Inaug.-Diss.--Halle-Wittenberg. / Vita.

Inaug.--Diss.--Heidelberg. / Lebenslauf. "Literatur": vi p. following p. 106.

Thesis--Eberhard-Karis-Universität zu Tübingen. / Vita. Includes bibliographical references.

Thesis (M. Sc.)--University of Hong Kong, 1961. / Type-written copy. Includes bibliographical references (leaves 100-104).

Suspended sediments were collected from 3 rivers, the Rheidol, Ystwyth and Tywi on a regular basis. The sediment samples were studied for a number of physical and chemical properties to identify characteristics which might distinguish between source areas. These properties included mineralogy, magnetic properties, sediment colour, and the trace element composition of .sediments. Point-sources for heavy metals can be identified for all three catchments, and the metal content, and partitioning was studied. The metal content, and the site on the sediment to which the metal is bound has been found to vary with discharge. The response to discharge is not the same for all metals, but in general, at peak discharge the lead was in the Fe/Mn oxide and organic fractions, and these fractions became important for cadmium. Prior to the flood peak however, up to 70% of the cadmium was in easily exchangeable form. The organic fraction was the most important for copper throughout the flood. j Sediments which had been deposited in floodplain features and in lakes were also studied, as they recorded the pattern of sediment transport during floods over an extended period. In the Ystwyth catchment the metal content and ratios between metals was used to determine the importance of two mining areas for the supply of sediment to a floodplain during the mining period (1860-1890). Sediments from profiles in fossil alluvial channels on a floodplain in the Rheidol showed a dramatic increase in metal .concentrations in the mid-profile, and recorded the period of concerted mining in the catchment. Mining operations in the Ystwyth catchment caused widespread lateral aggradation of floodplains, as established at Trawscoed. The historic record of catchment changes as contained in the sedimentary profile of reservoirs in the Mendips was studied. The record in Blagdon and Chew Valley lakes in the Mendips was deciphered using a combination of sedimentological, magnetic and chemical data. A catastrophic flood which is documented for the area, which occurred in July 1968, caused the mobilisation and transportation of sediments from dry valleys in the catchment. Sediments deposited during the floods were distinguished by their coarse nature, by an increase in magnetic susceptibility and the parameter'S', which indicated a topsoil source, and by a change in the partitioning of lead and zinc, where the organic fraction was absent.

published_or_final_version / Civil Engineering / Doctoral / Doctor of Philosophy

This work consists of two distinct and separate sections with differing aims and outcomes. Section A is the critical review, which provides a context for the published work presented in Section B. Section A, the critical review, summarises the current literature in the subject area of environmental aspects of river management and sets the research portfolio in context. The physical landscape has been increasingly influenced by a variety of economic and social changes in recent history. During these approximately 6,000 years, human impact on river systems has been substantial and widespread. The environmental changes, that have stressed river systems, have also damaged their value for both human use and environmental functions. Some of these stresses include changes in water quality and quantity, morphological modifications of the channel and/or floodplain, decline of native species and the introduction of alien species. This has led to changes in current river management philosophy, issues and techniques. Section n, the research portfolio of published work submitted as part of this thesis, focuses on the management of some aspects of environmental change within river systems, particularly problems relating to river rehabilitation and the introduction of invasive riparian plant species. Each published paper has its own individual abstract, summarising the key finding of the research. The research portfolio can be divided into three broad sections. The first section, the main published work, includes those papers approved by the University Research Committee as part of the Registration for the degree of PhD by Published Work in May 2000. These papers form the key part of the thesis. They are introduced in the critical review in the blue text boxes, in order of significance to the text discussed and not in chronological order. They are presented in full in the appendices (numbers as shown below) printed on light blue paper. The next section relates to additional relevant published work. These papers are also submitted as part of this thesis, but are considered to be secondary as a result of their date of publication, i.e. prior to 1995. In order to distinguish these papers clearly from the key papers, they are introduced in the critical review in green text boxes and presented in full in the appendices (numbers as shown below), printed on light green paper. The candidate's approximate contribution and the nature of this contribution to each joint paper have been indicated as a percentage and are shown on the title page of each appendix. The third part is other supporting material and is referred to in the critical review through inclusion of orange-brown text boxes. Since these publications consist mainly of consultancy reports and edited books, they are NOT submitted as part of this thesis and are not presented in full in the appendices. If necessary, this material can be made available to the internal and/or external examiners before or during the viva.

This thesis aims to assess the response of invertebrate assemblages to fine sediment in the River Isábena, a catchment generating unusually high sediment loads due to the presence of badland areas. First, a spatial approach was taken in order to assess evidence of sedimentrelated patterns in invertebrate assemblages across the catchment under stable flow conditions. This provided a “snap‐shot” of the spatial variability on the assemblages in relation to sediment stored in the channel, as well as the relative influence of physical habitat conditions of different spatial scales. Fine sediment proved to be driving invertebrate assemblages distribution, structure and trait composition, filtering‐out sensitive species and thus promoting nested patterns across the catchment. Strong interactions between spatial scales were found, emphasizing the existence of catchment scale controls (such as geology and rainfall) on reach physical habitat conditions and, in turn, on invertebrate assemblages distribution. This spatial approach was complemented with the study of the temporal dynamics, which focussed on understanding the dominant drivers in the catchment hydrosedimentary behaviour and the relationship between physical processes and the ecological response. Temporal distribution of rainfall (and thus the occurrence of floods), along with the sediment availability (both in the source and in the channel) were driving the intra‐annual suspended sediment dynamics. Turnover of invertebrate assemblages was jointly driven by taxon life cycles and the temporal dynamics in hydrosedimentary conditions (i.e. occurrence of floods and episodes of high sedimentation). Despite the intra‐annual variability, assemblages proved to be resilient, most likely as a result of the selection of pre‐adapted traits. Overall, these findings emphasize the importance of assessing the geomorphological and hydrological processes that control sediment production, delivery and permanence in the channel, in order to better understand the ecological impacts of excessive fine sediment loads.

The Northwest Himalaya is a region of rapid rock uplift and a strongly erosive climate that allows the competing influences on drainage development to be assessed in the Indus River. This study used U-Pb zircon and K-feldspar Pb isotope analysis together with conventional heavy minerals to reconstruct flood plain drainage patterns from ~20 ka. Furthermore, clay mineralogy is used to reconstruct changes in chemical weathering. U-Pb ages for zircon grains from the Indus tributaries were compared with available bedrock data in order to constrain the source of the sediment reaching the Arabian Sea. The trunk Indus is typified by <200 Ma zircons, in contrast with >800 Ma in the eastern tributaries, eroded from Himalayan sources. A significant population of grains <200 Ma in Thar Desert indicates monsoon-related eolian transport from the lower Indus. Modelling of modern delta sand that is rich in >1700 Ma zircons contrasts with modern water discharge which is dominated by the trunk Indus indicating preferential Lesser Himalayan erosion before ~7 ka. Pb isotope compositions of K-feldspars were used as an additional provenance constraint. The eastern tributaries show a clear Himalayan provenance, contrasting with radiogenic grains in the trunk Indus. Thar desert sands show isotopic values that suggest significant erosion from Karakoram, consistent with the zircon dating, as well as heavy mineral data. In turn Holocene river sands from the western edge of the desert indicate increasing reworking from the dunes prior to ~4.5 ka, linked to climatic drying. XRD clay mineralogy shows increasing smectite in the delta at 13–7.5 ka, indicating stronger chemical weathering as the summer monsoon intensified. In contrast, the upper flood plains show evidence for increased chemical weathering after ~7 ka linked to the cessation of fluvial transport under the influence of a weakening monsoon.

The character of montane channel morphology and associated hierarchical controls was investigated in the Dee catchment, Cairngorm Mountains, north-east Scotland. Montane channel morphology in Scotland is of considerable importance given its relatively undisturbed condition in a UK context, variety and for providing habitat for several important lotic species. Nine distinctive sub-catchments were chosen to investigate the linkages between landscape controls and channel morphology distribution. The distribution of channel morphology at the reach scale was mapped using an expanded version of a process-based classification system originally developed in the Pacific northwest, USA. Continuous mapping revealed a wide variety and irregular distribution of channel morphology that is influenced primarily by a suite of glacigenic valley bottom controls. Differences in channel morphology distribution were apparent between catchments reflecting the influence of unique landscape evolution histories. In addition, fifty reaches exhibiting a variety of morphology and associated geomorphic setting, were surveyed in the field to explore in more detail controls on channel morphology. The results generally confirm the relevancy of the typology in the region and the dominant control exerted by slope. However the importance of scale, local controls and the regional geomorphic context was also highlighted. Finally, GIS approaches to channel characterisation at the catchment scale were tested to assess their potential usefulness for catchment management applications. By using a combination of the previous results, GIS models were calibrated and tested to predict the distribution of channel type and Atlantic salmon spawning habitat. The accuracy of channel type predictions were compromised by the model criteria and quality of geospatial data used. However the potential utility of the spawning habitat model as a first order method for screening habitat suitability over large areas was demonstrated.