Search

here is conflict between the use of organic matter as a fuel and as a fertiliser in rural areas of sub-Saharan Africa. Increasing the proportion used for fuel has a direct impact on food security. Anaerobic digestion can address this conflict as it provides a clean fuel and converts dung into bioslurry which can be applied to soil. Anaerobic digestion is limited by various factors including shortage of available water and substrate material, and lack of desire to take up new cooking practices. This research found that in rural Ethiopia, 92 % households were willing to try biogas from anaerobic digestion for cooking and 54 % of households had adequate dung for substrate. Based on the removal of chemical oxygen demand, experiments with digestate filtering and recycling found that 1.39 GJ of useful energy per year could be provided by one cow, while 1.61 GJ could be achieved using undiluted dung. Both methods reduce the water demand by at least 75 % and can increase the reach of anaerobic digestion in water limited conditions. Anaerobic digestion also removed Escherichia coli and total coliforms by over 99 %, but effects on antibiotic resistance genes were varied, suggesting more work is required within resistances to specific antibiotic groups. In circumstances where households compost organic material and burn little dung, bioslurry may not be as beneficial for the accumulation of soil carbon as current practices. However, based on the national Ethiopian average of burning 60 % of dung as fuel, a transition to anaerobic digestion would improve soil properties. Anaerobic digestion therefore has a strong potential for improving lives in sub-Saharan Africa, but lack of follow-up and maintenance on recent digester installations have resulted in up to 80 % of digesters being non-functional, indicating different implementation strategies are urgently required.

While carbon fractionation approaches for mineral soils are well established, methods suitable for organic soils are lacking. This research addresses this gap and attempts to devise methods for comparison to model pools, such as used in RothC. It examines impacts on rate of decomposition of burning, as this is likely to increase the inert fraction and impact rate of decomposition. Field sites were established in a Scottish heathland and a Tasmanian buttongrass moorland, both sites being managed by regular burning. Surface samples were taken from plots burned at varying times in the past, and labile fractions were measured by fractionating with hot water and sulphuric acid. There was a greater amount of labile-carbon extracted in plots not recently burnt at both sites, suggesting labile-carbon was either lost or converted to recalcitrant-carbon by fire. The fractions obtained were compared to RothC pools derived by an equilibrium-run, the usual model initialisation approach. There was a larger proportion of labile-carbon measured than predicted by the model, suggesting the equilibrium-run is overestimating inert carbon. Attempts were made to isolate inert carbon in charcoal using chemical and visual methods, but results from both methods were highly variable, so no direct comparison could be made with model pools. Respiration measurements of incubated samples were used to derive carbon pools by inverse-modelling. Again, results showed more labile-carbon than predicted by the equilibriumrun and less labile-carbon in most recently burned treatments. Inverse-modelling shows great promise to provide more accurate model initialisation, so a MicroResp approach was developed as a rapid alternative to gas chromatography and verified using known additions of charcoal. Addition of 1% charcoal lowered respiration rates, whereas 5% and 10% increased them, suggesting the presence of labile carbon in charcoal. Future work should quantify the labile carbon in charcoal and further develop the promising MicroResp approach.

Forest soils contain large amounts of carbon, which can be lost through forest operations or changing environmental conditions. As forests are perennial with infrequent disturbance, soil organic carbon (SOC) accumulation differs from highly disturbed land uses, with significant accumulation occurring in deeper soil horizons (subsoil). This thesis investigated changes in SOC in both topsoil and subsoil under the main native broadleaved and coniferous species in the UK, oak and Scots pine. The overall results showed that prolonging forest rotations could allow more carbon (C) to be sequestered in soils. In Scots pine (Chapter 3), increasing rotation age would allow greater accumulation of the top F and H horizons, while in oak stands (Chapter 2), the most pronounced changes would be in the mineral subsoil horizons. In the oak forest, we demonstrated importance of sampling subsoil horizons (Chapter 2). Including subsoil measurements showed that young forest subsoil had similar amounts of SOC as ancient forests. Long-term RothC simulations of pine forests (Chapter 3) suggested that the lowest soil C stocks occur, not immediately after harvesting, but closer to the middle of the rotation. This shows the importance of including 40 to 50 year old stands in sampling campaigns to capture the magnitude of these losses. The simulations suggested that converting current grassland sites to forest would allow accumulation of similar amounts of SOC as in ancient forest sites. Model SOC pools were compared to measurable fractions for the 220 year old oak stands. Discrepancy between measured fractions and simulated pools (Chapter 4) suggested a significant difference in soil processes in the topsoil and subsoil. Current modelling approaches might not be sufficient to represent factors that control SOC dynamics with depth.