-167. 155. Kittrick, J. A. 1965. Electron microscope techniques. In C. A. Black (ed.) Methods of soil analysis. Part 1. Amer. Soc. Agron. Madison, Wis. p. 568. 156. . and M. L. Jackson. 1956. Electron -micros cope observations of the reaction of phosphate with minerals, leading to a unified theory of phosphate fixation in soils. J. Soil Sci. 7:81-89. 15?. Klinge, H. 1965. Podzol soils in the Amazon basin. J. Soil Sci. 16:95-103.

PAGE 234

219 158. Koch, J e T., E. R. Orchard, and M. E. Sumner. 1970. Leaf composition and yield response of corn in relation to quantityintensity parameters for potassium. Soil Sci. Soc. Araer. Proc. 34:94-98. 159. Kononova, M. H. 1966. Soil organic matter. 2nd English ed. Pergamon Press, New York. 160. Krauss, F. G. 1932. The pigeon pea (Cajanus indicus), its improvement, culture, and utilization in Hawaii. Hawaii Agr. Exp. Sta. Bull. 64. 161. Kunze, G. W. 1965. Pretreatment for raineralogical analysis. In C. A. Black (ed.) Methods of soil analysis. Part 1. Amer. Soc. Agr on., Madison, Wis. P, 568. 162. Lance, J. 0., and R. W. Pearson. 1969. Effect of low concentrations of aluminum on growth and water and nutrient uptake by cotton roots. Soil Sci Soc. Amer. Proc. 33*95-98. 163. Leaver, J. P., and E. W. Russell. 1957 « The reaction between phosphate and phosphate-fixing soils. J. Soil Sci. 8:13-126. 164. Lehr, J. J., and J. Ch. van Wesemael. 1952. The influence of neutral salts on the solubility of soil phosphate; with special reference to the effects of nitrates of sodium and calcium. J. Soil Sci. 3*125-135. 165. , and . 1956. Variations in the uptake by plants of soil phosphate as influenced by sodium nitrate and calcium nitrate. J. Soil Sci. 7 » 148-155. 166. LeMare, P. H. i960. Observations of the phosphate potential of some tropical soils. Int. Congr. Soil Sci. Trans, 7th. (Madison, Wis.) 111:600-603. 167. Levesque, M., and M. Schnitzer. 1966. Effects of NaOH concentration on the extraction of the organic matter and of major inorganic constituents from a soil. Can. J. Soil Sci. 46:7-13. 168. t and _____• 1967. Organo-metallic interactions in soils. Z~. Preparation and properties of fulvic acid-metal phosphates. Soil Sci. 103:183-190. 169. Lewis, T. E., and F. E. Broadbent. 196l. Soil organic mattermetal complexes. 4. Nature and properties of exchange sites. Soil Sci. 91:393-399. 170. Lewis, D. G., and J. P. Quirk. 1967. Phosphate diffusion in soil and uptake by plants. 1. Self diffusion of phosphate in soils. Plant and Soil 26:99-118.

PAGE 235

220 171 Lewis, D. G., and J. P. Quirk. 1967. Phosphate diffusion in 7 soil and uptake by plants. II. Phosphate uptake by wheat plants. Plant and Soil 26:119-128. 172 and ^ . 1967. Phosphate diffusion in ' soil and uptake "plants. III. &P movement and uptake by plants as indicated by 32p autoradiography. Plant and boil 26:445-^53. 173. Lin, C„ and K. T. Coleman, i960. The measurement of exchangeable aluminum in soils and clays. Soil Sci. Soc. Amer. Proc. 24:444-446. 174. Lindsay, W. L., and A. W. Taylor, i960. Phosphate reaction products in soil and their availability to plants. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 111:580-589. 176. Unp Ong, H., and R. B. Bisque. 1968. Coagulation of humic colloids by metal ions. Soil Sci. 106:220-224. 176 Low, P. F., and C 9 A. Black. 1943. Phosphate induced decomposition of kaolinite. Soil Sci. Soc. Araer. Proc. (19'+7). 12*180-184. 177 t an d _. 1950. Reactions of phosphate with kaolinite. Soil Sci. 70:2?3~290. 178. Lucas, L. N. 1969. Phosphorus availability in alluvial rainforest soils from Costa Rica as affected by phosphorus fertilization and soil amendments. Fh« D. Dissert., University of Florida. 179« Mahilum, B. C, R. L. Fox, and J. A. Silva. 1970. Residual effects of liming volcanic ash soils in the humid tropics. Soil Sci. 109:102-109. 180. Malcolm, R. L., and R. J. McCracken. 1968. Canopy drxp: A source of mobile soil organic matter for mobilization ox iron and aluminum. Soil Sci. Soc. Amer. Proc. 32:834-838. 181. Martin, A. E., and R. Reeve. 1957. Chemical studies of podzolic illuvial horizons. II. The use of acetylacetone as an extractant of translocated organic matter. J. Soil Sci. 8:277-286. l82# , and . 1958. Chemical studies of podzolic illuvial horizons. III. Titration curves of organic matter suspensions. J. Soil Sci. 9:89-100. _33 # t and . I960. Chemical studies of podzolic illuvial horizons. IV. The flocculation of humus by aluminum. J. Soil Sci. 11:369-381.

PAGE 236

221 18*4-. Martin, A. E. t and R. Reeve. 19 60. Chemical studies of podzolic illuvial horizons. V. Flocculation of humus byferric and ferrous iron and by nickel. J. Soil Sci. 11:302-393, 105. lias on, B., and L. G. Berry, 1968, Elements of mineralogy. W. H, Freeman and Co., San Francisco, 186, Mayers, J. M, 1969. An econ of the Ebini Livestock Station, Guyana. Dapt. Agr, Scon, and Farm Management, U. W. I,, Trinidad. 18?. McGeorge, W, T, 19; ic compounds associated with base exchange reactions in soils. Arizona Agr, Exp, Sta. Tech, Bull. 31. 188. McKeague, J. A., J. E. Brydon, and N, M, Miles. 1971, Differentiation of forms of extractable iron and aluminum in soils. Soil Sci. Soc. Amer. Proc. 35:3; 189. McLean, F, T., and B. E, Gilbert. 1928, Aluminum toxicity. Plant Pnysiol. 3:293-301 190. McLean, E, 0., D, C. Reicosky, and C. Lakshmanan, 1965. Aluminum in soils. VII. Interrelationships of organic matter, liming, and extractable aluminum with "permanent charge" (KCl), and pH dependent cation -exchange capacity of surface soils. Soil Sci. Soc. Amer. Proc. 29:37^-378. 191. McLeod, L c B., and L. P. Jackson. 1967. Water soluble and exchangeable aluminum in acid soils as affected by limins; and fertilisation. Can. J. Soil Sci. 47:203-210. 192. Mehlich, A. 1942. Base unsaturation and pH in relation to soil type. Soil Sci. Soc, Amer. Proc. (19^1). 6:150-156. 193 • . 1946. Soil properties affecting the proportionate amounts of calcium, magnesium and potassium in plants and in HC1 extracts. Soil Sci. 62:393-409. 194. • 1950. Cation ex change -anion exchange ratios of kaolinite soil colloids. Int. Congr. Soil Sci. Trans. 5th. (Amsterdam, Holland) 1: 133-135 . 195» , 1952. Effect of Fe and Al oxides on the release of Ca and on the cation -anion exchange properties of soils. Soil Sci. 73:361-37^. 196. . i960. Gharge characterization of soils. Int. Congr. Soil Sci. Trans 7th. (Madison, Wis.) 11:292-302.

PAGE 237

222 197. Hehlich, A. 1964. Influence of sorted hydroilfate on neutralisation of soil acidity. Soil Sci. Soc. Amer. Proc. 28:492-496. 193. _____ . 1964. Influence of sorbed hydroxy 1 and sulfate on liming efficiency, pH, and conductivity. Soil Sci. Soc. Amer. Proc. 28s496-499. 199. „___»__» ™& "• E » Colwell. 19^4, Influence of nature of soil colloids and degree of base saturation on growth and nutrient uptake by cotton and soybeans . Soil Sci. Sec. Amer. Proc. (1943) 8; 179-184. 200. __ _, and J. F. Reed. 1950. Hydrogen in soils in its relation to mineral content of plants. Int. Congr. Soil Sci. Trans. 4th. (Amsterdam, Holland) 1:251-254. 201. Mendenhall, W. 1963. Introduction to linear models and the design and analysis of experiments. Wadsworth Publishing Co., Inc., California. 202. Monteith, N. H., and G. D. Sherman. 1962, A cc parison of use of liming materials on a hydrol hui Lc lal humic ferruginous latosol in Hawaii,, Int. Soc. Soil S (Palmerston N., New Seal.) Trans, Coram. IV and V„ p. 584-587. 203. Mortensen, J. L. I965. Partial extraction of organic matter. In C. A. Black (ed.) Methods of soil analysis. Part II. Amer. Soc. Agron. Madison, Wis. p . 1404, 204 __________ . and F. L. Himes. 1964. Soil organic matter In fTs. Bear (ed.) Chemistry of the soil. Reinhold Publishing Corporation, New York. 205. Moschler, W. W., G. D. Jones, and tf. G. Thomas, i960. Lime and soil acidity effects on alfalfa growth in a red-yellow podzolic soil. Soil Sci. Soc. Amer. Proc. 24:507-509. 206. Muljadi, D., A. K. Posner, and J. P. Quirk. 1964. The adsorption of phosphate by kaolinite and aluminum oxides. Int. Congr. Soil Sci. Trans. 8th. (Bucharest, Romania) IV:509-523. 207. , , and ____. 1966. The mechanism of phosphate adsorption by kaolinite, gibbsite, and pseudoboehmite. 1. The isotherms and the effect of pH on adsorption. J. Soil Sci. 17:212-229. 208 • _, , and . 1966. The mechanism of phosphate adsorption by kaolinite, gibbsite, and pseudoboehmite II. The location of the adsorution sites. J. Soil Sci. 17:230237.

PAGE 238

223 209. Muljadi, D., A. M. Posner, and J. P. Quirk. 1966. The mechanism of phosphate adsorption by kaolinite, gibbslfe, and pseudoboehmite. III. The effect of temperature on adsorption. J. Soil Sci. 17*238-2^7. 210. Murraman, R. P., and H. Peech. 1969. Effect of pH on labile and soluble phosphate in soils. Soil Sci. Soc. Amer. Proc. 33:205-210. 211. Nichols, R. 1964. Studies on the major element deficiencies of the pigeon pea (Caja: ) in sand culture. 1. Foliar symptoms of the major ei ciencies. Plant and Soil 21:377-387. 212. . 1965. Studies on the major element deficiencies of the pigeon pea (CaJ aj culture. II. The effects of major e] :cies on nodulation, and mineral composition. Plant and Soil 22s 112-126. 213. Nightingale, H. I., and R. L. Smith. 1967.Evic the Dresence of calcium-organic complexes in sodic soils. ! Sci. 1035 261-264. 214. Niu Ching Wgn. 1961. Mature of humus in soils of the tropics and humid sub-tropics, Soviet Soil Sci. 506-512. 215. Norris, D. 0. 1959. The role of calcium and m .'urn in the nutrition of Rhizobium. Aust. J. Agric. Res. 10 1651-698. 216. _ . 1965. Acid production by Rhisobium. A unifying concept. Plant and Soil 22:143-166. 217. liutman, P. S, 1965. Symbiotic nitrogen fixation. In tf. V. Bartholomew, and F. E. Clark (ed.) Soil nitrogen. Amer. Soc. Agron. Madison, His. p. 360. 218. Nye, P., D. Craig, N. T. Coleman, and J. L. Ragland. I96I. Ion exchange equilibria involving aluminum. Soil Sci, Soc. Amer. Proc. 25:14-17. 219. Nye, P. H. I961. Organic matter and nutrient cycles under moist tropical forest. Plant and Soil 13:333-346. 220. , and M. H. Bertheux. 1957. The distribution of phosphorus in forest and savannah soils of the Gold Coast, and its agricultural significance. J. Agri. Sci. 49:145-149. 221. , and D. J. Greenland, i960. The soil under shifting cultivation. Tech. Comm. No. 51. Comm. Bureaux of Soils. Comm. Agric. Bureaux. Bucks., England.

PAGE 239

224 222. Oliveira, V., A. E e Ludwick, and M. T. 3 catty. 197-1. Potassium removed from some southern Brazilian soils by exhaustive cropping and chemical extraction methods. Soil Sci. Soc. Amer. Proc. 35»763-?67. 223. Olphen, H, van. 1963. An introduction to clay colloid chemistry. Interscience (Wiley), New York. 224* Ouellette, G. J., and L. Dessureaux. 1958. Chemical composj of alfalfa as related to degree of tolerance to manganese and aluminum. Can. J. Plant Sci. 38:206-2 225. Ozanne, P. G. I962. Some nutritional problems characteristic of sandy soils. Int. Soc. Soil Sci. (Palmerston N., Zeal.) Trans. Coram. IV and V. p. 139-143* 226. Patrick, W. H. 1964. Extractable iron and phosphorus in a submerged soil at controlled redox potentials. Int. Congr. Soil Sci. Trans. 5th (Bucharest, Romania) p. 605-609. 227. Peaslee, D. E., end H. E. Phillips, 1970* £ 1 of P in kaolinite treated with Ca or NHj, salts , h an organic dye. Soil Sci. Soc. Amer. Proc. 34$ 198-201. 228. Peech, H. 1965. Exchange acidity. In C. A. Black (ed.) Methods of soil analysis. Part 2. Amer. Soc. Agron., Madison, Wisconsin, p. 905. 229. Phillips, R. E e , G. A. Place, and D. A. Brown. 1968. Selfdiffusion of phosphorus in clays and soils: I, The effects of phosphorus rate. Soil Sci. Soc. Amer. Proc. 32:41-44. 230. Pioncke, H. B., and R. B. Corey. I967. Relations between acidic aluminum and soil pH, clay and organic matter. Soil Sci. Soc. Amer. Proc. 31 * 749-752. 231. Place, G. A., R. E. Phillips, and D. A. Brown. 1968. Selfdiffusion of phosphorus in clays and soils: II. The effect of pH. Soil Sci. Soc. Amer. Proc. 32:657-660. 232. Plucknett, D. L., and G. D. Sherman. 1963. Extractable aluminum in some Hawaiian soils. Soil Sci. Soc. Amer. Proc. 27:39-41. 233* Ponomareva, V. V. 1962. Role of humus substances in the formation of brown forest soil. Soviet Soil Sci. 1335-1344. 234. Popenoe, H. i960. Some soil cation relationships in an area of shifting cultivation in the humid tropics. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 11:303-311.

PAGE 240

225 235. j ' H * 1 ^ lK Titration curves of humic acid. Int. ( , Soil Sci. Trans. 8th. (Bucharest, Romania). Ill i 16117H. 23 5 __ , 1966. The humic acids extracted by various reagents" fr"om a soil. I. Yield, inorganic compounds, and rves. J. Soil Sci. 17:65-78. 237 Pratt P. F. 1961. Fhosr>horus and aluminum interactions in Lcation of soils. Soil Sci. Soc. Amer. Proc. 25s^67-^69. 238. Puri, A. N., and R. D. Swarnakar. 1969. Determining available p h ic acid in soils. Geoderma 2:85-93. 239. Ragland, J. L. f and N. T. Coleman. 1959The effect of soil soli and calcium in root growth. Soil Sci. 00c. . 2. 23:355-357. 240. , and __• I960. The hydrolysis of aluminum salts in clay and soil systems. Soil Sci. Soc. Amer. oc. 24:457-^60. 24] Rains, D. W„ W. E. Schmid, and E. Epstein. 1964. Absorption of cations bv roots, Effects of hydrogen ions and essential role of calcium. Plant Physiol. 39:27^-278. 24?. Randall, P. J., and P. B. Vose. I963. Effect of aluminum on id translocation of phosphorus^ by perennial ryegrass. P] hysiol. 38:^03; i09. 243 Ranney, R. W. 1969. An organic carbon-organic matter conversion equation for Pennyslvania surface soils. Soil Sci. Soc. Amer. Proc. 33:809-811. 244. Ramulu, U. S. S., P. F. Pratt, and A. L. Page. 19o7. Phosphorus fixation bv soils in relation to extractable iron oxides and mineralogical composition. Soil Sci. Soc. Amer. Proc. 31*193196. 245 Recs, W. J., and G. H. Sidrak. 1961. Inter-relationship of aluminum and manganese toxicities towards plants. Plant and Soil 14:101-117. 246. Reeve, N. G., and K. E. Sumner. 1970. Effects of aluminum toxicity and phosphorus fixation on crop growth on Oxisols in Natal. Soil Sci. Soc. Amer. Proc. 34:2o3-2o7. 2^7 , and . 1970. Lime requirements of "Natal Cxisols based on exchangeable aluminum. Soil Sci. Soc. Amer. Proc. 3^*595-598.

PAGE 241

226 248. Reeve, N. G., and N. E. Sumner. 1971. Cation exchange capacity and excl :;oil Sci. Soc. Amer. Proc. 35«38-42. 249. Reifenberg, A., and S. J. Buckwold. 195^» The release of silica from soils by the orthophosphate anion. J. Soil Sci. 5:106-115. 250. Rich,. C. I. 1968. Applications of : in soil chemistry and fertf science and engineering. Soil Sci. Soc. Amer, Froc. Madison, Uis. 251. Richards, P. W. 1952. The tropical rain forest. An ecolo study. Cambridge Univ. Press, Cambridge, England. 252. Richburg, J. S., and F. Adams. 1970. Solubility of aluminum in soil solutions and saturation-paste extracts. Soil Sci. Soc. Amer. Proc. 34 « 729-73'+. 253. Riley, D., and S. A. Barber. 1970. Salt accumulation at soybean (Glycine max. (L) Merr.) root-soil interface. Sci. Soc. Amer. Proc. Jki 154-155. 254. __, and . 1971« Effect of nitrate fertilization on phosphorus uptake r . to rootsoil interface. Soil Sci. Soc. Amer. Proc. 3'' 255. Rixon, A. J., and G. D. Sherman. 1962. Effects of heavy lime applications to volcanic ash soils in the humid tropics. Soil Sci. 9^ « 17-27. 256. Rorison, I. H. 1958. The effect of aluminum on legume nutrition, In E. G. Hallsworth (ed) Nutrition of the legumes. Butterworths, London, p. 43. 257. » c « D » Sutton, and E. G. Hallsworth. 1958. The effects of climatic conditions on aluminum and manganese toxicities. In E. G. Hallsworth (ed.) Nutrition of the legumes. Butterworths, London, p. 62. 258. Ross, G. J., and R. C. Turner. 1971. Effect of different anions on the crystallization of aluminum hydroxide in partially neutralized aqueous aluminum salt systems. Soil Sci. Soc. Amer. Proc. 35'389-392. 259. Rovira, A. D., and B. M. McDougall. 1967. Microbial and biochemical aspects of the rh5"osphere. In A, D. McLaren and G. H. Peterson (ed.) Soil biochemistry. Marcel Dekker, Inc., New York. p. 417.

PAGE 242

227 260. Russell, J. S. I960. Soil fertility changes in the lone-tern ital plots at Kybobolite, South Australia. 1. Chin pH, total nitrogen, organic carbon, and bulk density, Aust. J. Agr. Res. 118 902-926. 261. . I960, Soil fertility changes in the long-term experimental plots at Kybobolite, South Australia. 2, Changes in phosphorus. Aust. J. Agr. Res. 11:927-947. 262. , . 196I. Soil fertility changes in the longterm experimental plots at Kybobolite, South Australia. 3. Changes in cation exchange capacity and exchangeable cations. Aust. J. Agr. Res. 12:273-285. 263. S _ f D. Clarkson, and D. D. Davies. 1965. D N A syi in aluminum-treated roots of barley. Science. 148:1476-1477. 264. Samson, K. R, 1952. Fluoride adsorption by clay minerals \ hydrated alumina. Clay Minerals Bulletin 1:267-270. 265. Schofield, R. K. 1949. Effect of pH on electric charges carried by clay particles, J. Soil Sci. 1:1$. .. 266 • —_._-__-__._«____» and H. R. Samson. 1953* The c ;flocculation of kaolinite suspensions and the accompanying d ": over from positive to negative chloride adsorption. Clay Minerals Bulletin 2:45-50. 267. Schnitzer, M. 1969« Reactions between fulvic acid, a soil humic compound and inorganic soil constituents. Soil Sci. Soc. Proc. 33s 75-81. 268. _. .. _, and J. G. Desjardins. 19o2. Molecular and equivalent weights of the organic matter of a podzol. Soil Sci, Soc. Amer. Proc. 26:362-365. 2 ^9» _____„_____» and S. I. M. Skinner. I963. Organo-metallic interactions in soils*. 1, Reactions between a number of metal ions on the organic matter of a podzol Bh. horizon. Soil Sci. 96:86-93. 270. _, and . 1964. Organo-metallic interactions in soils: 3. Properties of iron and aluminum organic matter complexes, prepared in the laboratory and extracted from a soil. Soil Sci. 98:197-203. 271. , , and _. I965. Organo-metallic interactions in soils: 4. Carboxyl and phenolic hydroxyl groups in organic matter and metal retention. Soil Sci. 99 » 2 78-284.

PAGE 243

228 272. Schnitzer, M., and S. I. M. Skinner. 196?. Organo-metallic_ + .tactions In soils: 7, Stability constants of Pb2-' f ', Ki, Mn~ , Co, Ca, and Mg=fulvic acid complexes. Soil Sci. 103 l 2^7-252 e 273. Schuffelen, A. C, and H. burg. 195^. Structural deterioration of lateritic soils through liming. Int. Congr. Soil Sci. Trans. 5th. (Leopoldville, Belgian Congo) 11:158165. 27^, Schwertmann, U., and M. L. Jackson. 1964. Influence of hydr< aluminum ions on pH titration curves of 1 -aluminum > clays. Soil Sci. Soc. Amer. Proc. 2^:179-183. 275. Shelton, J. E., and N. T. Coleman. 1968. Ino] ihosphorus fractii their relationship to res3 applications of phosphorus on high phosphorus fixing i Soil Sci. Soc. Amer. Proc. 32:91-94. 276. Sherman, C, D. f and C c K. Fujimoto. 1946. The effect of the use of lime, soil fumigants, and mulch on the solubility of iganese in Hawaiian soils. Soil Sci. Soc t Ai Proc. II? 206-210. 277. Simonson, R. If. 1970. Loss of nutrient elements during soil formation. In Nutrient mobility in soils. Accumulation and losses. Soil Sci. Soc. Amer. Madison, Wis. ] 278. Small, J. 1946. pH and plants. Van Nostrand, ...ondon. 279* Smith, P. F. 1962. Mineral analysis of plant tissues. Ann. Rev. Plant Physiol. 13:81-108. 280. Synder, G. H., D. C. Reicosky, E. 0. McLean, and R. E. Franklin. 1968. Clay retention, activities, and excised root uptake of ions in bentonite suspensions and dialyzates. Soil Sci. Soc. Amer. Proc. 32:476-480. 281. Soileau, J. 11., 0. P. Engelstad, and J. B. Martin, Jr. 1969. Cotton growth in an acid fragipan subsoil: II. Effects of soluble calcium, magnesium and aluminum on roots and tops. Soil Sci. Soc. Amer. Proc. 33s919-924. 282. . and R. J. McCracken. 1967. Free iron and coloration in certain well-drained coastal plain soils in relation to their other properties and classification. Soil Sci. Soc. Amer. Proc. 31*248-255.

PAGE 244

229 283. Sombroek, W. G. 1966. Amazon soils A reconnaisance of ti soils of the on region. Wageningeni Centre for Agricultui tion. 284. Spencer, W. F. 1954. 1 < >0 cation 1 jactions on retention and availability of cations in sandy soils. Soil Sci. 77:129-136. 285. Stelly, •-.., and R. Ricaud. i960, ' L exi -actants as a diagnostic aid in • Levels of "available" soil nhosphorus. Int. Congr. Soil Sci. Trans, 7th. (Madison, Wis.) Ill i60-71. 285. Stevenson, F. J. I967. Organic acids in soil. In A. D„ McLaren and 'eterson (ed.) Soil bioches Lstry, Dekker, Inc., New York. 287. Stewart, tf. D. 1966. Nitrogen fixation in plants, Univ., London, Athlone Press, London. 288. Sumner, H, E. I963. Effect of iron oxides on positive negative charges in clays and soils. Clay Mine 5:218-226. 289. Swain, F 1 , M. 19'; ' ochemistry of 'humus. .' Jreger (ed.) Organic Geochemistry. Lllan Company, New York. 290. Tan, K. H e 1955, On the pedogenetic role of organic matter in volcanic ash soils under trcracal conditions. Soil Sci. and Plant Nutrition 12:80-33. 291. , and F. E. Clark. 1969« Polysaccharide constituents in fulvic and humic acids extracted from soil. Geoder 2:2^-5-255. 292. _, and J. Van Schuylenborgh, I96I. On the organic matter in tropical soils. Neth. J. Agr. Sci. 9:174-180. 293. Tandon, K. L. S. 1970. Fluoride extractable Al in soils: 2. As an index of phosrhate retention by soils. Soil Sci. 109:1318. 294. Taylor, A. W., S. L. Gurney, and W. L. Lindsay. 1950. An evaluation of some iron and aluminum phosphates as sources of phosphates for plants. Soil Sci. 90:25-31. 295. » . and E. C. Moreno. 1964. Precipitation of phosphate from calcium phosphate by iron oxide and aluminum hydroxide. Soil Sci. Soc. Amer. Proc. 28:49-52.

PAGE 245

297. 230 296. Thomas, G. W. I960. Effects of electrolyte imbibition cation exchange behaviour of soils. Soil Sci. Soc. / Proc. 24s 329-332. , and N. T. Coleman. 1959= A chromatographic approach to the leaching of fertilizer salts in soils. Soil Sci. Soc. Amer. Proc. 23*113-116. 298. Thorp, J., L. E. Strong, and E. Gamble. 1957° Experiments in soil genesis The role of leaching. Soil S. Proc. 21 s 99-102. 299. Tit ova, N. A. 1962. Iron humus complexes of certain soils. Soviet Soil Sci. 1351-1356. 300. Tokudome, S,, and I. Karmo. 1968. Nature of Japanese soils. Int. Congr. Soil Sci. Trans. 9th. (Adelaide, Aust.) His 163-173. 301. Tweneboah, C. K., D. J. Greenland, and J. M. Oades. 1967. Changes in charge characteristics of soils after treatment with 0.5 M calcium chloride at pH 1.5. Aust. J. Soil Res. 5»24?-26l. 302. Ulrich, A., and F. J. Hills. 1967. Principle plant s. In Soil testing and plant i Plant analysis. Soil Sci. Soc. Amer. Inc., Madison, Wis. p. 11, 303. Vann, J. H, 1969. Landforms, vegetation, and sea level changes along the Guiana coast of South America. Tech, Report #3. State Univ. Coll., Buffalo, N. Y, 304. Van Reevwijk, L. P., and J. H. DeVilliers, 1968. Potassium fixation by amorphous alumino-silica gels. Soil Sci. Soc. Amer. Proc. 32 1 238-2^0. 305. Van Schreven, D, A. 1958. Some factors affecting the uptake of nitrogen by legumes. In E. 0. Hallsworth (ed.) Nutrition of the legumes. Butterworths, London, p. 137 • 306. Van Schuylenborgb, J, i960. The electrokinetic behaviour of the sesquioxide hydrates and its bearing on the genesis of clay minerals. Inr. Congr. Soil Sci. Trans. 4th. (Amsterdam, Holland) I $89-92. 307. . 1965. The formation of sesqui oxides in soils. In S. G. Hallsworth and D. V. Crawford (ed.) Experimental pedology. Proc. 11th Easter School in Ag. Sci. Univ., Nottingham. Butterworths, London, p. 113«

PAGE 246

231 308. Van Weseraael, J. C, and J. J. Lehr. 195^* The influence ability of hate in iron ri Int. Congr. Soil Sci. Trans. 5th. (Leopoldville, Belgian Congo), II : 273-279. 309. Vincent, J. M. ilcium and magrowth of Rhizobium. J. Gen. Microbiol. ->3. 310. _, 1965. Environmental factors in the fixation of nitrogen by " In W. V. Bartholom (ed.) Soil nitrogen. Amer. ! 311. Vincents -Chandler, J. I966. The role of fertilisers in humid tropical pastures. Soil and Crop Sci. Soc. Fla. Proc. 26 $ 328360. 312. Viands, J. 3-953* Acid soil infertility as related to soil solution and solid-phase effects. Soil Sci. 755 333-: 313. Volk, G. M., and C. E. Bell. 19^5. Some major factors in lea< Lei' Lum, sulfur, and 1 rrom sandy soils. A lysimeter study. Fla. Ag. Exp. Sta. Bull. *H6. 31k t Vose, P. B., and P. J. Randall, 1962. Resistance to 3 and manganese toxicities related to variety and cation exch capacity. Nature (London). 196:85-86. 315. Wagenaar, G. A. W. 1965. Report on the soil survey project British Guiana. Hi Part II. Soil management. F. A. 0. Rome. 316. Wagner, G. H«, and F. J. Stevenson. 1965. Structural arrangement of functional groups in soil humic acid as revealed by infra-red analysis. Soil Sci, Soc. Amer. Proc. 29:«VH*8» 317. Waksman, S. A. 1925« What is humus? Proc. Nat'l. Acad. Sci. lliU63-568. 318. Walker, T. W. 1965. The significance of phosphorus in pedogensis. In E. G. Hallsworth and D. V. Crawford (ed) Experimental pedology. Proc. 11th Easter School in Agr. Sci., Univ. of Nottingham. Butterworths, London, p. 295. 319. Weaver, R. M. t J. K. Syers, and K. L. Jackson. 1968. Determination of silica in citrate-bicarbonate-dithionite extracts of soils. Soil Sci. Soc. Amer. Proc. 32:497-501. 320. Weir, C. C, and R. J. Soper. I963. Interaction of phosphate with ferric organic complexes. Can. J. Soil Sci. ^3* 393-399* 321. Went, F. W., and N. Starke. 1968. Mycorrhiza. Bioscience 18 1 1035-1039.

PAGE 247

232 322. Whittig, L„ D. 1965. X-ray diffraction techniques for mineral identification and mineralogical composition. In C. A. Black (ed.) Methods of soil analysis. Part I. Amer. Soc. Agron. Madison, Wis. p. 67I. 323. Whyte, R. C, G. Nilsson-Leissner, and H. G. Trumble. 1953. Legumes in agriculture. F. A, 0. Italy. 324. Wiklander, L. i960. Influence of liming on adsorption and desorption of cations in soils. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 11:283-291. 325. __. 1964. Cation and anion exchange phenomena. In F. E. Bear (ed.) Chemistry of the soil. 2nd edition. Reinhold Publishing Corporation, New York. 326 e . and S e K. Ghosh. 1970 » Continuous ion exchange I. Replacement of magnesium and calcium in kaolinite, illite, and montmorillonite. Acta Agriculturae Scandinavica 20:105-116. 327. Wild, A. 19*1-9. The retention of phosphate by soil. A review. J. Soil Sci. 1:331-238. 328. . 1950. The effect of neutral salts on the retention of P by clay at various pH values. Int. Congr. Soil Sci. Trans. 4th. (Amsterdam, Holland) It 146 -149. 329. Williams, C. H. f and C. M. Donald. 1957. Changes in organic matter and pH in a podzolic soil as influenced by subterranean clover and superphosphate. Aust. J. Agr. Res. 8:179-189. 330. Williams, E. G. i960. Some aspects of phosphate retention and availability in soils. Int. Congr. Soil Sci. Trans. 7th. (Madison, Wis.) 111:604-611. 331. _. 1962. Chemical tests as an aid to increased productivity. Int. Soc. Sci. (Palmerston N., New Zeal.) Trans. Comm. IV and V. p. 820-834. 332. Wright, K. S. 1937. Effects of phosphorus and lime in reducing aluminum toxicity of acid soils. Plant Physiol. 12:173-181. 333. . 1945. Aluminum toxicity: Microchemical tests for inorganically and organically bound phosphorus. Plant Physiol. 20:310-312. 334. . 1943. Internal precipitation of phosphorus in relation to aluminum toxicity. Plant Physiol. 18:708-712.

PAGE 248

233 335° Wright, K. E. t and B. A. Dona-hue. 1953. Alu I :icity studies with radioactive phosphorus. Plant Physiol. 23:6?4-680. 336. Wright, J. R., and M. Schnitzer. i960. functional groups in the organic matter of the A Q and B n horizons of a podzol. Int. Congr. Soil Sci. Trans. 7th. ison, Wis.) lis 120-127. 337. Yassoglou, N. J., C. Nobel, and S. C. Vrahamis. 1969. A study of some biosequences and lithosequences in the zone forest soils in Northern lological, properties. Soil Sci. Soc. Amer. Proc. 3 338. Younge, C R., and D. L. Plucknett. 19 tching the high phosphorus fixation of Hawaiian latosols. Soil Sole Soc. Amer". Proc. 30:653-655. 339. Younts, S. 3. 19?1. Trends in soil fertilJ Lant nutrition. In J. D. Eastin and R. D. Munson (ed.) Moving of the yield" plateau. Amer. Soc. Agron. Kadis on, Wis, p. 69. 340. Yuan, T. L. 1963* Some relationships among ' and pH in solution and soil systems. So .-163. 3^1* . 1964. Comparison of reagents for s; matter extraction and effect of pH on subsequent separation of humic and fulvic acids. Soil Sci.' 98 : 113-141. 342. . 1965. A survey of aluminum status in Florida soils. Soil and Crop Sci. Soc. Fla. Proc. 25* 1*3-152. 3^3. ____-„ • 1966. Characteristics of surface and spodic horizons of some spodosols. Soil and Crop Sci. Soc. Fla. Proc. 26:16217*. 3^4. . 1969, Composition of the amorphous materials in the cla»y fraction of some Sntisols, Inceptisols, and Spodosols. Soil Sci. 107:242-248. 3*5* t and J « G « A « Fiskell. 1959. Aluminum studies II. The extraction of aluminum from some Florida soils. Soil Sci. Soc. Amer. Proc. 23:202-205. y&» • N Gammon, Jr., and R. G. Leighty. 1967. Relative contribution of organic and clay fractions to cation exchange capacity of sandy soils from several soil groups. Soil Sci. 104:123-128.

PAGE 249

234 3'.i?. Yuan, T. L., W, K. Robertson, and R. J. Neller. .I960. Forms of ee aoid sandy soils. Soil Sex. 3ii8. V. 1952. Principles of forest, soil classification . the U. 3. S. R. Int. Soc, Comra. IV and V. (Palmerston N., .) p. 808-819.

PAGE 250

BIOGRAPHICAL SKETCH Alfred V3 3 11, 1937. at Lichfield Village, West C British Guiana, to Clinton E. and Ruth A. Downer. He received his I iry and secondary education in British Guiana, then attended the University of Reading, Berkshire, England, fr duated in 196l 1 degree of .;• of Science in Agricultv He worked in the Ext< •vies cf the Ministry of Agriculture, British Guiana, from July, 1961, to December, 19&5, and then with the Research Division cf the Ministry. He enrolled at Tuskegee Institute, ma, in February, 1967, and completed the degree of Mastc Science under the guidance of Dr. K. S. Chahal in August, 1968. He came to the University of Florida in September, I968, with the award of a graduate assistantship from the Center for Tropical Agriculture and the Department of Soils, and since then has worked toward the degree of Doctor of Philosophy in Soil Science. Alfred Victor Downer is married to the former Loretts F. Henry, and is the father of three children. He is an associate member of Sigma Xi honor society, and a member of the Soil Science Society of America, the American Society of Agronomy, and the American Institute of Biological Sciences. 235

PAGE 251

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. William G. Blue, Chairman Professor of Soils I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. ^fu^^ Jly U).J/ A^ Tzu L. Yua Associate Professor of Soils I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. Victor W. Carlisle Associate Professor of Soils I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. VLSfh "lAMte— /J Ann E. Moore \Associate Professor of Animal Science

PAGE 252

I certify that I have read this study and that in my opinion it conforms to acceptable standards of scholarly presentation and is fully adequate, in scope and quality, as a dissertation for the degree of Doctor of Philosophy. . -.&~ Gerald 0. Mott Professor of Agronomy This dissertation was submitted to the Dean of the College of Agriculture and to the Graduate Council, and was accepted as partial fulfillment of the requirements for the degree of Doctor of Philosophy. March, 1972 Qi z_ n. College of Agriculture Dean, Graduate School

PAGE 253


xml version 1.0 encoding UTF-8
REPORT xmlns http:www.fcla.edudlsmddaitss xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.fcla.edudlsmddaitssdaitssReport.xsd
INGEST IEID ECSO4IS6Y_5INE7I INGEST_TIME 2017-07-14T23:15:20Z PACKAGE UF00097607_00001