AbstractsGeography &GIS

The impact of forest loss on catchment watershed functions has been the subject of scientific and public debate since the 1840’s in the temperate world and since the 1920’s in the tropical world. The thesis seeks to explore this debate and the impact of tropical native forest loss and soil degradation on catchment hydrologic response and the related river basin hydrological regimes. Evidence on these hydrological topics has been sought from a literature review, wet tropical case study and a combined bio-physical science, social survey and numerical analysis and modelling approach. The thesis identifies the impact of forest loss on aquifer recharge, low flows, flood patterns and flow seasonality, and the existence of the soil degradation-induced ‘infiltration trade-off’ effect (Bruijnzeel, 1988 ; Bruijnzeel, 1989), these being the main disputed issues of the science and debate. The stakeholders in the debate are identified to have a mismatch of perceptions stemming from their science or observations having been derived from sites with different hydroclimatic and soil conditions and different spatial and temporal scales. The rift exists between: ‘mainstream’ predominantly temperate hydrologists who had focussed on short-term and small-scale research studies on non-degraded soils; and ‘soil-focused’ tropical hydrologists and the tropical public, agency practitioners and policy makers who have focussed on long-term, larger-scale studies with degraded soils. A case study of the wet tropical Singkarak-Ombilin river basin in West Sumatra, Indonesia was conducted in search of answers to disputed questions and to identify whether field evidence supports the ‘soil-focused’ hydrologist’s and tropical public and policy maker’s perceptions. Social survey, data analysis and parametric and non-parametric modelling methodologies were employed in search of a forest loss and land degradation impact on hydrologic response across five study catchments at varying scales (150 – 2,200 km2). Daily time-step rainfall-runoff modelling was applied. A 38 year discharge record was analysed representing a period over which land use change may have registered a hydrologic impact. As input to the models, rainfall across the case study area was represented by a daily rainfall surface generated from 94 monthly and 73 daily rainfall station records. This combination of new approaches and analytical methodologies proved to be a successful approach to hydrological analysis in data sparse tropical study conditions. The social conceptual models identified that forest loss had led to widespread land degradation, with evidence of increased erosion and sediment transport in basins from 150 – 1100 km2 scale and reduced groundwater recharge in basins of < 100 km2 scale. The combined social conceptual models, data analysis and parametric and non-parametric models indicated that forest loss and land degradation had increased flood peaks and reduced low flows (baseflow) in basins of < 320 km2 scale. At larger scale, combined forest loss - land degradation and…