Crusting, soil strength and seedling emergence in Botswana
This thesis gives the results of an investigation of the strengths and particularly the crust forming potential of arable soils from Botswana and the relationship to seedling emergence of sorghum, the most important crop in Botswana and one that because of its small seed weight (about 2 0 mg) can fail to emerge through a hard soil crust. A review of the literature suggested that the soil factors which had to be considered were dispersibility of clay and factors which affect this, and the bulk density of the soils. Soils which are found in many tropical and sub-tropical regions, with low organic matter and inactive clays can set hard after a simple wetting and drying cycle. In these soils, the strength is very strongly dependent on the water content, showing a hyperbolic or exponential relationship-, and the strengths when dry may be very greatly increased by remoulding the wet soil. The crust strength required to prevent seedling emergence varies with the size of the seedling and for cotton (seed weight about 80 mg), 1-3 MPa penetration resistance measured with a penetrometer is sufficient to prevent emergence. Seedlings exert a total force proportional to their number. In the experimental programme, seedlings' forces were measured, seedling emergence observed in a field experiment under crusting conditions, and the strength characteristics of a group of soils, representative of arable soils in Botswana, studied. A sorghum seedling was found to exert a maximum force of about 1 N or dividing by the area of the plumule, a pressure of about 0,5 MPa. The field experiment showed that much better emergence was obtained from planting 15 seeds together than from planting 4 seeds together when a crust formed after planting. A study of 32 soils, most of them sand to sandy loam in texture but with a few clays and hydroirorphic soils, from arable areas in Botswana showed the sandy to sandy loam soils to have high bulk densities ( 1,45-1 ,75 Mg/m3) and extremely low organic carbon contents (0,12-0,85 g/100g). The bulk densities of all the soils were inversely related to the organic carbon content and this was itself related to the clay content of the soils. The bulk densities of the sands were dependent on the grading of the sand fraction. Many of the soils were sensitive to remoulding in the Emerson test and the sands to loany sands had 0,4-1,0 g/100 g water dispersible clay. Measurements of tensile strength on air-dry samples showed that all the soils, except for one sand, set hard after a wetting and drying cycle, giving for vacuum wet samples indirect tensile strengths 1,0-14,4 kPa. For the sands to sandy loams this strength was related to the water dispersible clay content. Samples wet at atmospheric pressure were weaker than the vacuum wet samples, the reduction in strength was related to the air porosity of the non-vacuum wet soils prior to drying. Remoulding the soils prior to drying them increased the strength by a factor of up to 50 times, giving strengths from 4 kPa to 600 kPa. The strength after remoulding was dependent on the Emerson index. Compacting the soils increased their strength greatly and to an extent that agreed with the hypothesis that the strength obtained was proportional to the area of contact between the particles. Experiments on penetration resistance at a range of water contents were performed on a few soils. A hyperbolic relationship between water content and penetration resistance of the surface soil was found for sand to sandy loam soils, with the maximum resistance of dry soils above 2 MPa. The penetration resistance of the sandy loam soil was Increased three times by disturbing it when wet. Sprinkler wetting the sieved soils was not found to affect the penetration resistance by a large amount compared and other methods of welting. Penetration resistance was measured on air-dry samples of most of the main group of soils following varying degrees of wetting with a rainfall simulator. The clays and hydromorphic soils gave very low values of penetration resistance under these conditions, showing that at organic carbon contents of about 1% and clay contents from 20 to 30%, the decreased bulk density and tendency to form aggregates' on drying overcame the tendency to set hard. The mean values for the sands to sandy loams were from 1 to 6 MPa so all these soils could offer significant resistance to a sorghum seedling. The penetration resistance of the sands and loam/ sands depended on their bulk densities and water dispersible clay contents, while the penetration resistance of the sandy loams depended only on the water dispersible clay content.