Mitchell, Nicholas: Rainforest change analysis in Eastern Africa : A new multisourced, semi-quantitative approach to investigating more than 100 years of forest cover disturbance. - Bonn, 2012. - Dissertation, Rheinische Friedrich-Wilhelms-Universität Bonn.
Online-Ausgabe in bonndoc:
author = {{Nicholas Mitchell}},
title = {Rainforest change analysis in Eastern Africa : A new multisourced, semi-quantitative approach to investigating more than 100 years of forest cover disturbance},
school = {Rheinische Friedrich-Wilhelms-Universität Bonn},
year = 2012,
month = jan,

note = {Forest change and disturbance of the past strongly influence the state of today’s forests and their biodiversity. However, knowledge of former forest landscape states can be subject to misunderstanding and the practical management of forests requires the establishment of correct narratives of forest cover change. This thesis therefore investigates the long-term forest change and anthropogenic factors at work within three tropical rain forests of high biodiversity and high use value in Kenya and Uganda.
A wide range of data sources are employed for a semi-quantitative analysis. Starting from an existing time series of satellite imagery classifications the research incorporates the visual interpretation of historical aerial photography, forestry records, maps of both topographic and thematic type, archive documents, oral histories, place name meanings, and fossil pollen evidence. GIS is used as the means to manage and focus the evidence and to analyse the wide range of data.
In combination the sources allow the building of a narrative characterised by variation across both space and time. The localised reality of forest change is reflected in the inclusion of case studies from which forest narratives of each of the three main forest areas are subsequently constructed. The forest cover time series are extended back to around 1910 for each of the forests and thus to a pre-commercial exploitation state; they reveal losses of 60% and 43% of the forests of Kakamega-Nandi and Mabira respectively. These losses have been arrested in recent years while Budongo Forest has shown negligible change across the full period with the first losses recently occurring outside the forest reserves.
The long-term approach has revealed fluctuations in forest cover, most notably in Mabira Forest across the 20th century and in parts of the Kakamega-Nandi area both across decades and across millennia. A landscape view shows these areas to have long-existed as mosaics of forest, woodland and grassland, and the loss of grassland over the last century has exceeded that of forest. The study identifies an historic role for disease and tribal conflict in the creation and protection of forest cover in East Africa but also traces a development in the underlying causes of forest cover change towards commercial and governance factors. The creation of a population time series demonstrates that population density cannot be described as the main driver of deforestation. Two spatially-explicit indices distinguish between locally and commercially-driven disturbances and are compared with an index of forest cover change. Results reveal a localised pattern and that commercial disturbance has played an especially large role in the degradation and fragmentation of the Kakamega-Nandi forests while local disturbance is shown to be most dramatic in Mabira Forest. Most of Budongo Forest has been persistently degraded by systematic commercial exploitation.
It is suggested that these forests should be managed with recognition of their mosaic heritage but also as dynamic and changing entities. The study concludes that while the heterogeneity found within forest landscapes is often due to human disturbance, ecologists should also consider natural processes, including variations in past climate, for explanations. The cumulative nature of disturbance is highlighted with the recommendation that past commercial exploitation should be included in any assessment of forest degradation. The use of GIS and the creation of disturbance indices is recommended as a viable means of quantitatively assessing forest degradation and of distinguishing between the contributions of different types of disturbance. The most under-used resources available for researching long-term forest change are stated to be topographic maps and forestry archives. The quantitative data they provide can be usefully supported by qualitative information, most flexibly provided by forest history interviews.},

url = {}

The following license files are associated with this item: