Some aspects of the chemistry and mineralogy of soil potassium in Sri Lankan acid tea soils and Scottish soils under a range of crops
The Sri Lankan acid tea soils, collected from six different agro-ecological regions and from the long-term field experiments in St. Coombs, were subjected to a detailed mineralogical investigation with particular reference to the understanding of the chemistry of soil potassium and K fertiliser applications. These extremely weathered soils contained predominantly kaolinite, Al-chlorite, gibbsite and goethite, but K-fixing minerals like smectite and vermiculite were absent. Small quantities of micaceous minerals were only found in the fine and coarse sand fractions of these soils, except in the Hantana soil which contained substantial quantities of micaceous minerals in the clay, silt and sand fractions. The clay and silt fractions of the Scottish soils of the Countesswells and Foudland soil series comprised 60% to 70% of mica and mica-weathered products of interstratified mica/vermiculite and vermiculite/chlorite. The rate of transformation of the micaceous minerals and the concentration of hydroxy-Al 'islands' in the Countesswells soil series increased with decreasing soil pH. The Foudland soil contained more weatherable micaceous minerals, and spring barley, winter wheat, potatoes and ryegrass/clover in a crop rotation field experiment produced optimum crop yields in the absence of K fertilisers, due to the release of substantial amounts of non-exchangeable K. The Countesswells soil was unable to supply sufficient non-exchangeable K for ryegrass/clover growth in order to achieve the same yield as that with K fertiliser. A mixed NH4- and Cl- ion exchange resin method for extraction of soil K was developed, which was capable of estimating the amounts of short-, medium- and long-term K reserves that are available to ryegrass, grown in different soil types covering a wide range of soil pH and K-bearing minerals, without destroying the mineral matrix, but K-bearing minerals in both soils and pure minerals were found to either transform or dissolve progresively as mineralogical changes took place at the soil/root interface. A new perspex cell was constructed to hold a 4mm layer of soil or mineral, for in situ studies of root morphology, and for study of the mineralogical changes in soils or pure K-bearing minerals, due to the uptake of both exchangeable and non-exchangeable K by ryegrass, within a 2mm radius of the root surface. This study showed that there were more total, primary and lateral ryegrass roots growing in a soil of higher K status, compared with one of lower K status. The numbers of dead or decomposed roots were greater in the lower K status soil. Interstratified smectite, mica/vermiculite and vermiculite/chlorite minerals were formed in the soil within a 2mm radius of the root surface, due to K uptake by ryegrass. These mineralogical changes were not observed in the same soil under intensive cropping with ryegrass in the glasshouse in bigger cylindrical pots or after 21 years of continuous growth of ryegrass in the field. Dissolution of the 2:1 layer silicate mineral structure occurred in soils and vermiculite in the perspex cells, due to the production of H3O- ions, particularly where root growth was greatest. This dissolution mechanism may override the diffusion mechanism of K release to plant roots within a 2mm radius of the root surface, and may help explain the wide range of diffusion coefficients of soil K, (10-10 to 10-23)cm2 sec-1, which appear in the literature. The new techniques of extraction of soil K by use of a mixed NH4- and C1- ion exchange resin and of studying root growth in a perspex cell, which were developed in this thesis, merit further investigation.